Jump to
Abstract
The plasma kinetics of recombinant human cholesteryl ester transfer protein (rCETP) were studied in six rabbits before and after cholesterol feeding (0.5% wt/wt). The rCETP, labeled with the use of the Bolton Hunter reagent, was shown to retain neutral lipid transfer activity. After intravenous infusion, labeled rCETP associated with rabbit lipoproteins to an extent similar to endogenous rabbit CETP (62% to 64% HDL associated). The plasma kinetics of CETP, modeled with the use of SAAM-II, conformed to a two-pool model, likely representing free and loosely HDL-associated CETP (fast pool) and a tightly apo (apolipoprotein) AI–associated (slow pool) CETP. The plasma residency time (chow diet) of the fast pool averaged 7.1 hours and of the slow pool, 76.3 hours. The production rate (PR) into and the fractional catabolic rate (FCR) of the fast pool were 20 and 10 times the PR and FCR, respectively, of the slow pool. In response to cholesterol feeding, CETP PR, FCR, and plasma mass increased by 416%, 60%, and 230%, respectively. There was a strong correlation (r=.95, P=.003) between the increase in rabbit plasma CETP and the modeled increase in CETP PR in response to cholesterol feeding, suggesting that labeled human rCETP is a satisfactory tracer for rabbit plasma CETP. CETP is catabolized by distinct pools, likely corresponding to an apo AI–associated (slow) pool and a free and/or loosely HDL-associated (fast) pool. Factors that alter the affinity of CETP for HDL would be predicted to result in altered CETP catabolism. The effect of dietary cholesterol on plasma CETP mass can be explained largely by the effects on CETP synthesis, consistent with the observed effects of cholesterol on tissue mRNA levels.
- Received February 28, 1996.
- Revision received May 15, 1996.
Cholesteryl ester transfer protein is a hydrophobic glycoprotein that mediates neutral lipid exchange between apo A-I– and apo B–containing lipoproteins.1 Interactions between CETP and plasma lipoproteins are based on weak hydrophobic and ionic forces, with the majority of CETP in human plasma found in loose association with HDL. No data are available on the plasma kinetics of CETP in any species. CETP gene expression is upregulated by cholesterol,2 and plasma CETP increases rapidly in response to cholesterol feeding in humans3 and many other species.4 5 6 7 8 Lipopolysaccharide administration decreases hepatic CETP mRNA levels in CETP transgenic mice, an effect that is apparently mediated by adrenal glucocorticoids.9 Within 8 hours of lipopolysaccharide injection, there is a 50% decrease in plasma CETP that closely parallels the decline in hepatic CETP mRNA. This study suggests that removal of CETP from the plasma pool is rapid. Since plasma clearance of the major apolipoproteins is less rapid, we have hypothesized that the majority of CETP is catabolized from a non-lipoprotein-associated pool.
CETP is present in plasma at low concentrations (2 to 6 μg/mL in different species) and hence isolation and labeling of endogenous CETP or the study of CETP kinetics by stable isotope technology is not feasible. We used purified human recombinant CETP to study the plasma kinetics of CETP in the rabbit. The rabbit was chosen as an animal model because plasma CETP levels are high, tissue sites of CETP synthesis are similar to those of humans, and rabbit CETP has an overall sequence homology of 81% compared with human CETP.10 We studied the plasma kinetics of CETP in rabbits sequentially fed a chow diet and a cholesterol-rich diet.
Methods
Animals
Pathogen-free adult male New Zealand White rabbits weighing 3.0 to 3.5 kg were obtained from Charles River Laboratories. Animals were acclimated for 2 weeks with daily handling before the start of the study. Six animals were fed a chow diet for 4 weeks followed by 4 weeks on a high cholesterol diet. Two animals were studied in reverse order (cholesterol-rich diet followed by chow diet). The study was approved by the Animal Care Committee of this institution.
Diets
Purina pelleted rabbit chow (chow) with or without the addition of 0.5% wt/wt cholesterol (chol) was obtained from Purina Test Diets. Diet was stored under refrigerated conditions and used within 4 months of preparation.
Preparation of Recombinant Human CETP
Plasmids containing a cloned CETP cDNA and the dihydrofolate reductase (dhfr) gene were cotransfected into a dhfr-deficient Chinese hamster ovary (CHO) cell line (ATCC #CLR9096). CHO cells incorporating the plasmid were selected by culturing the cells in hypoxanthine- and thymidine-free Ham's F-12 with 10% dialyzed fetal calf serum selection medium. Surviving cells were tested for CETP production, and stable CETP-secreting CHO clones were isolated. The level of CETP expression was then amplified by subjecting the cells to progressively increasing levels of methotrexate, and the CHO clone producing the highest level of CETP was chosen for seeding into the extracapillary space of a 1.1-m2 hollow-fiber bioreactor (Cellex Biosciences Inc). Tissue culture supernatant from CHO grown in the bioreactor in serum-free medium was collected and filtered. Protease inhibitors were added, and the cell media was concentrated 40-fold with the use of an Amicon stirred cell concentrator with a 30 000–D MW cutoff. The concentrated cell media was made to 1 mol/L in NaCl, and any precipitate was removed by centrifugation. Purified CETP was obtained by passage on a hydrophobic resin (Toyopearl Butyl-650M) equilibrated with 750 mmol/L NaCl/2 mmol/L EDTA. The column was washed successively with 750 mmol/L NaCl/2 mmol/L EDTA and 200 mmol/L NaCl/2 mmol/L EDTA until the OD280 absorbance returned to baseline. The rCETP was eluted with 2 mmol/L EDTA, and fractions showing lipid transfer activity were pooled and loaded onto a mono-Q Sepharose (Pharmacia) anion exchange column equilibrated with 10 mmol/L K2PO4, pH 7.4. The rCETP was eluted with a 10 to 200 mmol/L K2PO4 gradient. Lipid transfer–active fractions were again pooled and concentrated as described above. Purified rCETP was >98% pure as judged by Coomassie-stained SDS gels.11
Labeling of rCETP
For each turnover study, freshly prepared rCETP was used. An aliquot of rCETP was labeled with the use of the di-iodinated Bolton Hunter12 reagent (Amersham) to yield an average iodination rate of 1 lysine per CETP molecule. Free reagent was removed by passage on a Sephadex G-25 column.
Verification of Biological Activity of Labeled rCETP
To ensure that the labeled protein retained normal biological activity, the specific activity of cholesteryl ester transfer of labeled rCETP was determined as the ability of labeled rCETP added to CETP-depleted human plasma (obtained by passage on a TP-2 immunoaffinity column and shown to be free of CETP by Western blot) to promote transfer of 3H-cholesteryl ester13 from HDL to LDL as described previously.14 The specific activity of cholesteryl ester transfer of labeled rCETP was compared with that of native plasma CETP.
Verification of Association of Labeled Human rCETP With Rabbit Lipoproteins
To determine whether labeled rCETP associated with plasma lipoproteins in a manner analogous to that of native rabbit CETP, labeled rCETP was combined with autologous plasma and intravenously infused into a rabbit. Plasma was obtained 30 minutes later, and lipoprotein fractions were separated by gel filtration with the use of Superose 6 and 12 columns in tandem. The column was standardized with the use of rabbit plasma. VLDL, LDL, and HDL peaks were identified by agarose gel electrophoresis with lipid staining and Western blot with the use of anti-rabbit apo B and apo A-I (kindly supplied by Dr J.C. Fruchart) and anti-rabbit apo E (a gift from Dr K. Weisgraber). Lipoprotein association of labeled human rCETP was determined by counting individual fractions and was compared with that of native rabbit CETP, determined by solid-phase RIA.15 Lipoprotein distribution of endogenous CETP in rabbit plasma was also compared with that of endogenous CETP in normal human plasma after standardization of the column with human lipoproteins.
Study Protocol
These studies were carried out in animals on each of the chow and chol diets. For each turnover study, animals were pretreated with KI to prevent thyroid uptake of iodine, and the sterility of the preparation was ensured by the use of sterile technique during the preparation of rCETP-containing plasma and Millipore filtration (22 μm) of the final preparation. Although labeled rCETP was shown to have normal lipid transfer activity and was able to associate with rabbit lipoproteins, a biological screen was performed before injection of labeled rCETP into the study animal to remove any potentially denatured molecules of rCETP. Labeled rCETP was equilibrated with rabbit plasma, and a 400-μL sample containing ≈20 μg labeled human rCETP (an amount equivalent to <2% rabbit plasma pool of CETP) was injected into a donor rabbit. With the use of a femoral cut-down technique, this animal was exsanguinated under fentanyl anesthesia 30 minutes later, and the plasma was isolated under sterile conditions. Twenty milliliters of this heterologous rabbit plasma containing labeled human rCETP was then injected over a 2-minute period into each of two study rabbits (10% of total plasma volume of the recipient rabbit). Blood samples were collected from an ear vein at frequent intervals over the first hour and continued over 5 days. Before the study, rabbits were handled on a daily basis to reduce stress-related artifacts. Each sample was limited to 0.75 mL of whole blood to prevent loss of >5% of total blood volume throughout the entire study. Whole plasma samples were counted and CETP mass and plasma lipids determined.
Plasma Lipoproteins
Plasma lipids were determined by automated enzymatic techniques. HDL cholesterol was measured after precipitation of apo B lipoproteins with the use of high-molecular-weight dextran sulfate/Mg2+.16 This laboratory is standardized with the Center for Disease Control in Atlanta, Ga, with respect to lipoprotein lipid measurements.
Plasma Kinetics of rCETP
CETP residency time (1/FCR) was determined from the area under the plasma radioactivity decay curve by the SAAM-II multiexponential curve fitting technique.17 The PR was calculated as (CETP concentration×plasma volume)/(RT×body weight), with the plasma volume being determined by dividing the total quantity of radioactivity injected by the radioactivity per unit volume determined in the 10-minute postinjection sample.
Statistical Methods
Student's t test for paired samples was used to compare changes in plasma CETP, lipoproteins, and kinetic variables before and after cholesterol feeding. Pearson correlation coefficients were calculated for the relations between changes in plasma CETP mass and changes in CETP PR and FCR after cholesterol feeding.18
Results
rCETP was labeled under gentle conditions with the use of the Bolton Hunter reagent, which adds a hydroxysuccinamide group to a lysine residue through an amide linkage.12 The cholesteryl ester transfer activity14 of labeled rCETP in CETP-depleted human plasma was 49±6 nmol cholesteryl ester·mL−1·h−1 per microgram CETP compared with 51±5 for endogenous plasma CETP.
Since recombinant rabbit CETP was not available, we elected to use the recombinant human protein as a kinetic marker for rabbit CETP. Transgenic studies have demonstrated that human CETP has much lower affinity for mouse HDL compared with HDL in human apo AI transgenic mice.19 However, mouse apo AI has only 66% homology to the human apo AI sequence,20 whereas rabbit apo AI has 78% homology to human apo AI.21 In addition, unlike the mouse, the rabbit has high levels of CETP in plasma, and there is 81% homology between human and rabbit CETP.10 Preliminary studies were carried out to determine whether labeled human rCETP would associate with rabbit HDL in a manner similar to endogenous rabbit CETP. Gel filtration was carried out with the use of Superose 6 and 12 columns in tandem, for which the elution profile of rabbit lipoproteins was determined. Native rabbit CETP in plasma of animals fed either a chow or a high cholesterol diet quantified by RIA was found to be associated with apo AI lipoproteins of a small to intermediate range of particle sizes. A total of 62% of rabbit CETP was lipoprotein associated on the chow diet and 64% on the chol diet, with the remainder eluting after HDL. The percent lipoprotein association of labeled human rCETP in rabbit plasma after intravenous injection was similar, with 62% to 64% of rCETP eluting with HDL on each of the two diets. Within the HDL fraction, the particle distribution of labeled rCETP differed from native rabbit CETP. The majority of endogenous CETP was present in a single broad HDL peak, whereas labeled human rCETP tended to distribute over two HDL peaks. On the chol diet, there was increased association of rCETP with large HDL, and a small amount of labeled rCETP was associated with apo B lipoproteins. On the chow diet, the distribution of labeled human rCETP was closer to that of native rabbit CETP, with the majority of rCETP present in a single broad HDL peak, while a minor fraction eluted with large apo AI–containing particles. (Fig 1⇓, a and b). Thus, this study demonstrates that human rCETP, in comparison to rabbit CETP, has a similar general affinity for rabbit HDL; approximately one third of CETP in rabbit plasma is not lipoprotein associated. In contrast, using identical gel filtration methodology and standardization of the column with human lipoproteins, we have demonstrated that in human plasma, a much larger proportion of CETP (89.5%) coelutes with HDL (Fig 2⇓).
a, Gel filtration elution profile of endogenous rabbit CETP and labeled human rCETP after intravenous infusion in rabbit on high cholesterol diet. Labeled rCETP, combined with autologous plasma, was intravenously infused into a rabbit; plasma was obtained 30 minutes later and lipoprotein fractions were separated by gel filtration through the use of Superose 6 and 12 columns in tandem. The column was standardized with the use of rabbit plasma. VLDL, LDL, and HDL peaks were identified by agarose gel electrophoresis with lipid staining and Western blot with the use of specific monoclonal antibodies to identify rabbit apo E (1D7), rabbit apo B (MB47), and apo A-I (2F1). VLDL, containing apo E and apo B, corresponds to fractions l9 to 24; LDL, containing apo B but little apo E, corresponds to fractions 24 to 26; HDL, containing apo AI, eluted in fractions 26 to 31. The albumin peak extends from fraction 31 to 35. Lipoprotein association of labeled human rCETP (□) was determined by counting individual fractions and was compared with that of native rabbit CETP (♦), determined by solid-phase RIA.15 Similar percentage association with apo AI particles (64%) is demonstrated for native and rCETP. Labeled human rCETP shows preferential association with larger HDL particles. The counts in the VLDL/LDL density range likely represent association of a small amount of CETP with β-VLDL during cholesterol feeding. b, Gel filtration elution profile of endogenous rabbit CETP and labeled human rCETP after intravenous infusion in rabbit on chow. This study was performed in an identical fashion to that illustrated in a. Since this study was done some months later, the Superose 6/12 column was restandardized with the use of rabbit plasma and dur to interval use of the column for other purposes and a slight shift in lipoprotein elution profiles was noted. VLDL eluted in fractions 21 to 23, LDL in fractions 24 to 27, and HDL in fractions 28 to 32. Albumin eluted in fractions 32 to 36. Lipoprotein association of labeled human rCETP was determined by counting individual fractions and was compared with that of native rabbit CETP, determined by solid-phase RIA.15 In the animal on a high cholesterol diet (a), a similar percentage association with apo AI particles (62%) is demonstrated for native and rCETP, with human rCETP showing only a small amount of rCETP showing preferential association with larger HDL species.
Gel filtration elution profile of endogenous human CETP. Lipoprotein distribution of endogenous CETP in human plasma was determined after standardization of the Superose 6/12 column with human lipoproteins. The lipoprotein elution profile is indicated. for each of VLDL, LDL, and HDL. The albumin peak consisted of fractions 29 to 34. In contrast to endogenous CETP in rabbit plasma (Fig 1⇑), the majority (89.5%) of CETP in human plasma coelutes with HDL.
The diets were well tolerated, with similar weight gain on each of the two diets. Cholesterol feeding resulted in a significant increase in total cholesterol and HDL cholesterol and more than a twofold increase in plasma CETP (Table 1⇓). Six animals were studied on a chow diet for 4 weeks followed by a cholesterol-rich diet, and two animals were studied in reverse order. As a probable result of accumulation of cholesterol stores, plasma cholesterol and CETP levels did not return to baseline when the 4-week chow diet followed the cholesterol diet, and the data for these two animals are presented separately (Table 3⇓).
Plasma Lipid and CETP Concentrations During Chow and Chol Diets
The infusion of rCETP resulted in no adverse effects. Body temperature remained normal throughout the study. Since adrenal steroids decrease plasma CETP levels, the rabbits were handled frequently before the study to minimize the stress response. Plasma CETP concentrations remained constant throughout the turnover studies.
Plasma kinetics of CETP were modeled with the use of SAAM-II. A typical decay curve for one animal on each of the two diets is shown in Fig 3⇓. The data conform best to a two-pool model. The assumption was made that the rapidly-turning-over pool of CETP (representing two thirds of total CETP in the model) was CETP, which was non–lipoprotein associated or in loose association with HDL. This probably represents the pool of CETP active in neutral lipid transfer. The more slowly-turning-over pool, with an FCR similar to apo A-I, was assumed to represent CETP, which is tightly associated with apo AI. The PR into the fast pool was 20 times that of the PR into the slow pool. Similarly, the FCR of CETP from the fast pool was >10 times as great as the FCR from the slow pool. The total FCR of plasma CETP was high (mean CETP plasma residency time, 9.9 hours on a chow diet and 6.1 hours on a cholesterol-rich diet). The residency time of the slow-pool CETP averaged 76.3 hours, similar to that reported for rabbit apo A-I,22 consistent with our proposition that the slow pool of CETP may be tightly bound to apo AI. Cholesterol feeding resulted in a more than a twofold increase in plasma CETP and a fourfold increase in the CETP production rate. The FCR of CETP also increased in response to cholesterol feeding by a mean of 60%. The data for individual animals are illustrated in Table 2⇓. The direction of change in CETP PR and FCR was similar for the two animals fed the cholesterol-rich diet first followed by chow but, as expected, the magnitude of change was less (Table 3⇓).
Data and SAAM-II–modeled fits for plasma decay of CETP. The decay curves illustrated are for a typical rabbit (I) studied on chow diet followed by chol diet. The data conform best to a two-pool model, likely representing a free exchangeable pool of plasma CETP and tightly HDL-associated pool of CETP.
CETP Kinetic Parameters for Six Rabbits Studied on Chow Diet Followed by Chol Diet
CETP Kinetic Parameters for Two Rabbits Studied on Chol Diet Followed by Chow Diet
The change in plasma CETP in response to cholesterol feeding correlated well with the change in each of CETP total PR (r=.95, P=.003), CETP fast PR (r=.94, P=.003), and CETP slow PR (r=.91, P=.002) (Fig 4⇓). These data strongly suggest that labeled human rCETP is an appropriate tracer for endogenous rabbit CETP. Changes in CETP synthesis in response to cholesterol ingestion appear to account for the increase that occurred in the plasma concentration of CETP. The observed correlations also suggest that there is equilibrium between the slow and fast pools of CETP because the labeled rCETP injected into plasma equilibrated with both pools.
Relationship between change in plasma CETP and change in CETP production rate in response to cholesterol feeding. There were significant correlations between individual rabbit change in plasma CETP and change in each of total CETP PR (r=.949, P=.003) and change in CETP fast PR (r=.94, P=.003) and change in CETP slow PR (r=.91, P=.002). These data demonstrate that most of the change in plasma CETP in response to cholesterol feeding can be explained by changes in CETP synthetic rates and also indicate that labeled human rCETP is a suitable tracer for endogenous rabbit CETP.
The FCR of CETP increased significantly as plasma CETP increased in response to cholesterol feeding. Plasma CETP increased to a greater extent than did HDL cholesterol; hence the availability of HDL as a transport protein for CETP was relatively decreased. Nonetheless, there was no change in the percentage of plasma CETP in the slow versus fast pools, and the FCR of both pools increased (+39% for the slow pool and +76% for the fast pool). The increase in CETP FCR correlated with the increase in plasma CETP. The Pearson correlation coefficients were as follows for change in plasma CETP versus change in: total FCR, r=.55 (P=.01); FCR of fast pool, r=.65 (P=.02); and FCR of slow pool, r=.81 (P=.005) (Fig 5⇓).
Relationship between change in plasma CETP and change in CETP FCR. The Pearson correlation coefficient for change in plasma CETP versus change in CETP total FCR was r=.55, P=.01, suggesting that CETP catabolic pathway(s) are upregulated in response to cholesterol feeding or increased plasma CETP concentration. There were also significant correlations between change in plasma CETP mass and change in CETP fast FCR (r=.65, P=.02) and change in slow FCR (r=.81, P=.005).
Discussion
We have determined the plasma kinetics of CETP in the rabbit, using labeled human rCETP as a marker for endogenous rabbit CETP. There is considerable homology between human CETP and rabbit CETP10 and between human apo AI and rabbit apo AI,21 and in preliminary studies, we demonstrated that labeled rCETP retains neutral lipid transfer activity and associates with rabbit HDL, with a gel filtration elution profile similar although not identical to that of endogenous rabbit CETP. Somewhat more of the labeled rCETP coeluted with the largest apo AI–containing lipoproteins on each of the two diets (Fig 1⇑, a and b). There are a number of possible reasons for these differences in HDL particle association. The plasma samples were collected soon after injection of rCETP, and equilibration with lipoproteins may not have been complete. It is also possible that human rCETP, compared with rabbit CETP, has a greater affinity for large HDL1 lipoproteins and β-VLDL, which occur during cholesterol feeding in rabbits, and this could explain the particular association of rCETP with large HDL and, to a lesser extent, of rCETP with apo B–containing lipoproteins on the high cholesterol diet. Further studies are under way to address these questions. A biological screen of the labeled protein was carried out to ensure removal of any molecules of rCETP that may have been structurally altered during the labeling process. Our results are compatible with the hypothesis that labeled human rCETP is a suitable tracer for endogenous rabbit CETP. We have demonstrated very strong correlations between the increase in rabbit plasma CETP mass and the changes in the modeled PR of CETP (into each of the slow and fast pools) in response to cholesterol feeding.
This study demonstrates that the turnover of CETP in the plasma compartment is rapid in comparison to kinetic parameters reported for the major apolipoproteins, with the exception of apo E4.23 24 25 In this study, the mean plasma residency time for CETP in animals fed a chow diet was 9.9 hours. These data are in accord with the recent observation that injection of lipopolysaccharide, which induces a glucocorticoid-mediated suppression of CETP gene expression, results in a precipitous fall in plasma CETP within 8 hours of infusion.9
The plasma kinetics of CETP conform best to a two-pool model. The plasma residency time of the rapidly-turning-over pool is short (7.1 hours) and may represent the pool of CETP that is active in neutral lipid transfer. This pool likely consists of free plasma CETP in rapid equilibrium with CETP, which is loosely bound to HDL or other lipoproteins. In contrast, the plasma residency time of the slow pool (76 hours) is similar to that of apo AI and probably represents CETP, which is tightly associated with apo AI. The majority of plasma CETP leaves the plasma compartment through the fast pool.
CETP has affinity for all lipoprotein particles.26 27 Because of the higher particle number, the majority of CETP in fasting plasma has been found to be associated with HDL by a variety of techniques.26 27 28 29 30 31 The mechanism of neutral lipid transfer in plasma may involve a collision complex of HDL, CETP, and another lipoprotein32 or a shuttle mechanism,33 but in either case, the CETP/lipoprotein association is dependent on weak, hydrophobic bonding.34 35 Although the majority of CETP in normal plasma is associated with lipoprotein (Lp) AI,29 Lp AI/AII clearly participates to a similar extent in the CETP-mediated neutral lipid transfer reaction, suggesting that CETP dissociates readily from its major carrier lipoprotein, Lp AI.28 36 Studies by Ohnishi et al37 and Epps and colleagues38 demonstrate that there is rapid equilibration (<3 hours) of cholesteryl ester and triglyceride between lipid microemulsions in the presence of CETP. Their data suggest that CETP interacts with lipoproteins through a dissociation/diffusion/adsorption/exchange mechanism and that the association/dissociation of CETP and lipoproteins is rapid. The present kinetic data support the hypothesis that there is a pool of CETP in close association with apo AI that has a long plasma residency time, similar to that of apo AI, and a second pool of CETP that is rapidly cleared from plasma. The fast pool probably represents a pool of CETP active in neutral lipid transfer and in rapid association/dissociation with HDL and other lipoproteins.
The rapid catabolism of CETP from the latter pool suggests that factors favoring the dissociation of CETP from lipoproteins may enhance its catabolism. We have demonstrated using gel filtration that in the rabbit, a considerable amount of plasma CETP (35%) is not lipoprotein associated. In contrast, in humans, 90% of endogenous CETP coelutes with HDL and VHDL (Fig 2⇑). The reason for this difference in lipoprotein association is not known but may be related to HDL apolipoprotein content and surface charge. Weinberg et al11 have demonstrated that rCETP is surface active and binds to egg phosphatidylcholine monolayers with affinity similar to apolipoproteins. The maximal surface pressure generated by CETP is considerably less than that of apolipoproteins, suggesting that CETP, which has a very stable surface structure, does not penetrate into the lipid surface. On the basis of the estimated dissociation constant for rCETP and the plasma concentrations of CETP in normals, their data would predict that lipoprotein surfaces in vivo are saturated with CETP. In their studies, the exclusion pressure of rCETP was less than that of apo AI and slightly lower than the calculated surface pressure of HDL, supporting the conclusion that the binding of CETP to lipoproteins is labile and susceptible to small changes in surface pressure.11 Apo AIV increases the surface pressure of lipoproteins and may facilitate CETP dissociation from HDL.39 Apo AIV has been shown to enhance the neutral lipid transfer process in vitro.40 Displacement of CETP from HDL by apo AIV might also be expected to promote its catabolism in vivo. The rapid catabolism of plasma CETP in the rabbit may be related in part to its poorer association with plasma lipoproteins. In humans, less CETP is present in the non–lipoprotein-associated state; hence, the rate of CETP catabolism may differ.
Considerable available data demonstrate that CETP gene expression is upregulated by cholesterol. Cholesterol feeding increases plasma CETP mass in various species including humans,3 nonhuman primates,7 hamsters,5 rabbits,4 and mice bearing the human CETP transgene.6 8 41 Plasma CETP levels are elevated in hypercholesterolemia14 and are normalized by reduction of LDL cholesterol by 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor therapy (R. McPherson, unpublished data). We have also demonstrated in vitro that cholesterol loading of human adipose tissue, by incubation with chylomicron remnants, increases CETP mRNA levels and promotes secretion of CETP into the medium.2
In the present study, cholesterol feeding resulted in a 2- to 3-fold increase in the plasma mass of CETP and a 20-fold increase in the production rate of CETP. The change in plasma CETP correlated strongly with the change in CETP production rate, suggesting that the increase in plasma CETP in response to cholesterol feeding can be explained largely by the increase in CETP synthesis, consistent with molecular data on the effects of cholesterol on CETP gene expression. Similarly, Warren et al42 demonstrated in cholesterol-fed rabbits that there was a strong temporal relationship between the rise in plasma CETP activity and the increase in hepatic CETP mRNA abundance, suggesting that the cholesterol-induced rise in plasma CETP was due to increased CETP synthesis. The fractional catabolic rate of CETP also increased in response to cholesterol feeding, although to a smaller extent (+60%). There was a significant correlation between the increase in plasma CETP mass and the increase in CETP FCR. Although CETP mass in plasma increased substantially and to a greater extent than HDL cholesterol, the kinetic data did not reveal any decrease in the percentage of CETP present in the tightly apo AI–associated slow pool during the chol diet. The FCR of the fast pool increased by 76% in response to cholesterol feeding, whereas the increase in the FCR of the slow pool was only 39%. The increased FCR of CETP during the cholesterol-rich diet may thus represent upregulation of a specific clearance pathway for CETP. The mechanism(s) by which CETP is cleared from plasma is not known. In studies under way in this laboratory, we are using tyramine cellobiose–labeled rCETP43 to determine the tissue site(s) of CETP degradation.
Selected Abbreviations and Acronyms
| apo | = | apolipoprotein |
| CETP | = | cholesteryl ester transfer protein |
| chol | = | 0.5% cholesterol |
| FCR | = | fractional catabolic rate |
| PR | = | production rate |
| RIA | = | radioimmunoassay |
| RT | = | residence time |
Acknowledgments
This study was supported by the Heart and Stroke Foundation of Ontario (A2394; R.M., P.L.), Medical Research Council of Canada (group grant; R.M., P.L.) HL-22682 and HL-21006 (A.R.T.).
References
This Issue
Jump to
Article Tools
Share this Article
- Plasma Kinetics of Cholesteryl Ester Transfer Protein in the RabbitPermalink:
