Regulation of Human Plasma Lecithin:Cholesterol Acyltransferase Activity by Lipoprotein Acceptor Cholesteryl Ester Content*

Very low density lipoproteins and low density lipoproteins attain maximal cholesteryl ester contents dur- ing the incubation of human plasma and, under these conditions, both lecithinxholesterol acyltransferase and transfer proteins are inhibited. These lipoproteins provide the major part of free cholesterol for the leci-thin:cholesterol acyltransferase reaction, and are the major acceptors of cholesteryl ester generated by lecithinxholesterol acyltransferase and transported to the lipoprotein acceptors by the transfer protein. The results obtained indicate that the concentration of accep- tor limits esterification and transfer in plasma, and that in vivo these acceptors contain close to their max- imal cholesteryl ester content. Human plasma end product acceptor lipoproteins have a composition similar to that of the ester-rich large low density lipopro- tein characteristic of primate models of experimental atherosclerosis.

V e r y low density lipoproteins and low density lipoproteins attain maximal cholesteryl ester contents during the incubation o f h u m a n p l a s m a and, under these conditions, both lecithinxholesterol acyltransferase and t r a n s f e r proteins are inhibited. These lipoproteins provide the m a j o r part of free cholesterol f o r the lecithin:cholesterol acyltransferase reaction, and are the m a j o r acceptors of cholesteryl ester generated b y lecithinxholesterol acyltransferase and transported to the lipoprotein acceptors by the t r a n s f e r protein. The results obtained indicate that the concentration of acceptor limits esterification and t r a n s f e r in plasma, and that in vivo these acceptors contain close to their maximal cholesteryl ester content. H u m a n p l a s m a end product acceptor lipoproteins have a composition similar to that of the ester-rich large low density lipoprotein characteristic of primate models of experimental atherosclerosis.
In single-walled vesicles of lecithin and cholesterol, the activity of isolated 1ecithin:cholesterol acyltransferase is limited by the accumulation of cholesteryl ester in the lipid bilayer (1). Inhibition is relieved by addition of nonsubstrate acceptor liposomes which, in the presence of cholesteryl ester transfer protein (2), can store the transported cholesteryl ester (1). Beyond this point both LCAT' and transfer reactions are inhibited in the medium. Two reports (3, 4) have indicated structural association of LCAT and transfer protein in human plasma. The major part of cholesteryl ester is located in the low density lipoprotein class which is not a direct substrate for LCAT, but which accepts preformed esters via transfer protein activity. In the present paper we have investigated whether in plasma LCAT-mediated synthesis of cholesteryl esters is similarly limited by the ability of acceptor lipoproteins to bind and store cholesteryl esters and, if so, what is the maximal cholesteryl ester content of these end product lipoproteins.
* This work was supported by Grant HL 23738 and Arteriosclerosis SCOR Grant HL 14237 of the United States Public Health Service. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

EXPERIMENTAL PROCEDURES
Assay of LCAT-Transferase activity was determined as the rate of decrease of plasma-free cholesterol mass as a function of time during incubation at 37'C. Free and ester cholesterol were measured enzymatically (5) as previously described (3). Loss of free sterol under these conditions was inhibited >95% by the LCAT inhibitor, dithiobis-2-nitrobenzoic acid ( 6 ) .
Assay of Cholesteryl Ester Transfer Activity-Transfer activity was determined as the rate of increase of cholesteryl ester mass in very low and low density lipoprotein classes, precipitated with heparin-MnC12 (7) when plasma was incubated at 37"C, as previously described (3).
Zmmunoaffinity Chromatography of Apoprotein B-Apo B, the major apoprotein of LDL, was obtained from LDL isolated centrifugally from human plasma (8). The apo B protein was precipitated with tetramethyl urea (5056, v/v). Antibody to apo B was raised in rabbits (3) and the y-globulin fraction, purified by DEAE-cellulose chromatography (9), was complexed with Sepharose-CNBr (10). Columns (1 X 20 cm) of immobilized anti-apo B were equilibrated with 0.15 M NaC1-0.001 M disodium EDTA, pH 7.4. Plasma from freshly drawn blood was passed through at a flow rate of 5-10 ml/h; fractions of eluate containing detectable protein were pooled and removal of antigen was determined by radial immunoassay (11). The assay was previously described (3).

RESULTS
LCAT and Transfer Activities of Human Plasma-When plasma from freshly drawn blood was incubated at 37"C, the initial linear rate of cholesterol esterification was maintained for 30-40 min (Fig. 1). This initial rate of LCAT activity was 0.9-1.2 pg of cholesterol esterified min" ml-l plasma (mean 1.05 pg, 12 experiments). The initial rate of transfer of cholesteryl ester to VLDL and LDL was 80-95% (mean 90%, 5 experiments) that of LCAT activity in the same plasma. Both LCAT and transfer rates decreased proportionately with further incubation, to the point that after

3-4 h there was
essentially no further decrease in plasma-free cholesterol, increase in plasma cholesteryl ester, or further transfer of cholesteryl ester to VLDL and LDL (Fig. 2). When net esterification was inhibited with 1.4 mM DTNB, transfer protein activity was unchanged (Table I) and the cholesteryl ester transferred to VLDL and LDL under these conditions was derived from the supernatant of heparin-MnC12-precipitations, which contained the high density lipoprotein fractions (3). Inhibition by DTNB was fully reversible (Fig. 1).
When plasma was incubated until esterification and transfer were inhibited and then mixed with VLDL and LDL from unincubated plasma, the initial rate was completely restored (Table 11), indicating that inhibition was not the result of irreversible inactivation of the esterification and transfer systems. On the other hand, addition of the same lipoproteins to fresh plasma during the initial linear phase of esterification and transfer was without discernible effect. The release from inhibition might have resulted from either an increase in the supply of substrate, free cholesterol, and lecithin, or addition of acceptor for product cholesteryl ester. To distinguish between these possibilities, VLDL and LDL were enriched with cholesteryl ester, without depletion of substrate lipids, by incubation in the presence of DTNB (Table I). These lipoproteins were without ability to relieve the inhibition of LCAT and transfer protein activities in incubated plasma.
Effects of Acceptor Concentration on LCAT and Transfer Protein Activities-The apo B content was varied from 10 to 90% of the initial plasma value by immunoaffinity chromatog-

Minutes Incubation
Minutes Incubation Minutes Incubation FIG. 1 (left). Initial rate of cholesterol esterification in hu-centrifugation (2000 X g for 10 min) to remove lipoproteins that man plasma. Citrated plasma-0.01 M Tris-HC1 (pH 7.4) was incu-contained apo B. LCAT was assayed as the rate of decrease in plasmabated at 37°C either directly ( -) or after inhibition of activity free cholesterol (FC) in the absence of DTNB. In both cases, quadwith DTNB (final concentration 1.4 mM). The plasma was maintained ruplicate samples were taken for chemical analysis. Values are exfor 30 min on ice, then the inhibition reversed by addition of a 5-fold pressed in terms of the change in free or esterified cholesterol relative molar excesa of 2-mercaptoethanol (6)  VLDL + LDL was obtained by precipitation from initial or incubated plasma and after washing, resolublized as described under "Experimental Procedures." Ester/free cholesterol mass ratios were 2.2 for the initial precipitate and 2.4 and 3.0 for the end product lipoproteins obtained in the presence and absence of DTNB, respectively. Each lipoprotein was added at a concentration sufficient to double the endogenous level of apo B in the plasma. Final VLDL and LDL were obtained after incubation for 3 h at 37'C.

I11 Effects ofpartial removal of lipoproteins that contained apo B on the maximum of cholesteryl ester accumulated in plasma
Plasma was passed through immobilized anti-apo B antibody. The initial apo B content after dilution with citrate and Tris buffer was 625 p g / d and after affinity chromatography, was 302 pg/ml (corrected to the same plasma protein content). Original and apo Bdepleted plasmas were incubated for 3 h at 37°C. Initial and final levels of free and ester cholesterol were determined and after precipitation of plasma samples with heparin-MnCl,, the ester-free weight ratio was determined on the lipoproteins of the precipitated fraction. raphy, and the mass of cholesteryl ester synthesized and transferred to acceptor lipoproteins before inhibition of LCAT activity was determined. Removal of apo B was without effect on the level of either LCAT or transfer proteins in plasma (3). As shown in Table 111, the ability of plasma to generate and transfer cholesteryl ester was reduced in proportion to its apo B content. However, the cholesteryl ester content of the end product lipoprotein acceptors was the same in each case.
Source of Free Cholesterol for the LCAT Reaction-The experiments above indicated that in human plasma, VLDL and LDL receive the major part of cholesteryl ester generated by the LCAT reaction. The source of free cholesterol for LCAT was determined by assaying the free cholesterol content of lipoproteins during the linear initial phase of the esterification reaction, Le. when incubation time was short relative to the known rates of exchange of cholesterol between these lipoproteins (12). As shown in Fig. 3, essentially all of the substrate for LCAT in plasma was derived from VLDL and LDL. These fractions contain no LCAT (3) and such substrate must therefore be esterified by the minor lipoprotein FC, free cholesterol,

TABLE IV
Composition of initial and end product VLDL and LDL from human plasma Triglyceride, protein, and phospholipid were assayed as previously described (16). Values given are the means of duplicate analyses which differed <5.0% for all procedures; VLDL and LDL were precipitated from plasma before and after incubation (4 h, 37°C) with heparin-MnCL as detailed under "Experimental Procedures" then the washed precipitate was redissolved and separated into VLDL (d < 1.019 g/cm3) and LDL (1.019 < d < 1.063 g/cm:'). In the experiment illustrated the initial plasma contained 232 pg ml" of LDL-free cholesterol and 61 pg ml" of VLDL-free cholesterol, and ester cholesterol contents of 701 and 78 pg ml", respectively, based on this separation. Loss of VLDL free cholesterol was 30 p g ml ' original plasma and gain of sterol as cholesteryl ester was 35 pg ml" plasma; the equivalent loss and gain for LDL was 68 and 85 pg ml-', respectively. In three other experiments, the cholesteryl ester content of VLDL (percentage by weight) was initially 9.8%, 13.8% and 22.0%, and finally 20.996, 23.1%. and 28.1%; and of LDL was initially 38.3%, 40.2%, and 45.8%, and finally, 46.8%, 50.0%. and 50.356, respectively. Other h i d changes were comuarable to those illustrated. Further information on the limiting composition of human plasma VLDL and LDL, and on their contributions to LCAT substrate and product acceptor functions, was obtained by chemical analysis of the initial and end product acceptor lipoproteins. As shown in Table IV, LDL cholesteryl ester increased about 10% above its proportion in fresh plasma before further transfer of cholesteryl ester was inhibited. The major increase of cholesteryl ester mass was in LDL, although a greater change in composition was found in the smaller mass of VLDL in the plasma. The mass of free cholesterol supplied by each lipoprotein class for the esterification was almost the same as that taken back, as ester, via the transfer protein reaction.

DISCUSSION
The central core of cholesteryl ester in plasma lipoprotein particles, particularly LDL, is much greater than can be dissolved in the lipid vesicle bilayer. Nevertheless, there is convincing evidence that this core may be in equilibrium with an ester pool in the lipoprotein surface (13) accessible to transfer. We previously showed that in a defined vesicle system with pure LCAT and transfer protein, these activities became limited by the capacity of the acceptors to take up cholesteryl ester. The present studies strongly suggest that a similar mechanism limits the accumulation of cholesteryl ester in human plasma. LCAT activity in incubated plasma was stimulated by acceptor lipoproteins but not by substrate, and the extent of esterification was a function of acceptor concentration in plasma. The results of these experiments support the concept of end product lipoproteins, which contain a maximal cholesteryl ester content and accordingly no longer permit ongoing esterification of cholesterol in the complex containing LCAT and transfer protein. The increase in LDL and VLDL cholesteryl ester content that can occur in plasma before inhibition is comparatively small, about 10% of original mass of this lipid fraction. This requires that the level of acceptor lipoproteins must be closely integrated with the rate of esterification of cholesterol in plasma. The composition of end product LDL obtained in this study is very similar to that of the atherogenic "large LDL" found in cholesterol fed primates (14) and to the atypical LDL of some hypercholesterolemic human subjects (15). All of these LDL particles may represent the end products of LCAT and transfer protein activities, in vivo or in vitro, that contain maximal contents of cholesteryl ester. These studies also indicate that VLDL and LDL receive back, in ester form, almost the same mass of sterol as they contributed to the LCAT and transfer reactions.
This appears to provide further evidence of a close integration between LCAT and transfer reactions ( 3 ) . The same is also shown by the close proportionality throughout of these activities in incubated plasma. It is noteworthy that little cholesteryl ester in whole plasma is taken up by HDL. These results seem most consistent with the concept that transferase and transfer functions are associated together with only a small pool of substrates, which turn over rapidly to supply the large pool of plasma VLDL and LDL ester. I.