834
Lipid Transfer Protein-Mediated Distribution of
HDL-Derived Cholesteryl Esters Among Plasma
Apo B-Containing Lipoprotein Subpopulations
Carol A. Marzetta, Tom J. Meyers, and John J. Albers
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
The substrate specificity of lipid transfer protein has been examined in whole plasma in vitro by following
the redistribution of high density lipoprotein (HDL)-derived [3H] cholesteryl ester (CE) into apolipoprotein (apo) B-containing lipoproteins using density gradient ultracentrifugation. HDL-derived [3H]CEs
were incubated with plasma or isolated lipoprotein classes (very low density lipoprotein, intermediate
density lipoprotein, and low density lipoprotein [LDL] subpopulations from the HDL donor) with and
without lipoprotein lipase for 0.5-6 hours at 37°C. After incubation, lipoproteins were separated into 38
fractions after density gradient ultracentrifugation, and radioactivity, protein, and cholesterol were
monitored across the profiles. These studies indicate that 1) lipid transfer protein activity varied among
the individuals as well as within an individual; 2) the majority of the [3H]CE was associated with LDL;
3) in most individuals (71%), more HDL-derived [3H]CE distributed within the buoyant LDL density
region; and 4) the distribution of HDL-derived [3H]CE was similar to the distribution of lipoprotein
lipase-derived "remnant" particles within buoyant LDL. These in vitro studies support the hypothesis
that HDL-derived [3H]CEs vary in their distribution among apo B-containing particles and that more
HDL-derived [3H]CEs are transferred to lipoproteins within the buoyant LDL density range. Additional
studies suggest that lipoprotein heterogeneity within this density range, such as the presence of
remnant-like lipoproteins, may contribute to the selective distribution of HDL-derived [3H]CE into
buoyant LDL. (Arteriosclerosis and Thrombosis 1993;13:834-841)
KEY WORDS • cholesteryl ester transfer protein • LDL • HDL • VLDL
C
holesteryl ester (CE) transfer activity was first
recognized by Nichols and his colleagues in the
mid 1960s12 and has stimulated much interest
and speculation on its physiological role. It has been
suggested that this lipid transfer protein (LTP) activity is
involved in a sequence of events in which free cholesterol
from peripheral tissue is taken up by high density lipoprotein (HDL), esterified by lecithin :cholesterol-acyi transferase (LCAT), and then transferred to very low density
lipoprotein (VLDL) or low density lipoprotein (LDL) in a
process called reverse cholesterol transport.34 Although
reverse cholesterol transport is thought to be an antiatherogenic event by which cholesterol is removed from
peripheral tissue to be transported and cleared from the
circulation by the liver, it is unclear whether the transfer of
CE from HDL to apolipoprotein (apo) B-containing
lipoproteins is atherogenic56 or antiatherogenic.7 If the
ultimate outcome of such a process results in the net
movement of cholesterol from peripheral tissue to the
liver, then LTP activity would appear to be beneficial.
From the Department of Medicine (C.A.M., J.J.A.), University
of Washington, and the Northwest Lipid Research Laboratories
(TJ.M., J.JA.), Seattle, Wash.
Supported by National Institutes of Health grant HL-30086 and
Individual National Research Service Award HL-07797-01
(C.A.M.).
Address for correspondence: Carol A. Marzetta, PhD, Pfizer
Central Research, Eastern Point Road, Groton, CT 06340.
Received November 23, 1992; revision accepted February 19,
1993.
However, the transfer of CE to VLDL and LDL may
contribute to the formation of CE-enriched apo B-containing particles, which have been associated with an
increased risk for developing atherosclerosis.8-9
Little is known about the distribution of LTP-mediated
HDL-derived CE among the various apo B-containing
lipoproteins. What determines the ability of a lipoprotein
to accept an HDL-derived CE is likewise not well understood. Both surface and core lipids are thought to influence LTP activity. The phospholipid moiety, the presence
of core lipids, the ratio of trigryceride (TG) to CE within
the core, the species of the CEs, the free cholesterol-tophospholipid ratio, the free cholesterol concentration, and
the presence of fatty acids in lipoproteins have all been
shown to influence the activity of LTP.'°-|S These studies
have been done primarily with isolated lipoproteins and
semipurified LTP.
The present studies were designed to examine in
detail the distribution of HDL-derived CE among all
apo B-containing lipoproteins in whole plasma by use
of density gradient ultracentrifugation (DGUC). In
addition, comparisons were made between the in vitro
formation of VLDL "remnants" and the distribution of
HDL-derived [3H]CE among apo B-containing lipoproteins in whole plasma.
Methods
Materials
[Cholesteryl-l,2,6,7-3H(«)]cholesteryl oleate (75.3 Ci/
mmol; Dupont, Boston) or [cholesteryl-4-14C]cho-
Marietta et al HDL-Derived CEs Among LDL Subpopulations
lesteryl oleate (52.5 mCi/mmol; New England Nuclear,
Boston), glycerol-tri-[l-14C]oleate (Amersham, Arlington Heights, 111.), and [125I]NaI (New England Nuclear)
were used. Purified bovine milk lipoprotein lipase (LpL)
was isolated as described previously.16 The paraoxon
was a gift from Dr. Gerhard Schrader.
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
Isolation and Radiolabeling Lipoproteins
All subjects used in these studies were volunteers.
The subjects were normolipidemic17 except for subject
6, who was mildly hypertriglyceridemic (195 mg/dL),
and subjects 10 and 13, who had familial combined
hyperlipidemia according to the criteria described previously.18 Subject 10 had been taking 1 g/day of niacin
for approximately 6 months before the study. No other
subject was taking medications known to alter lipid or
lipoprotein metabolism. The study protocol was approved by the Human Subjects Review Committee of
the University of Washington, and informed consent
was obtained from each subject.
HDL-derived [3H]CE. The density of 50 mL of whole
plasma was raised to 1.080 g/mL with solid potassium
bromide. The 1.080-g/mL whole plasma was divided
into two 39-mL quick-seal tubes (Beckman, Palo Alto,
Calif.), then overlayered with a 1.080-g/mL solution and
subjected to ultracentrifugation in a 50-Ti rotor for 18
hours at 50,000 rpm at 4°C. The d> 1.080 g/mL lipoproteins were dialyzed to 1.006 g/mL in 0.9% NaCl containing 0.01% EDTA and 0.01% NaN3. Dithiobisnitrobenzoic acid (DTNB; final concentration of 1.4
mmol/L) was added to 12 mL of the d> 1.080 g/mL
lipoproteins and then incubated at 37°C in a shaking
water bath for 6-9 hours in a 15-mL glass conical tube
that had a thin layer of [cholesteryl-l,2,6,7-3H(«)]cholesteryl oleate (8-500xlO 6 cpm) dried to the wall.19
After incubation, the density of the d> 1.080 lipoproteins was raised to 1.21 g/mL with potassium bromide,
poured into two ultracentrifugation tubes, overlayered
with a 1.21-g/mL solution, and subjected to ultracentrifugation in an SW-41 rotor for 40 hours at 10°C and
41,000 rpm. The concentrated HDL was obtained by
tube slicing, and an average of 24.5 ±14% (mean±SD,
n=8) of the [3H]CE was incorporated into HDL.
The d> 1.080 plasma fraction was used for the HDL
radiolabeling methods because this fraction includes
active LTP and HDL that had been subjected to minimal ultracentrifugation. The HDL CE radiolabeling
method was done in the presence of DTNB to inhibit
LCAT and minimize compositional changes of HDL
during the 6-9-hour incubations. Except in studies
requiring substantial amounts of radiolabeled HDL
(e.g., multiple-subject screening studies), fresh plasma
was obtained on the day of the incubation study from
the same subject who served as the donor for the
HDL-derived [3H]CE.
VLDL [I4C]TG. VLDL was isolated from plasma at
1.006 g/mL after ultracentrifugation in an SW-41 rotor
for 16 hours at 10°C and 41,000 rpm. The isolated
VLDL was concentrated further in a dialysis bag using
carboxymethyl cellulose (Sigma, St. Louis, Mo.). The
concentrated VLDL was then added to 10 mL plasma
(from the same VLDL donor) in the presence of DTNB
(final concentration, 1.4 mmol/L) and incubated in
tubes that had a thin layer of glycerol-tri-[l-uC]oleate
(4-10xlO 6 cpm) dried to the wall as described for the
835
HDL-derived [3H]CE. The radiolabeled VLDL was
isolated after incubation by ultracentrifugation as described earlier. Approximately 6.1±4% (mean±SD,
n=4) of the [14C]TG was incorporated into VLDL.
VLDL l25I-protein. VLDL was isolated from plasma
and concentrated as outlined above. The concentrated
VLDL was then radioiodinated with 125I (New England
Nuclear) by the method of McFarlane20 as modified by
Bilheimer et al.21
Incubation Procedures
Lipid transfer activity assay. Radiolabeled HDL-derived [3H]CE (70,000-580,000 cpm) was added to an
aliquot of freshly isolated plasma and incubated for 30
minutes to 6 hours at 37°C in a shaking water bath. The
assays were stopped by placing the samples on ice or ice
and paraoxon when LpL had been added to the incubation. The postincubation samples were then separated and characterized by DGUC. The density of 2 mL
of each sample was raised to 1.21 g/mL and underlayered into a discontinuous salt gradient consisting of 3.7
mL of a 1.006 g/mL solution, 6.5 mL of a 1.063 g/mL
solution, and 2.3 mL of the 1.21 g/mL whole plasma
postincubation sample. The samples were then subjected to ultracentrifugation in an SW-41 rotor at 41,000
rpm for 24 hours at 15°C. After ultracentrifugation, the
samples were drained through a flow-cell UV monitor
into a fraction collector as described previously.22 The
distribution of HDL-derived [3H]CE among the apo
B-containing lipoproteins was determined by counting
an aliquot of each fraction and correcting for background radioactivity and quench. Recovery of radioactivity after DGUC averaged 77.3 ±10% (mean±SD of
five representative experiments).
When the movement of HDL-derived [3H]CE to apo
B-containing particles was examined, in most studies
the incubation conditions were achieved by use of
freshly isolated plasma in which LTP and LCAT were
active to mimic in vivo conditions as closely as possible.
In some experiments, total cholesterol per fraction
after separation by DGUC was measured enzymatically
with a Total Cholesterol Kit (Boehringer Mannheim,
Indianapolis, Ind.) or by the method of Allain et al.23
Lipid transfer activity is given as the percent of total
HDL-derived [3H]CE radioactivity isolated within the
various apo B-containing lipoprotein subpopulations.
Paired t tests were used to compare data sets and corrected for multiplicity by the method of Bonferroni.24
Lipid transfer activity assay in the presence of LpL.
Whole plasma, HDL-derived [3H]CE, 1,000 units of
LpL (nanomoles of free fatty acids released per minute
per milliliter at 37°C), and 150 ^ig human serum albumin (Sigma) were incubated at 37^C in a shaking water
bath for 30-90 minutes. In some studies, 125I-VLDL or
VLDL-[14C]TG were added to the whole plasma. The
incubations were stopped by placing the samples on ice
in tubes containing paraoxon (final concentration, 2
mmol/L). An aliquot of each sample was taken for free
fatty acid determinations. Lipoproteins were then separated and characterized by DGUC as described above.
Aliquots were taken for radioactivity and corrected for
background, quench, and 125I radioactivity in the tritium
window.
Studies using isolated lipoprotein subpopulations. The
density of freshly isolated whole plasma was raised to
Arteriosclerosis and Thrombosis
Vol 13, No 6 June 1993
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
1.080 g/mL with potassium bromide, overlayered with a
1.080 g/mL solution, and subjected to ultracentrifugation in a 60-Ti rotor at 40,000 rpm for 18 hours at 10°C.
The d< 1.080 lipoproteins were isolated by tube slicing
and dialyzed to 1.030 g/mL. The dialyzed lipoproteins
were then underlayered into a discontinuous salt gradient consisting of 8 mL of a 1.006 g/mL solution, 20 mL
of the lipoprotein sample at 1.030 g/mL, and 11 mL of
a 1.063 g/mL solution. The samples were then subjected
to 5.5 hours of ultracentrifugation at 50,000 rpm and
20°C and fractionated as described previously.22 Material within specific fractions was pooled and dialyzed to
1.006 g/mL, and aliquots were taken for cholesterol
determinations. Three milliliters of the d> 1.080 lipoproteins was dialyzed to 1.006 g/mL and incubated in a
tube coated with 47.0xlO 6 cpm [3H]CE for 5 hours at
37°C as described above. The density of the [3H]CE
d> 1.080 lipoproteins was raised to 1.21 g/mL, and
radiolabeled HDL was isolated by ultracentrifugation in
a 40.3 -Ti rotor at 36,000 rpm for 24 hours at 10°C.
Incubation conditions consisted of 1) 2 mL of freshly
isolated whole plasma and HDL-derived [3H]CE
(207,500 cpm) for the whole-plasma control; 2) 378 /xg
VLDL cholesterol, 980 fig LDL! cholesterol, 208.5 /xg of
d> 1.080 unlabeled lipoproteins (HDL, LTP, LCAT,
and plasma proteins), and 207,500 cpm of HDL-derived
[3H]CE for the LDLt sample; 3) 378 /xg VLDL cholesterol, 1,522 fig LDL: cholesterol, 208.5 /xg of d>1.080
unlabeled lipoproteins, and 207,500 cpm of HDL-derived pH]CE for the LDL 2 sample; and 4) 378 /xg
VLDL cholesterol, 980 /xg LDL, cholesterol, 1,522 fig
LDL2 cholesterol, 208.5 /ig of d>1.080 unlabeled lipoproteins, and 207,500 cpm of HDL-derived [3H]CE for
the LDL] and LDL2 samples. Total volumes were made
equivalent (3.9 mL) with saline. All samples were
incubated for 2 hours at 37°C in a shaking water bath.
Incubations were stopped by placing the samples on ice,
and then each sample was subjected to DGUC in the
following discontinuous salt gradient: 3.5 mL of 1.006
g/mL, 5.7 mL of 1.030 g/mL, and 2.4 mL of the
incubation sample at 1.21 g/mL. The samples were
ultracentrifuged and fractionated as described earlier.
Results
An -example of the distribution of HDL-derived
[3H]CE among apo B-containing lipoproteins separated
and characterized by DGUC before and after incubation
is shown in Figure 1A. The percentages of HDL-derived
fHJCE isolated within VLDL+intermediate density lipoprotein (IDL) and LDL and HDL were determined by
summing the radioactivity within VLDL+IDL, LDL,
and HDL based on the DGUC profile. After 6 hours of
incubation at 37°C, the distribution of the HDL-derived
[3H]CE was found primarily associated with LDL (42%)
and was essentially superimposable with the distribution
of LDL cholesterol, whereas <10% of the [3H]CE was
isolated within the VLDL+IDL fraction (Figure IB).
The distribution of HDL-derived [3H]CE among
LDL was not usually superimposable with the plasma
LDL mass. Shown in Figure 2 are three examples of
DGUC profiles in which the HDL-derived [3H]CE was
distributed primarily within the buoyant LDL density
range after incubation. As shown in these examples, the
peak of the [3H]CE radioactivity occurred at a lower
density compared with the peak of the LDL protein
II
iioac ity
836
*
—o—
4 C control
6hrit37°C
3000
•a
A
f\
/ \\
A /
2000-
6
1000
5400
3600 •§,
1800
X
10
20
Fraction #
30
FIGURE 1. Panel A: Graph showing density gradient ultracentrifugation cholesterol and [3H]cholesteryl ester (CE)
radioactivity profiles in subject 12. Plasma was incubated with
HDL-derived [3H]CEfor 1-6 hours at either 4°C (control) or
37"C. A, Distribution of [3H]CE radioactivity in the 4°C
control; a, distribution of [3H]CE radioactivity 6 hours after
incubation at 37°C; •, distribution of cholesterol among the
plasma lipoproteins (6-hour sample). Panel B: Graph showing
changes with time in the percentage of total [3H]CE radioactivity associated with VLDL+IDL (; fractions 1-9), LDL
(*, fractions 10-20), and HDL (m, fractions 21-39) from
subject 12. HDL, high density lipoprotein; LDL, low density
lipoprotein; VLDL, very low density lipoprotein; IDL, intermediate density lipoprotein.
mass (1.037 versus 1.039 g/mL, 1.033 versus 1.036 g/mL,
and 1.038 versus 1.040 g/mL for panels A, B, and C,
respectively). Among the studies in which HDL-derived
[3H]CEs were incubated with freshly isolated plasma,
the distribution of the [3H]CE was primarily in the
buoyant region of LDL in 10 of 14 studies (71.4%),
compared with three of 14 studies (21.4%) in which the
radiolabeled CE was superimposable with the LDL
protein or cholesterol mass. In one subject (subject 10),
the [3H]CE distributed preferentially into a more dense
region of the LDL.
On the basis of each subject's unique DGUC LDL
protein peak, LDL was divided equally into buoyant
(LDL[) and dense (LDL2) subpopulations, and the
percentage of [3H]CE radioactivity isolated within each
lipoprotein subpopulation was determined (Table 1).
On average, 17.9±6% of the total [3H]CE radioactivity
Marietta et al
HDL-Derived CEs Among LDL Subpopulations
837
TABLE 1. Percent Distribution of HDL-Derived [3H]CholesteryI
Esters Among Plasma Lipoproteins After Incubation at 37 °C*
Subject
Control a
Control b
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
I
1
1.2
1.2
6.1
5.3
9.6
12.1
9.0
11.3
13.4
16.7
10.7
14.9
17.4
9.3
8.0
5
u
0.9
11.3
4
•8
2.2
1
3b
1
LDL2
2.3
2
3at
I
LDL,
5.2
VLDL+IDL
6
7
7.9
6.7
12.6
10.4
9.2
9.3
10.9
14.1
12.8
7.5
8at
9.4
9.8
9.1
8b
14.7
9.9
8.6
9
8.0
9.1
10
11
14.8
4.2
8.4
6.3
5.0
3.5
3.2
12
7.7
13.8
13.0
11.8±4t
9.5±3§
8.4±3
Mean±SD
HDL
90.3
96.7
77.3
69.3
71.4
72.0
60.3
70.8
66.7
55.7
71.7
66.8
74.5
74.7
88.3
65.5
70.4±8
FIGURE 2. Graphs showing density gradient ultracentrifugation profiles of the distribution of relative protein mass (optical
density at 280 nm, solid line) and [3H]cholesteryl ester (CE)
radioactivity (o, 4°C control and m, after 90 minutes of
incubation at 37°C) in subject 1 (panel A), subject 2 (panel B),
and subject 3 (panel C). For subjects 1 and 3, LDL, was
defined as fractions 10-14 (1.031<d<1.039), and LDL2 was
defined as fractions 15-19 (1.039<d<1.054); for subject 2,
LDL, was defined as fractions 9-13 (1.026<d<1.037), and
LDL2 was defined as fractions 14-18 (1.037<d<1.051).
LDL, low density lipoprotein; HDL, high density lipoprotein.
HDL, high density lipoprotein; VLDL, very low density lipoprotein; IDL, intermediate density lipoprotein; LDL, low density
lipoprotein.
•Between 50,000 and 580,000 cpm of HDL-derived [3H]cholesteryl esters were added to freshly isolated plasma. Samples were
incubated for 90 minutes at 37°C except for 3b, 8b, and 12, which
were incubated for 2 hours. After incubation, each sample was
characterized by density gradient ultracentrifugation (DGUC)
and subdivided into VLDL+IDL, LDL,, and LDL2 on the basis of
each subject's DGUC profile. LDL was divided at the protein peak
into equal buoyant (LDL!) and dense (LDL2) subpopulations.
Control a was kept at 4°C for 90 minutes, and control b was kept
at 4°C for 2 hours. Data presented as percent of total radioactivity.
flncubations done on the same subjects 2 and 3 months apart,
respectively.
^Significantly different from total LDL; /?<0.0009 (and
/?<0.0045 when corrected for multiple t tests).
SSignificantly different from LDL2; /?<0.0054 (and p<0.0027
when corrected for multiple t tests).
was isolated within the LDL density range and
11.8±4% of the [3H]CE radioactivity within the
VLDL+IDL region after 1.5-2 hours of incubation at
37°C (/><0.005, unpaired t test). In addition, significantly more [3H]CE radioactivity was isolated within the
buoyant LDLi subpopulation compared with the more
dense LDL 2 subpopulation (9.5±3% versus 8.4±3%,
respectively; p<0.005, unpaired t test).
To examine potential differences in the rate of movement of HDL-derived [3H]CE into the various apo
B-containing subpopulations, time-course experiments
were carried out in the presence of LpL to enhance the
LTP reaction. Freshly isolated plasma from two subjects
was incubated with HDL-derived [3H]CE, LpL, and
human serum albumin for 30, 60, and 90 minutes at
37°C. After incubation, each sample was characterized
by DGUC, and changes in the total amount of the
radioactivity isolated within VLDL, IDL, LDL,, LDL2,
and HDL were plotted as a function of time (Figure 3).
In both subjects, the rate of increase of HDL-derived
radioactivity was greater in the buoyant LDL, subpopulation than in the other apo B-containing lipoprotein
subpopulations. The rates of change of radioactivity
(calculated as the percentage of change per minute
between 30 and 90 minutes of incubation) into LDL,
and LDL 2 for subject 13 (Figure 3A) were 0.233 and
0.137, respectively. The rates of change of radioactivity
into LDL, and LDL 2 for subject 1 (Figure 3B) were
0.083 and 0.045, respectively. That is, the rate of movement of HDL-derived [3H]CE into LDL, was approximately 2.5 times greater than VLDL, 4.4 times greater
than IDL, and 2.6 times greater than LDL2 in both
subjects.
Studies were done in which LDL subpopulations
(buoyant and dense) were first isolated and then incubated separately or together with radiolabeled HDL to
test the hypothesis that HDL-derived CE would transfer preferentially into the buoyant LDL. Plasma LDL
was first subfractionated by DGUC in a vertical rotor
and divided into VLDL+IDL, LDL,, and LDL 2 on the
basis of the protein profile. An aliquot of HDL
(d> 1.080) was radiolabeled with [3H]CE and incubated
at the donor's physiological concentrations of VLDL
cholesterol (d< 1.006 fraction; 9.7 mg/dL); LDL, cholesterol (25.2 mg/dL), LDL 2 cholesterol (39.1 mg/dL),
or both; unlabeled HDL cholesterol (d> 1.080; 53.8
10
20
30
Fraction #
838
Arteriosclerosis and Thrombosis
Vol 13, No 6 June 1993
100
80
60
40
HDL
LDL,
«•>
20
10
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
VLDL
20
40
60
80
100
Minutes at 37° C
100
B
10
20
30
Fraction #
HDL
20
40
60
80
100
Minutes at 37° C
FIGURE 3. Graphs showing changes in the high density
lipoprotein (HDL) -derived [3H]cholesteryl ester (CE) radioactivity with time among the various lipoprotein subpopulations in subject 13 (panel A) and subject 1 (panel B). Plasma
from each subject was incubated with HDL-derived [3H]CE
for 30, 60, and 90 minutes in the presence of lipoprotein lipase
(LpL). After incubation, each sample was separated by
density gradient ultracentrifugation, and radioactivity within
each lipoprotein subpopulation was measured. The data at 0
minutes of incubation represent the radioactivity in each
lipoprotein subpopulation of the 4°C control (without LpL).
LDL, low density lipoprotein; IDL, intermediate density
lipoprotein; VLDL, very low density lipoprotein.
FIGURE 4. Density gradient ultracentrifugation profiles of
the percentage distribution of very low density lipoproteinderived 123I-protein (A), high density lipoprotein-derived
[3H]cholesteryl ester (CE) (%), and relative protein (solid line;
optical density at 280 nm) of whole plasma incubated at either
4°C (upper panel) or after 1 hour of incubation at 37°C in the
presence of lipoprotein lipase (LpL) (lower panel).
mg/dL); and radiolabeled HDL-derived [ 3 H]CE (total
cholesterol, 8.8 mg/dL) for 2 hours at 37°C and characterized by D G U C as described earlier. Table 2 gives the
calculated specific activities for each L D L subpopulation. U n d e r all conditions, the buoyant LDL, fraction
had a higher specific activity than did the more dense
L D L 2 fraction ( 2 5 . 4 ± 8 and 1 5 . 3 ± 5 , respectively,
m e a n ± S D ; n = 3 each).
Additional studies were done to test whether the
distribution of HDL-derived [ 3 H]CE into buoyant L D L
could be due, at least in part, to the presence of
TABLE 2. Specific Activities for LDL, and LDL2 After
Incubation at 37 °C for 2 Hours
CE acceptor
Plasma
LDL,
LDL2
LDL,+LDL2
Lipoprotein
fraction
LDL,
LDL2
LDL,
LDL 2
LDL,
LDL2
SA*
(cpm/CE)
25.3
13.3
33.3
21.5
17.6
11.2
LDL, low density lipoprotein; CE, cholesteryl ester.
*SA, specific activity; cholesteryl ester (jig per subpopulation)
calculated as 0.75 x total cholesterol of each fraction.
Marietta et al
HDL-Derived CEs Among LDL Subpopulations
839
LDL (fractions 11-19) was found to be associated with
the TG moiety.
O
00
cs
I
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
i
1
Pi
20
30
Fraction #
FIGURE 5. Density gradient ultracentrifugation profiles of
the percentage distribution of very low density lipoproteinderived [14C]trigtyceridcs (TG) (A.), high density lipoproteinderived [3H]cholesteryl ester (CE) (•), and relative protein
(solid line; optical density at 280 nm) in plasma incubated at
4°C (upper panel) or after 90 minutes at 37°C in the presence
of lipoprotein lipase (LpL) (lower panel). The I4C radioactivity represents the percentage distribution through fraction 32
only, which was 32.4% of the total I4C radioactivity. The
remaining 67.6% was associated with albumin (as free fatty
acids) at the bottom of the gradient (fractions 33-37).
"remnant-like" particles. First, iodinated VLDL and
HDL-derived [3H]CE were added to plasma in the
presence of LpL to generate remnant-like particles, and
the distribution of 125I-protein and [3H]CE was examined simultaneously. As shown in Figure 4, after 1 hour
of incubation at 37°C in the presence of LpL, the
distribution of the m I-protein was very similar to the
distribution of the [3H]CE within the apo B-containing
particles. Consistent with previous observations, the
radioactivity distributed primarily in the buoyant region
of the LDL.
Second, experiments were done by incubating VLDL
[14C]TG and HDL-derived [3H]CE to plasma, again in
the presence of LpL. Again, after 90 minutes of incubation at 37°C, both isotopes distributed similarly within
the buoyant region of LDL (Figure 5). Although the
percentage of 14C radioactivity that was associated with
free fatty acids could not be determined for each
fraction within LDL, approximately 50% of the total
radioactivity isolated within the fractions pooled within
Discussion
The movement of HDL-derived CE into VLDL and
LDL has been described previously; however, those
studies have been done with isolated lipoprotein classes
in a buffer system.25-28 The studies presented here
describe the transfer of HDL-derived [ H]CE into various apo B-containing lipoproteins in whole plasma
under in vitro conditions, which more closely mimic the
physiological state (in plasma with LCAT and LTP
active). After incubation, the lipoproteins were separated and characterized by DGUC and/or gel filtration
so that the distribution of the HDL-derived radiolabeled CE into all apo B-containing lipoproteins could
be defined in more detail. A number of informative
observations have been made from these studies. First,
at physiological concentrations of plasma lipoproteins,
significantly more HDL-derived [3H]CE was transferred
to LDL than to VLDL. Second, in most subjects, more
HDL-derived [3H]CE distributed into the buoyant LDL
subpopulation compared with the dense LDL subpopulation. And finally, studies in which remnant-like particles were produced in vitro suggested that these
lipoprotein particles are avid acceptors for HDL-derived [3H]CE and may contribute to the skewed distribution of [3H]CE among buoyant LDL.
Studies characterizing the distribution of HDL CE
among apo B-containing lipoproteins have been inconsistent. Contrary to the observations reported here,
Barter et al28 and Eisenberg26 suggested that plasma
LDL are inefficient HDL CE acceptors in in vitro
incubation systems. In Eisenberg's studies, however,
isolated LDL and HDL were incubated at equivalent
CE concentrations in artificial buffer systems. Since
normolipidemic subjects generally have between five
and nine times more CE mass in LDL than in VLDL,17
it is difficult to extrapolate the results of these studies
back to the physiological state. Barter et al28 examined
the movement of HDL-derived [3H]CE into VLDL and
LDL after incubation in whole plasma and noted that
the specific activity of the LDL was approximately four
times lower than the plasma VLDL. In other studies
using whole plasma19-29 or whole sera,1-2-30 however, the
amount of radiolabeled CE or CE mass transferred into
LDL was at least equivalent to, if not greater than, into
VLDL.
In the present studies, the transfer of HDL-derived
[3H]CE into apo B-containing lipoproteins was examined in whole plasma. At each subject's physiological
concentrations of lipoproteins, significantly more HDLderived [3H]CEs were transferred to LDL than to
VLDL+IDL. The inconsistent results among the various published reports may reflect differences in the
incubation conditions (concentrations of donors and
acceptors, isolated lipoproteins versus plasma or sera,
enzyme inhibitors used, etc.), length of time of the
incubations, and/or differences in study subjects.
Although we assumed that the amount of the HDLderived [3H]CE associated with apo B-containing lipoproteins would be directly related to the lipoprotein
mass,31 in most of the subjects studied, the HDLderived [3H]CE distributed preferentially within the
buoyant region of the LDL. This preference for buoyant
840
Arteriosclerosis and Thrombosis
Vol 13, No 6 June 1993
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
LDL as the CE acceptor was supported by additional
studies in which the rate of transfer of HDL-derived
[3H]CE was greater into LDLj than LDL 2 , and the
specific activities of isolated LDLt were greater than
isolated LDL2. Recently, Phair et al32 have tested
various kinetic models describing "reverse cholesterol
transport." In their theoretical models, which include
the transfer of HDL cholesterol to apo B-containing
lipoproteins, particles between 2.5 and 3.5 xlO 6 g//imol
were the predicted preferred acceptors for HDL-derived CE. That is, our observation suggesting large,
buoyant LDL as preferred acceptors for HDL-derived
CE is consistent with their model predictions. In addition, recent work by Gambert et al33 has also been
consistent with the present observations. In a series of
studies, they demonstrated that when LDL and HDL
were incubated together in the presence of LTP, the
distribution of esterified cholesterol within LDL was
reisolated in larger particles.
The studies presented here offer insight into several
forms of lipoprotein heterogeneity. As expected, LTP
activity varies among individuals. However, these data
also suggest that variability in the transfer of HDLderived [3H]CE among the apo B-containing particles
exists within an individual. Finally, various subpopulations of lipoproteins within the IDL-LDL density
range, such as remnant-like particles, may also contribute to the distribution of HDL-derived [3H]CE among
the apo B-containing lipoproteins. This last observation is consistent with studies by Sammett and Tall15
showing that during LpL-mediated TG hydrolysis,
HDL-derived [3H]CE are transferred to the large LDL
remnant-like particles produced in vitro. Although the
notion of metabolic heterogeneity within the LDL
density range is certainly not new,18-34-39 the factors that
influence this phenomenon are still poorly understood.
In conclusion, the present studies suggest heterogeneity among apo B-containing lipoproteins as acceptors of HDL-derived CE. In most subjects studied, more
of the HDL-derived [3H]CE distributed within LDL
than VLDL. In addition, in most subjects, more HDLderived pHJCE distributed within buoyant LDL than
dense LDL. This distribution of HDL-derived [3H]CE
into buoyant LDL may be related, in part, to the
presence of remnant-like particles that may reside
within this density region.
Acknowledgments
The authors thank Martha Kimura and Steve Hashimoto for
their excellent technical expertise and Dr. John Brunzell for
the purified LpL and for his support throughout these studies.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
References
1. Rehnborg CS, Nichols AV: The fate of cholesteryl esters in human
serum incubated in vitro at 38°C. Biochim Biophys Acta 1964;84:
596-603
2. Nichols AV, Smith L: Effect of very low-density lipoproteins on
lipid transfer in incubated serum. / Lipid Res 1965;6:206-210
3. Gkxnset JA, Norum KR: The metabolic role of lecithin: cholesterol
acyl transferase: Perspectives from pathology. Adv Lipid Res 1973;1:
1-65
4. Tall AR, Small DM: Body cholesterol removal: Role of plasma
high-density lipoproteins. Adv Lipid Res 1980,17:1-51
5. Son YC, Zilversmit DB: Increased lipid transfer activities in hyperlipidemic rabbit plasma. Arteriosclerosis 1986;6:345-355
6. Tail AR, Granot E, Brocia R, Tabas I, Hesler C, Williams K,
Denke M: Accelerated transfer of cholesteryl esters in dyslipi-
27.
28.
29.
30.
demic plasma: Role of cholesteryl ester transfer protein. J Clin
Invest 1987;79:1217-1225
Fielding PE, Fielding CJ, Havel RJ, Kane JP, Tun P: Cholesterol
net transport, esterification, and transfer in human hyperlipidemic
plasma. J Clin Invest 1983;71:449-460
Rudel LL: Plasma lipoproteins in atherosclerosis in nonhuman primates, in Kalter SS (ed): Use of Nonhuman Primates in Cardiovascular
Research. Austin, Tex, University of Texas Press, 1980, pp 37-57
Mahley RW: Atherogenic hyperlipoproteinemia: The cellular and
molecular biology of plasma lipoproteins altered by dietary fat and
cholesterol. Med Clin North Am 1982;66:375-402
Pattnaik NM, Zilversmit DB: Interactions of cholesteryl ester
exchange protein with human plasma lipoproteins and phospholipid vesicles. J Biol Chem 1979;254:2782-2786
Morton RE, Zilversmit DB: Inter-relationship of lipids transferred
by the lipid-transfer protein isolated from human lipoproteindeficient plasma. J Biol Chem 1983;258:11751-11757
Harmony JAK, Jackson RL, Ellsworth JL, Demel RA: Interaction
of a human plasma lipid transfer protein complex with lipid monolayers. Biochim Biophys Acta 1982;690:215-223
Noel S-P, Dupras R, Vezina C, Marcel YL: Comparisons of verylow-density lipoproteins isolated from rat liver perfusate, rat serum
and human plasma as acceptors for cholesteryl ester transfer.
Biochim Biophys Acta 1984;796:277-284
Fielding CJ, Reaven GM, Lui G, Fielding PE: Increased free
cholesterol in plasma low and very low density lipoproteins in
non-insulin-dependent diabetes mellitus: Its role in the inhibition
of cholcsteryl ester transfer. Proc Natl Acad Sci USA 1985;81:
2512-2516
Sammett D, Tall AR: Mechanisms of enhancement of cholesteryl
ester transfer protein activity by lipolysis. / Biol Chem 1985;260:
6687-6697
Iverius P-H, Ostlund-Lindqvist AM: Lipoprotein lipase from
bovine milk: Isolation procedure, chemical characterization and
molecular weight analysis. J Biol Chem 1976;251:7791-7805
The Upid Research Clinics Population Studies Data Book, Volume 1:
The Prevalence Study. US Dept of Health, Education, and Welfare
publication No. (NIH) 80-1527. Washington, DC
Marzetta CM, Foster DM, Brunzell JD: Relationships between
LDL density and kinetic heterogeneity in subjects with normolipidemia and familial combined hyperlipidemia using density gradient ultracentrifugation. J Upid Res 1989^30:1307-1317
Thomas MS, Rudel LL: 3[H]-Cholesteryl ester labeling and transfer among human and nonhuman primate plasma lipoproteins.
Anal Biochem 1984;130:215-222
McFarlane AS: Efficient trace labeling of proteins with iodine.
Nature 1958; 182:57
Bilheimer DW, Eisenberg S, Levy RI: The metabolism of very low
density lipoprotein proteins: Preliminary in vitro and in vivo observations. Biochim Biophys Acta 1972;260:212-221
Marzetta CA, Rudel LL: A species comparison of low density
lipoprotein heterogeneity in nonhuman primates fed atherogenic
diets. J Upid Res 1986;27:753-762
Allain CC, Poon LS, Chan CSG, Richmond W, Fu PC: Enzymatic
determination of total serum cholesterol. Clin Chem 1974^20:470-475
Godfrey AM: Statistics in practice. N Engl J Med 1985;313:
1450-1456
Tall AR, Sammett D, Vita GM, Deckelbaum R, Olivecrona T:
Lipoprotein lipase enhances the cholesteryl ester transfer proteinmediated transfer of cholesteryl esters from high density lipoproteins to very low density lipoproteins. J Biol Chem 1984;259:
9587-9594
Eisenberg S: Preferential enrichment of large-sized very low density lipoprotein populations with transferred cholesteryl esters.
J Lipid Res 1985;26:487-494
Sniderman A, Teng B, Vezina C, Marcel YL: Cholesterol ester
exchange between human plasma high and low density lipoproteins mediated by a plasma protein factor. Atherosclerosis 1978;31:
327-333
Barter PJ, Hopkins GJ, Gorjatschko L, Jones ME: A unified model
of esterified cholesterol exchanges between human plasma lipoproteins. Atherosclerosis 1982;44:27-40
Dullaart RPF, Groener JEM, Erkclens DW: Effect of the composition of very low and low density lipoproteins on the rate of
cholesteryl ester transfer from high density lipoproteins in man,
studied in vitro. EurJ Clin Invest 1987;17:241-248
Akanuma Y, Glomset J: In vitro incorporation of cholesterol-"C
into very low density lipoprotein cholesteryl esters. J Lipid Res
1968;9:620-626
Marietta et al
31. Chajek T, Fielding CJ: Isolation and characterization of a human
serum cholesterol ester transfer protein. Proc NatlAcad Sci U SA
1978;75:3445-3449
32. Phair RD, Hazzard WR, Applebaum-Boden D: Theoretical consequences of hypotheses concerning reverse cholesterol transport,
in Miller NE (ed): High Density Upoproteins and Atherosclerosis II.
New York, Elsevier, 1989, pp 313-320
33. Gambert P, Bouzerand-Gambert C, Athias A, Farnier M, Lallemant C: Human low density lipoprotein subfractions separated by
gradient gel electrophoresis: Composition, distribution, and alterations induced by cholesteryl ester transfer protein. J Upid Res
1990^1:1199-1210
34. Fisher WR, Zech LA, Bardalaye P, Warmke G, Berman M: The
metabolism of apolipoprotein B in subjects with hypertriglyceridemia and polydisperse LDL / Upid Res 1980;21:760-774
35. Teng B, Sniderman AD, Soutar AK, Thompson GR: Metabolic
basis of hyperapobetalipoproteinemia: Turnover of apolipoprotein
HDL-Derived CEs Among LDL Subpopulations
36.
37.
38.
39.
841
B in low density lipoprotein and its precursors and subfractions
compared with normal and familial hypercholesterolemia. / Clin
Invest 1986;77:663-672
Vega GL, Grundy SM: Kinetic heterogeneity of low density lipoproteins in primary hypertrigryceridemia. Arteriosclerosis 1986;6:
395-406
Foster DM, Chait A, Albers JJ, Failor RA, Harris C, Brunzell JD:
Evidence for kinetic heterogeneity among human low density lipoproteins. Metabolism 1986^5:685-6%
Marzetta CA, Foster DM, Brunzell JD: Relationships between
LDL density and kinetic heterogeneity in subjects with normolipidemia and familial combined hyperlipidemia using density gradient ultracentrifugation. / Upid Res 1989;30:1307-1317
Marzetta CA, Foster DM, Brunzell JD: Conversion of plasma
VLDL and IDL precursors into various LDL subpopulations
using density gradient ultracentrifugation. J Upid Res 1990;31:
975-984
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
Lipid transfer protein-mediated distribution of HDL-derived cholesteryl esters among
plasma apo B-containing lipoprotein subpopulations.
C A Marzetta, T J Meyers and J J Albers
Arterioscler Thromb Vasc Biol. 1993;13:834-841
doi: 10.1161/01.ATV.13.6.834
Arteriosclerosis, Thrombosis, and Vascular Biology is published by the American Heart Association, 7272 Greenville
Avenue, Dallas, TX 75231
Copyright © 1993 American Heart Association, Inc. All rights reserved.
Print ISSN: 1079-5642. Online ISSN: 1524-4636
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://atvb.ahajournals.org/content/13/6/834
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Arteriosclerosis, Thrombosis, and Vascular Biology can be obtained via RightsLink, a service of the Copyright
Clearance Center, not the Editorial Office. Once the online version of the published article for which permission
is being requested is located, click Request Permissions in the middle column of the Web page under Services.
Further information about this process is available in the Permissions and Rights Question and Answerdocument.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Arteriosclerosis, Thrombosis, and Vascular Biology is online
at:
http://atvb.ahajournals.org//subscriptions/