Effect of hyperapo B LDL on cholesterol esterification in THP-1 macrophages

ATHEROSCLEROSIS

Atherosclerosis

102 (1993) 23-36

Effect of hyperapo B LDL on cholesterol esterifkation THP- 1 macrophages

in

Stephanie D. Kafonek *, Inna Raikhel, Paul S. Bachorik, Peter 0. Kwiterovich, Jr. Departments of Pediatrics and Medicine, Lipid Research Atherosclerosis Unit. The Johns Hopkins University School qf Medicine. Baltimore, MD 2120.5, USA

(Received

23 December

1991;

revision received IO March 1993; accepted 29 March 1993)

Abstract

Hyperapobetalipoproteinemia (hyperapo B), a common disorder associated with coronary artery disease, is characterized by an increased number of small, dense, low density lipoprotein (LDL) particles. The cellular mechanisms responsible for early atherosclerosis in hyperapo B are unknown. We tested the hypothesis that hyperapo B LDL may be preferentially metabolized through an LDL receptor independent pathway promoting the accumulation of cellular cholesteryl ester (CE). THP-1 macrophages have little inducible LDL receptor activity after differentiation with phorbol esters and are, therefore, suitable for assessing non-LDL receptor mediated uptake of lipoproteins. LDL isolated from hyperapo B donors was found to have significantly lower total cholesterol to protein ratio (P = 0.03) higher average density (P = 0.0001) and smaller particle diameter (P = 0.016) compared with normal (control) LDL. LDL (250 pg lipoprotein-protein/ml) from normal (n = 11) and hyperapo B (n = 18) subjects were incubated for 24 h with THP-1 macrophages. The mean (S.D.) CE accumulation was 6.2 (3.6) for the normal and 6.4 (2.6) for the hyperapo B LDL (P = 0.84). CE accumulation in cells incubated with malondialdehyde modified (MDA) LDL, or without added lipoprotein, was 18.2 (2.0) and 0.6 (0.7), respectively. CE mass accumulation was significantly correlated with time (6-48 h) of incubation and concentration (100-500 &ml) of LDL protein (P < 0.05); no differences were observed between normal and hyperapo B LDL. Similarly, when the major LDL species was isolated by density gradient ultracentrifugation, mean (S.D.) CE was similar for the normal and hyperapo B LDL (8.7 (1.2) vs. 6.9 (1.5)). There were no differences in the mean (SD.) incorporation of [‘4C]oleate into CE (nmolimg cell protein per 6 h) in THP-1 macrophages incubated with normal or hyperapo B LDL (0.238 (0.045) vs. 0.211 (0.046)); results were comparable in human monocyte-derived (HMD) macrophages (0.298 (0.037) vs. 0.258 (0.022)). Also, mean (SD.) cellular uptake and degradation (ng ‘251/mg cell protein per h) in THP macrophages of normal and hyperapo B LDL were similar (uptake: 18 (14) vs. 12 (6.0); degradation: 58 (32) vs. 44 (8)). In summary: (1) hyperapo B LDL did not stimulate the accumulation of cellular CE via LDL receptor independent pathways in THP-1 macrophages, (2) normal and hyperapo B LDL stimulation of CE synthesis is similar in THP-1 and HMD macrophages and (3) no differences in cellular uptake and degradation of normal and hyperapo B LDL were observed in THP macrophages. Key worris:

* Corresponding

Low density lipoprotein;

Hyperapo B; THP-1 macrophages; Human macrophages

author.

0021-9150193/$06.00 0 1993 Elsevier Scientific SSDI 002 l-9 150(93)05060-1

Publishers

Ireland

Ltd. All rights reserved

24

1. Introduction Hyperapobetalipoproteinemia (hyperapo B) is a lipoprotein disorder commonly found in patients with premature coronary artery disease (CAD) [ 1,2]. Hyperapo B is characterized by an elevated level of the major apolipoprotein of low density lipoprotein (LDL), apo B, and by an increased number of smaller, denser LDL particles that contain less cholesterol than normal [3]. Patients with hyperapo B may be normolipidemic, hypertriglyceridemic, or when the number of LDL particles is sufficiently elevated, hypercholesterolemic [4]. The elevated number of small, dense LDL particles in hyperapo B results from the hepatic overproduction of apo B and very low density lipoproteins (VLDL) [5]. The activity of the LDL receptor is normal in hyperapo B [6]. In some families, hyperapo B may reflect the presence of familial combined hyperlipidemia [7]. The cellular mechanism(s) responsible for the increased prevalence of coronary atherosclerosis in hyperapo B has not been identified. At least two hypotheses have been proposed: (1) decreased aflinity of small, dense LDL particles for the LDL receptor and (2) enhanced removal of such particles through an LDL receptor-independent (scavenger) pathway. The first hypothesis is supported by a number of observations. First, the fractional catabolic rate of small, dense LDL particles in vivo in one study appeared to be slower than that of larger, more buoyant LDL particles [5]. Second, in vitro studies demonstrate that smaller, denser LDL have a decreased reactivity to a monoclonal antibody against the LDL receptor binding domain of apo B [8]. Third, recent studies of the displacement of lz51-LDL from the LDL receptor in cultured libroblasts at 4°C indicate that small, dense LDL displaces 12?-LDL to a significantly lesser extent than normal sized LDL [9]. Earlier studies in cultured fibroblasts [6,10] and macrophages [lo] had failed to find any significant differences between the binding and degradation of small, dense LDL and normal LDL, an observation probably related to the 3-fold variability in LDL binding and degradation at 37°C in cultured libroblasts. Less information is available relevant to the sec-

S.D. Kajionek et al. /Atherosclerosis

102 (1993) 23-36

ond hypothesis, namely, that smaller, denser LDL may be preferentially taken up through LDL receptor-independent pathways. THP- 1 cells, a human monocytic leukemia cell line, manifest many properties of human monocyte-derived (HMD) macrophages and are a convenient system for testing hypotheses related to lipoprotein metabolism [ 1l-131. These cells are maintained as monocytes in suspension, but can be induced to differentiate into adherent macrophages by the addition of phorbol ester. Previous studies have demonstrated that the rate of cholesterol esterification does not increase when phorbol ester treated THP-1 cells are exposed to native LDL, suggesting that LDL receptor activity may be down-regulated in these cells [ 111. Modified LDL (such as acetyl-LDL), however, markedly enhances cholesterol esterilication in such cells through an LDL receptor independent pathway [ 1l- 131. We therefore used phorbol ester treated THP-1 cells as a model system to test the hypothesis that hyperapo B LDL promotes abnormal cholesterol esterification through LDL receptor independent pathways. 2. Materials and methods 2. I. Human subjects Sixteen normolipidemic (7 male, 9 female) and 22 hyperapo B (13 male, 9 female) subjects between 25 and 70 years of age were studied (see Table 1). All subjects were free of diabetes mellitus, CAD and liver, kidney and thyroid disease. None were taking lipid-lowering medications. Two hyperapo B subjects and one control subject were taking oral contraceptive agents. Five other hyperapo B subjects were taking a single medication each including flecainide, verapamil, allopurinol, dyazide and metaprolol. One normolipidemic subject was taking capozide. No other medications were reported by the subjects. Normolipidemia was defined as (1) absence of type III hyperlipoproteinemia (dysbetalipoproteinemia), (2) triglyceride and LDL cholesterol concentrations below the 75th percentile (age- and sex-specific) [14], (3) HDL cholesterol above the 25th percentile (age- and sex-specific) [14] and (4) plasma LDL apo B < 130 mg/dl. Hyperapo B was

S. D. Kafonek et al. /Atherosclerosis

102 (I 993) 23-36

defined as (1) absence of type III hyperlipoproteinemia (dysbetalipoproteinemia), (2) absence of familial hypercholesterolemia (FH) as judged by normal or near-normal LDL cholesterol concentrations, absence of xanthomas, and family studies, (3) plasma LDL apo B ~130 mg/dl, and (4) LDL cholesterol ~95th percentile, age- and sex-specific [ 141. 2.2. Lipid, lipoprotein cholesterol and triglyceride measurements Blood was collected after a 12-h fast. The samples were drawn into evacuated tubes containing solid disodium ethylene-diamine tetraacetate (EDTA), final concentration 1.5 mg/ml. The measurement of cholesterol and triglycerides in plasma and in isolated lipoproteins was performed enzymatically (Boehringer Manheim Diagnostics, Indianapolis, IN). VLDL, LDL and HDL cholesterol were separated and measured using the methods of the Lipid Research Clinic’s Program, modified as reported earlier [7]. 2.3. Apolipoprotein B assay

The concentration of LDL apo B was measured in plasma and isolated lipoproteins by radial immunodiffusion (RID), using the M-Partigen Apolipoprotein B Kit (Calbiochem-Behring Corp., La Jolla, CA), as previously described [7]. 2.4. Isolation of LDL

LDL (density (d) = 1.O19- 1.063 g/ml) were isolated from plasma by sequential preparative ultracentrifugation after adjusting the density using solutions of KBr-NaCl containing 0.001 M EDTA [7,15]. LDL was determined to be free of albumin and other lipoproteins by agarose gel electrophoresis. In some samples, the density of isolated LDL was then adjusted to 1.050 g/ml by dialysis against KBr (d = 1.050 g/ml) in preparation for equilibrium density gradient ultracentrifugation (EDGU). 2.5. Gradient gel electrophoresis

Electrophoresis of unfractionated plasma was performed as described by Krauss and Burke [ 161. Samples were applied to polyacrylamide gradient gels (2-16%) and subjected to electrophoresis (125

25

V, 25 h, 12-14”C), using Tris (0.09 M)-boric acid (0.8 M) Naz EDTA (0.003 M) buffer, pH 8.3. Gels were pre-run at 125 V for 2 h. Samples (5 ~1) were applied to each lane and the voltage was gradually adjusted (20 V for 15 min, 70 V for 15 min, then 125 V for 24 h). To detect lipid, the gradient gels of unfractionated plasma were stained with 0.05% Oil Red 0 in 60% ethanol overnight at room temperature [ 171. The bands were marked with black India Ink. Proteins were stained in 50% methanol containing 10% acetic acid and Coomassie Brilliant Blue R-250. Gels were destained with 20% methanol containing 9% acetic acid. Migration distances for each band were measured. A mixture of apoferritin, thyroglobulin and carboxylated latex beads were analyzed in parallel and used as standards. 2.6. Equilibrium density gradient ultracentrijiigation

A discontinuous density gradient was constructed and LDL was subjected to EDGU and the relative position (R,) of the major LDL band determined as described previously [7]. Fractions (0.2 ml) were collected from the bottom of the tube and the absorbance of each was measured at 280 nm. Selected fractions were pooled to provide either the major LDL subfraction or LDL preparations of light, intermediate and heavy densities (see Results). All LDL preparations were dialyzed against 0.9% NaCl, containing 0.05% EDTA. 2.7. Preparation of MDA-LDL Malondialdehyde bis-dimethyl acetal (MDA, Aldrich) was acid hydrolyzed to malondialdehyde, diluted to 0.2 M with a sodium phosphate buffer and adjusted to pH 6.5 with 10 N NaOH [ 181.The mixture was added in a 1:l ratio (by volume) to LDL (5-10 mg protein/ml) in 0.01 M sodium phosphate buffer (pH 7.4) and incubated at 37°C for 3 h. The reaction was terminated by placing the solution on ice and the MDA-LDL was dialyzed overnight at 4°C against 0.01 M sodium phosphate buffer containing 0.01% EDTA. The protein content was determined by the method of Lowry et al. [19]. To confirm the completeness of the reaction, the mobility of MDA-LDL was compared with

26

S.D. Kafonek et al. /Atherosclerosis

unmodified LDL by electrophoresis using a 0.6% agarose gel. No unmodified LDL was detected in the MDA-LDL preparations. 2.8. Cell culture THP-1 cells. THP- 1 monocyte macrophage cells were obtained from the American Type Culture Collection (Rockville, MD). The cells were grown in suspension in RPMI-1640 medium containing 10% fetal calf serum, streptomycin (100 &ml) and penicillin (100 units/ml) and maintained at 1 x lo6 cells/ml. Two days prior to use, cells were seeded (3 x lo6 cells per 35 mm culture dish) and maintained in medium containing lo-’ M phorbol 12-myristate, 13-acetate (PMA) [l 11, during which time they differentiated into macrophages. The cells were washed twice with phosphate buffered saline (PBS) prior to incubation with LDL. Human monocyte-derived macrophages. HMD macrophages were isolated from a white cell concentrate generated as a by-product of hemapheresis. The cells were diluted with RPMI- 1640 medium, distributed among six 50-ml tubes and with Ficoll-paque (Pharmacia, underlayered Piskataway, NJ). The tubes were centrifuged at 1600 rev./min for 40 min at room temperature. The white cells were removed, washed with RPMI1640 medium supplemented with 100 pg of streptomycin and penicillin (medium A) and 10% commercially prepared AB serum, and incubated with RPMI-1640 medium and 30% AB serum. After 1 h, the cells were washed with PBS and cultured in RPMI-1640 medium containing 5% AB serum for 5 days before use. Incorporation of [14C]oleate into cellular lipids.

Cells were transferred to RPMI-1640 medium without serum containing 0.1 mM [ “C]oleate (10 000 dpmnmol) complexed to bovine serum albumin (BSA) (molar ratio of oleate/albumin, 5:1), and incubated in the presence or absence of LDL. When the incubation was complete, the cellular lipids were extracted by adding 2 ml hexane/isopropanol (3:2, v/v) to each tissue culture flask for 40 min, followed by two washes (1 ml). The extracts and washes were pooled. Tritiated cholesteryl ester and triolein were added as recovery standards and the extracts were dried under a nitrogen stream. The cellular lipids were separated

IO2 (1993) 23-36

by thin-layer chromatography, visualized with 12 vapor, and scraped into scintillation vials. The amount of radiolabelled lipid incorporated into the individual lipid class was determined by liquid scintillation spectrometry. The amount of radioactivity in each well of cells was normalized to the amount of cell protein as measured by Lowry et al. 1191. Measurement of the mass of cellular lipids. The mass of free and esterilied cholesterol was determined by gas chromatography using the method of Ishikawa et al. [20]. For measurement of free cholesterol, the lipid extract was dried under a nitrogen stream, resolubilized in carbon disulfide and injected onto an HP-17 cross-linked 50% phenyl methyl silicone capillary column at 270°C using a helium flow of 30 ml/min. Total cholesterol was measured similarly following saponification. Stigmasterol was used as an internal standard. The mass of esterified cholesterol (EC) was calculated by subtracting the mass of free cholesterol from the mass of total cholesterol and multiplying by 1.68 [7]. The mass of triglyceride was determined enzymatically, using a commercially available kit (Triglyceride Reagent Set, Dow Diagnostics, Cat. No. 47161). Total triglyceride mass was determined directly from an aliquot (20 ~1) of the total (100 ~1) lipid extract. The mass obtained in the aliquot was multiplied by a factor of five and expressed as hg TG mass per mg cell protein. Radioiodination of LDL. LDL protein (l-2 mg) was radioiodinated with Na 1251(Amersham International, Amersham, Bucks., UK) using the method of McFarlane [21] as modified by Bilheimer et al. [22]. Excess 1251was removed by a Sephadex G-25 column (Pharmacia PD-10) and extensive dialysis against 0.15 M NaCl and 0.1% EDTA, pH 7.4. The mean (S.D.) specific activity (ng ‘251/mg LDL protein) of the normal LDL preparations (n = 4) was 341 (2 15) and 209 (75) for the hyperapo B LDL preparations (n = 4). Greater than 99% of the radiolabelled LDL was precipitable by TCA and less than 2% of the labelled iodine was found in the lipid fraction of the lipoprotein for all preparations. Uptake and degradation of LDL. THP-1 macrophages were seeded at a concentration of 3 x lo6 cells/ml in RPMI-1640 containing 10% FCS and

SD. Kafonek et al. /Atherosclerosis

102 (1993) 23-36

21

lop7 M phorbol ester. On Day 3, the medium was aspirated, the cells washed extensively and then incubated with 5 pg radiolabelled lipoproteinprotein in the presence and absence of 50 pg unlabelled lipoprotein. Wells without cells were similarly incubated to assess non-specific uptake and degradation. After incubation (2-8 h), the medium was removed and the proteins were precipitated with 10% TCA (final concentration); the cell monolayer was washed extensively and dissolved in 1 N NaOH. The radioactivity was measured in an aliquot of the TCA supernate and LDL degradation was expressed relative to the amount of cellular protein; cell-associated LDL was determined from an aliquot of the redissolved cells. The extent of high-affinity uptake and degradation at 37°C was determined by incubating THP-1 macrophages (5 pg LDL protein/ml medium) in the presence and absence of a 1O-fold excess of unlabelled lipoprotein (50 pg LDL protein/ml medium); high-affinity uptake and degradation are calculated by subtraction.

variance (ANOVA) was used to compare the extent of CE accumulation resulting from normal or hyperapo B LDL subfractions (see Table 4) and a one-way ANOVA was used to compare heavy hyperapo B LDL to intermediate and light LDL. 3. Results 3.1. Study subjects

The normal and hyperapo B groups demonstrated significantly different mean triglyceride, and total and LDL cholesterol (P < 0.05, Tukey’s SRT) (Table 1). There were no differences in HDL cholesterol between the two groups. The groups were selected on the basis of differences in plasma apo B and this variable was not included in tests of significance. There were no differences in the mean BMI (24.7 vs. 26.3 kg/m2, P = 0.15) or mean age (45 vs. 49 years, P = 0.32) between the normal and hyperapo B subjects, respectively. 3.2. Characterization of LDL preparations

2.9. Statistical analyses Characteristics of the study population (Table 1) were compared using a multivariate analysis of variance. Data relevant to LDL composition and cellular lipid accumulation were compared using a two-tailed Student’s T-test. A two-way analysis of

Table I Mean lipid, lipoprotein normal and hyperapo

Characterization of the LDL preparations isolated from the normal and hyperapo B donors is summarized in Table 2. The LDL preparations from the hyperapo B donors were depleted in cholesterol and relatively enriched in protein com-

Table 2 Characterization cholesterol B donors”

and apo B concentrations

in

Parameter

of normal

and hyperapo

P-value

Donors Normal

Subjects

n

TCb,’

TGC

HDL C

LDL Cc

16 22

173 241

Mean (SD.) 94 I98

55 47

96 I41

109 I88

LDL cholesterol LDL protein RPa

dGroups were significantly different using ANOVA, Wilk’s lambda, P < 0.0001. bTC, total cholesterol; TG, triglyceride; HDL C, lipoprotein cholesterol; LDL C, low density cholesterol; Apo B, apolipoprotein B. ‘Tukey’s Studentized range test, P < .05 between

Hyperapo

B

Mean (SD.)

n

Apo B

(Wdl) Normal Hyperapo B

B LDL preparations

multivariate Diameter high density lipoproteingroups.

(A)

n

I.45 (0.167)

II

I.31 (0.160)

18

0.0343

0.770 (0.042) 245.3 (5. IO)

I3

0.676 (0.033) 235.8 (6.25)

14

0.0001

8

0.016

5

‘Rp, position of the major LDL fraction after EDGU (position of the band relative to the bottom of the tube/length of tube).

S.D. Kafonek et al. /Atherosclerosis

28

Isolation

IO2 (1993) 23-36

of LDL Subfractions

o Normal LDL ?? HyperapoB LDL

Fraction Bottom

Number TOP

Fig. I. Isolation of LDL subfractions by EDGU. LDL (d I .019- 1.063 g/ml) from one normal and from one hyperapo B donor were subjected to EDGU. After ultracentrifugation, the bottom 1.5ml of each tube was collected (fraction I). The remaining gradient was collected from the bottom of the tube in I3 fractions, each containing 0.3 ml of LDL. The absorbance (280 nm) was determined on each fraction and used to select those fractions corresponding to three peak LDL protein density bands as follows: heavy LDL, fractions 3-6; intermediate LDL, fractions 7-10; light LDL, fractions I I-14.

pared with those from the normals. The mean (S.D.) LDL cholesterol/LDL protein ratio was 1.45 (0.167) in the normal controls compared with 1.31 (0.160) in the hyperapo B donors (P = 0.034). As judged by the position of the major LDL fraction isolated by EDGU, LDL from the hyperapo B patients demonstrated significantly higher average density than that of the normals. Buoyant density is inversely related to the Rp value. The mean (S.D.) position of the major LDL band relative to the length of the centrifuge tube (R,) was 0.770 (0.042) in the normal controls compared with 0.676 (0.033) in the hyperapo B subjects (P = 0.0001). Fig. 1. illustrates the patterns observed when LDL was subjected to EDGU. When LDL subfractions were collected from light, intermediate and heavy density ranges following EDGU, the hyperapo B LDL from a given density range was denser (less buoyant) and had a lower

cholesterol/protein ratio compared with the normal LDL. In one of the experiments (13 subjects), plasma from normal and hyperapo B donors was subjected to gradient gel electrophoresis (GGE). The mean (S.D.) diameter (angstrom) was 245.3 (5.10) in the normal controls compared with 235.8 (6.25) in the hyperapo B subjects (P = 0.016). 3.3. LDL metabolism in THP-I macrophage cells Differentiated, phorbol ester treated THP- 1 cells were used to determine whether the metabolism of hyperapo B LDL differed from normal LDL in macrophages with predominately scavenger-receptor activity. Studies were performed with isolated normal (control) and hyperapo B LDL preparations to assess differences in (1) the extent of cellular mass accumulation of cholesteryl ester (CE) and triglyceride (TG), (2) the rate of in-

SD. Kafonek et al. /Atherosclerosis

29

102 (1993) 23-36

corporation of [14C]oleate into cellular CE and TG and (3) the rates of cellular association and degradation of radiolabelled LDL preparations.

Table 3 Cellular lipid accumulation in THP macrophages incubated with LDL

3.3.1. Cellular mass accumulation of CE and TG LDL (d 1.019-l .063 mg/ml) were isolated from 11 normal controls and 18 hyperapo B subjects. Triplicate wells of THP-1 macrophages were incubated with each LDL preparation (250 pg LDL protein/ml medium) for 24 h (Table 3). The mean (SD.) CE mass (pg CE/mg cell protein) was 6.21 (3.56) in the normal controls and 6.43 (2.62) in the hyperapo B subjects. The mean CE mass in cells incubated with LDL was significantly higher (P = 0.001) than those incubated in the absence of LDL and significantly lower (P = 0.001) than those incubated with MDA-LDL. The mean (SD.) TG mass (pg TG/mg cell protein) was 160 (42) compared with 144 (44) with normal and hyperapo B LDL, respectively. These values were not significantly different. We considered that a difference in CE mass accumulation between normal and hyperapo B LDL might become apparent at different LDL protein concentrations or might manifest a time dependence. To address this question, we performed similar studies after incubating the cells for varying periods (6 to 48 h) and varying concentrations of LDL protein (100-500 pg LDL protein/ml medium). For these studies, pools of normal and hyperapo B LDL were prepared by combining isolated LDL preparations from normal and hyperapo B donors, respectively. The extent of CE mass accumulation was highly correlated with the length of incubation with LDL (r = 0.99, P = 0.017) for all preparations; there were no differences in cellular CE mass accumulation over 6-48 h between normal and hyperapo B LDL (Fig. 2). The mean (SD.) CE mass (pg/mg cell protein) of all LDL pools (both normal and hyperapo B) was 1.5 (0.35) at 6 h, 4.4 (0.94) at 24 h and 8.0 (0.97) at 48 h. Similarly, CE mass accumulation was also highly correlated with increasing concentration of LDL protein (mean r = 0.85, P < 0.05) (Fig. 3), but there were no differences between normal and hyperapo B LDL except at the highest concentration (500 &ml) of LDL in which hyperapo B LDL stimulated significantly less

Cellular lipid

Lipoprotein Mean (S.D.) pgimg cell protein per 24 h None Normal LDL (n = I I)

Cholesteryl ester

I.77 6.21* (1.3) (3.56)

Triglyceride

59 (2)

l60* (42)

Hyperapo B MDA-LDL LDL (n = 18) 6.43* (2.62) 144: (44)

l8.2$ (2) 122 (17)

tP = 0.001 between cells with and without added lipoprotein. SF’= 0.001 between LDL and MDA-LDL. *P = N.S. between normal and hyperapo B LDL.

(P = 0.02) CE accumulation than normal LDL. The extent of CE mass accumulation (&mg cell protein) ranged from a mean (S.D.) of 3.1 (0.99) at the lowest concentration (100 pg lipoproteinprotein/ml) to 8.6 (0.72) at the highest concentration (500 pg lipoprotein-protein/ml) for the normal LDL and 2.4 (0.16) and 6.2 (0.44) for the hyperapo B LDL (Fig. 3). We next examined whether the heterogeneity of LDL species within the d 1.019- 1.063 g/ml density range may have attenuated any effect of the major LDL species from hyperapo B donors compared with normal controls. For this experiment, the major LDL species was isolated by EDGU from four normal controls and six hyperapo B subjects. The major LDL band was collected from the gradient by pooling the four sequential fractions with the highest absorbance at 280 nm. These fractions corresponded to the visible major band of the tube used to calculate the Rp (relative position) values cited in Table 2. THP-1 macrophages were incubated with 50 kg lipoprotein-protein/ml medium for 24 h and the mass (&mg cell protein) of CE determined. Mean (S.D.) CE mass was 8.7 (1.2) for the normal controls and 6.9 (1.5) for the hyperapo B subjects.

SD. Kafonek et al. /Atherosclerosis

30

‘ .-;; aJ

5

0

NORMAL LDL

0

HYPERAPOB

102 (1993) 23-36

I

LDL

10

k

-

z

F’ iY

0

W

TIME (HOURS) Fig. 2. CE mass in THP-I cells incubated with LDL as a function of time. THP-1 cells were seeded at a concentration of 3 x IO6 cells/ml in RPMI-1640 with 10% FCS and IO-’ M phorbol ester. On Day 3, the medium was aspirated, the cells washed extensively and then incubated with RPMI-1640 in the presence of normal (n = 2) or hyperapo B LDL (n = 2), (250 pg lipoprotein-protein/ml), for 6, 24 and 48 h. Triplicate wells were assayed for each condition. The mean (SD.) mass (&mg cell protein) of cholesteryl ester (CE) is presented.

‘Ok

@ NORMAL LDL v HYPERAPOB LDL

6-

0

100

200

LDL-PROTEIN

300

(us/ml

400

500

600

medium)

Fig. 3. CE mass in THP-1 cells incubated with increasing concentrations of LDL protein, THP-1 cells were seeded at a concentration of 3 x IO6 cells/ml in RPMI-1640 with 10% FCS and IO-’ M phorbol ester. On Day 3, the medium was aspirated, the cells washed extensively and then incubated with RPMI-1640 in the presence of normal or hyperapo B LDL (100-500 pg lipoprotein-protein/ml) for 24 h. Triplicate wells were assayed for each condition. The mean (S.D.) mass of cholesteryl ester (CE) is presented.

SD. Kafonek et al. /Atherosclerosis

31

102 (1993) 23-36

3.3.2. Incorporation of [14C]oleate in cellular CE and TG

Next, the rate of lipid synthesis was determined by measuring the rate of [‘4C]oleate incorporation into cellular CE and TG in macrophages incubated with LDL. THP-1 macrophages were incubated with LDL (250 pg LDL protein/ml medium) from three normal and five hyperapo B subjects for 6 h. The mean (S.D.) rate of CE synthesis (nmol/mg cell protein) was 0.238 (0.045) in the normal controls and 0.211 (0.046) in the hyperapo B subjects. These were not significantly different. In the absence of lipoprotein, the mean (S.D.) rate was 0.164 (0.008). Thus, hyperapo B LDL did not differ from control LDL in the rate of cellular CE synthesis or the extent of CE mass accumulation. Similarly, no significant differences were observed in cellular TG synthesis. Mean (SD.) rates of [ “C]oleate incorporation into cellular TG (nmol/mg cell protein) were 134 (83) and 123 (35) for normal and hyperapo B LDL, respectively (P = N.S.); the mean rate in the absence of lipoprotein was 66.8.

Table 4 Cholesteryl ester (CE) and triglyceride (TG) synthesis in THP-1 cells incubated with LDL subfractions Cellular lipid

Normal LDL

Hyperapo B LDL

Mean (S.E.) CE (nmol/mg cell protein per 24 h) Light LDL 1.116* (0.024) Intermediate LDL 0.993** (0.113) Heavy LDL NSQ TGS (nmolimg cell protein per 24 h) Light LDL 155 (16) Intermediate LDL 136 (16) Heavy LDL NSQ

0.986* (0.034) 0.934** (0.081) 0.5961 (0.055) 142 (3) 130 (7) II2 (II)

*P = 0.04, **p = 0.77. tP = 0.007 for heavy hyperapo B LDL compared with intermediate and light hyperapo B LDL, Newman-Keuls test, SP = N.S. for all TG comparisons.

Effects of LDL subfractions on [‘4Cjoleate incorporation into CE and TG. We also examined the

possibility that LDL subfractions might differentially stimulate CE synthesis. Three LDL subfractions designated as light, intermediate and heavy LDL, were isolated after EDGU from one normal and one hyperapo B subject (Fig. 1). The amount of LDL protein from the normal donor in the heaviest density range was inadequate for study. Within a given density range, the hyperapo B LDL subfractions, compared with the normal, had lower ratios of cholesterol to protein (light LDL, 1.80 vs. 2.29; intermediate LDL, 1.67 vs. 1.92) indicating a shift toward a higher average density consistent with the presence of cholesterol depleted particles in the hyperapo B subjects. THP-1 macrophages were incubated with the normal and hyperapo B LDL subfractions (50 pg/ml) in the presence of 0.1 mM [r4C]oleate/albumin for 24 h (Table 4). The light LDL fraction in the normal subject stimulated significantly more (P = 0.04) incorporation of [t4C]oleate into CE compared with the hyperapo B LDL; rates were similar for the intermediate LDL from the normal and hyperapo B donor. The

heavy LDL fraction from the hyperapo B donor stimulated significantly less (P < 0.005) [14C]oleate incorporation into CE than the intermediate or light LDL fraction from either donor. No differences (P > 0.5) in TG synthesis were observed between the normal and hyperapo B LDL subfractions (Table 4). 3.3.3. LDL uptake and degradation We next examined the uptake and degradation of normal and hyperapo B LDL in THP-1 macrophages. In the first experiment, LDL from two normal (1 male, 1 female) and two hyperapo B donors (1 male, 1 female) was studied; in the second experiment, pooled LDL from male or female normal controls and from male or female hyperapo B donors was studied. There were no differences in LDL cellular association and degradation based on gender and the data are presented by donor phenotype without respect to gender.

S. D. Kafonek et al. /Atherosclerosis

0

NORMAL

0

HYPERAPOB

102 (1993) 23-36

preparations (n = 2) was 18.0 (14.0) compared with 12.0 (6.0) for the hyperapo B LDL (n = 2). The mean (S.D.) rate of degradation for the normal LDL (n = 2) was 58 (32) compared with 44 (8) for the hyperapo B LDL (n = 2). In the second experiment, the mean (S.D.) rate (ng 1251/mgcell protein per h) of uptake for the normal LDL preparations (n = 2) was 6.5 (1.4) compared with 8.0 for one hyperapo B preparation. Corresponding degradation rates were 200 (10) for the normal LDL preparation (n = 2) compared with 188 (88) for the hyperapo B LDL preparations (n = 2).

LDL LDL

3.4. Normal Hh4D macrophages 0 0

0

NORMAL LDL HYPERAPOB LDL

2

4

6

a

10

TIME (HOURS) Fig. 4. Cellular association and degradation of LDL. THP-I cells were seeded at a concentration of 3 x IO6 cells/ml in RPMI-1640 with 10% FCS and IO-’ M phorbol ester. On Day 3, the medium was apirated, the cells washed extensively and then incubated with 5 pg radiolabelled lipoprotein-protein in the presence or absence of 50 gg unlabelled lipoprotein-protein. After incubation (2-8 h), the medium was removed to TCA for measurement of degradation; the remaining monolayer of cells was dissolved in I N NaOH and an aliquot was assayed for cellular uptake. Triplicate wells were assayed for each condition The mean (S.D.) uptake and degradation is presented.

Finally, we considered the possibility that the findings in THP-1 cells might not reflect those in HMD macrophages. To examine this question, we measured the rate of incorporation of [ i4C]oleate into CE in HMD macrophages isolated from normal subjects and incubated with normal or hyperapo B LDL. LDL (d 1.019- 1.063 g/ml) was isolated from pooled plasma obtained from six normal or six hyperapo B subjects. Normal HMD macrophages were incubated for 6 h with RPMI1640 containing LPDS (4 mg/ml) and 0.1 mM [‘4C]oleate/albumin complex, in the presence or absence of normal LDL or hyperapo B LDL (250

Table 5 Mean (S.D.) rates of [t4C]oleate incorporation TG in normal HMD macrophages LDL in medium (250 &ml)

TGa

(nmohmg cell protein per 6 h) Normal pool LDL Hyperapo B pool LDL

No differences in high-affinity uptake or degradation were observed between the normal and hyperapo B LDL preparations in either experiment; data are shown in Fig. 4 from the first experiment. The mean (S.D.) rate (ng 1251/mg cell protein per h) of uptake for the normal LDL

CEa

into CE and

MDA-LDL No lipoprotein

0.298 (0.037) 0.258 (0.022) 0.651 (0.037) .0.134 (0.027)

137 (7) II6 (5) 109 (3) 110 (12)

aP = N.S. between normal and hyperapo B LDL P= N.S.

SD. Kafonek et al. /Atherosclerosis

33

102 (1993) 23-36

Normal

HMD Macrophages

2.5

2.0 0.6

0 m m m

no LDL

normal hyperapoB MDA-LDL

LDL LDL

Fig. 5. Rate of [‘4C]oleate incorporation into CE in normal HMD macrophages Normal HMD macrophages were cultured for 5 days in RPMI-1640 and 5% AB serum. Cells were washed extensively with PBS and incubated in RPMI-1640 and 0.1 mM [‘4C]oleate, with and without normal or hyperapo B LDL (125 &ml), for 6 h. Duplicate flasks were assayed for each condition, The mean (SD.) incorporation of [‘4C]oleate into CE is presented.

pg lipoprotein protein/ml). MDA-LDL prepared from normal LDL was used as a positive control. There were no differences in the rates of CE or TG synthesis in cells incubated with either normal or hyperapo B LDL (Table 5); CE synthesis was stimulated 2-fold in the presence compared with the absence of LDL. Cells incubated with MDA-LDL stimulated CE synthesis Sfold, presumably due to unregulated uptake through the scavenger receptor [18]. No significant effect was observed on TG synthesis with any of the LDL preparations. To assess individual variability, CE synthesis in normal HMD macrophages was determined following incubation of the cells with normal or hyperapo B LDL isolated from individual donors. LDL was isolated from 5 hyperapo B and 4 normal donors. Very little cholesterol esterification occurred in the absence of LDL. LDL from normal and hyperapo B donors stimulated CE synthesis 6- to g-fold following a 6 h incubation (Fig. 5). MDA-LDL (positive control) stimulated CE

synthesis lo- to 15-fold. The incorporation of [‘4C]oleate into cellular CE was similar in cells exposed to either normal or hyperapo B LDL, but the cells incubated with hyperapo B LDL consistently demonstrated a trend toward less CE synthesis than those exposed to normal LDL. The LDL protein concentration to which the cells were exposed was the same (4.88 x lo-’ M) for the normal and hyperapo B LDL preparations. However, because of the lower cholesterol/protein ratios of hyperapo B LDL, the mean (SE) LDL cholesterol concentrations were 9.438 (0.48) x 10m4M for hyperapo B LDL compared with 9.754 (0.29) x 10m4 M for normal LDL. The rate of cellular CE synthesis was highly correlated with the LDL cholesterol concentration in the medium (n = 9, r = 0.704, P < 0.05). Thus, the rates of CE synthesis depended, in large part, on the amount of cholesterol available in the LDL particle, rather than the origin (normal versus hyperapo B subjects) of the LDL.

34

4. Discussion This study was designed to assess the effect of hyperapo B LDL on cellular cholesterol esterification in cultured human macrophages. Our major findings were (1) hyperapo B LDL did not promote abnormally high cholesteryl ester deposition through the scavenger pathway, as judged by results in THP-1 macrophages, (2) hyperapo B LDL was no more effective than normal LDL in promoting CE synthesis in THP-1 macrophages or normal HMD macrophages and, (3) the cellular association and degradation of hyperapo B LDL was similar to normal LDL in THP-1 macrophages. The LDL isolated from the donors with the hyperapo B phenotype was significantly different from that isolated from the normal controls. Hyperapo B LDL, compared with control LDL, had significantly lower total cholesterol to protein ratio, significantly higher average density as judged by EDGU, and significantly lower particle diameter, as judged by GGE. The human monocytic leukemia cell line, THP1, has emerged as a useful model for studying in vitro foam cell formation [ 1 l- 131. In the presence of phorbol esters, these cells differentiate into adherent macrophages and demonstrate reduced LDL receptor activity; scavenger receptor activity increases. In our studies, no differences in cellular CE or TG mass accumulation were found in THP1 macrophages incubated with hyperapo B compared with normal control LDL. This was observed regardless of the length of incubation or LDL protein concentration. The mass of free and esterified cholesterol increased in differentiated THP-I cells incubated with either normal or hyperapo B LDL, indicating some uptake of LDL particles through non-LDL receptor pathways, or perhaps through residual LDL receptors. The extent of mean cellular CE accumulation after incubation with MDA-LDL was three times as high (18.2 vs. 6.4) as that for hyperapo B LDL indicating that hyperapo B LDL was not equivalent to MDA-LDL in stimulating CE deposition through scavenger receptor activity. The sample size in our studies was sufficient to determine if hyperapo B LDL resembled MDA-LDL rather

S. D. Kafonek et al. /Atherosclerosis

IO.2 (1993) 23-36

than normal LDL in the ability to stimulate CE accumulation. A difference of greater than 2.2 pg CE/mg cell protein per 24 h between LDL preparations would be required to demonstrate a statistical difference with the sample size we studied and the inherent variability of the preparations and the cell system. The observed difference between the normal and hyperapo B LDL preparations did not approach 2.2 making it unlikely that the sample size was too small to detect statistically significant differences. Several studies performed in vivo [5] and in vitro using cultured cells [23] have suggested differences in the metabolism of LDL subfractions. Some studies suggest that small, dense LDL in humans are cleared more slowly than the more buoyant LDL fraction [5] although others do not [24,25]. Delayed clearance of dense LDL from the circulation could facilitate an abnormal interaction between the lipoprotein and the arterial cell wall rendering the particle atherogenic. Excessive cellular accumulation of CEs in macrophages can result in foam cell formation, the earliest hallmark of the atherosclerotic plaque. Early studies found no differences in the binding and degradation of heavy and light subfractions of LDL by cultured fibroblasts and HMD macrophages [ 10,261. However, more recent work indicates that small, dense LDL competes less effectively than normal LDL for the LDL receptor [9,23]. Similarly Hara and Howard [23] used a mutant line of Chinese hamster ovary cells which had been transfected with the full lengt h cDNA for the normal human LDL receptor to study small differences in the binding of light and dense LDL subfractions [23]. Among six LDL subfractions isolated by density gradient ultracentrifugation from normal human pooled plasma, the heaviest LDL bound to the cells with the lowest affinity. We demonstrated no differences in uptake and degradation of unmodified hyperapo B LDL compared with normal control LDL in THP-1 macrophages, cells with predominately non-LDL receptor activity. The susceptibility of hyperapo B LDL to modification and subsequent uptake by scavenger receptors has yet to be studied. We considered that the heterogeneity of LDL particles isolated in the d 1 .O 19- 1.063 g/ml density

S.D. Kafonek et al. /Atherosclerosis

102 (1993) 23-36

range might mask any differences in the ability of ‘light’ and ‘heavy’ LDL to stimulate cholesterol esteritication. LDL within the conventional d 1.019- 1.063 g/ml density range was subfractionated by EDGU to obtain the major LDL species. No differences were found in the extent of CE mass accumulation in cells incubated with the major LDL species of hyperapo B compared with normal control LDL. Further, no differences in esterification were found in THP-1 cells incubated with light- and intermediate-density LDL subfractions from normal and hyperapo B plasma, but the densest subfraction (heavy LDL) from the hyperapo B donor stimulated CE synthesis only one-half as much as the light- and intermediatedensity LDL subfractions (P < 0.005). This may suggest a differential affinity of LDL subfractions for the LDL receptor as suggested by Hara and Howard [23], or alternatively, LDL subfractions in the heaviest densities may provide less cholesterol per particle for intracellular esteritication. In our studies, no differences were found in the rates of degradation of intermediate and dense subfractions from hyperapo B donors incubated with THP-1 macrophages (data not shown). The findings in THP-1 macrophages were also observed for human monocyte-derived macrophages. The mean rates (nmol/mg cell protein per 6 h) of [‘4C]oleate incorporation into CE and TG with normal and hyperapo B LDL were comparable with the two cell systems. For normal LDL, rates were 0.238 and 0.298, respectively, for THP-1 and human macrophages; for hyperapo B LDL, rates were 0.211 and 0.258, respectively. There was a trend for lower esterification rates with hyperapo B compared with normal LDL for cellular CE and TG in both cell systems, but the changes did not reach statistical significance. Shi et al. [27] demonstrated increased LDL degradation in peripheral blood mononuclear leukocytes (MNL) isolated from subjects with CAD compared with subjects without CAD. In a later publication by the same authors [28], dense LDL from subjects with CAD demonstrated decreased ability to down-regulate LDL receptor activity in HL-60 promyelocytic leukemic cells compared with control LDL from subjects without CAD. They also demonstrated decreased CE synthesis as

35

measured by the incorporation of [ “C]oleate into cellular CE with dense compared with control LDL. The authors concluded that the cholesteroldepleted (dense) LDL had diminished ability to down-regulate LDL receptor activity because of lower cholesterol delivery to the cell. While care must be taken when extrapolating studies in cultured cells to metabolic processes in vivo, our findings suggest that the apparent increased atherogenicity of small, dense LDL does not result from an inherent factor in unmodified hyperapo B LDL leading to an abnormal accumulation of CE in macrophages. Accelerated atherogenesis in hyperapo B patients probably results from other factors, such as prolonged residence of an increased number of small, dense LDL particles in the plasma, susceptibility of these particles to potentially atherogenic processes such as oxidation [29], or perhaps to pathophysiologic effects of human serum basic proteins [30]. 5. Acknowledgments We wish to acknowledge Mary Ann Palmer and Debra S. Lally for assistance in preparing the manuscript. This work was supported by National Research Service Award No. 5-F32HL07541-02 and National Institutes of Health Grant No. 3 1497. Computational assistance was received from Dr. Alexander J. Seidler and the GCRC Computing Center sponsored by National institutes of Health Grant No. 00035. 6. References Sniderman, A.D.. Shapiro, S., Marpole. D.. Malcom. 1.. Skinner, B., Teng, B. and Kwiterovich. P.O.. The association of coronary atherosclerosis and hyperapobetalipoproteinemia (increased protein but normal cholesterol content in human plasma low density lipoprotein). Proc. Natl. Acad. Sci.. USA. 77 (1980) 604. Sniderman, A.D.. Wolfson. C.. Teng. B., Franklin. F., Bachorik, P.S. and Kwiterovich, P.O., Association of hyperapobetalipoproteinemia with endogenous hypertriglyceridemia and atherosclerosis. Ann. Int. Med.. 97 (1982) 833. Teng. B.. Thompson, G.R., Sniderman. A.D.. Forte. T.M.. Krauss. R.M. and Kwiterovich. P.O.. Cornpositron and distribution of low density lipoprotein fracttons tn normolipidemia. and hyperapobetalipoproteinemia.

SD. Kafonek et al. /Atherosclerosis

36

4

5

6

7

8

9

10

11

12

13

14

15

16

familial hypercholesterolemia, Proc. Natl. Acad. Sci., USA, 80 (1983) 6662. Brunzell, J.D., Sniderman, A.D., Albers, J.J. and Kwiterovich, P.O., Apoproteins B and A-I and coronary artery disease in humans, Arteriosclerosis, 4 (1984) 79. Teng, B., Sniderman, A.D., Soutar, A.R. and Thompson, G.R., Metabolic basis of hyperapobetalipoproteinemia, J. Clin. Invest., 77 (1986) 663. Ladias, J.A.A., Kwiterovich, P.O., Smith, H., Miller, M., Bachorik, P.S., Forte, T. and Lusis, A.J., Apolipoprotein B-100 Hopkins (Arginine@t, - Tryptophan). A new apolipoprotein B-100 variant in a family with premature atherosclerosis and hyperapobetalipoproteinemia, J. Am. Med. Assoc., 262 (1989) 1980. Kwiterovich, P.O., White, S., Forte, T., Bachorik, P.S., Smith, H. and Sniderman, A., Hyperapobetalipoproteinemia in a kindred with familial combined hyperlipidemia and familial hypercholesterolemia, Arteriosclerosis, 7 (1987) 21 I. Teng, B., Sniderman, A.D., Krauss, R.M., Kwiterovich, P.O., Milne, R.W. and Marcel, Y.L., Modulation of apolipoproteinB antigenic determinants in human low density lipoprotein subclasses, J. Biol. Chem., 260 (1985) 5067. Arad, Y., Rumsey, SC., Galeano, N.F., Al-Haideri, M., Walsh, M.T., Marcel, Y.L., Milne, R.W., Kwiterovich, P.O. and Deckelbaum, R.J., LDL has optimal size for maximal binding to the LDL receptor, Circulation, 84 (1992) 551. Knight, B.L., Thompson, G.R. and Soutar, A.K., Binding and degradation of heavy and light subfractions of low density lipoproteins by cultured Iibroblasts and macrophages, Atherosclerosis, 59 (1986) 301. Hara, H., Tanishita, H., Yokoyama, S., Tajima, S. and Yamamoto, A., Induction of acetylated low density lipoprotein receptor and suppression of low density lipoprotein receptor on the cells of human monocytic leukemia cell line (THP-I cell), Biochem. Biophys. Res. Commun., 146 (1987) 802. Banka, CL., Black, AS., Dyer, C. and Curtiss, L.K., THP-1 cells form foam cells in response to coculture with lipoproteins but not platelets, J. Lipid Res., 32 (1991) 35. Via, D.P., Pons, L., Dennison, D.K., Fanslow, A.E. and Bernini, F., Induction of acetyl-LDL receptor activity by phorbol ester in human monocyte cell line THP-I, J. Lipid Res., 30 (1989) 1515. Lipid Research Clinic, NIH Publication No. 80-1527. US Department of Health and Human Services, Washington DC, 1980. Have], R.J., Eder, H.A. and Bragdon, J.H., The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum, J. Clin. Invest., 34 (1955) 1345. Krauss, R.M. and Burke, D.J., Identification of multiple subclasses of plasma low density lipoproteins in normal humans, J. Lipid Res., 23 (1982) 97.

I7

18

19

20

21 22

23

24

25

26

27

28

29

30

102 (1993) 23-36

Bautovich, G.J., Dash, M.J., Hensley, W.J. and Turtle, J.R., Gradient gel electrophoresis of human plasma lipoproteins, Clin. Chem., I9 (1973) 415. Fogelman, A.M., Sheeter, I., Seager, J., Hokom, M., Child, J.S. and Edwards, P.A., Malondialdehyde alteration of low density lipoproteins leads to cholesteryl ester accumulation in human monocyte-macrophages, Proc. Natl. Acad. Sci., USA, 77 (1980) 2214. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J., Protein measurement with the folin phenol reagent, J. Biol. Chem., 193 (1951) 265. Ishikawa, T.T., MacGee, J., Morrison, J.A. and Glueck, C.J., Quantitative analysis of cholesterol in 5 to 20 pl of plasma, Lipid Res., 15 (1974) 286. McFarlane, A.S., Efficient trace-labelling of proteins with iodine, Nature (London), 182 (1958) 53. Bilheimer, D.W., Goldstein, J.L., Grundy, SM. and Brown, M.S., Reduction in cholesterol and low-density lipoprotein synthesis after portacaval shunt surgery in a patient with homozygous familial hypercholesterolemia, J. Clin. Invest., 56 (1975) 1420. Hara, H. and Howard, B.V., Characterization of LDL binding in TR 715-19 cells, Atherosclerosis, 83 (1990) 155. Kesaniemi, Y.A. and Grundy, SM., Overproduction of low density lipoproteins associated with coronary heart disease, Arteriosclerosis, 3 (1983) 40. Vega, G.I., Beltz, W.F. and Grundy, S.M., Low density lipoprotein metabolism in hypertriglyceridemic and normolipidemic patients with coronary heart disease, J. Lipid Res., 26 (1985) 115. Swinkels, D.W., Demacker, P.N.M., Hak-Lemmers, H.L.M., Mol, M.J.T.M., Yap, S.H. and Vant Laar, A., Some metabolic characteristics of low-density lipoprotein subfractions, LDL-1 and LDL-2: in vitro and in vivo studies, Biochim. Biophys. Acta, 960 (1988) 1. Shi, F., Hurst, P.G. and McNamara, D.J., Increased degradation of low density lipoproteins by mononuclear leukocytes associated with coronary artery disease, Atherosclerosis, 85 (1990) 127. Shi, F., Jouni, Z.E. and McNamara, D.J., Reduced regulatory capacity of low density lipoproteins from patients with coronary artery disease, Atherosclerosis, 91 (1991) 217. de Graaf, J., Hak-Lemmers, H.L.M., Hectors, M.P.C., Demacker, P.N.M., Hendriks, J.C.M. and Stalenhoef, A.F.H., Enhanced susceptibility to in vitro oxidation of the dense low density lipoprotein subfraction in healthy subjects, Arterioscler. Thromb., 11 (1991) 298. Kwiterovich, P.O., Motevalli, M., Miller, M., Bachorik, P.S., Kafonek, SD., Chatterjee, S., Beaty, T. and Virgil, D., Further insights into the pathophysiology of hyperapobetalipoproteinemia, Clin. Chem., 37 (1991) 317.