Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html
patient-oriented research

Downloaded from www.jlr.org by guest, on December 19, 2016

Rapid turnover of apolipoprotein C-III-containing triglyceride-rich lipoproteins contributing to the formation of LDL subfractions

Chunyu Zheng,* Christina Khoo,1,* Katsunori Ikewaki,† and Frank M. Sacks2,*
Department of Nutrition,* Harvard School of Public Health, Boston, MA 02115; and Division of Cardiology,† Department of Internal Medicine, Jikei University School of Medicine, Tokyo, Japan 105-8461

Abstract The atherogenicity theory for triglyceride-rich lipoproteins (TRLs; VLDL 1 intermediate density lipoprotein) generally cites the action of apolipoprotein C-III (apoC-III), a component of some TRLs, to retard their metabolism in plasma. We studied the kinetics of multiple TRL and LDL subfractions according to the content of apoC-III and apoE in 11 hypertriglyceridemic and normolipidemic persons. The liver secretes mainly two types of apoB lipoproteins: TRL with apoC-III and LDL without apoC-III. Approximately 45% of TRLs with apoC-III are secreted together with apoE. Contrary to expectation, TRLs with apoC-III but not apoE have fast catabolism, losing some or all of their apoC-III and becoming LDL. In contrast, apoE directs TRL flux toward rapid clearance, limiting LDL formation. Direct clearance of TRL with apoC-III is suppressed among particles also containing apoE. TRLs without apoCIII or apoE are a minor, slow-metabolizing precursor of LDL with little direct removal. Increased VLDL apoC-III levels are correlated with increased VLDL production rather than with slow particle turnover. Finally, hypertriglyceridemic subjects have significantly greater production of apoC-IIIcontaining VLDL and global prolongation in residence time of all particle types. ApoE may be the key determinant of the metabolic fate of atherogenic apoC-III-containing TRLs in plasma, channeling them toward removal from the circulation and reducing the formation of LDLs, both those with apoC-III and the main type without apoC-III.—Zheng, C., C. Khoo, K. Ikewaki, and F. M. Sacks. Rapid turnover of apolipoprotein C-III-containing triglyceride-rich lipoproteins contributing to the formation of LDL subfractions. J. Lipid Res. 2007. 48: 1190–1203.
Supplementary key words kinetics & stable isotopes & apolipoprotein B-100 & apolipoprotein E & low density lipoprotein
Epidemiological studies demonstrate that apolipoprotein C-III (apoC-III) and the apoB lipoproteins that have apoC-III as a component independently predict coro-

nary heart disease (1–3). ApoC-III is present on ?40–80% of triglyceride-rich lipoproteins (TRLs) and ?5–10% of LDLs in plasma (4–6). The mechanisms by which apoC-III causes hypertriglyceridemia and atherosclerosis are incompletely understood.
Experiments in vitro show that apoC-III can inhibit lipoprotein lipase (7, 8) and hepatic lipase (9) and retard the clearance of VLDL by interfering with the binding of apoB-100 (10) or apoE to hepatic receptors (11, 12). Direct evidence supporting a role of high apoC-III level in abnormal TRL metabolism has come from transgenic animal studies. Overexpression of apoC-III in mice causes hypertriglyceridemia (13–17), whereas apoC-III deficiency protects against it (18, 19). In these studies, impaired particle clearance via LDL receptors (14–16), reduced binding affinity to cell surface proteoglycans (15, 17), inhibition of lipolysis (17, 19), and overproduction of VLDL triglyceride (14, 16) have all been implicated as mechanisms for the hypertriglyceridemic effect of apoC-III. In humans, there is also evidence for apoC-III affecting TRL metabolism. Patients with combined deficiency of apoC-III and apoA-I experience rapid VLDL clearance (20). In a kinetic study, the plasma concentrations and secretion rates of VLDL apoC-III were correlated with those of VLDL triglyceride (21). In another kinetic study in a similar group of humans, VLDL apoC-III concentrations and secretion rates were correlated with VLDL triglyceride and VLDL apoB residence times but not with secretion rates (22). Thus, although in vitro experiments, animal models, and human kinetic studies all provide evidence that apoC-III adversely affects the metabolism of TRL, the studies do not form a consensus about which mechanisms are dominant.
The observation that high levels of apoC-III cause hypertriglyceridemia by retarding VLDL catabolism has

Manuscript received 22 September 2006 and in revised form 3 January 2007 and in re-revised form 31 January 2007.
Published, JLR Papers in Press, February 21, 2007. DOI 10.1194/jlr.P600011-JLR200

1 Present address of C. Khoo: Hill’s Pet Nutrition, Topeka, KS 66617. 2 To whom correspondence should be addressed.
e-mail: fsacks@hsph.harvard.edu The online version of this article (available at http://www.jlr.org) contains supplementary data.

1190 Journal of Lipid Research Volume 48, 2007

Copyright D 2007 by the American Society for Biochemistry and Molecular Biology, Inc. This article is available online at http://www.jlr.org

Downloaded from www.jlr.org by guest, on December 19, 2016

Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html

led to a kinetic theory for the atherogenicity of TRL: inhi-

TABLE 1. Baseline characteristics of the participants

bition by apoC-III of TRL catabolism increases the residence times for TRLs and their atherogenic remnants. Characteristic

Normolipidemic Subjects (n 5 6)

Hypertriglyceridemic Subjects (n 5 5)

This view follows the long-established kinetic theory for LDL atherogenicity: reduced LDL receptor activity causes long residence times, high concentrations, and prolonged exposure of vascular wall cells to circulating LDLs. However, there is no direct evidence that apoC-III-containing TRLs, the particles that strongly predict coronary heart disease, have long residence times in plasma.
In contrast to the actions of apoC-III, apoE assists in the clearance of apoB lipoproteins by binding to cell surface receptors (23, 24) and proteoglycans (25). Overexpression of apoE in mice corrects the hypertriglyceridemia produced by the overexpression of apoC-III (16). Humans who are deficient in apoE, or who are homozygous for a defective isoform of apoE, have retarded metabolism of TRL and abnormal TRL composition (26, 27). Thus, apoC-III and apoE may be viewed functionally as antagonists in apoB lipoprotein metabolism.
The objective of the current study is to understand in as much depth as feasible apoB lipoprotein metabolism, focusing on the types of particles that are likely candidates

Age (years)

58 6 11

Gender (female/male) Body mass index (kg/m2)

3/3 28 6 6

Plasma lipids and apolipoproteins (mg/dl)

Total apoB

76 6 20

Total triglyceride

113 6 53

Total cholesterol

168 6 24

VLDL cholesterol

863

Intermediate density

10 6 4

lipoprotein cholesterol

LDL cholesterol

94 6 17

LDL1 LDL2 LDL3 HDL cholesterol

24 6 4 32 6 10 38 6 12 57 6 18

VLDL apoB

3.0 6 0.7

VLDL apoE

1.3 6 0.7

VLDL apoC-III

2.1 6 0.8

52 6 4 2/3
28 6 4
84 6 19 262 6 54a 175 6 33
20 6 5b 18 6 5b
104 6 21 18 6 3 26 6 6 60 6 14b 33 6 5b 5.8 6 2.0b 2.2 6 0.8 7.9 6 4b

apoB, apolipoprotein B. Values are means 6 SD. Concentrations
were measured in the samples after an overnight fast. LDL1, density of 1.025–1.032 g/ml; LDL2, density of 1.032–1.038 g/ml; LDL3, density of 1.038–1.050 g/ml.
a P , 0.01 between the two groups. b P , 0.05 between the two groups.

for kinetic distinctness and that are likely relevant to

remnant lipoprotein metabolism and atherogenic dyslipidemia. A central aim is to learn about apoC-III-containing apoB lipoproteins. Because most actions on lipoprotein

mittees at the Harvard School of Public Health and Brigham and Women’s Hospital. All participants gave informed consent.

metabolism of apoC-III and apoE are contrary, if not directly antagonistic, the concomitant presence of apoC- Controlled dietary intake

III and apoE on the same TRL particle presents an ob-

There was a 3-week controlled dietary period before the kinetic

stacle to the understanding of the normal role of either apolipoprotein, as we found in a previous kinetic study (28). Thus, we consider it essential to distinguish the effects of apoC-III and apoE on the metabolism of apoB lipoproteins, and we prepared from plasma distinct apoB lipoprotein types with either apoC-III or apoE, both of these proteins, or neither of them. This approach gains

study. The entire diet was provided to the participants as outpatients, and they were asked not to consume alcoholic beverages or any other source of caloric intake. Dietary energy levels were adjusted to reports of hunger or satiety and to trends in body weight that were measured every other day. The diet was prepared at the metabolic kitchen of the General Clinical Research Center at Brigham and Women’s Hospital (Boston, MA). The diet consists of 37% energy from fat (8% from saturated fat, 24%

clinical relevance from two lines of evidence. First, the from monounsaturated fat, and 5% from polyunsaturated fat),

apoB concentration of TRLs that have apoC-III is an ex- 48% from carbohydrate, and 15% from protein.

ceptionally strong risk factor for cardiovascular disease,

eclipsing the effect of plasma triglyceride (3). Second, we Tracer infusion

recently demonstrated that apoC-III by itself and as a

Participants were admitted to the General Clinical Research

component of TRL and LDL is directly involved in athero- Center of Brigham and Women’s Hospital in the evening before

genesis by activating adhesion molecules in monocytes and endothelial cells and by activating proinflammatory nuclear factor-kB (29–31). For these reasons, we aim for better understanding of the mechanisms that produce or sustain high concentrations of apoB lipoproteins with apoC-III.

study. After an overnight fast, they received a priming dose of 4.2 mmol/kg [D3]L-leucine (Tracer Technologies, Cambridge, MA), followed by a constant infusion of [D3]L-leucine at 4.8 mmol/kg/h for 15 h. A bolus injection of [D5]L-phenylalanine (1.2 mg/kg) was also administered at the same time. Blood
samples were collected at baseline, every 20 min in the first 2 h

after the infusion, and hourly thereafter. For the first 4 h, partici-

pants were restricted to noncaloric drinks. After this, they were

METHODS

given a standardized lunch and supper that had no fat, leucine or phenylalanine and that contained 60% of the calories required

Subjects Five hypertriglyceridemic and six normolipidemic participants

for maintenance, to avoid discomfort and the abnormal metabolic effects of a prolonged fast.

were recruited into the study. All were nondiabetic and had normal serum chemistry profiles (Table 1). Exclusion criteria in- Sequential immunoaffinity chromatography

cluded secondary hyperlipidemia, (APO)E2/E2, E4/E4, or E2/

Blood was collected into EDTA coated vacuum tubes and pro-

E4 genotype, and the use of medications that affect lipid me- tease inhibitors were added. Plasma was separated immediately

tabolism. The study was approved by the Human Subjects Com- after sampling by centrifugation at 2,500 rpm for 20 min, divided

Kinetics of apoB lipoproteins according to apoC-III and apoE 1191

Downloaded from www.jlr.org by guest, on December 19, 2016

Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html

into aliquots in polypropylene vials, immediately sealed under tive chemical ionization and selective ion monitoring. ApoB mass

nitrogen gas, and kept frozen at 280jC until analysis. Separation was measured by comparing the ratio between the area under the

of lipoproteins by apoE and apoC-III content was carried out with leucine curve and the area under the norleucine curve with a

affinity-purified polyclonal antibodies against apoE and apoC-III standard curve of various leucine/norleucine ratios (34, 35).

(Academy Biomedical, Houston, TX) coupled to Sepharose 4B Plasma total apoB concentration was measured by ELISA and

resin, as described previously (5, 6). Plasma was first incubated applied equally among the individual lipoprotein fractions to

with anti-apoE immunoaffinity resin, and the unbound fraction correct their apoB masses for loss. Plasma volume (liters) was

(E2) was collected by gravity flow. The bound fraction (E1) was assumed to be 4.4% of body weight (kg).

eluted by incubation with 3 M NaSCN and was immediately

desalted and dialyzed. Both bound and unbound fractions from the anti-apoE column were then incubated with anti-apoC-III Model development and kinetic analysis

immunoaffinity resin. This sequential column procedure sepa-

Tracer enrichment and apoB were measured in 24 apoB lipo-

rated plasma into four immunofractions: those with both apoE and apoC-III (E1CIII1), those with apoE but not apoC-III (E1CIII2), those with apoC-III but not apoE (E2CIII1), and those without apoE or apoC-III (E2CIII2). The efficiency of the

protein fractions (four immunofractions further separated into six density fractions) for each sample (Fig. 1). A multiplecompartment model was used to find the best fit to the observed data using SAAM II software (SAAM Institute, Seattle, WA).

procedure, defined as the percentage of ligand (apoE or apoC- Figure 2 shows the diagram of the model. A plasma amino acid-

III) that is removed from plasma by the resin, is 98% for apoC-III and 97% for apoE. In addition, reincubation of eluted E1 and CIII1 fractions on the respective resin resulted in complete re-

forcing function is followed by a hepatic intracellular delay compartment, accounting for the time required for the synthesis and secretion of apoB-100 into plasma. The model is developed by

binding and no recovery of cholesterol or apoB in the unbound designating every physically separated lipoprotein fraction to a

fractions, suggesting that any possible dissociation of apoE or single compartment. Enrichment curves for light VLDL show a

apoC-III from apoB lipoproteins occurring during NaSCN elution, desalting, or concentrating procedures did not convert E1 or CIII1 particles to E2 or CIII2. We also evaluated the effect of

slow-turnover tail and thus require an additional slow-turnover compartment. Compartments 11–17 correspond to apoB lipoproteins without apoE and apoC-III (E2CIII2). Among them,

freezing and storage. Column efficiency for the anti-apoE and compartments 11 and 12 are fast- and slow-turnover components

anti-apoC-III columns was 98% and 97% for fresh samples and of light VLDL, compartment 13 is for dense VLDL, compartment

99% and 98% for frozen samples. Moreover, the percentages and plasma concentrations of apoB lipoproteins in the four immunofractions were similarly unaffected. Therefore, freezing

14 is for IDL, and compartments 15–17 are for LDL1, LDL2, and LDL3, respectively. Similarly, compartments 21–27 correspond to E2CIII1 apoB lipoproteins, compartments 31–34 cor-

and storage of plasma under these conditions do not affect the respond to E1CIII2, and compartments 41–47 correspond to

subsequent immunoaffinity columns.
Ultracentrifugation
The four immunofractions described above were then ultracentrifuged separately at 25,000 rpm on a Ti 25 rotor in an L870M instrument (Beckman, Brea, CA) to isolate light VLDL (Svedberg units of flotation, 60–400), dense VLDL (Svedberg units, 20–60), intermediate density lipoprotein (IDL; 1.006– 1.025 g/ml), LDL1 (1.025–1.032), LDL2 (1.032–1.038), and LDL3 (1.038–1.050) using a modification of the methods of Lindgren, Jensen, and Hatch (32). A density of 1.050 g/ml was selected as the cutoff point for LDL to avoid contamination by lipoprotein [a], which is concentrated at densities between 1.050 and 1.080 g/ml (33).

E1CIII1. ApoB masses in three fractions (E1CIII2 LDL1, LDL2, and LDL3) are too low to measure; thus, these three fractions are not included.
The model allows direct secretion and direct removal of apoB
into and out of every lipoprotein fraction. The model allows
stepwise delipidation among apoB lipoproteins of the same apoE and apoC-III composition (e.g., E1CIII1 VLDL Y E1CIII1 IDL Y E1CIII1 LDL). Lipolysis pathways for IDL directly to LDL2 and LDL3, bypassing LDL1, are also tested, and found to be required (i.e., having a nonzero parameter value in the best-fit model) only for the E2CIII2 fraction. Another important fea-
ture of our model is selective conversion pathways among apoB
lipoproteins of different apoE and apoC-III composition (e.g., E1CIII1 light VLDL Y E1CIII2 dense VLDL). When plasma apoB lipoproteins undergo lipolysis, their apoE and apoC-III

Determination of lipids and apolipoproteins
Triglyceride and cholesterol concentrations were determined enzymatically on a Cobas MIRA Plus Autoanalyser (Roche, Nutley, NJ). ApoE, apoC-III, apoB, apoC-I, and apoC-II concentrations were determined by sandwich ELISA procedures using affinity-purified antibodies (Academy Biomedical). Intra-assay coefficients of variation for lipid and apolipoprotein measurements were between 2% and 6%, and interassay coefficients of variation were between 4% and 8%.

content becomes lower (4), indicating a loss of apoE and apoCIII. When a TRL particle completely loses its apoC-III or apoE content, its phenotype, based on the presence of apoE and apoCIII, changes accordingly. On the other hand, when apoB lipoproteins free of apoC-III or apoE interact with HDL or another apoB lipoprotein, they may gain apoC-III or apoE and also change their phenotype. When evaluating potential conversion pathways between fractions with different apoE and apoC-III composition, we first compare the enrichment curves of the originating and destination compartments to eliminate pathways

Measurement of tracer enrichment and pool size

that strongly prohibit a precursor-product relationship. Next, pathways that are compatible with the tracer enrichment curves

After ultracentrifugation, apoB was precipitated from the lipo- are added to the existing model one by one and are fitted to the

proteins with isopropanol, and the precipitate was free of apoli- data. Pathways for which rate constants are zero or negligible are

poproteins other than apoB, as determined by high-sensitivity eliminated. Average particle size and triglyceride content are

SDS-PAGE with silver staining. Norleucine internal standard was also taken into consideration when selecting a pathway. After

added, and the mixture was converted to volatile heptafluoro- the model structure was established by the mean tracer and

butyric acid derivatives as described previously (28). Tracer tracee data of all 11 subjects, each participant’s data were fitted

enrichment was measured on a 5890 gas chromatograph/5988A individually to obtain the parameter values. Tracer and tracee

mass spectrometer (Hewlett-Packard, Palo Alto, CA) using nega- data for both [D5]L-phenylalanine bolus and [D3]L-leucine primed,

1192 Journal of Lipid Research Volume 48, 2007

Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html

Downloaded from www.jlr.org by guest, on December 19, 2016

Fig. 1. Tracer-tracee ratios of D3-leucine (A, B) and D5-phenylalanine (C, D) in VLDL, intermediate density lipoprotein (IDL), and LDL subfractions in normolipidemic (NTG; n 5 6) and hypertriglyceridemic (HTG; n 5 5) participants. E1CIII1, apolipoprotein B (apoB) lipoproteins with both apoE and apoC-III; E2CIII1, particles with apoC-III but not apoE; E1CIII2, particles with apoE but not apoC-III (concentrations of E1CIII2 LDL were too low to measure, and they were not modeled); E2CIII2, particles without apoE or apoC-III. Data points represent average leucine and phenylalanine tracer-tracee ratios in each group. Lines represent model-derived curves fitted to the data. Phenylalanine data are presented on a logarithmic scale. See Methods for details. ApoB mass and model predicted pool size data are available in the supplementary data.

continuous infusions were combined in a parallel multiplecompartment model and were solved simultaneously by making the rate constants equal for leucine and phenylalanine experiments. Thus, for every study participant, a single set of rate constants was produced. This model is able to generate excellent fits to tracer and tracee data for both leucine and phenylalanine. Fitting of apoB mass data is also excellent, and the results are shown in the supplementary data. The coefficients of variation

for most parameter estimates were ,30% and those for the major pathways were ,15%.
Statistical analysis Data were analyzed using the SAS software (SAS Institute, Cary,
NC) and are presented as means 6 SD. An unpaired t-test was used for between-group comparisons and a paired t-test was

Kinetics of apoB lipoproteins according to apoC-III and apoE 1193

Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html

Downloaded from www.jlr.org by guest, on December 19, 2016

Fig. 2. Multicompartmental model. Plasma tracer-forcing function supplies a delay compartment responsible for the production and secretion of apoB-100. Compartments 11–47 represent plasma pools of apoB-100. VA, light VLDL; VB, dense VLDL; LDL1, density of 1.025–1.032 g/ml; LDL2, density of 1.032–1.038 g/ml; LDL3, density of 1.038–1.050 g/ml. See Methods for details.

used for within-person comparisons, unless specified otherwise. P < 0.05 was considered statistically significant. Spearman correlation analysis and regression analysis were also conducted to study the associations between apoE, apoC-III, hypertriglyceridemia, and VLDL metabolism.
RESULTS
Demographic information and baseline measurements of lipoproteins for the participants are summarized in Table 1. Hypertriglyceridemic and normolipidemic participants have similar age, gender, and body mass index. The majority of them are overweight but not obese. Concentrations of plasma total apoB, total cholesterol, and LDL cholesterol are similar in the two groups. The hypertriglyceridemic participants have significantly higher concentrations of plasma total triglyceride, VLDL apoB, VLDL cholesterol, and VLDL apoC-III, higher LDL3 cholesterol, and lower HDL cholesterol.
ApoC-III is secreted on ?90% of VLDL and 15% of LDL Figure 3 shows secretion rates of apoB-100 for each parti-
cle type. (The numerical data and data for LDL subfractions are available online.) The vast majority of VLDL particles are secreted as apoC-III-containing (90% in both hypertriglyceridemic and normolipidemic subjects). VLDL without apoE or apoC-III constitutes ,10% of total VLDL secretion. In IDL, the percentage secreted as E2CIII2 increases substantially to 36% in normolipidemic subjects and 43% in hypertriglyceridemic subjects, although the majority are still secreted as apoC-III-containing. The secretion pattern is reversed in LDL, most secreted without apoC-III.

ApoE is secreted with apoC-III in TRL
Approximately 40% of TRLs that are secreted with apoC-III also contain apoE (Fig. 3). In contrast, nearly all of the apoB lipoproteins secreted with apoE have apoC-III. Less than 2% of TRLs are secreted with apoE but not apoC-III in both groups. Similar to apoC-III, secretion of apoE-containing apoB lipoproteins (E1CIII2 and E1CIII1) is concentrated within TRLs (41% of total VLDL secretion, 24–28% of IDL). Less than 8% of LDL is secreted with apoE.
ApoC-III-containing TRLs and LDLs have fast turnover rates
ApoC-III-containing apoB lipoproteins have faster rates of appearance of tracer than their counterparts without apoC-III (Fig. 1). With the injection of the bolus phenylalanine tracer, the peak tracer-tracee ratio is higher and the rate of disappearance is faster for apoC-III-containing lipoproteins. These patterns are true for all 11 participants. Accordingly, apoC-III-containing VLDL, IDL, and LDL have fast fractional catabolic rates (FCRs): 40–150% higher than those without apoC-III (Fig. 4). FCRs of E1CIII2, E2CIII1, and E1CIII1 TRL and LDL, although all significantly higher than those of E2CIII2, are generally not statistically different from each other. However, as shown below, the presence of apoE and apoC-III each appears to affect flux through specific catabolic pathways.
TRLs without apoC-III or apoE are almost exclusively lipolyzed into LDL
In both hypertriglyceridemic and normolipidemic subjects, a substantial proportion of E2CIII2 VLDL and IDL

1194 Journal of Lipid Research Volume 48, 2007

Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html

Downloaded from www.jlr.org by guest, on December 19, 2016

Fig. 3. Liver secretion rates of apoB lipoproteins. Data represent model-predicted secretion rates (mg/ day/kg body weight; means 6 SD) of light VLDL, dense VLDL, IDL, and LDL according to their apoC-III and apoE composition. Average total apoB secretion rates in normotriglyceridemic and hypertriglyceridemic participants are 11 and 12 mg/day/kg, respectively (P 5 0.77). Secretion of E1CIII2 constitutes ,2% of total apoB lipoproteins in both groups and thus is not plotted. * P , 0.05 between hypertriglyceridemic and normolipidemic groups for corresponding fractions; † P , 0.05 in total secretion rates of light VLDL, IDL, or LDL between hypertriglyceridemic and normolipidemic groups. See supplementary data for exact
values and LDL subfractions.

is formed from apoC-III-containing TRLs by loss of apoC-III from E2CIII1 associated with the lipolytic conversion of larger to smaller particles. VLDLs that do not contain apoE or apoC-III are poor candidates for direct removal and are almost exclusively lipolyzed to IDL. Subsequently, .90% of E2CIII2 IDL is hydrolyzed to LDL, about half going to LDL1 and the rest going directly to LDL2 and LDL3.
TRLs with apoC-III but not apoE are the major contributors to plasma LDL formation
Approximately 70% of E2CIII1 TRLs are hydrolyzed to LDL, either E2CIII1 LDL (20%) or E2CIII2 LDL (50%); the remaining 30% are removed from the circulation as E2CIII1 TRLs (Fig. 5). Thus, lipolytic conversion from E2CIII1 TRLs to E2CIII2 LDLs is a major pathway responsible for LDL formation in plasma. Indeed, E2CIII1 TRLs are precursors for more than half (55%) of total plasma E2CIII2 LDLs among hypertriglyceridemic subjects, generating twice as much LDL as direct liver secretion (27%) (Fig. 6; detailed data are available in the supplementary data). In addition, the majority of E2CIII1 LDLs are also eventually converted to E2CIII2 LDLs. In all, E2CIII1 TRLs and LDLs provide 45–61% of the total production of E2CIII2 LDLs.

TRLs with apoE but not apoC-III are rapidly removed from the circulation
E1CIII2 TRLs constitute ,2% of liver secretion of apoB lipoproteins. Instead, the majority of these particles are products of E1CIII1 TRLs after the loss of apoC-III during lipolysis (Fig. 5). E1CIII2 VLDLs are ?10 times more likely to undergo direct removal than conversion to IDLs. The rate constants for removal are higher in E1CIII2 than for any other type of TRL. We could not detect any measurable amount of LDL particles with apoE but not apoC-III, presumably because of the fast removal of E1CIII2 IDL.
Metabolism of TRLs with both apoE and apoC-III is a function of both apolipoproteins
Ninety percent of E1CIII1 TRLs are cleared from the circulation before reaching LDL size (Fig. 5). Most are removed as E1CIII1 VLDL or IDL, and some are converted to E1CIII2 before leaving the circulation. Only 10% of E1CIII1 TRLs are converted into LDLs. The contribution of E1CIII1 TRLs to LDL formation is very low, producing only 5–8% of total plasma LDL compared with 45–61% for E2CIII1 TRLs (P , 0.01) (Fig. 6). This disparity between these two TRL particle types as precursors for LDL contrasts with their similar secretion rates

Kinetics of apoB lipoproteins according to apoC-III and apoE 1195

Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html

Downloaded from www.jlr.org by guest, on December 19, 2016

Fig. 4. Fractional catabolic rates (FCRs) of apoB lipoproteins according to contents of apoE and apoC-III. * P , 0.05 between hypertriglyceridemic subjects (HTG; n 5 5) and normolipidemic controls (NTG; n 5 6); † P , 0.05 compared with the corresponding
E2CIII2 fraction in each density fraction. Data are means 6 SD.

(i.e., 3.0 mg/kg/day for E1CIII1 vs. 3.9 mg/kg/day for E2CIII1 among normolipidemic subjects; P 5 0.48) (Fig. 3). Rate constants for direct removal for E1CIII1 TRLs are intermediate between E2CIII1 and E1CIII2 (i.e., faster than E2CIII1 but slower than E1CIII2) (Fig. 5). These findings strongly implicate apoE in redirecting TRL catabolism toward the direct removal as well as apoC-III in inhibiting this pathway.
LDL subfraction kinetics
LDL without apoE or apoC-III constitutes .90% of LDL in the circulation (Fig. 5). They are either products of TRL lipolysis in plasma or are directly secreted from the liver. Figure 6 summarizes the contribution of each pathway to LDL production. Overall, significantly more LDLs come from TRL lipolysis than from direct liver secretion (?3:1 in hypertriglyceridemic subjects and 2:1 in normolipidemic subjects; both P , 0.01). Among E2CIII2 LDLs formed by lipolysis, E2CIII1 TRLs contribute by far the most (75% and 63% in hypertriglyceridemic and normolipidemic subjects, respectively), followed by E2CIII2 TRLs (19% and 25%) and E1CIII1 TRLs (5% and 12%).
ApoC-III-containing LDL subfractions demonstrate different metabolic patterns than LDL that does not have apoC-III. First, both E2CIII1 and E1CIII1 LDL subfractions have higher FCRs than their E2CIII2 counterparts (Fig. 4). Second, .80% of E2CIII1 LDL1, LDL2, and LDL3 lose apoC-III during circulation in plasma to form

E2CIII2 LDL3, similar to the metabolic fate of E2CIII1 TRLs (Fig. 5). Finally, among both hypertriglyceridemic and control subjects, .50% of E1CIII1 LDL1 and LDL2 are removed from the circulation before being lipolyzed into E2CIII1 LDL2 or LDL3, presumably as a result of the positive effect of apoE on LDL receptor-mediated clearance. In contrast, only 30–40% of E2CIII2 LDL1 and LDL2 are removed from the circulation before reaching LDL3.
VLDL apoC-III and apoE levels are determined mainly by increased VLDL secretion, not by retarded catabolism
Concentrations of VLDL apoC-III and apoE are strongly and positively associated with VLDL triglyceride (r 5 0.98, P , 0.001 for apoC-III and r 5 0.88, P 5 0.002 for apoE). Concentrations of apoC-III and apoE found in VLDL are strongly associated with the secretion rates of VLDL apoB but not with the FCR (Fig. 7, far left and middle left panels). Among VLDL particles that contain apoC-III with or without apoE, the number of apoC-III molecules per particle is not associated with altered FCRs (Fig. 7, upper middle right panel) or rate constants for lipolysis or clearance (data not shown). Among VLDLs that contain both apoC-III and apoE, the molar ratio of apoE to apoB is significantly correlated with their FCRs (r 5 0.70, P 5 0.03) (Fig. 7, lower middle right panel). Overall, apoE/apoB ratios in apoE-containing VLDLs are positively associated with rate constants for direct removal (r 5 0.68, P 5 0.03 in E1CIII1 VLDL).

1196 Journal of Lipid Research Volume 48, 2007

Downloaded from www.jlr.org by guest, on December 19, 2016

Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html
Fig. 5. Flux of apoB-100 in hypertriglyceridemic (A) and normolipidemic (B) participants. Oval boxes represent apoB lipoprotein fractions separated by apoC-III and apoE content and by density; numbers inside indicate pool sizes (mg). Arrows out of the liver represent direct liver secretion; percentages next to the arrows indicate the percentage of total liver secretion into each fraction. Arrows out of lipoprotein compartments represent conversion and direct removal; percentages above the arrows indicate the relative proportion of flux out of each compartment, and numbers in parentheses indicate the rate constant for each pathway (pools/day). VA, light VLDL; VB, dense VLDL. Data are means 6 SD. † The parameter is significantly higher in hypertriglyceridemic than in normolipidemic subjects; * the parameter is significantly lower in hypertriglyceridemic subjects (P , 0.05).
Kinetics of apoB lipoproteins according to apoC-III and apoE 1197

Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html

Downloaded from www.jlr.org by guest, on December 19, 2016

Fig. 6. Contribution of lipolysis pathways and direct liver secretion to LDL formation. Data represent
production rates (mg/day/kg; means 6 SD) of different types of LDL particles from direct liver secretion and various lipolytic conversion pathways. * P , 0.05 in corresponding fractions between hypertriglyceridemic and normolipidemic groups; † P , 0.05 in total E2CIII1 LDL production rates between hyper-
triglyceridemic and normolipidemic groups. See supplementary Table II for exact numerical data. TRL,
triglyceride-rich lipoprotein.

ApoC-III-containing TRLs are enriched in apoC-II and apoC-I
Consistent with our earlier observations (5), there is a trend toward a higher content of triglyceride and cholesterol per particle among apoC-III-containing TRLs and

LDLs compared with those without apoC-III (Fig. 8). However, the most striking finding is that apoB lipoproteins with apoC-III contain significantly more apoC-II and apoC-I molecules per particle than those without apoC-III. Within every density fraction, E1CIII1 apoB lipoproteins

Fig. 7. Relationship between VLDL apoC-III, VLDL apoE, HDL cholesterol, and kinetic parameters. Linear regression lines are shown for each relationship in all 11 subjects [open diamonds, hypertriglyceridemic subjects (HTG); closed diamonds, normolipidemic subjects (NTG)]. Spearman correlation coefficients (r) are shown together with P values. SR, direct liver secretion rates (mg/day/kg); FCRs are measured in pools/day. Molar ratios are calculated using concentrations of VLDL apoC-III, apoE, and apoB in the stated fractions and their respective molecular weights. The y axes in the far right panels indicate rate constants for lipolytic conversion pathways between E2CIII1 dense VLDL (VB) and E2CIII2 IDL (top) and between E2CIII1 IDL and E2CIII2 LDL1 (bottom).
1198 Journal of Lipid Research Volume 48, 2007

Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html

Downloaded from www.jlr.org by guest, on December 19, 2016

Fig. 8. Contents of triglyceride, cholesterol, apoC-II, and apoC-I in light (L-) and dense (D-) VLDL, IDL,
and LDL types defined by the presence of apoC-III and apoE. Molar ratios of lipids and apolipoproteins
per apoB in each particle type are calculated based on their concentrations and those of apoB-100 in the samples taken before the infusion (i.e., time 0). Data shown are averages from all 11 participants. * P , 0.05 between corresponding E1CIII1 and E1CIII2 fractions; † P , 0.05 between corresponding E2CIII1 and
E2CIII2 fractions.

have the highest content of apoC-II and apoC-I, followed closely by E2CIII1 and then E1CIII2, whereas E2CIII2 fractions always have the lowest content. In addition, within each type of particle (E2CIII2, E1CIII2, etc.), apoC-II and apoC-I contents are highest in VLDL, progressively diminishing as particle density increases.
Slow TRL metabolism and overproduction of apoC-III-containing light VLDLs are two hallmarks of hypertriglyceridemia
Hypertriglyceridemic subjects have significantly slower FCRs than normolipidemic subjects in all dense VLDL and IDL fractions, regardless of their apoE and apoC-III composition (Fig. 4). This suggests that slow TRL FCR in hypertriglyceridemia involves factors other than apoC-III and apoE. Hypertriglyceridemic subjects also have lower rate constants for some conversion pathways from E2CIII1 TRLs to E2CIII2 IDLs and LDLs (Fig. 5), resulting in retention of these apoC-III-containing TRLs in plasma and higher flux to E2CIII1 LDLs. This may be attributable to low levels of HDLs among hypertriglyceridemic subjects, because HDLs are significantly correlated with rate constants for these conversion pathways (Spearman correlation coefficient: r 5 0.64, P 5 0.03 between HDL cholesterol

levels and rate constants for the E2CIII1 dense VLDL Y E2CIII2 IDL pathway; r 5 0.57, P 5 0.06 for the E2CIII1 IDL Y E2CIII2 LDL1 pathway) (Fig. 7, far right panels). Another major metabolic perturbation in hypertriglyceridemia is the overproduction of light VLDL (Fig. 3). Although secretion rates for total apoB lipoproteins are similar in both groups, there is a major difference in the secretion pattern. Secretion of light VLDL E2CIII1 constitutes 22% of total apoB secretion in hypertriglyceridemic subjects, more than three times as high as in normolipidemic subjects (P , 0.01). Together with increased secretion of E2CIII1 light VLDL, hypertriglyceridemic subjects have a .2-fold increase in the secretion of total light VLDL (P , 0.05). Overall, the combination of these factors, slow metabolism of the majority of TRLs and overproduction of apoC-III-containing light VLDLs, results in significantly increased plasma levels of apoC-III-containing TRLs and LDLs found among hypertriglyceridemic patients.
DISCUSSION
This study reveals substantial heterogeneity in TRL and LDL metabolism according to whether they have apoC-III

Kinetics of apoB lipoproteins according to apoC-III and apoE 1199

Downloaded from www.jlr.org by guest, on December 19, 2016

Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html

or apoE. ApoC-III-containing particles dominate VLDL plasma by immunoaffinity chromatography before ultracen-

metabolism in both hypertriglyceridemic and normolip- trifugation, ensuring that the native apoC-III-containing

idemic subjects, whereas those with neither apoC-III nor particles are segregated before further procedures that

apoE dominate LDL metabolism. ApoB lipoproteins con- could affect their properties. Finally, we are finding the

taining apoC-III have more rapid tracer enrichment, same pattern of tracer enrichment in the same participants

higher peak tracer-tracee ratios, and faster disappearance studied a second time under different nutritional con-

than particles without apoC-III in both hypertriglycer- ditions and in another group of 12 participants (6 normo-

idemic and normolipidemic participants. These patterns lipidemic and 6 hypertriglyceridemic) studied on a

of the tracer enrichment curves indicate that possession of different diet (unpublished observations).

apoC-III is a marker for a short residence time in plasma

This result contrasts with those from apoC-III transgenic

for VLDL, IDL, or LDL. Multiple-compartment modeling animal studies and provokes several thoughts unantici-

further showed that the presence of apoC-III is associated pated at the outset of this study. First, impaired lipolysis

with relative suppression of direct removal of these lipo- has been proposed as the main defect behind apoC-III-

proteins from the circulation and enhancement of their induced hypertriglyceridemia, because in vitro experiments

conversion to smaller and denser particles. ApoE, in con- suggest that apoC-III may be a direct noncompetitive in-

trast, apparently modulates the direction of flux in favor of hibitor for lipoprotein lipase (8). However, it has been

direct removal. TRLs that do not have apoC-III or apoE, called into question whether this is the case in vivo, as mice

constituting a minor portion of TRL secretion (,20%), overexpressing human apoC-III do not have reduced

are mostly stepwise converted to smaller particles down to lipoprotein lipase activity (14, 16, 17) and TRLs isolated

LDLs, and there is little direct removal from the circu- from these apoC-III transgenic mice do not retard lipase

lation before that end point. LDLs thus arise from two action in vitro (14, 16). In support of this view, our study

sources, plasma VLDLs and IDLs that have apoC-III but finds no significant correlation between the number of

not apoE and direct hepatic secretion.

molecules of apoC-III per particle and the rate constants

In general, these observations are consistent with some for lipolytic conversion among apoC-III-containing VLDLs.

known actions of apoC-III and apoE. First, considering Alternatively, it has been reported that VLDLs from apoC-

E2CIII1 VLDL, the fraction with the highest secretion III transgenic mice may have reduced binding to cell sur-

rate, conversion to IDL and LDL dominates direct removal face heparin sulfate proteoglycan as another, indirect,

by 4:1 or 5:1. This could be attributable to the inhibitory mechanism for retarded lipolysis in these animals (15, 17).

action of apoC-III on particle clearance by way of apoB-100 However, reduced binding to heparan sulfate proteoglycan

interacting with the LDL receptor or other hepatic re- of VLDL apoC-III could also operate at the hepatocyte sur-

ceptors (10). Second, considering E1CIII1 VLDL metab- face and be a mechanism for decreased receptor-mediated

olism, the rate constants for particle clearance are higher uptake and reduced clearance, as found in our study. In

and the rate constants for conversion to IDL and LDL are another study, in vitro, apoC-III content of apoB lipo-

lower than those of E2CIII1 VLDL, reflecting the role of proteins in humans was associated with increased binding

apoE in enhancing receptor-mediated clearance (23, 24). affinity to vascular proteoglycan biglycan (36). Thus, these

Correlation analysis also suggests that VLDL apoE content, divergent findings from distinct model systems do not pro-

the molar ratio of apoE to apoB, is positively related to rate vide a conclusion regarding whether VLDL metabolism is

constants for the direct removal of VLDL. Consistent with affected by apoC-III interacting with vascular proteoglycan.

these observations, apoC-III transgenic animal studies Evidence for a possible species difference in how apoC-III

show that retarded tissue uptake of VLDL or chylomi- affects VLDL metabolism comes from lipoprotein compo-

cron remnants in these animals can be corrected by add- sition analysis. VLDLs separated from apoC-III transgenic

ing exogenous apoE or by coexpressing apoE transgene mice have significantly lower contents of both apoE (14–16)

(15, 16). Conversely, the absence of apoE could explain and apoC-II (14, 16) per particle than those from wild-type

the preference for lipolysis over direct removal observed mice. However, apoC-III-containing TRLs and LDLs in hu-

among E2CIII2 and also E2CIII1 particles.

mans in this and previous studies (28) have high contents

A major unresolved question raised by this study is why of these apolipoproteins. Thus, the displacement of other

the presence of apoC-III is not associated with a retarded apolipoproteins by apoC-III overexpression in mice con-

metabolism of apoB lipoproteins. We are confident that trasts directly with the apparent coexistence of consider-

this is true for several reasons. First, rapid turnover for able amounts of apoC-III, apoC-II, and apoE on TRLs in

apoC-III-containing TRLs and LDLs is clear from the humans. In addition, the species difference may be re-

tracer enrichment curves and is not a result of any par- lated to the lack of cholesteryl ester transfer protein and

ticular features of the kinetic model. Second, all 11 par- the low levels of apoB lipoproteins relative to the amount

ticipants had the general pattern of tracer enrichments of apoC-III expressed in the mouse models.

shown for the group, particularly the characteristics of

Nonetheless, the kinetic hypothesis for the link between

fast tracer appearance and disappearance for apoC- apoC-III and coronary heart disease, requiring that apoC-

III-containing particles. Third, fast turnover of apoC- III-rich remnant lipoproteins be slowly metabolized in

III-containing particles relative to those without apoC-III plasma, needs revision, because these lipoproteins are me-

occurred for all density classes. Fourth, the apoC- tabolized as fast or faster than their apoC-III-free counter-

III-containing particles were extracted completely from parts and only some can be defined as remnants. The

1200 Journal of Lipid Research Volume 48, 2007

Downloaded from www.jlr.org by guest, on December 19, 2016

Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html

strong correlation between apoC-III and risk of coronary that undoubtedly occur among particles that already con-

heart disease among population studies (1–3) may be tain these apolipoproteins, because they would not cause

attributable to apoC-III’s direct involvement in atheroge- a phenotype change under immunoaffinity chromatogra-

nesis. We recently reported that apoC-III alone or as a phy. Third, because the current study was conducted in

component of VLDL and LDL stimulates the adhesion of the fasting state, it does not exclude the possibility that

monocytes to vascular endothelial cells (29–31). ApoC-III deviation from the observed pattern could occur under

activates b1-integrin in human monocytic cells and vas- other physiological conditions. In fact, during alimentary

cular cell adhesion molecule-1 and intercellular cell ad- lipemia, it has been reported that apoE is able to redis-

hesion molecule-1 in endothelial cells. Upstream pathways tribute from HDL to VLDL (40). Fourth, the results could

include the activation of protein kinase C, RhoA, and be specific to the nutritional condition of this study: a high-

nuclear factor-kB, all of which could have proinflamma- monounsaturated-fat diet. A high-carbohydrate diet re-

tory effects on cells in developing atheromatous lesions. duces the conversion of VLDL to IDL (41), a process that

Furthermore, apoC-III may also increase the binding of this study links to the content of both apoE and apoC-III.

apoB lipoproteins to vascular proteoglycan biglycans (36),

Studies of apoC-III and apoE kinetics have not been

an action that would increase the retention of apoB able to resolve to what extent they are freely exchangeable

lipoproteins in the arterial wall. Thus, the abundance of among apoB lipoproteins, and between apoB lipoproteins

apoC-III on TRLs may be a crucial factor in explaining and HDL, with some finding evidence of separate ex-

their atherogenicity. In addition to the content of apoC- changeable and nonexchangeable pools in VLDL or HDL

III, their high potential for LDL formation further renders and limited transfer (42, 43) and others favoring quick

the E2CIII1 apoB lipoproteins possibly the most athero- exchange to equilibrium producing a single homoge-

genic of lipoproteins.

neous pool (39, 44). Because apoC-III-containing VLDLs

This study demonstrates a global reduction in FCR for have on average ?50 copies of apoC-III per particle, the

TRLs in hypertriglyceridemia, regardless of particle apoC- coexistence in the circulation of these apoC-III-laden par-

III and apoE composition. Previous studies have found ticles with those free of any apoC-III, and the coexistence

slow VLDL turnover among hypertriglyceridemic subjects of apoE-containing VLDLs with VLDLs free of apoE, are

(37, 38). This has at times been attributed to the action of compatible only with conditional, and not free, exchange

a high apoC-III concentration in hypertriglyceridemic sub- of apoC-III and apoE between VLDL and HDL. Our

jects (22). However, we also find that hypertriglyceridemia finding that VLDL particles not having apoC-III or apoE,

status is independently associated with reduced TRL FCR, the smallest type of VLDL (5, 6), do not acquire apoE

indicating that factors other than apoC-III or apoE retard during circulation is supported by in vitro experiments

TRL catabolism in hypertriglyceridemia. Our results also showing that apoE does not adhere readily to small apoB

show that hypertriglyceridemic subjects have a three-fold lipoproteins (45). However, this does not explain why

greater secretion rate of light VLDL particles containing VLDL E2CIII1, which are very large particles, do not

apoC-III without apoE than normolipidemic subjects. acquire apoE. Perhaps apoE does not attach to VLDLs that

VLDL apoC-III levels are strongly correlated with both are undergoing rapid lipolysis, the principal metabolic

VLDL secretion rates and VLDL triglyceride concentra- pathway for VLDL E2CIII1.

tions but have no relationship with VLDL FCR. These ob-

ApoC-III especially, but also apoE, detaches from apoB

servations are compatible with the concept that apoC-III lipoproteins during lipolysis, and they are likely taken up

may serve as a regulator for hepatic triglyceride produc- by HDL. The magnitude of these lipolytic conversion path-

tion. Overproduction of VLDL triglyceride but not apoB ways among apoB lipoproteins may depend on the avail-

has been reported in some apoC-III transgenic animal ability of HDLs as acceptors for apoC-III and apoE. Our

models (14, 16). In humans, production of VLDL apoC-III results demonstrate significant correlations between HDL

correlates strongly with the secretion of VLDL triglyceride cholesterol and rate constants of conversion pathways

(21), consistent with the findings of our study. It is possible from E2CIII1 TRLs to E2CIII2 LDLs. This may explain

that apoC-III could stimulate the secretion of triglyceride- why hypertriglyceridemic subjects in our study, having low

rich VLDL at the cellular level, and the exact mechanisms HDL concentrations, accumulate atherogenic E2CIII1

warrant further investigation.

apoB lipoproteins as a result of the impaired ability to con-

This study finds no evidence for the acquisition in plasma vert these particles to E2CIII2. Finally, the movement of

of apoC-III or apoE by TRLs that do not already contain apoE and apoC-III from TRLs to HDLs may be a way that

them when they enter the systemic circulation. This is sur- apoB lipoprotein metabolism affects HDL metabolism,

prising, because a considerable proportion of plasma apoC- because apoC-III and apoE are shown to interact with he-

III and apoE is found on HDLs, and studies have suggested patic lipase, ABCA1, scavenger receptor class B type I, and

that apoC-III and apoE can exchange between VLDL and other regulators of HDL metabolism (24, 46).

HDL (39, 40). This may reflect limitations of the design of

In conclusion, by uniting concepts of apoB lipoprotein

this study. First, transfer of apoC-III or apoE from HDL or metabolism related to particle size and apoC-III and apoE

other TRLs to nascent TRLs at the hepatocyte surface is content, a new structure for lipoprotein physiology is re-

compatible with our data and would be seen by our system vealed. A lipoprotein system that has low potential for

as direct hepatic secretion. Second, our system also does atherogenesis and optimal provision of lipids for nutrition

not recognize the transfers or exchanges of apoC-III or apoE would secrete triglycerides on VLDLs that have both apoC-

Kinetics of apoB lipoproteins according to apoC-III and apoE 1201

Downloaded from www.jlr.org by guest, on December 19, 2016

Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html

III and apoE. ApoC-III directs flux away from direct removal by the liver to allow lipoproteins to circulate longer and provide peripheral tissues with lipids. As apoC-III is

uptake of rat lipoproteins containing apoproteins B and E. J. Biol. Chem. 255: 10464–10471. 12. Sehayek, E., and S. Eisenberg. 1991. Mechanisms of inhibition by apolipoprotein C of apolipoprotein E-dependent cellular metabo-

lost from the particles during lipolysis, lifting its inhibitory effect on clearance, the influence of apoE comes into play, directing the remnant particles to the liver for clear-

lism of human triglyceride-rich lipoproteins through the low density lipoprotein receptor pathway. J. Biol. Chem. 266: 18259–18267. 13. Ito, Y., N. Azrolan, A. O’Connell, A. Walsh, and J. L. Breslow. 1990. Hypertriglyceridemia as a result of human apo CIII gene expres-

ance. Ideally, the liver would produce few apoB lipoproteins with apoC-III that do not also have apoE, particles that are not only themselves atherogenic but also major precursors for producing LDL. Deviation from optimal settings (e.g., the overproduction of VLDL with apoC-III but not apoE) could give rise to not only hypertriglyceridemia but also hypercholesterolemia and hyperapobetalipoproteinemia.

sion in transgenic mice. Science. 249: 790–793. 14. Aalto-Setala, K., E. A. Fisher, X. Chen, T. Chajek-Shaul, T. Hayek,
R. Zechner, A. Walsh, R. Ramakrishnan, H. N. Ginsberg, and J. L. Breslow. 1992. Mechanism of hypertriglyceridemia in human apolipoprotein (apo) CIII transgenic mice. Diminished very low density lipoprotein fractional catabolic rate associated with increased apo CIII and reduced apo E on the particles. J. Clin. Invest. 90: 1889–1900. 15. Aalto-Setala, K., P. H. Weinstock, C. L. Bisgaier, L. Wu, J. D. Smith,

Our results also call for renewed commitment to the study of apoC-III, an enigmatic component of most TRLs that is strongly predictive of coronary heart disease.

and J. L. Breslow. 1996. Further characterization of the metabolic properties of triglyceride-rich lipoproteins from human and mouse apoC-III transgenic mice. J. Lipid Res. 37: 1802–1811. 16. de Silva, H. V., S. J. Lauer, J. Wang, W. S. Simonet, K. H. Weisgraber,

R. W. Mahley, and J. M. Taylor. 1994. Overexpression of human

The authors are grateful to the volunteers who participated in

apolipoprotein C-III in transgenic mice results in an accumulation of apolipoprotein B48 remnants that is corrected by excess apolipo-

this study. The authors also express thanks to Helena Judge Ellis, Jeremy Furtado, and Glen Daly for their technical and managerial assistance. This work was supported by Grants R01 HL-34980, R01 HL-56210, and RR-02635 from the National

protein E. J. Biol. Chem. 269: 2324–2335. 17. Ebara, T., R. Ramakrishnan, G. Steiner, and N. S. Shachter. 1997.
Chylomicronemia due to apolipoprotein CIII overexpression in apolipoprotein E-null mice. Apolipoprotein CIII-induced hypertriglyceridemia is not mediated by effects on apolipoprotein E. J. Clin.

Heart, Lung, and Blood Institute, National Institutes of Health (Bethesda, MD).

Invest. 99: 2672–2681. 18. Maeda, N., H. Li, D. Lee, P. Oliver, S. H. Quarfordt, and J. Osada.
1994. Targeted disruption of the apolipoprotein C-III gene in mice

results in hypotriglyceridemia and protection from postprandial

hypertriglyceridemia. J. Biol. Chem. 269: 23610–23616.

REFERENCES

19. Jong, M. C., P. C. Rensen, V. E. Dahlmans, H. van der Boom, T. J. van Berkel, and L. M. Havekes. 2001. Apolipoprotein C-III

deficiency accelerates triglyceride hydrolysis by lipoprotein lipase

1. Alaupovic, P., W. J. Mack, C. Knight-Gibson, and H. N. Hodis. 1997.

in wild-type and apoE knockout mice. J. Lipid Res. 42: 1578–1585.

The role of triglyceride-rich lipoprotein families in the progression 20. Ginsberg, H. N., N. A. Le, I. J. Goldberg, J. C. Gibson, A.

of atherosclerotic lesions as determined by sequential coronary

Rubinstein, P. Wang-Iverson, R. Norum, and W. V. Brown. 1986.

angiography from a controlled clinical trial. Arterioscler. Thromb.

Apolipoprotein B metabolism in subjects with deficiency of

Vasc. Biol. 17: 715–722.

apolipoproteins CIII and AI. Evidence that apolipoprotein CIII

2. Sacks, F. M., P. Alaupovic, L. A. Moye, T. G. Cole, B. Sussex, M. J.

inhibits catabolism of triglyceride-rich lipoproteins by lipoprotein

Stampfer, M. A. Pfeffer, and E. Braunwald. 2000. VLDL, apolipo-

lipase in vivo. J. Clin. Invest. 78: 1287–1295.

proteins B, CIII, and E, and risk of recurrent coronary events in the 21. Cohn, J. S., B. W. Patterson, K. D. Uffelman, J. Davignon, and G.

Cholesterol and Recurrent Events (CARE) Trial. Circulation. 102:

Steiner. 2004. Rate of production of plasma and very-low-density

1886–1892.

lipoprotein (VLDL) apolipoprotein C-III is strongly related to the

3. Lee, S. J., H. Campos, L. A. Moye, and F. M. Sacks. 2003. LDL

concentration and level of production of VLDL triglyceride in

containing apolipoprotein CIII is an independent risk factor for

male subjects with different body weights and levels of insulin sen-

coronary events in diabetic patients. Arterioscler. Thromb. Vasc. Biol.

sitivity. J. Clin. Endocrinol. Metab. 89: 3949–3955.

23: 853–858.

22. Chan, D. C., G. F. Watts, M. N. Nguyen, and P. H. Barrett. 2006.

4. Alaupovic, P. 1996. Significance of apolipoproteins for structure,

Apolipoproteins C-III and A-V as predictors of very-low-density

function, and classification of plasma lipoproteins. Methods Enzymol.

lipoprotein triglyceride and apolipoprotein B-100 kinetics. Arterioscler.

263: 32–60.

Thromb. Vasc. Biol. 26: 590–596.

5. Campos, H., D. Perlov, C. Khoo, and F. M. Sacks. 2001. Distinct 23. Kowal, R. C., J. Herz, K. H. Weisgraber, R. W. Mahley, M. S. Brown,

patterns of lipoproteins with apoB defined by presence of apoE or

and J. L. Goldstein. 1990. Opposing effects of apolipoproteins E

apoC-III in hypercholesterolemia and hypertriglyceridemia. J. Lipid

and C on lipoprotein binding to low density lipoprotein receptor-

Res. 42: 1239–1249.

related protein. J. Biol. Chem. 265: 10771–10779.

6. Khoo, C., H. Campos, H. Judge, and F. M. Sacks. 1999. Effects of es- 24. Mahley, R. W., and S. C. Rall, Jr. 2000. Apolipoprotein E: far more

trogenic oral contraceptives on the lipoprotein B particle system de-

than a lipid transport protein. Annu. Rev. Genomics Hum. Genet. 1:

fined by apolipoproteins E and C-III content. J. Lipid Res. 40: 202–212.

507–537.

7. Brown, W. V., and M. L. Baginsky. 1972. Inhibition of lipoprotein 25. Ji, Z. S., S. Fazio, Y. L. Lee, and R. W. Mahley. 1994. Secretion-

lipase by an apoprotein of human very low density lipoprotein.

capture role for apolipoprotein E in remnant lipoprotein metabo-

Biochem. Biophys. Res. Commun. 46: 375–382.

lism involving cell surface heparan sulfate proteoglycans. J. Biol.

8. Wang, C. S., W. J. McConathy, H. U. Kloer, and P. Alaupovic. 1985.

Chem. 269: 2764–2772.

Modulation of lipoprotein lipase activity by apolipoproteins. Effect 26. Ikewaki, K., W. Cain, F. Thomas, R. Shamburek, L. A. Zech, D.

of apolipoprotein C-III. J. Clin. Invest. 75: 384–390.

Usher, H. B. Brewer, Jr., and D. J. Rader. 2004. Abnormal in vivo

9. Kinnunen, P. K., and C. Ehnolm. 1976. Effect of serum and

metabolism of apoB-containing lipoproteins in human apoE

C-apoproteins from very low density lipoproteins on human post-

deficiency. J. Lipid Res. 45: 1302–1311.

heparin plasma hepatic lipase. FEBS Lett. 65: 354–357.

27. Schaefer, E. J., R. E. Gregg, G. Ghiselli, T. M. Forte, J. M. Ordovas,

10. Clavey, V., S. Lestavel-Delattre, C. Copin, J. M. Bard, and J. C.

L. A. Zech, and H. B. Brewer, Jr. 1986. Familial apolipoprotein E

Fruchart. 1995. Modulation of lipoprotein B binding to the LDL

deficiency. J. Clin. Invest. 78: 1206–1219.

receptor by exogenous lipids and apolipoproteins CI, CII, CIII, and 28. Tomiyasu, K., B. W. Walsh, K. Ikewaki, H. Judge, and F. M. Sacks.

E. Arterioscler. Thromb. Vasc. Biol. 15: 963–971.

2001. Differential metabolism of human VLDL according to content

11. Windler, E. E., P. T. Kovanen, Y. S. Chao, M. S. Brown, R. J. Havel,

of apoE and apoC-III. Arterioscler. Thromb. Vasc. Biol. 21: 1494–1500.

and J. L. Goldstein. 1980. The estradiol-stimulated lipoprotein 29. Kawakami, A., M. Aikawa, P. Alcaide, F. W. Luscinskas, P. Libby, and

receptor of rat liver. A binding site that membrane mediates the

F. M. Sacks. 2006. Apolipoprotein CIII induces expression of

1202 Journal of Lipid Research Volume 48, 2007

Supplemental Material can be found at: http://www.jlr.org/content/suppl/2007/02/21/P600011-JLR20 0.DC1.html

vascular cell adhesion molecule-1 in vascular endothelial cells and 38. Dunn, F. L., S. M. Grundy, D. W. Bilheimer, R. J. Havel, and

increases adhesion of monocytic cells. Circulation. 114: 681–687.

P. Raskin. 1985. Impaired catabolism of very low-density lipoprotein-

30. Kawakami, A., M. Aikawa, P. Libby, P. Alcaide, F. W. Luscinskas,

triglyceride in a family with primary hypertriglyceridemia. Metabo-

and F. M. Sacks. 2006. Apolipoprotein CIII in apolipoprotein

lism. 34: 316–324.

B lipoproteins enhances the adhesion of human monocytic cells 39. Nguyen, M. N., D. C. Chan, K. P. Dwyer, P. Bolitho, G. F. Watts,

to endothelial cells. Circulation. 113: 691–700.

and P. H. Barrett. 2006. Use of Intralipid for kinetic analysis of

31. Kawakami, A., M. Aikawa, N. Nitta, M. Yoshida, P. Libby, and F. M.

HDL apoC-III: evidence for a homogeneous kinetic pool of apoC-

Sacks. 2007. Apolipoprotein CIII-induced THP-1 cell adhesion to

III in plasma. J. Lipid Res. 47: 1274–1280.

endothelial cells involves pertussis toxin-sensitive G protein- and 40. Blum, C. B. 1982. Dynamics of apolipoprotein E metabolism in

protein kinase Calpha-mediated nuclear factor-kappaB activation.

humans. J. Lipid Res. 23: 1308–1316.

Arterioscler. Thromb. Vasc. Biol. 27: 219–225.

41. Abbott, W. G., B. Swinburn, G. Ruotolo, H. Hara, L. Patti,

32. Lindgren, F. T., L. C. Jensen, and F. T. Hatch. 1972. The isolation

I. Harper, S. M. Grundy, and B. V. Howard. 1990. Effect of a

and quantitative analysis of serum lipoproteins. In Blood Lipids

high-carbohydrate, low-saturated-fat diet on apolipoprotein B

and Lipoproteins: Quantitation, Composition, and Metabolism.

and triglyceride metabolism in Pima Indians. J. Clin. Invest. 86:

G. J. Nelson, editor. John Wiley-Interscience, New York. 181–274.

642–650.

33. Ehnholm, C., H. Garoff, K. Simons, and H. Aro. 1971. Purification 42. Batal, R., M. Tremblay, P. H. Barrett, H. Jacques, A. Fredenrich,

and quantitation of the human plasma lipoprotein carrying the

O. Mamer, J. Davignon, and J. S. Cohn. 2000. Plasma kinetics of

Lp(a) antigen. Biochim. Biophys. Acta. 236: 431–439.

apoC-III and apoE in normolipidemic and hypertriglyceridemic

34. Su, W., H. Campos, H. Judge, B. W. Walsh, and F. M. Sacks. 1998.

subjects. J. Lipid Res. 41: 706–718.

Metabolism of apo(a) and apoB100 of lipoprotein(a) in women: 43. Bukberg, P. R., N. A. Le, H. N. Ginsberg, J. C. Gibson, A.

effect of postmenopausal estrogen replacement. J. Clin. Endocrinol.

Rubinstein, and W. V. Brown. 1985. Evidence for non-equilibrating

Metab. 83: 3267–3276.

pools of apolipoprotein C-III in plasma lipoproteins. J. Lipid Res.

35. Maugeais, C., K. Ouguerram, R. Frenais, P. Maugere, B. Charbonnel,

26: 1047–1057.

T. Magot, and M. Krempf. 2001. Effect of low-density lipoprotein 44. Huff, M. W., and P. J. Nestel. 1982. Metabolism of apolipoproteins

apheresis on kinetics of apolipoprotein B in heterozygous familial

CII, CIII1, CIII2 and VLDL-B in human subjects consuming high

hypercholesterolemia. J. Clin. Endocrinol. Metab. 86: 1679–1686.

carbohydrate diets. Metabolism. 31: 493–498.

36. Olin-Lewis, K., R. M. Krauss, M. La Belle, P. J. Blanche, P. H. 45. Dong, L. M., C. Wilson, M. R. Wardell, T. Simmons, R. W. Mahley,

Barrett, T. N. Wight, and A. Chait. 2002. ApoC-III content of apoB-

K. H. Weisgraber, and D. A. Agard. 1994. Human apolipoprotein E.

containing lipoproteins is associated with binding to the vascular

Role of arginine 61 in mediating the lipoprotein preferences of

proteoglycan biglycan. J. Lipid Res. 43: 1969–1977.

the E3 and E4 isoforms. J. Biol. Chem. 269: 22358–22365.

37. Fisher, W. R., L. A. Zech, P. Bardalaye, G. Warmke, and M. Berman. 46. Shachter, N. S. 2001. Apolipoproteins C-I and C-III as impor-

1980. The metabolism of apolipoprotein B in subjects with hyper-

tant modulators of lipoprotein metabolism. Curr. Opin. Lipidol. 12:

triglyceridemia and polydisperse LDL. J. Lipid Res. 21: 760–774.

297–304.

Downloaded from www.jlr.org by guest, on December 19, 2016

Kinetics of apoB lipoproteins according to apoC-III and apoE 1203