Effects of gemfibrozil on very-low-density lipoprotein composition and low-density lipoprotein size in patients with hypertriglyceridemia or combined hyperlipidemia

atherosclerosis Atherosclerosis 126 (1996) 105- II6

Effects of gemfibrozil on very-low-density lipoprotein composition and low-density lipoprotein size in patients with hypertriglyceridemia or combined hyperlipidemia Chao-Yuh Yang *, Zi-Wei Gu, Yong-Hong Xie, Natalya V. Valentinova, Manlan Yang, Daniel Yeshurun, Jun Anthony V. Quion, Antonio M. Gotto, Jr Department of Medicine. 13aylor College of Medicine and The Methodist Hospital, Houston TX 77030, USA

6565 Fannin Street, MS/AhOl,

Received 1 November 1995; revised 1 May 1996; accepted,,1 May 1996

Abstract

To examme the effects of gemfibrozil on very-low-density lipoprotein (VLDL) composition and low-density lipoprotein (LDL) size, five men with hypertriglyceridemia (HTG) alone and five men with HTG and hypercholesterolemia (combined hyperlipidemia, CHLP) were randomized for 8 weeks to Lopid SR (slow-release gemfibrozil; two 600-mg tablets once per day) or placebo in a crossover study. Drug therapy versus placebo significantly decreased plasma triglyceride (680/o),and VLDL (77%) and significantly increased high-density lipoprotein cholesterol (25%); total cholesterol, apolipoprotein B and lipoprotein[a] concentrations did not change significantly. With drug, mean total apoE in pla.sma was 53% lower in patients with HTG and 39% lower in patients with CHLP. Gemfibrozil significantly affected VLDL composition: protein increased 26%, molar ratio of apoE to apoB reduced 48%, apoC-II increased 19%, an.d apoC-III decreased9%. LDL cholesteryl ester significantly increased with drug treatment. VLDL subfractions were separated and classified as heparin binding (VLDL,, apoE rich) or nonbinding (VLDLNn., and VLDL,,,, both apoE poor). All VLDL subfractions were significantly lower with drug therapy, and the differences for total VLDL and for VLDL subfractions were greater in patients with HTG. With placebo, VLDL, accounted for 41.8% of VLDL in HTG and 49.0% of VLDL in CHLP, reduced to 27.6% and 38.6%, respectively, with gemfibrozil. Taken together, these results suggest that treatment with gemfibrozil reduces plasma concentrations of VLDL and alters the apoprotein composition of VLDL in a manner that may favor LDL- and VLDL-receptor-mediated clearance of the apoE-rich VLDL subfraction, thereby reducing TG-rich particle concentrations, and possibly reducing risk for coronary heart disease. Keywords: Gemnbrozil; Plasma lipoproteins; Hyperlipidemia; VLDL subfractions: VLDL composition; LDL size; Apolipoprotein E

* Corresponding author. Tel.: ~- 1 713 7984210; fax: + 1 713 7984121 0021-9150/96/$15.000 1996 Elsevier Science Ireland Ltd. All rights reserved PZI SO02 l-9 150(96)05899-6

106

C.-Y. Yang et al. / Atherosclerosis

1. Introduction When plasma triglyceride (TG) levels are elevated, a large proportion of the total plasma cholesterol is carried in very-low density lipoprotein (VLDL) particles, which also carry most of the TG. Epidemiologic studies show an association between abnormalities in TG metabolism and the development of coronary heart disease [l-5]. Studies in HTG patients show that gemfibrozil decreases VLDL-TG synthesis and stimulates their clearance with increased lipoprotein lipase (LPL) activity [6-81. How gemfibrozil affects VLDL subfractions with different apoE content in hyperlipidemic patients has not been reported. Whereas several laboratories previously identified a fraction of circulating VLDL containing no detectable apoE [9-121, we have demonstrated, using heparin-Sepharose affinity chromatography, three VLDL fractions; one of these (a heparinbinding fraction designated VLDL,) contains considerable apoE and two others (nonbinding fractions, designated VLDL,,., and VLDL,,-,) contain little apoE [13]. Becauseof the significant role of apoE in the function and metabolic fate of VLDL [14- 171 and because the lipid-lowering fibrates are known to affect lipoprotein composition and apolipoprotein abundance [18-231, we undertook the present study to determine whether gemfibrozil would affect the distribution of these VLDL fractions. Our focus was the effects of drug on VLDL composition and apoE fraction abundance, and how these effects might differ in hypertriglyceridemia (HTG) alone and in combination with hypercholesterolemia (combined hyperlipidemia; CHLP). We also examined effects of drug treatment on LDL composition and diameter, and on plasma lipid, lipoprotein, and apolipoprotein concentrations.

126 (1996) 105- I16

Metabolism Clinic of The Methodist Hospital, Houston, Texas. Patients with HTG alone (N = 5) were defined by fasting TG concentrations greater than 500 mg/dl. Patients with CHLP (N = 5) were defined by elevation of fasting TG and total cholesterol in comparable proportion; all 10 patients had elevated VLDL and decreased highdensity lipoprotein (HDL) cholesterol concentrations. TG values for all patients ranged from 240 to 1500 mg/dl with total cholesterol ranging from 168 to 346 mg/dl. None of the patients had diabetes mellitus or disorders of the gastrointestinal tract, liver, kidneys, or endocrine system. All gave written, informed consent to a research protocol approved by the Baylor College of MedicineAffiliated Hospitals institutional review board. Plasma for control assessmentswas obtained from eight healthy male volunteers not taking any drug known to affect levels of plasma lipids or lipoproteins. 2.2. Study design After an initial physical examination and laboratory tests, patients qualifying for the study were randomly assigned for 8 weeks to gemfibrozil 1200 mg daily (slow-release formulation, Lopid SR; two 600-mg tablets once a day) or to placebo, and then crossed over to the other group for another 8 weeks. During the study, the patients consumed their usual diets and maintained stable body weights. Each patient visited the clinic five times: an initial visit before starting on the medication or placebo, and subsequent visits at 4-week intervals. At each clinic visit, 50 ml of fasting blood was obtained. For evaluation of effects of treatment on plasma lipid, lipoprotein, and apolipoprotein concentrations, samples from the third and fifth visits, representing samples with or without drug, were used. 2.3. Plasma analysis

2. Methods 2.1. Subjects Ten male outpatients, aged 30-63 years, were recruited from the Atherosclerosis Lipid

Blood was drawn into tubes containing disodium-EDTA and centrifuged for 30 min at 4°C to recover plasma. A portion of the plasma was divided into aliquots and stored at - 20°C until assayed.Concentrations of total cholesterol, HDL

C.-Y. Yang et al. / Atherosclerosis

cholesterol, and TG in plasma were determined by Roche Biomedical Laboratories (Houston, TX) using an Olympus 5000 analyzer. 2.4. Separation ,oj’ VLDL, VLDL ApoE fractions, and LDL To separate VLDL, VLDL apoE fractions, and LDL from plasma, aprotinin (50 TIU/ml), sodium azide (0.6 mg/ml), and EDTA (0.6 mg/ ml) were added to plasma. VLDL (density < 1.006 g/ml) was obtained by preparative ultracentrifugation at 45 000 rev./min at 4°C for 18 h in a Beckman 50.2 Ti rotor (Beckman Instruments, Palo Alto, CA), concentrated, and then dialyzed against 10 mM Tri-HCl, 0.18 M NaCl, 1 mM EDTA, and 0.05% NaN,, pH 7.4 (buffer A). Affinity chromatography of VLDL was performed on a 1.5- x 16-cm heparin-Sepharose CL-6B column (Pharmacia, Piscataway, NJ) equilibrated with buffer A as described elsewhere [13]. The collection rate was 0.52 ml/min at 4°C. With a discontinuous elution, the fractions VLDL,,-, (nonbinding), and VLDLNR.2 (nonbinding), and VLDL, (heparin binding) were obtained at 0.18 M NaCl, 0.25 M NaCl, and 1.00 M NaCl, respectively. Percentage recovery for VLDL and its fractions with or without drug was determined according to absorption at 280 nm. LDL was recovered from the supernatant after a density adjustment to d = 1.063 g/ml and ultracentrifugation. 2.5, Analysis of lipid and total protein content qf iipoprotein,Y Determinations of TG, cholesterol, and cholesteryl ester content of lipoproteins were made enzymatically (Boehringer Mannheim, Germany). Phospholipid content was determined enzymatically using a reagent kit supplied by the Wako Company (Osaka, Japan). Total protein content of lipoproteins was determined by the rnethod of Lowry et al. [24] with sodium dodecyl sulfate (SDS) at a final concentration of 0.1%.

126 (1996) 105 - I16

107

2.6. Purljication of apolipoproteins ApoE was purified from total VLDL apolipoproteins. After VLDL delipidation, the total VLDL apolipoprotein content was fractionated chromatographically on Sephacryl S-200, and the apoE was separated from C apolipoproteins [25]. The purified apoE was used as antigen for production of antibody and as a primary standard to calibrate the apolipoprotein concentration in Omega Standard Serum (Technicon Instruments Co., Tarrytown, NY), which was used as a secondary standard for apolipoprotein quantification. The standard for quantification of lipoprotein[a] (Lp[a]) was kindly provided by Dr Joel D. Morrisett of Baylor College of Medicine. 2.7. Preparation of antibodies Polyclonal anti-human apoB and anti-human apoE antibodies were raised in goats and purified as described elsewhere [26]. 2.8. Enzyme-linked immunosorbent assa?’ (ELISA) for apolipoprotein analysis ApoB, apoE, and Lp[a] from plasma as well as apoB and apoE from VLDL were measured by ELISA, using the antibodies prepared as previously described [26] with the exception of the goat anti-human apo[a] antibody, which was a product of the Alpha Biomedical Lab (Bellevue, WA). To determine apoB, apoE, and Lp[a] in plasma or VLDL, 96-well polystyrene plates (Corning, Corning, NY) were coated with 100 /II/well of optimal concentrations of each antibody, which had been determined earlier in a dilution series using a checkerboard system. The concentrations of goat anti-human LDL, apoE, and apo[a] were 5, 6, and 2.5 ,ug/ml, respectively. Coating was performed overnight in a phosphate-buffered saline (PBS) buffer (75 mM phosphate, 75 mM NaCl, pH 7.2, 0.01% NaN,). Next, the plates were washed twice with PBS containing 1% bovine serum albumin (PBS-BSA) and incubated at 37°C for 1 h. After the plates were washed with PBSBSA containing 0.1% Tween-20 (PBS-BSATween), 100 111of sample or standard was added,

108

C.- Y. Yang et al. / Atherosclerosis

followed by incubation at 37°C for 2 h. The PBS-BSA-Tween buffer for sample dilution and washing was used to minimize nonspecific binding to the plates. Samples and standards were diluted with an electronic diluter (Micro Lab 1000, Hamilton, Switzerland). After the plate was washed four times with PBS-BSA-Tween, the conjugate in PBS-BSA-Tween (100 PI/well) at the optimal dilution of 1:7000 for apoB and apoE and of 1:500 for apo[a] was added, followed by incubation at 37°C for 2 h. The plates were washed, and then 160 ,~l @phenylenediamine/H,0, (substrate solution) was pipetted into each well of the washed plates. Color development was allowed to continue at room temperature for 20 min, at which time 40 ~1 of 5 N H,SO, was added to stop the reaction. A Dynatech MR5000 microtiter reader was used for data collection and analysis. 2.9. Quantitation of C apolipoproteins in VLDL The C apolipoproteins in VLDL were delipidated with isopropanol as described [27,28]. In brief, 200- to 300~~1VLDL samples at a concentration of l-2 mg/ml were mixed with an equal volume of pure isopropanol. After several minutes, the samples were centrifuged in an Eppendorf centrifuge at room temperature for 10 min. After the supernatant was removed, another 100 ~1 isopropanol was added to each tube for re-extraction. Both supernatants were combined and transferred to a 15ml polypropylene tube. Sequential washing with 2 ml methanol, 3 ml chloroform, and 5 ml diethyl ether was performed. The C apolipoproteins were precipitated by chilling at 4°C for 1 h. The precipitates (C apolipoproteins) were centrifuged at 2000 rev./min for 2 min and then solubilized in 200 ~1 solvent containing 40 ~1 3 M guanidine-HCl and 160 ~10.1% trifluoroacetic acid (TFA). A Beckman high-performance liquid chromatography (HPLC) system equipped with a Vydac Cl8 column (4.6 x 250 mm) and TFA buffer system [29] was used to isolate C apolipoproteins. A linear gradient from 25% to 55% buffer B at 1% per min with a flow rate of 1.5 ml/min was applied. The column temperature was set at 50°C. A typical HPLC pattern for C apolipoprotein

126 (1996) 105- I16

separation is shown in Fig. 1. Individual peaks isolated by HPLC were characterized by automatic sequenceanalysis and quantitated by amino acid analysis. The quantitated C protein mixture was used as a standard to determine the C apolipoprotein composition in VLDL according to the procedure described by Hughes et al. [30]. A Hewlett Packard automatic HPLC system with a Vydac Cl8 column (2.1 x 200 mm) was used to quantitate the unknown samples. 2.10. Electrophoresis of LDL and evaluation To determine the average diameter of LDL particles, gradient gel electrophoresis under nondenaturing conditions was used. The separating gel contained 2.5-6% and the stacking gel contained 2.5% polyacrylamide. LDL, isolated in density range 1.019-1.063 g/ml, was dialyzed against the same phosphate buffer containing 167.5 mg/l NaH,PO$H,O, 379 mg/l Na,HPO,,

Minutes Fig. 1. High-performance liquid chromatography separation of C apolipoproteins. The C apolipoproteins in VLDL were prepared by isopropanol extraction and then solubilized in 200 ~1 solvent containing 40 ~1 3 M guanidine-HCl and 160 ~1 0.1% TFA (see Section 2). A Vydac Cl8 column (4.6 x 250 mm) and TFA buffer system with a linear gradient from 25% to 55% at 1% per min and a flow rate of 1.5 ml/min were applied. ApoC-I, apoC-II, and apoC-III were eluted and separated as indicated.

C.-Y. Yang et al. i Atherosclerosis

109

126 (1996) 105-l I6

Table I Plasma lipid, lipoprotein, and apolipoprotein values in hyperlipidemic subjects (N = 10) according to treatment assignment Drug TG TC HDL-C VLDL LPkal ApoE ApoB ApoE/ApoB

253 238 35 1.3 13.8 4.4 109 0.7

i: + + f+ f * *

130 41 11 0.8 20.4 .4 28 0.3

Placebo

P-value

Drug/placebo

812 260 28 5.7 7.5 8.3 110 1.2

0.001 NS

0.32 0.91 1.25 0.23 1.84 0.53 0.95 0.58

k + + * * f k +

412 51 9 3.6 5.5 3.5 29 0.7

0.028 0.001 NS 0.001 NS 0.004

All values are given in mg/dl except VLDL units per dl (u/dl) and protein ratio. NS, not significant, P > 0.05.

0.01% Na,EDT.A, and 0.01% thimerosal. The protein concentration (mg/ml) was estimated as optical density (280 nm)/l.6 and adjusted to 2 mg/ml. The LDL solution was mixed with an application buffer (60 mM Tris-HCl, pH 8.6, 10% glycerol) in a 1:l ratio, and 20 ~1 of this sample was applied to the gel. In addition, 20 ~1 of standard proteins (HMW calibration kit, Pharmacia, Piscataway, NJ) and 25 ,ul of carboxylated latex beads (Duke Scientific, Palo Alto, CA) were applied to the gel. Latex beads were dialyzed against the same phosphate buffer as used with LDL except the latter contained 5% glycerol and 0.005% SDS. The electrophoresis buffer contained 3 g/l Tris and 14.4 g/l glycine. Gels were run for 22 h at 40 mA per two slabs. Gels were fixed in 20% ethanol, 10% glacial acetic acid, and 2.5% 5-sulfosalicylic acid and stained with Coomassie blue R-250. SepraScan 2001 1-D Densitometry Software was used to scan gels and to calculate the LDL average diameter. Thyroglobulin (with a hydrated molecular diameter of 17 nm) and latex beads (38 _+ 4 nm) served as standards 2311.An LDL score was calculated as described by Campos et al. [32]. In brief, a score from 1 (largest particles) to 7 (smallest particles) was given to each of seven LDL subpopulations described by Krauss and Burke [31]. LDL particles larger than 27.8 nm in diarneter were i.ncluded and given a score of 1. Each LDL peak was assigned to one of the subpopulations according to the position of its maximum value. The score of a subpopulation was multiplied by the relative area occupied by

the peak (total area = l), and the sum of these products for all peaks yielded the resulting LDL score. In addition, LDL diameter was estimated from LDL chemical composition, with the assumption that each particle contained only one molecule of apoB-100. The molar ratios of the other components in the particle were calculated. The volumes of the LDL particles were calculated on the basis of the lipid content in each particle and their respective partial specific volumes using the method of Luzzati et al. [33]. 2.11. Statistical analysis A two-tailed paired t-test was used to estimate differences between results in patients given drug or placebo, and between control and experimental values as indicated in each table. Significance was accepted at a level of P < 0.05 (two-tailed test). All statistical analyses were performed using SigmaPlot 5.0 (San Rafael, CA). All data are expressed in terms of mean values and standard deviation (S.D.).

3. Results

3.1. Plasma lipid concentrations Mean plasma lipid, lipoprotein, and apolipoprotein values of patients after 8 weeks of drug or placebo treatment are shown in Table 1. In gemn-

110

C.-Y. Yang et al. /Atherosclerosis

126 (1996) 105-116

Table 2 Plasma lipid, lipoprotein, and apohpoprotein values according to treatment assignments and hyperlipidemia classification Hypertriglyceridemia (N = 5) Drug TG TC HDL-C VLDL Wal ApoE ApoB ApoE/ApoB

293 231 31 1.6 19.2 4.4 103 0.7

Placebo k 88 k 45 + 10 f 0.8 _+ 29.3 * 0.9 k 27 * 0.3

1175 272 24 8.9 7.6 9.4 105 1.5

f. + & k + f * *

112 68 5 1.6 8.0 3.8 35 0.7

Combined hyperlipidemia (N = 5) Dr/Pl

=‘r%

0.25 0.85 1.29 0.18 2.53 0.47 0.98 0.47

213 245 38 1.0 8.2 4.4 114 0.6

Placebo + f + f + f. + k

163 40 11 0.8 2.8 2.0 32 0.2

449 249 33 2.6 7.5 7.2 116 1.0

& k * + * If: k *

Control (N = 8)

Dr/Pl 202 31 10 1.5 2.0 3.2 25 0.3

0.47 0.98 1.16 0.39 1.11 0.61 0.98 0.65

106 It 137 * 40 + 0.6 k 5.2 + 2+1 59 * 0.61 f

49 21 4 0.4 3.5 12 0.24

All values mg/dl except VLDL (u/dl) and protein ratio.

brozil recipients, plasma TG was 68% lower, HDL cholesterol was 25% higher, apoE was 47% lower, and VLDL was 77% lower, all significant differences compared with placebo. Differences in total cholesterol, apoB, and Lp[a] were not statistically significant. The difference between the ‘drug and placebo groups in plasma TG concentration was much greater in patients with HTG (Table 2): 75% lower in patients with HTG compared with 53% lower in patients with CHLP. The effect of drug on total cholesterol concentration was also greater in this group: total cholesterol 15% lower in HTG patients, compared with 2% lower in patients with CHLP. After gemfibrozil treatment, plasma concentrations of apoE declined by 53% in HTG and 39% in CHLP. Although there was a substantial reduction in plasma apoE concentrations in gemfibrozil-treated subjects, they were still significantly elevated compared to normal subjects (Table 2). 3.2. Composition and protein content of VLDL Data on VLDL composition after 8 weeks of treatment by gemfibrozil or placebo are shown for all patients in Table 3 and by hyperlipidemia classification in Table 4. For all subjects, VLDL protein was a significant 26% higher in drug recipients; there were no significant differences in phospholipid, TG, free cholesterol, or cholesteryl ester. HPLC determination of the apoC composition of VLDL showed apoC-II to be significantly higher

(19%) and apoC-III to be significantly lower (9%) in drug recipients, with no significant difference in apoC-I content of VLDL between all drug and placebo recipients. The apoE/apoB molar ratio in VLDL was a significant 48% lower with drug therapy in all patients. Total protein in VLDL was 39% higher in HTG patients and 12% higher in CHLP patients for drug versus placebo (Table 4). The drug-versus-placebo difference for apoC-II was greater in HTG patients ( + 29%) than in CHLP patients ( + 11%); the two hyperlipidemia groups were comparable in treatment effect on apoC-I and apoC-III content in VLDL. The apoE/apoB ratio was 65% lower in HTG drug recipients than in HTG placebo recipients; the corresponding difference in CHLP was - 12%. 3.3. VLDL ApoE fractions For all 10 patients, lower VLDL fraction concentrations (Table 3) in the drug group paralleled the significant 77% lower total VLDL concentration (Table 1): VLDL, (apoE rich and heparin binding), VLDLNRml, and VLDL,,-, (both apoE poor and not heparin binding) were 73%, 74%, and 82% lower than in placebo recipients, all significant differences. The decreasewith drug was also seen in both hyperlipidemia groups (Tables 2 and 4): in patients with HTG, - 82%, - 78%, - 77%, and - 88% for total VLDL, VLDLNRMI, VLDL,,,, and VLDL, and - 61%, - 37%, - -

C.-Y.

Yang et al. AlAtherosclerosis

126 (1996)

III

105--l 16

Table 3 VLDL composition and ApoE fraction concentrations in hyperlipidemic subjects (N = 10) according to treatment assignment ~______ Placebo P-value Drug/placebo Drug Proteind Phospholipid” TG” Free cholesterol” Cholesteryl ester“ ApoC? ApoC-Ilb ApoC-III ApoC-lII/ApoC-II ApoC-III/ApoC-I ApoE/ApoB VLDL,,., (u/dl) VLDL,,.2 (wdl) VLDL, (u/dI)

10.3 14.8 60.5 3.4 I I.1 :!6.7 :!5.8 47.3 1.9 1.8 2.4 36 53 42

+ * * f 5 2 f k k * * f f +

0.7 1.0 3.0 0.8 2.3 3.8 4.5 6.4 0.6 0.4 0.7 35 30 25

8.2 14.3 61.7 3.3 12.5 26.3 21.7 51.9 2.5 2.0 4.6 132 203 240

* ) i * & & k i & k k f f )

1.8 I.0 3.4 0.3 2.9 4.0 3.8 6.5 0.8 0.6 2.7 117 I17 144

0.0120 NS NS NS NS NS 0.0142 0.0002 0.0082 0.0292 0.0250 0.023 0.0012 0.0016

1.26 1.04 0.9X 1.03 0.89 1.02 1.19 0.91 0.76 0.90 . 0.52 0.27 0.26 0.18

“Percentage of total particle weight (lipid + protein = lOOX) “Percentage of total ApoC content. NS. not significant ( P > 0.05).

65%, and -- 66% in patients with CHLP. Distribution of VLDL fractions according to hyperlipidemia group and treatment assignment is shown in Table 5. Even without drug treatments, the mean abundance of VLDL,, was lower in both hyperlipidemia groups than in control subjects (41.8% of VLDlL in HTG patients, 44.0% in CHLP patients, 54.4% in controls). Drug treatment reduced the proportion for men, more substantially in HTG patients (to 27.6%) than in CHLP patients (to 38.6%). With drug treatment in HTG, the proportion of VLDL accounted for by the apoE-poor fractions increased for both fractions, most substantially for VLDL,,-,. With drug treatment in CHLP, the proportion of VLDL,,-, decreasedslightly, so that all the redistribution from VLDL, was to the VLDL,,., fraction.

terified cholesterol content. LDL diameter as calculated from LDL chemical composition was greater in drug recipients (0.44 f 0.48 nm), although the increase was not statistically significant. LDL size as determined by gel electrophoresis did not change with treatment. LDL heterogeneity varied among individuals but after gemfibrozil treatment, LDL particles tended to more homogeneous. As described in Section 2, LDL pattern was also characterized by an LDL score [32]. A trend toward a lower LDL score (larger particles) with gemfibrozil treatment was observed in patients with HTG (8.4%) and in all patients (6.1%) (data not shown); however, these differences were not statistically significant.

4. Discussion 3.4. LDL compo,rition and diameter

Assessment of LDL composition after drug or placebo treatment showed significantly lower protein and TG content ( - 4% and - 36%) and significantly higher cholesteryl ester content ( + 8%) in drug recipients (Table 6). Treatment did not cause L,DL to vary in phospholipid or unes-

Gemfibrozil is a widely used, potent lipid-lowering drug. It is highly effective in reducing plasma TG and VLDL levels and in increasing HDL cholesterol in HTG, type III hyperlipidemia, and CHLP [Zl-23,34-381. It has been reported that gemfibrozil increases plasma level of HDL by stimulating its synthesis and inducing transport of HDL by increasing tissue cholesterol

112

C.- Y. Yang et al. / Atherosclerosis 126 (1996) 10% 116

Table 4 VLDL composition and ApoE fraction concentration according to treatment assignment and hyperlipidemia classification Hypertriglyceridemia (N = 5) Drug Protein PL TG FC CE ApoC-I ApoC-II ApoC-III ApoC-III/ApoC-II ApoC-III/ApoC-I ApoE/ApoB VLDL,,., (u/dl) VLDL,,., WI) VLDL, (u/dl)

10.4 14.9 60 3.1 11.5 25.7 24.7 49.6 2.2 1.9 2.2 49 68 45

* & + f + + k + k * & f + k

0.8 1.3 3.4 0.9 2.4 3.0 6.2 8.3 0.8 0.5 0.9 40 26 24

Combined hyperlipidemia (N = 5)

Placebo

Dr/Pl

Drug

7.5 * 1.0 13.7 + 1.0 64 k 2.7 3.1 * 0.2 11.5 k 1.6 25.9 k 5.9 19.2 + 2.8 54.9 + 8.2 3.0 * 0.9 2.2 + 0.8 6.2 k 3.1 225 k 94 297 + 60 365 k 55

1.39 1.09 0.94 1.00 1.00 0.99 1.29 0.90 0.73 0.86 0.35 0.22 0.23 0.12

10.1 14.7 61.0 3.6 10.7 28.1 26.9 45.0 1.7 1.7 2.6 24 38 39

* + + + f f f + + + + + * *

0.4 0.9 3.2 0.7 2.4 4.8 2.0 3.3 0.1 0.4 0.5 28 28 30

Placebo

Dr/Pl

9.0 $- 2.2 14.9 & 0.8 59.2 + 1.6 3.4 If 0.4 13.5 + 3.6 26.7 + 1.8 24.2 k 3.0 48.9 & 2.3 2.1 * 0.4 1.8 It 0.2 3.0 + 0.4 38 * 20 108 + 70 115 * 66

1.12 0.99 1.03 1.06 0.79 1.05 1.11 0.92 0.81 0.94 0.88 0.63 0.35 0.34

All values mg/dl except protein ratio and VLDL (u/dl).

removal [6]. In our hyperlipidemic patients, treatment with gemfibrozil was associated with significantly higher HDL-cholesterol, significantly lower plasma TG and VLDL, and changes in the VLDL content of apoE. The effects of gemfibrozil on HTG subjects are more efficient than on CHLP subjects. In this study, the average Lp[a] concentration increased insignificantly after gemfibrozil treatment, but remained within normal limits ( < 30 mg/dl) [39]. The mechanism by which gemfibrozil reduces plasma TG concentration is believed to involve decreased VLDL synthesis and increased catabolism of VLDL by lipoprotein lipase (LPL) Table 5 Distribution of VLDL ApoE fractions according to hyperlipidemia classification and treatment assignment VLDL,,-, HTG Placebo 24.6 (N = 5) Drug 28.5 Placebo 14.6 CHLP (N = 5) 23.8 Dws 10.5 Control (N = 8) All values in percent.

VLDL,,-,

VLDL,

33.6 43.9 41.4 37.6 35.1

41.8 21.6 44.0 38.6 54.4

and hepatic lipase. Kesaniemi and Grundy [7] have reported that the fractional catabolic rate of VLDL-TG is increased and the synthesis of VLDL-apoB and TG is decreased after gemfibrozil treatment. Gemfibrozil reduces VLDL and apoB production in the liver and simultaneously stimulates LPL activity, which results in increased clearance of TG-rich particles. Experiments in fat-fed animals indicate that remnants of TG-rich lipoproteins are atherogenic [40]. Several case-control studies have indicated that post-prandial lipemia is a significant CAD risk factor [41-431. Other studies have reported that post-prandial TG-rich lipoproteins are metabolized more efficiently after gemfibrozil treatment [44]. The effect of gemfibrozil on post-prandial lipoprotein levels parallels its effect on fasting levels [45]. The C apolipoproteins in human plasma are an integral part of TG-rich VLDL and play a central role in VLDL metabolism initiated by the LPL. ApoC-II activates LPL 1461;when in vitro assay systems contain LPL but no apoC-II, no triglyceride hydrolysis occurs. In addition, apoC-II may play an inhibitory role in the hepatic uptake of TG-rich lipoproteins [47]. It has been suggested that apoC-III inhibits the premature clearance of

C.-Y. Yang et al. / Atherosclerosis

113

126 (19V6) IOC-116

Table 6 LDL composition and diameter in hyperlipidemic subjects (N = 10) according to treatment assignment Placebo

Drug Protein (‘x) Phospholipid (XI) TG (“4) Free cholesterol (‘%I) Cholesteryl ester” (“/I) Calculated diameter” (nm) Diameter by gel (nm)

20.1 k 2.7 15.9 + 4.9 f 4.8 + 54.3 k 20.71 1 24.93 i:

2.0 2.1 1.3 3.2 1.0 0.6

20.9 + 16.4 f 7.1 * 4.6 k 50.4 i 20.32 k 24.83 k

2.6 1.4 3.2 1.0 3.8 1.0 0.6

Drug/placebo

P-value

0.96 0.97 0.64 1.04

0.046 NS 0.027 NS 0.009 0.042 NS

1.08 I .02 1.01

“Cholesteryl ester was calculated as the difference between total cholesterol and free cholesterol concentrations, multiplied by 1.67 to correct for the difference between cholesterol and cholesteryl ester molecular weights. bDiameters calculated from compositional data according to the method of Luzzati et al. [33]. NS, not,significant (P 2 0.05). -

TG-rich VLDL until sufficient TG has been hydrolyzed from the core [4?]. ApoC-III also inhibits LPL activity [48], raising the possibility that the apoC-III/apoC-II ratio of a TG-rich lipoprotein plays a critical role in regulating TG hydrolysis. Overproduction of apoC-III in transgenie mice results in HTG with a highly significant linear relation between plasma apoC-III and plasma TG concentrations [49]. In the present study, the ratio of apoC-III to apoC-II in VLDL was significantly lower in drug recipients than in placebo recipients (Table .3). These results are consistent with the view that gemfibrozil increases the catabolism of TG [7,50]. ApoE plays a central role in the catabolism of VLDL. It has been postulated that functional apoE is required for the lipolytic conversion of VLDL into LDL to proceed normally [51] and that the amount of apoE, and in particular the location of apol3 on the particle, is important for the VLDL, receptor interaction [52]. A partial precursor-product relation between apoE-poor and apoE-rich VLDL has been studied in hyperlipidemic subjects [53]. The apoE-rich VLDL is transformed into intermediate-density lipoprotein (IDL) [54] and is more effective than apoE-poor VLDL in competing with LDL for binding to the B/E receptor of human fibroblasts [ll]. VLDL receptor only binds the apoE-containing lipoproteins VLDL, IDL, p-VLDL [55,56]. In normolipidemic subjects, VLDL particles are quantitively converted from TG-rich lipoproteins

into cholesteryl ester-rich LDL [57,58]. In contrast to studies in normolipidemic subjects, VLDL kinetic studies in HTG subjects have demonstrated a marked incomplete conversion of VLDL to LDL [59-611. Recently, Evans et al. [14] demonstrated that the apoE-poor VLDL subfraction is resistant to lipolysis by LPL compared to its apoE-rich counterpart. In the present study, levels of total VLDL and all VLDL subfractions decreased significantly with drug treatment (Tables l-3). The mean total VLDL, (apoE-rich) significantly decreasedfrom 41.7% to 32.1%. This result differs from that of normal subjects in which apoE-rich VLDL accounts for a higher proportion of total VLDL than apoE-poor fractions [ 131. The explanation for this result may be that there are still substantial amounts of VLDL,,., and VLDL,,e, (apoE-poor VLDL which are resistant to lypolysis) existing after drug treatment. A few studies have reported that LDL particle size tends to normalize with administration of fibrates [62]; a preponderance of small, dense (protein-rich) LDL particles, which is associated with hypertriglyceridemia, correlates with increased risk for myocardial infarction [63]. In our study, there was only slight evidence for improved LDL subclass distribution. LDL diameter as calculated from LDL chemical composition was 2% greater with gemfibrozil administration compared with placebo, and LDL score as developed by Campos et al. [32] in which the score rises with an increased proportion of small, dense LDL, was

114

C.- Y. Yang et al. / Atherosclerosis

slightly but insignificantly lower after gemfibrozil treatment. There was no difference in the LDL diameters between the treatment groups as determined by gel electrophoresis. The effect of gemfibrozil treatment on LDL particle size may have been demonstrated with a larger number of patients. In summary, this study reaffirms that slow-release gemfibrozil (Lopid SR) is a potent lipid regulating drug that significantly lowers plasma TG and VLDL. In the present study, after gemfibrozil therapy, TG decreased an average of 68% and HDL-C increased an average of 25%. ApoE and VLDL also showed significant decreases.After gemfibrozil, the content of apoC-II and apoCIII in VLDL changed in opposite directions. Consequently, the C-III/C-II (the percentage ratio in VLDL) decreased. VLDL and its subfractions were reduced to a greater extent in the HTG compared to the CHLP patients. VLDL subfraction redistribution resulted in an almost 10% decrease in VLDL, (from 41.7% to 32.1%). These results suggest that Lopid SR therapy decreased VLDL synthesis, increased catabolism of VLDL by LPL and hepatic lipase, and altered the apoprotein composition of VLDL in a way that may favor the formation of LDL, assist VLDL receptor mediated clearance of VLDL subfraction, and reduce TG-rich lipoprotein particles.

Acknowledgements This work was supported by National Institutes of Health Grant HL-27341, by a research grant from Warner-Lambert/Parke Davis Co., and by grants from The Methodist Hospital Foundation and The DeBakey Heart Center Fund. We are grateful to John Gaubetz, M.S., for generously providing the lipoprotein[a] standard, to Juan G. Guevara, Ph.D., for assistance in performing the experiments with native gel electrophoresis and latex beads, to Suzanne Simpson and Joel D. Morrisett, Ph.D. for editing the manuscript, and to Renee Wells for assistance in preparing the manuscript.

126 (1996) 1OS- I1 6

References [I] Castelli WP. The triglyceride issue: a view from Framingham. Am Heart .I 1986;112:432. [2] Brunner D, Altman S, Loebl K, Schwartz S, Levin S. Serum cholesterol and triglyceride in patients suffering from ischemic heart disease and in healthy subjects. Atherosclerosis 1977;28:197. [3] Lewis B, Chait A, Oakley CMO, Wootton IDP, Krikler DM, Onitiri A, Sigurdsson G, February A. Serum lipoprotein abnormalities in patients with ischemic heart disease: comparisons with a control population. Br Med J 1974;3:489. [4] Goldstein JL, Hazzard WR, Schrott HG, Bierman EL, Motsky AG, Levenski MJ, Campbell ED. Hyperlipidemia in coronary heart disease. I. Lipid levels in 500 survivors of myocardial infarction. J Clin Invest 1973;52:1533. [S] Hodis HN, Mack WJ. Triglyceride-rich lipoproteins and the progression of coronary artery disease. Curr Opin Lipid 1995;6:209. [6] Saku KP, Gartside S, Hynd BA, Kashyap ML. Mechanism of action of gemfibrozil on lipoprotein metabolism. J Clin Invest 1985;75:1702. [7] Kesaniemi YA, Grundy SM. Influence of gemfibrozil and clofibrate on metabolism of cholesterol and plasma triglycerides in man. J Am Med Assoc 1984;251:2241. [8] Schonfeld G. The effects of finbrates on lipoprotein and nemostatic coronary risk factors, Atherosclerosis 1994;111:161. [9] Shelbourne FA, Quarfordt SH. The interaction of heparin with an apoprotein of human very low density lipoproteins. J Clin Invest 1977;60:944. [lo] Trezzi E, Calvi C, Roma P, Catapano AL. Subfraction of human very low density lipoprotein by heparin-Sepharose affinity chromatography. J Lipid Res 1983;24:790. [I l] Fielding PE, Fielding CJ. An apo-E-free very low density lipoprotein enriched in phosphatidylethanolamine in human plasma. J Biol Chem 1986;261:5233. [12] Yamada N, Shames DM, Shelbourne JB, Have1 RJ. Metabolism of lipoproteins containing apolipoprotein B100 in blood plasma of rabbits: heterogeneity related to the presence of apolipoprotein E. Proc Nat1 Acad Sci USA 1986;83:3479. [I31 Yang CY, Gu ZW, Valentinova N, Pownall HJ, Lee B, Yang M, Xie Y-H, Guyton JR, Vlasik TN, Fruchart JC, Gotto AM Jr. Human very low density lipoprotein structure: interaction of the C apolipoproteins with apolipoprotein B-100. J Lipid Res 1993;34:1311. [14] Evans AJ, Wolfe BM, Strong WLP, Huff MW. Reduced lipolysis of large apoE-poor very low density lipoprotein subfractions from type IV hypertriglyceridemic subjects in vitro and in vivo. Metabolism 1993;42:105- 115. [15] Gomez-Coronado D, Saez GT, Lasuncion MA, Herrera E. Different hydrolytic efficiencies of adipose tissue lipoprotein lipase on very low density lipoprotein subfractions separated by heparin-Sepharose chromatography. Biochim Biophys Acta 1993;1167:70&71.

C.-Y. Yang et al. 1 Atherosclerosis

P61Jingami H, Yamamoto T. The VLDL receptor: wayward

brother of the LDL receptor. Curr Opin Lipidol 1995;6:104. P71 Beisiegel U. Receptors for triglyceride-rich lipoproteins and their role in lipoprotein metabolism. Curr Opin Lipido1 1995;6:117. Deckelbaum J, Granot E, Oschry Y, Rose L, Eisenberg, S. Plasma triglyceride determines structure composition in low and high density lipoproteins. Arteriosclerosis 1984;4:225-232. [I91 Eisenberg S, Gavish D, Oschry Y, Fainaru M, Deckel-

baum RJ. Abnormalities in very low, low, and high density lipoproteins in hypertriglyceridemia: reversal toward normal with bezafibrate treatment. J Clin Invest 1984;74:470-482. PO1Franceschini G. Sitori M, Gianfrancheschini G, Frosi T, Montanari G, Sitori CO. Reversable increase of the apo CII/aoi Cl11 ratio in the very low density lipoproteins after fenofibrate treatment in hypertriglyceridemic patients. Artery 1985;12:363-381. Pll Manttari M, Koskinen P, Manninen V, Huttunen JK, Frick M, Nikkila E. The effect of gemfibrozil on the concentration and composition of serum lipoproteins. A controlled study with special reference to initial triglyceride levels. Atherosclerosis 1990;81:11-17. PI Tsai MY, Yuan J, Hunninghake DB. Effect of gemhbrozil on composition of lipoproteins and distribution of LDL subspecies. Atherosclerosis 1992;95:35. ~231Dachet C, Cavallero E, Martin C, Girardot G, Jacotot B. Effect of gemfibrozil on very low density and low density lipoprotein subfractions in hypertriglyceridemic patients. Atherosclerosis 1995;113:1. 1241 Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenal reagent. J Biol Chem 1951;193:265. 1251 Stanley CR Jr, Weisgraber KH, Mahley RW. Isolation and characte-rizdtion of apoE. Methods Enzymol 1986;128:273. WI Yang CY, Xie YH, Yang M, Quion JA, Gotto AM Jr. ELISA quantitation of apohpoproteins in plasma lipoprotein fractions: ApoE in apoB-containing lipoproteins (LpB:E) and apoB in apoE-containing lipoproteins (LpE:B). J Protein Chem 19$5;14:503. ~271 Holmquist L, Carlson K, Carlson LA. Comparison between the use of isopropanol and tetramethylurea for the solubilization :and quantitaticn of human serum very low density apolipoproteins. Anal Biochem 1978;88:457. 1281Fidge NH, Nestel PJ. Metabolism of apolipoprotein C. Methods Enzymol 1986;129:443. 1291Yang CY, Yang MT, Pownall HJ, Gotto AM Jr. Primary structure of rabbit apolipoprotein A-I-high liquid performance chromatography, PTC-amino acid analysis, and microsequencing. In: Advanced Methods in Protein Microsequence Analysis. Berlin Springer, 1986;320. [301 Hughes TA, Moore MA, Neame P, Medley F, Chung BH. Rapid quantitative apolipoprotein analysis by gradient ultracentrifugation and reversed phase high performance liquid chromatography. J Lipid Res 1983:29:363.

126 (1996) 105-I 16

115

[3l] Krauss RM, Burke DJ. Identification of multiple subclasses of plasma LDL in normal humans. J Lipid Res 1982;23:97. [32] Campos H, Blijlevens E, McNamdra JR, Ordovas JM, Posner BM, Wilson PWF, Castelli WP, Schaefer EJ. LDL particle size distribution results from the Framingham Offspring Study. Arterioscler Thromb 1992;12:1410. [33] Luzzati U, Tardieu A, Aggerbeck LP. Structure of serum low-density lipoprotein. 1. A solution X-my scattering study of a hyperlipidemic monkey low-density lipoprotein J Mol Biol 1979;1:435. [341 Lewis JE. Long-term use of gemfibrozil (Lopid) in the treatment of dyslipidemia. Angiology 1982;33:603. [351 Davignon J. Fibrates: a review of important issues and recent findings. Can J Cardiol 1994.10suppl B:6lB-71B. [361 Kashyap ML. The effect of gemfibrozil on plasma lipids and lipoproteins in man. Vast Med 1984;2:16. [371 Konttinen A, Kuisma I, Kuisma R, Ralli S, Pohjola. Ojala K. The effect of gemfibrozil on serum lipids in diabetic patients. Ann Clin Res 1979;11:240. [381 Fenderson RW, Deutsch S, Menachemi E, Chin B, Samuel P. Effect of gemfibrozil on serum lipids in man. Angiology I982;33:581. [391 Rodriguez CR, Seman LJ, Ordovas JM, Jenner J, Genest MSJ Jr, Wilson PWF, Schaefer EJ. Lipoprotein(a) and coronary heart disease.Chem Phys Lipids 1994:67/68:389. [401 Mahley RW. Atherogenic lipoproteins and coronary artery disease: concepts derived from recent advances in cellular and molecular biology. Circulation 1985:72:943. [411 Simpson HS, Williamson CM, Olivecrona T, Pringle S, Maclean J, Lorimer AR, Bonnefous R, Bogaievsky Y, Packard CJ, Shepherd J. Postprandial lipemia. fenofibrate and coronary artery disease.Atherosclerosis 1990385:193202. [421 Groot PHE, van Stiphout WAHJ, Krauss XH, Jansen H,

van To1 A, van Ramshorst E. Chin-On S. Hofman A, Cresswell SR, Havekes L! POStprdndidt lipoprotein metabolism in normolipidemic men with and without coronary artery disease. Arterioscler Thromb 1991;11:6533662. I431 Patsch J, Miesenbock G, Hopferwieser T, Muhberger V, Knapp E, Dunn JK, Gotto AM Jr, Patsch W. Relation of triglyceride metabolism and coronary artery disease. Studies in the postprandial state. Arterioscler Thromb 1992;12:1336-1345. [441 Syvanne M, Vuorinenmarkkola H, Hilden H, Taskinen MR. Gemfibrozil reduces postprandial lipemia in non-insulin-dependent diabetes mellitus. Arteriosclerosis Thromb 1993:13:286. [451 Jones PJ, Pownall HJ, Patsch W, Herd JA, Farmer JA. Payton-Ross C, Kimball KT, Gotto AM, Morrisett JD. Effect of gemfibrozil on levels of lipoprotein[a] in type II hyperlipoproteinemic subjects. J Lipid Res, in press. [461 Rosa JC. Levy RI, Herbert P, Lux SE, Fredrickson DS. A specific apoprotein activator for lipoprotein lipase. Biochem Biophys Res Commun 1970;41:57.

116

C-Y.

Yang et al. / Atherosclerosis

[47] Windler E, Chao Y, Have1 RJ. Relation of the hepatic uptake of triglyceride-rich lipoproteins in the rat. J Biol Chem 1980;255:8303. [48] Have1 RJ, Shore VG, Shore B, Bier DM. Role of specific glycopeptides of human serum lipoproteins in the activation of lipoprotein lipase. Circ Res 1970,27:595. [49] Ito Y, Azrolan N, O’Connell A, Walsh A, Breslow JL. Hypertriglyceridemia as a result of human apoC-III gene expression in transgenic mice. Science 1990;249:790. [50] Nikkila EA, Ylikahri R, Huttunen JK. Gemfibrozil: Effect on serum lipids, lipoproteins, post heparin plasma lipase activities and glucose tolerance in primary hypertriglyceridemia. Proc R Sot Med 1976;69(suppl 2):58. [51] Have1 RJ. The formation of LDL: mechanisms and regulation. J Lipid Res 1984;25:1570. [52] Shimano H, Namba Y, Ohsuga J, Kawamura M, Yamamoto K, Shimada M, Gotoda T, Harada K, Yazaki Y, Yamada N. Secretion-recapture process of apolipoprotein E in hepatic uptake of chylomicron remnants in transgenic mice. J Clin Invest 1994;93:2215-2223. [53] Barrett PHR, Baker N, Nestel PJ. Model development to describe the heterogeneous kinetics of apolipoprotein B and triglyceride in hypertriglyceridemic subjects. J Lipid Res 1991;32:743. [54] Nestel P, Billington T, Tada N, Nugent P, Fidge. Heterogeneity of very-low density lipoprotein metabolism in hyperlipidemic subjects. Clin Metab Exp 1983;32:810;41:57. [55] Beisiegel U, Weber W, Ihrke G, Herz J, Stanley K. The LDL-receptor-related protein, LRP, is an apolipoprotein E binding protein. Nature 1989;341:162-164.

126 (1996) 105-I 16

[56] Sakai J, Hishino A, Takahashi S, Miura Y, lshii H, Suzuki H, Kawarabayasi Y, Yamamoto T. Structure, chromosome location, and expression of the human very low density lipoprotein receptor gene. J Biol Chem 1994;269:2173-2181. [57] Eisenberg S. Preferential enrichment of large-sized very low density lipoprotein populations with transferred cholesteryl esters. J Lipid Res 1985;26:487-494. [58] Olivecrona T, Bengtsson-Olivecrona G. Lipoprotein lipase and hepatic lipase. Curr Opin Lipidol 1990;1:222230. [59] Sigurdsson G, Nicoll A, Lewis B. Metabolism of very low density lipoprotein in hyperlipidaemia. Studies of apolipoprotein B kinetics in man. Eur J Clin Invest 1976;61:167-177. [60] Sheppard J, Packard CJ. Metabolic heterogeneity in Am Heart J very low density lipoproteins. 1987;113:503-508. [61] Huff M, Breckenridge W, Strong W. Metabolism of apolipoproteins C-II, C-III and B in hypertriglyceridemic men: changes after heparin-induced lipolysis. Arteriosclerosis 1988;8:471-479. [62] Kleinman Y, Eisenberg S, Oschry Y, Gavish D, Stein 0, Stein Y. Defective metabolism of hypertriglyceridemic low-density lipoprotein in cultured human skin fibroblasts: normalization with bezafibrate therapy. J Clin Invest 1985;75:1796. [63] Slyper AH. Low-density lipoprotein density and atherosclerosis: unravelling the connection. J Am Med Assoc 1994;272:305.