Enhanced Cholesteryl Ester Transfer Protein Activities and Abnormalities of High Density Lipoproteins in Familial Hypercholesterolemia

Summary Cholesteryl ester transfer protein may play a role in the cholesteryl ester metabolism between high density lipoproteins (HDL) and apolipoprotein B-containing lipoproteins. To investigate relationship between H D L and cholesteryl ester transfer protein (CETP) activity in the development of atherosclerosis, the present study has focused on CETP activity in the patients with familial hypercholesterolemia (GH). HDL-C and HDL-C/apo A-I mass ratio in heterozygous F H were lower than those in normolipidemic controls. There was a 2-fold increase in total CETP activity in incubated FH serum compared with normolipidemic controls. Assays for CETP activity in the lipoprotein deficient serum (d > 1.215 g/ml) were carried out by measuring the transfer of radioactive cholesteryl ester from HDL (1.125 < d < 1.21 g/ml) to LDL (1.019 < d < 1.060 g/ml). CETP activities in heterozygous FH (79 + 4 nmol/ml/h) was significantly higher than those in normolipidemic controls (54 + 6 nmol/ml/h). The increased total cholesteryl ester transfer mainly results from increased CETP activity in the d > 1.215 g/ml, possibly reflecting an increase in CETP mass in serum. Increased CETP activity in the d > 1.215 g/ml was correlated positively with IDL-cholesterol/triglyceride mass ratio (r = 0.496, p < 0 . 0 1 ) , and negatively with HDLcholesterol/apo A-I mass ratio ( r = — 0.334, p < 0.05). These results indicate that the enhanced CETP activities may contribute to increase risk for developing atherosclerosis in F H by changing the distribution of cholesteryl ester in serum lipoproteins. Key words Familial Hypercholesterolemia — Cholesteryl Ester Transfer Protein Activity — High Density Lipoprotein (HDL) - H D L Cholesterol/Apolipoprotein A-I Mass Ratio — Intermediate Density Lipoprotein (IDL) Cholesterol/Triglyceride Mass Ratio

Introduction Most prospective epidemiologic studies have found that low level of high density lipoprotein (HDL) Horm. metab. Res. 24(1992) 284-288 © Georg Thieme Verlag Stuttgart-New York

cholesterol is a risk factor for coronary heart disease (Gordon and Rifkind 1989). Cholesteryl ester transfer protein (CETP) may regulate H D L levels as well as activities of lipoprotein lipase and hepatic lipase {Patsch, Prasad, Gotto and Patsch 1987). CETP is a hydrophobic glycoprotein with an apparent Mr of 74,000, containing 476 aminoacids (Tall 1986; Hesler, Swenson and Tall 1987; Drayna, Jarnagin, Mclean, Henzel, Kohr, Fielding and Lawn 1987). This protein facilitates exchange (homo-exchange process) and transfer (hetero-exchange process) of cholesteryl ester among lipoproteins. Incubation of human plasma results in a net transfer of cholesteryl ester from HDL to apo B-containing lipoproteins (i. e. very low density lipoprotein (VLDL), intermediate density lipoprotein (IDL), and low density lipoprotein (LDL)), with a reciprocal transfer of triglyceride from apo B-containing lipoproteins to HDL, causing alterations in composition and size of lipoproteins (Tall 1986). The importance of CETP in human HDL metabolism has been described in the studies of familial CETP deficiency (Koizumi, Mabuchi, Yoshimura, Michishita, Takeda, Itoh, Sakai, Sakai, Ueda and Takeda 1985; Brown, Inazu, Hesler, Agellon, Mann, Whitlock, Marcel, Milne, Koizumi, Mabuchi, Takeda and Tall 1989; Inazu, Brown, Hesler, Agellon, Koizumi, Takata, Maruhama, Mabuchi and Tall 1990). Homozygotes show markedly increased HDLcholesterol levels (175-248 mg/dl), including increased amount of HDL2 to HDL1 and apo E-enriched HDL; and a concurrent decrease in LDL cholesterol level. Studies of heterozygous CETP deficiency show that they have approximately half the amount of CETP activity and mass, an increased ratio of HDL2/HDL 3 and moderately increased HDL-cholesterol levels ( 6 6 + 1 5 (SD) mg/dl) (Inazu et al. 1990). Thus, low CETP activity appears to cause increase in HDL-cholesterol level. However, the relationship between high level of CETP activity and H D L level cross-sectional in other populations has not been established. In this study, we investigate the relationships between HDL-cholesterol, HDL-cholesterol/apo A-I mass ratio and CETP activity in patients with familial hypercholesterolemia (FH). There were abnormalities of HDL in the patients with heterozygous F H as well as homozygous FH (Takegoshi, Kametani, Oiwake, Imura, Takeuchi, Shinozaki, Nishino, Miyamoto and Mabuchi 1981; Breier, Lisch, Drexel, Sailer and Braunsteiner 1983; Goldberg, Fless, Baker, Joffe, Received: 6 Dec. 1990

Accepted: 10 Oct. 1991 after revision

Downloaded by: Georgetown University Medical Center. Copyrighted material.

A. Inazu, J. Koizumi, H. Mabuchi, K. Kajinami and R. Takeda The Second Department of Internal Medicine, School of Medicine, University of Kanazawa, Kanazawa, Japan

Cholesteryl Ester Transfer Activity in Famiolial Hypercholesterolemia

Material and Methods

Subjects FH were diagnosed according to following criteria: primary hypercholesterolemia (> 230 mg/dl) with tendon xanthomas and primary hypercholesterolemia with or without tendon xanthomas in a family with first-degree relative of FH {Mabuchi, Ito, Haba, Veda, Veda, Tatami, Kametani, Koizumi, Ohta, Miyamoto, Takeda and Takegoshi 1977). Overnight fasting (> 14 hours) serum was obtained from FH patients as well as normolipidemic controls (Cholesterol < 220 mg/dl and Triglyceride < 150 mg/dl). The patients were classified as having a IIa phenotype if serum triglyceride levels were less than 150 mg/dl, and a IIb phenotype if above 150 mg/dl. All subjects in this study were not taking any medication.

Lipoproteins Serum LDL (1.019 < d < 1.060 g/ml) of normolipidemic subject was isolated by sequential ultracentrifugation at 10 °C and 40,000 rpm in a Beckman Ti 50.3 rotor. [3H] cholesteryl ester (cholesteryl linoleate, [cholesteroly-l,2,6,7-3H(N)-j, NEN Research Products). Labeled HDL3 was prepared as described previously {Koizumi et al. 1985). The specific activity of labeled HDL3 was 45,000 dpm/ug cholesterol.

Total cholesteryl ester transfer assay incubated serum Incubation of fresh serum to determine the total activity of cholesteryl ester transfer in serum was begun within 2 hrs of venesection. 200 ul of serum were incubated with 20 ul of [3H]-HDLcholesteryl ester (10 ug cholesterol) at 37 °C in a metabolic shaker, in the presence of 2 mM dithionitrobenzonic acid (DTNB) {Tall, Sammett and Granot 1986). At each time point, an aliquot was removed and chilled on ice. The radioactivity in the supernatant was determined after precipitation of apo B-containing lipoproteins with heparin/MnCl 2 (final concentration: 1.3 mg/dl and 0.092 M, respectively). Total cholesteryl ester transfer activity was expressed as the transferred [ H]-cholesteryl ester (CE) radioactivities into the apo Bcontaining lipoproteins/total CE radioactivities in incubated serum.

Isotopic assay to measure cholesteryl ester transfer activity in the density > 1.215g/ml fraction Dialysed serum fraction of d > 1.215 g/ml was used as a source of cETP {Koizumi et al. 1985; Son and Zilversmit 1986). 5 ml of d =- 1.257 g/ml solution of potassium bromide was added to 1.0 ml of fresh serum (final density— 1.215 g/ml), and centrifugated at 10 °C and 40,000 rp for 48 hours. Bottom fraction of 2 ml was isolated by tube slicer, dialysed against 50 mM Tris-HCl; 150 mM NaCl, 2 mM EDTA, pH 7.4 at 4 °C for 24 hours. Assays for CETP activity were carried out using the modified reconstitutional system of the method of Tall et al. {Tall, Sammett and Granot 1986; Tall, Granot, Brocia, Tabas, Hester, Williams and Denke 1987). Since serum LDL is relatively homogenous compared to VLDL, LDL was used as acceptor lipoproteins in our reconstitutional assay. Lecithin: cholesterol acyltransferase (LCAT) activity in the d > 1.215 g/ml fraction was inhibited by incubation in the presence of 1.4 mM dithionitrobenzonic

285

acid (DTNB). The equal protein amount of the d > 1.215 g/ml fraction was incubated in a mixture containing [3H] cholesteryl esterlabeled HDL (10 ug of cholesterol) and unlabeled LDL (250 ug of cholesterol) in 500 ul of 50 mM Tris-HCl, 150 mM NaCl buffer pH 7.4, 2 mM EDTA for 2 hr at 37 °C. The reaction was stopped by chilling the tubes on ice for 30 min. CETP activity was estimated from the decrease in radioactivity in the HDL-containing supernatant after precipitation of apo B-containing lipoproteins with heparin-MnCh. This condition gave a stable and wide linear range of radiolabeled cholesteryl ester transfer from HDL3 to LDL (< 20 % transfer). All measurements were made in duplicate. The assay was found to be reproducible within a range ±10%.

Analytical methods Serum cholesterol and triglyceride levels were measured enzymatically {Allain, Poon, Chan, Richmond and Fu 1974). HDL-cholesterol was measured after heparin-calcium precipitation {Burstein and Scholnich 1973). Lipoproteins were separated by sequential ultracentrifugation into VLDL (d < 1.006 g/ml), IDL (d: 1.006 to 1.019 g/ml), and LDL (d: 1.019 to 1.063 g/ml). Apolipoprotein levels were measured by immuno-turbidimetry using goat antiserum against human apolipoproteins (Daiichi Chemical Co. Osaka, Japan) {Irish, Barrantes and Ledue 1987). Intra-assay and the between assays coefficient of variations (CV) at each apolipoprotein measurement in our laboratory were 3.78% and 2.96% for apo A-1,2.43 %> and 4.50%for apo A-II, 1.57% and 3.97% for apo B, 2.68% and 5.16% for apo C-II, 2.85% and 4.79% for apo C-III, and 1.95% and 5.08% for apo E (n = 10). Statistical calculations were performed with Student's t-test. Results The levels of serum lipids and apolipoproteins in 19 normolipidemic controls, 29 patients with heterozygous F H were shown in Table 1. IDL-C, LDL-C, apo B and apo E levels were significantly increased in heterozygous G H patients with phenotype Ila, and VLDL-C, IDL-C, LDL-C, apo b, apo C-II, apo C-III, and apo E levels were significantly

Table 1 Serum lipids and apolipoproteins in normolipidemic controls and patients with familial hypercholesterolemia. Normolipidemic Controls Age Sex (M/F) BMI (kg/m2) Cholesterol Triglyceride Phospholipid VLDL-C IDL-C LDL-C HDL-C Apo A-l Apo A-II ApoB Apo C-II Apo C-III Apo E

41+5 13/6 21.8 ±0.7 180 ± 5 91 ± 6 197 ± 6 13±3 6±1 94±4 48±3 121 ± 5 30 + 2 29±3 3.4 ±0.2 7.4 ±0.5 3.5 ±0.3

FH-hetero Ila

Ilb

47±3 8/9 23.0 ±0.8 334 ± 1 1 * * * 96±7 270±9*** 13±1 10±2** 227 ± 1 1 * * * 39 ±3* 109 ± 5 28 ± 1 162±7*** 3.1 ±0.3 7.3 ±0.5 5.0 ±0.3**

56 ±3* 5/7 27.5 ±2.1** 334 ±27*** 229 ±29*** 274+16*** 32±5** 29±7** 209±21*** 36±2** 109 ± 8 28±2 178±15*** 4.9 ±0.3*** 11.1±1.1** 7.1 ±0.8**

Values given are mean±SE. All lipoprotein and apolipoprotein values are mg/dl. Significantly different from results for normolipidemic controls: * p < ; **p 1.215 g/ml using a reconstitutional system, which reflects total CETP mass of whole serum (Quinet, Agellon, Kroon, Marcel, Lee, Whitlock and Tall 1990). CETP activities in heterozygous FH (79 + 4 nmol/ml/h) were significantly higher than those in normolipidemic controls (54 + 6) (p < 0.001). There were no significant differences in CETP activities between heterozygous FH with phenotype IIa and IIb (85 + 5 vs 73 + 8 nmol/ml/h) (Fig. 2). Correlational analysis between lipoproteins and CETP activity was performed among subjects within normolipidemic control group, heterozygous FH group and combined group of these two groups (Table 3). Increased CETP activity was correlated positively with VLDL-cholesterol (r = 0.458) in normolipidemic control group, and LDL-

Table 2 Comipositicin of HDL pari:icles in normolipidemic control and patients \ milial hyperchc)lesterolemia. with fa Hetero-FH N Ila lib TG content in HDL (% w/w)

9.0 ±0.8

6.9 ±0.5*

A-l/A-ll mass ratio

4.0 ±0.2

3.9 ±0.2

HDL-C/A-I mass ratio

0.39 ±0.01

0.35 ±0.01*

10.7 ±0.5

3.9 ±0.1

. 0.31 ±0.02*"

Values are mean + SE. Significantly different from normolipidemic controls (N): *p 1.215 g/ml in normolipidemic controls (N) and heterozygous familial hypercholesterolemia (FH) (Ila: closed circle, lib: open circle). Table 3 Pearson correlation coefficients for variables vs CETP activity determined in normolipidemic controls and patients with familial hypercholesterolemia. Normolipidemic control

Fig. 1 Total cholesteryl ester transfer activity from HDL to apo Bcontaining lipoproteins in incubated serum in patients with familial hypercholesterolemia (GH) and in normolipidemic controls (N). The values shown are the mean±SE from normolipidemic controls (n = 4) and the patients with heterozygous GH (n = 5). The difference of the means was significant at the 1 h and 3 h point (p < 0.03 and p < 0.02, respectively).

VLDL-C IDL-C LDL-C HDL-C apo A-I apo B apo E VLDL-C/TG IDL-C/TG LDL-C/TG HDL-C/TG apo A-l/A-ll HDL-C/apo A-I

-

-

Hetero-FH

Combined group

- 0.017 0.458* 5.38159E-03 0.147 0.149 0.261 0.165 - 0.033 0.134 0.078 0.312 0.086 0.221 0.070 0.157 0.558** 0.362* 0.386 0.265 0.017 0.282 0.131 0.291 0.148 - 0.277 0.282

0.246 0.230 0.493*** -0.189 - 0.060 0.460*** 0.465*** 0.300* 0.496*** 0.168 0.210 0.110 -0.334**

* p < 0 . 1 ; **p 1.21 g/ml fraction iso-

There was a marked decrease in HDLcholesterol/apo A-I mass ratio and increase in % TG content in HDL in heterozygous FH with phenotype IIb compared with phenotype Ha. These data suggest that an enhanced hetero-exchange of cholesteryl ester and triglyceride between HDL and VLDL (or IDL) rather than homo-exchange of cholesteryl esters occurred in phenotype IIb, and that this process results in a further reduction in HDL-cholesterol levels. Lipolysis of HDL triglyceride transferred from VLDL may generate much smaller HDL particles. Indeed, sequential activities of cETP and hepatic lipase play a role in interconversion of HDL2 to HDL3 in vitro (Tall 1990). Thus, neutral lipid

Downloaded by: Georgetown University Medical Center. Copyrighted material.

Fig. 3 IDL-cholesterol/triglyceride mass ratio plotted against the CETP activity measured in the d > 1.215 g/ml fraction. A positive correlation between the two variables was found for all groups shown (r = 0.496, n=36, p