Effects of probucol on low density lipoprotein removal and high density lipoprotein synthesis

203

Atheroscierosis, 38 (1981) 203-209 @ Elsevier/North-Holland Scientific Publishers, Ltd.

EFFECTS OF PROBUCOL ON LOW DENSITY LIPOPROTEIN REMOVAL AND HIGH DENSITY LIPOPROTEIN SYNTHESIS

PAUL J. NESTEL and TIMOTHY BILLINGTON Cardiovascular Metabolism and Nutrition Research Unit, Baker Medical Research Institute, Commercial Road, Melbourne 3181 (Australia) (Received 7 July, 1980) (Accepted 29 July, 1980)

Summary

The possible mechanisms of action of probucol on the metabolism of low density lipoprotein and of high density lipoprotein were studied in 5 hyperlipidaemic subjects. The kinetics of LDL-B protein and of HDL-AI protein were determined by 2-pool analysis of specific radioactivity-time curves after reinjection of 1311-labelled LDL and 12sI-labelled HDL at the end of placebo and treatment periods. Probucol increased the fractional removal rate of LDL in 4, an action probably linked to increased bile acid excretion which occurred in all 5 subjects. Nevertheless, the plasma cholesterol concentration fell significantly in only 3. The synthesis of HDL-AI protein fell substantially and was probably responsible for the consistent reduction in plasma apo-AI levels. Thus, probucol appears to enhance LDL-B protein removal and bile acid excretion but inhibits AI protein formation; all these effects may be determined in the small intestine. Key words:

Bile acid excretion -Lipoprotein

transport -ProbucoE

Introduction

Probucol is an established -cholesterol-lowering drug, being particularly effective when low density lipoproteins (LDL) are alone raised [l-3]. Triglyceride levels are not altered, but simultaneous reduction in high density lipoprotein (HDL) has been reported [1,3]. The only postulated mechanism of action in man is based on Miettinen’s observations of enhanced bile acid excretion [4]. This work was supported by grants from the National Heart Foundation Health and Medical Research Council.

of Australia and National

204

Although chemically quite distinct from the bile acid sequestering resins, probucol may therefore share their mode of action in lowering LDLcholesterol and, in particular, it is known that the drug is mostly not absorbed. In the present study, we have examined the drug’s influence on LDL metabolism more directly by determining the kinetic parameters of flux, removal and pool size, which can be obtained from studies of reinjected radiolabelled LDL. Because of the reduction in HDL concentration, HDL apoprotein kinetics were also determined. Experimental Methods Five hyperlipidaemic subjects were studied. Their relevant clinical characteristics are presented in Table 1. They were chosen because of previous failure to control their hyperlipidaemia with clofibrate, cholestyramine or a combination of both. They agreed to a single-blind trial of the drug versus placebo, each period lasting 8 weeks. During the final 2 weeks of each period LDL-B protein and HDL-AI protein kinetics were carried out. HDL studies were made in only 4 subjects. Bile acid excretion was measured at the same time. The placebo stldy preceded treatment with probucol, 500 mg twice daily. Blood samples for plasma lipids were taken 5 times at weekly intervals during the final 4 weeks of each period; lipoprotein lipids and plasma apo-AI concentrations were measured twice during the final 2 weeks. The subjects were studied as outpatients. Blood was taken for the isolation and radiolabelling of LDL and HDL with 1311and 12’I, respectively, as described previously [ 5,6]. An average of 75 /.&i was injected and 180 mg of potassium iodide was given daily in divided doses to inhibit thyroidal uptake of radioiodine. Samples of plasma for the specific radioactivity-time curves of LDL and HDL were obtained over the following 10 days for the calculations of LDL-B protein and HDL-AI protein kinetics. The methods have been described fully [ 5,6] : briefly 6 samples of plasma were obtained during the first 3 days after reinjection of lipoproteins and every second day thereafter; LDL and HDL were isolated by ultracentrifugation at densities of d 1.019-1.063 g/ml and 1.063-1.21 g/ml, respectively; LDL-B protein was precipitated with isopropanol; HDL-AI protein was isolated by polyacrylamide SDS gel electrophoresis [ 61. The specific activities of the isolated proteins were determined from radioassays and mass measurements of the separated proteins. Since both LDL-B protein and HDL-AI protein specific radioactivity timeTABLE

1

CLINICAL NO

DETAILS

Age

OF TEST SUBJECTS

SW

1 2 3 4 5

56 64 31 54 49

Weight

Lipoprotein phenotype

(kg)

(Yr) M F M F M

69 65 86 71 69

Combined hyperlipoproteinaemia Familial hypercholesterolaemi Familial hypercholesterolaemia Combined hyperlipoproteinaemia Combined hyperlipoproteinaemi

205

curves can be resolved into 2 exponential functions, the kinetics were analyzed by the 2-pool model of Gurpide et al. [7], as previously modified by us for apo-AI kinetics [6]. The analysis provides values for the flux (transport) through pool 1 (the rapidly equilibrating pool), the amount of protein in this pool and the fraction of this pool cleared per unit time. This model envisages 2 pools, exchanging freely, with input occurring predominantly into pool 1, the rapidly exchanging pool. Previous studies suggest that pool 1 is mainly within the plasma [6,8], LDL-B protein being derived through the intravascular catabolism of VLDL, and HDL-AI protein entering the plasma from the liver and within intestinal lipoproteins. The model is compatible with loss of material from both pools; it is possible to calculate the fraction of pool 1 transferring to pool 2 plus that leaving the system irreversibly (the fractional removal rate). Irreversible loss from pool 1 can be determined only if it is assumed that irreversible loss from pool 2 is negligible, an assumption that cannot be made until the nature of pool 2, regarded as being outside the plasma [ 681, is defined and until the sites of LDL and HDL removal are fully identified. During the final 8 days of each study period, subjects collected stools for measurements of bile acid excretion, the daily excretion rates being calculated from recoveries of chromium sesquioxide [9] taken during the collection period and for the preceding 10 days. The separation and quantification of acidic steroids by thin layer and gas chromatographic techniques have been described fully [lo]. Plasma cholesterol and triglyceride measurements were made by automatic enzymic techniques; plasma apo-AI was measured by electroimmunoassay [ll]; and HDLcholesterol was measured after precipitating the other lipoproteins with heparin-manganous chloride. To prevent losses, LDL-apo B was measured in the isolated lipoprotein without further washing (a procedure used for the specific activity estimations). Statistical evaluations were carried out by the paired t-test (control versus treatment). Results Table 2 shows the mean plasma cholesterol and triglyceride concentrations in each of the 5 subjects, for the placebo period (5 values) and probucol period (5 values). Plasma cholesterol concentrations only fell significantly with treatment in 3 subjects (P < 0.05 in each case; Nos. 1, 2 and 5), but the reduction for the group was not significant. Falls in both LDL- and HDLcholesterol contributed to the plasma cholesterol changes; significant falls in HDLcholesterol occurred for the group as a whole (P < 0.01); probucol treatment affected both HDLcholesterol and plasma apo-AI in every subject. Plasma triglyceride levels were not altered. Low density lipoprotein apo-B kinetics are shown in Table 3. The fractional removal rate was raised with treatment in 4 subjects, including the 3 in whom LDLcholesterol (and LDL apo-B) concentrations fell significantly. In the remaining 2 subjects (both overweight), treatment was associated with modest increases in LDL apo-B and LDL-cholesterol concentrations, despite the almost

206 TABLE 2 PLASMA LIPID CONCENTRATIONS Subject

a

Plasma cholesterol

Plasma triglyceride

HDL-cholesterol

I

I

I

II

(mgldl)

II

(m/W

II

(w/W

1 2 3 4 5

324 349 470 324 343

278 b 252 b 430 260 271 b

288 103 131 324 179

270 88 125 311 192

49 60 53 35 -

24 44 29 22 -

Mean + SEM

362 + 27

302 f 32

185 + 50

177247

40 f 5

29 f 6

a Plasma cholesterol and triglyceride concentrations are the means of 5 measurements during each period. I: placebo. II: probucol. HDL-cholesterol concentrations are the means of 2 measurements during final week of each period. b Significantly lower (P < 0.05) than mean concentration during placebo period.

doubling in the fractional removal rate in one (No. 3). Flux, or synthesis, did not change significantly with treatment: the 2 subjects (Nos. 3 and 4), whose LDL levels failed to fall, showing no consistent pattern in this respect. The change8 in the mass of apo-B in pool 1, the dimensions of which approximate to the plasma compartment, moved in the same direction as the LDL concentration. Thus, there was no uniform response to probucol, though removal was enhanced in 4 of the 5 subjects (not significantly for the whole group), but this resulted in a fall in LDL apo-B pool size and concentration in only 3. The findings for HDL-AI protein kinetics were consistent in all 4 subjects. The mass in pool 1 fell with probucol in proportion to the reductions in plasma apo-AI concentration (Table 4). These were very substantial in 2 subjects (Nos. 3 and 4) who, interestingly, showed small increases in LDL levels. The changes were clearly due to reductions in apo-AI synthesis, fractional removal rate8 remaining unaffected. With only 4 studies, mean differences in pool size and in flux gave P values of 0.05. Bile acid excretion rose with treatment in every subject, resulting in a significant increase for the group (P < 0.02; Table 5). Discussion This study is not so much an examination of the cholesterol-lowering potential of probucol, as an investigation of its modes of action. The numbers of subjects are, therefore, smaller than in a staightforward evaluation of lipid changes with treatment, which have previously shown that probucol is about as effective in lowering LDLcholesterol as clofibrate and the bile acid-sequestering resins [l-3]. Furthermore, the subjects had also been selected because their plasma cholesterol levels had failed to fall satisfactorily with other drugs. The reason for the reduction in LDL has been attributed to the effect of probucol on bile acid excretion [ 41. Although this study confirms Miettinen’s observations of increased bile acid excretion, this also occurred in the 2 sub-

128 f 1.47

Mean + SEM

Period I = placebo, Period II = probucol.

128 132 125 127

(mUdI)

I

79 f 21.9

95 128 70 24

II

AI KINETICS

Plasma ape-AI

LIPOPROTEIN

1 2 3 4

Subject

HIGH DENSITY

TABLE 4

Period I = placebo, Period II = probucol.

3.39 + 0.30

3.46 3.52 4.01 2.58

(&

I

Mass pool I

4.28 + 0.68

146 f 9.2

Mean + SEM

134 + 20.4

4.49 3.59 6.86 3.17 3.31

117 96 197 167 94

127 148 180 135 138

1 2 3 4 5

I

I (9)

Mass pool I

II

B KINETICS

LDL ape-B

LIPOPROTEIN-APO

(mr/dU

Subject

LOW DENSITY

TABLE 3

2.08 ?: 0.67

2.54 3.53 1.96 0.31

II

3.97 + 1.08

3.19 2.38 8.09 4.01 2.20

II

13.1 + 1.5

15.9 11.2 10.0 15.3

(mgFg/d)

I

Flux

16.5 f 1.4

20.1 12.4 14.4 18.0 17.4

(mg/kg/d)

I

Flux

7.1 + I.7

10.7 10.5 5.5 4.3

II

17.1 + 0.7

16.3 16.5 19.0 16.2 18.3

II

0.80 f 0.19

0.62 0.62 0.58 1.40

(d-I 1

I

0.77

0.62 0.44 0.56 1.47

II

+ 0.23

0.86 + 0.32

0.12 2.1 0.48 0.34 0.65

II

Fractional removal rate

0.46 f 0.05

0.50 0.46 0.29 0.60 0.46

(d-I)

I

Fractional removal rate

208 TABLE 5 BILE ACID EXCRETION Subject

Acid Steroids I

II

(w/kg/d) 1 2 3 4 5

2.9 0.9 2.0 3.3 3.4

6.6 1.6 4.3 5.1 5.7

Mean ? SEM

2.5 k 0.47

4.6 + 0.85

Period I = placebo, Period II = probucol.

jects in whom LDL levels rose. It has been assumed that increased loss of bile acids raises the requirements for cholesterol in the liver, part of which may be met from enhanced uptake of LDL by the liver. Receptor-mediated uptake of apo-B, and hence HDL, has now been demonstrated in the liver [12] ; and at least 2 other procedures, treatment with bile acid-sequestering drugs [13] and overfeeding with sucrose [14], appear to link increased removal of LDL to heightened bile acid loss. Shepherd et al. [15] have recently reported that cholestyramine appears to stimulate LDL removal via the apoB-receptor pathway. Our studies with probucol show at least that, whenever the fractional removal rate of LDL increased, bile acid excretion had also risen. Although the fractional removal rate of LDL-B protein rose in 4 of the 5 subjects treated with probucol, the concentration of LDL fell in only 3. This does not necessarily rule out increased LDL removal as the major reason for its lowered concentration, but would indicate that this can be negated by other factors. One exception in this study was an overweight woman with combined hyperlipoproteinaemia (subject No. 4). The other subject (No. 3), whose LDLB protein pool size actually rose with probucol despite a greater removal rate, was 22% above ideal body weight. We have previously observed that apo-B production in very low density lipoproteins (VLDL) is positively correlated with body weight [16], so that overproduction of LDL-B protein (which is derived from VLDL) may nullify the effectiveness of drugs that act through stimulating LDL removal. The action of probucol on HDL metabolism is profound. The drug clearly suppresses apo-AI synthesis, to a greater or lessr extent. The number of subjects studied is small, but it is interesting to note that the 2 subjects, in whom LDL levels failed to fall, also showed the most substantial changes in HDL metabolism. Interactions between LDL and HDL metabolism in man have been observed by us in other studies [6], but not to suggest, as with probucol, that LDL removal may be adversely affected when the HDL concentration is markedly reduced. The effects on LDL removal (through an increase in bile acid excretion) and on HDL synthesis may both be mediated in the small intestine. Probucol is a poorly absorbed drug and its major unwanted effect,

209

diarrhoea, also suggests interference with normal gut function. Bile acid reabsorption may become impaired and this needs to be tested. APO-AI is formed to a substantial degree in the small intestine and enters the circulation within intestinal triglyceride-rich particles [ 171. Although the unchanged fasting plasma triglyceride levels do not suggest abnormal fat transport, it is nevertheless possible for fat absorption to occur apparently normally at much lower plasma apo-AI concentrations [ 8 1. The lowering of the plasma cholesterol and of LDL in particular, is based on the expectation that a reduction in atherosclerotic arterial disease will follow. Whether a proportionately greater depression in the HDL concentration will cancel out any benefit derived from the LDL-lowering effect of probucol is a serious theoretical possibility, though a final answer may come only when the apparently protective role of HDL in atherosclerosis is understood. Acknowledgements The author thank technical assistance.

MS A. Everett,

MS M. O’Connor

and MS $3. Mensen for

References 1 Le Losier, J., Du-Breuil-Quidoz, S., Lussier-Cacan, S.. Huang, Y.S. and Davignon, J., Diet and probucol in lowering cholesterol concentrations, Arch. Int. Med., 137 (1977) 1429. 2 Parsons, Jr., W.B.. Effect of probucol ln hyperlipidemic patients during two years of administration, Amer. Heart J., 96 (1978) 213. 3 Riesen, W.F.. Keller, M. and Mordasinl. R.. Probucol in hypercholesterolemia -A double blind study, Atherosclerosis, 36 (1980) 201. 4 Miettinen. T.A., Mode of action of a new hypocholesterolaemic drug (DH-581) in familial hypercholesterolaemia. Atherosclerosis. 15 (1972) 163. 5 Reardon. M.F., Fidge, N.H. and Nestel, P.J., Catabolism of very low density lipoprotein B apoprotein in man, J. Clin. Invest., 61 (1978) 850. 6 Fidge, N.H., Nestel, P.J., Ishikawa, T.. Reardon, M. and Billington. T., Turnover of apoprotelns AI and AI1 of high density lipoprotein and the relationship to other lipoproteins in normal and hyperlipidemic individuals. Metabolism, 29 (1980) 643. 7 Gurpide, E., Mann, J. and Sandberg, E.. Determination of kinetic parameters in a two-pool system by administration of one or more tracers, Biochemistry, 3 (1964) 1250. 8 Langer, T., Strober, W. and Levy, R.I.. The metabolism of low density lipoprotein in familial type II hyperlipoproteinemia. J. Clln. Invest., 51 (1972) 1528. 9 Davlgnon. J., Slmmonds, W.J. and Ahrens, Jr., E.H., Usefulness of chromic oxide as an internal standard for balance studies ln formula-fed patients and for assessment of colonic function, J. Clin. Invest., 47 (1968) 127. 10 Grundy, S.M., Ahrens. Jr., E.H.. and Miettlnen, T.A., Quantitative isolation and gas-&quid chromatographic analysis of total fecal bile acids, J. Lipid, Res., 6 (1965) 397. 11 Curry, M.D., Alaupovlc, P. and Suentam, C.A., Determination of apolipoprotein A and its constitutive AI and AI1 polypeptides by separate electroimmunoassays, Clln. Chem., 22 (1976) 315. 12 Kovanen, P.T., Brown, M.S. and Goldstein, J.L., Increased binding of low density lipoprotein to liver membranes from rats treated with l’la-ethinyl estradiol, J. Biol. Chem., 254 (1979) 11367. 13 Levy, R.I. and Langer, T., Hyperlipidemic drugs and lipoprotein metabolism, Adv. Exp. Med. Biol.. 6 (1972) 155. 14 Nestel, P.J.. Reardon. M.F. and Fidge, N.H., Sucroseinduced changes in VLDL- and LDL-B apoprotein removal rates, Metabolism. 28 (1979) 531. 15 Shepherd, J., Packard, C.J., Bicker, S., Lawrie. T.D.V. and Morgan, H.G.. Cholestyramine receptormediated low-density-lipoprotein catabolism, N. Engl. J. Med., 302 (1980) 1219. 16 Nestel, P.J. Reardon, M.F. and Fidge, N.H.. Very low density lipoprotein B-apoproteln kinetics in human subjects, Clrc. Res., 45 (1979) 35. 17 Green, P.H.R.. Gllckmsn. R.M., Saudek. C.D., Blum, C.B. and Tall, A.R.. Human intestinal lipoproteins, J. Clin. Invest., 64 (1979) 233. 18 Schaefer, E.J. and Brewer, Jr., H.B.. Tangier disease - A defect in the conversion of chylomicrons to high density lipoproteins. Cllm. Res., 26 (1978) 532A.