- Email: [email protected]
In vivo metabolism of human apoprotein A-I-phospholipid complexes. comparison with human high density lipoprotein-apoprotein A-I metabolism
ClinicaChimicaActa, 131 (1983)201-210 Elsevier
201
CCA 2554
In vivo metabolism of human apoprotein A-I-phospholipid complexes. Comparison with human high density lipoprotein-apoprotein A-I metabolism Claude L. Malmendier *, Claude Delcroix and Jean P. Ameryckx Research Unit on Atherosclerosis, Laboratory of Chemical Parhology and Interdisciplinary Institute for Research on Human and Molecular Biology, Free University of Brussels, Brussels (Belgium) (Received September 7th, 1982; revision March 9th, 1983)
The metabolism of apolipoprotein A-I was investigated in healthy young adults. The apoprotein was injected intravenously in one of the four following forms: radiolabelled high density lipoproteins, HDL incubated with labelled apo A-I, labelled A-I-dipalmitoyl- or dimyristoylphosphatidylcholine complexes or free apoprotein A-I. The radioactivity-time curves of high density lipoproteins were followed for 15 days. The kinetic studies demonstrate similar fractional catabolic rates, half-lifes and rates of synthesis. Fractional catabolic rate calculated from plasma data was in reasonable agreement with the value derived from the urine/plasma activity ratio. When free apo A-I was injected, the unusual form of the U/P ratio curve is interpreted as the result of different fates of the injected molecules. This study indicates that the half-life of apo A-I seems not to be affected by the form in which the apoprotein is injected and that kinetics of apo A-I-phospholipid complexes may be used as a valid index of apoprotein A-I/HDL metabolism.
Introduction
The specific activity decay curves of two major apoproteins of high density lipoproteins (HDL), apoproteins A-I and A-II, are parallel to one another in normal * Correspondence.: Professor Claude L. Mahnendier, Research Unit on Atherosclerosis, Faculty of Medicine U.L.B., Bat D, 2 rue Evers, 1000 Brussels, Belgium. Abbreoiatiorw: ‘251-apo A-I, delipidated unbound apoprotein A-I labelled with lz51 [l]; ‘2sI-A-I/HDL, apoprotein A-I from ‘251-labelled apoprotein A-l incorporated in vitro into HDL [a]; ‘251-HDL/A-I, or DMPC, native HDL labelled with lz51 by the iodine monochloride method [l]; ‘3’I-A-I/DPPC “‘I-labelled apoprotein A-I incorporated into dipalmitoyl- or dimyristoylphosphatidylcholine vesicles.
0009-8981/83/SO3.00
0 1983 Elsevier Science Publishers B.V.
202
subjects studied either with a balanced diet or a carbohydrate-rich diet [I], as well as in Tangier heterozygotes [2]. But apolipoprotein A-I (apo A-I) is catabolised at a much greater fractional rate than apo A-II in Tangier homozygotes [2] whereas in h~ert~~yce~de~c subjects [3] a higher catabolic rate of apo A-II is associated with a lower rate of synthesis of apoprotein A-I. These studies used homologous or heterologous high density lipoproteins labelled in vitro by the iodine monochloride method of McFarlane. With this procedure of labelling all HDL apoproteins are iodinated and measurement of the specific activity of apo A-I and apo A-II requires therefore the previous isolation of HDL by preparative ultracentrifugation followed by either polyac~la~de gel electrophoresis ]2,3] or Sephadex c~omato~aphy [I]. In order to study the effect of a diet or a drug on apoprotein catabolism without using these time-consuming procedures, the turnover of ‘25I-high density lipoprotein, ‘3’I-A-I-dipalmitoylphosphatidylcholine complexes or free apoprotein A-I-‘25I was first tested in normal volunteers. Material and methods
The healthy volunteers participating in this study gave informed consent and the experimental procedure was approved by the Ethical Committee of the Faculty of Medicine and of the Medical Research Scientific Foundation. Each subject received for 3 days before and throu~out the study 400 mg of KI per day in two doses to prevent thyroidal uptake of radioiodide, Isolation of HDL Two hundred ml of venous blood were collected from normal donors after an ove~~t fast (14 h) into 0.1% disodium EDTA and plasma was isolated at 4°C and at 2500 X g in a J-6B Beckman refrigerated centrifuge. HDL (density 1.063- 1.210 g/ml) were isolated in a Beckman 60 Ti rotor at 12°C and at 260000 x g for 20 h at d 1.063, then for 20 h at 1.250 and finally for 20 h at d 1.210 g/ml [4]. They were then dialysed for 24 h at 4°C with several changes against 150 mmol/l NaCl, 0.01% EDTA, 0.01% thimerosal, pH 7.4. Purification of apoprotein A-I Isolated HDL were delipidated using ethanol/diethyl ether 3 : 2 (v/v). Apo A-I was fractionated on a Sephacryl S-200 column (2.6 x 100 cm) using Tris-HCl (200 mmol/l, pH 8.2) containing 6.0 mol/l urea, 2 mmol/l sodium dodecylsulphate and 0.02% sodium azide as elution buffer (rate of elution: 7.5 ml/m@. The Fraction III corresponding to apo A-I was dialysed for 24 h at 4°C against NH,HCO, (5 mmol/l, pH 8.0) and concentrated in an Amicon Cell (Diaflo UM2 membrane) prior to lyophilisation. Apo A-I was purified by ion-exchange chromatography on a DEAE-Sephacel(l.6 x 48 cm) colurnr~ as described previously [5]. Apo A-I was then dialysed against ammonium bicarbonate, concentrated and lyophilised as above and finally stored at -20% The purity of the isolated apo A-I was confirmed by polyacrylamide gel electrophoresis, amino acid composition and combined im-
munodiffusion/immunoelectrophoresis. Rabbit monospecific antisera were prepared as described by Schonfeld and Pfleger [6]. L.&ding
of HDL apoprotein A-I (1) Samples of HDL were radioiodinated with 125I by a modification of the iodine monochloride technique [7]. Unbound iodine was removed by dialysis at 4°C against 6 1 of 150 mmol/l NaCl, 0.01% EDTA and 0.01% thimerosal, pH 7.4. Ninety-nine percent of the radioactivity remaining after dialysis was attached to the lipoprotein. Before injection, HDL was filtered on a Millipore filter (0.45 nm, Millipore Corp., Bedford, MA, USA). The material was tested for the presence of bacteria and pyrogens prior to use. A maximum of 96 h elapsed before the taking of the initial blood sample and reinjection. (2) Lyophilised apo A-I was dissolved in a glycine-NaOH buffer (1 mol/l, pH 10) at a concentration of 3 g/l and radiolabelled with 13’1by the iodine monochloride method as adapted by Shepherd et al [8]. Free iodine was eliminated by dialysis for 48 h at 4OC against Tris (10 mmol/l) in NaCl, 150 mmol/l, pH 8.6. (3) Samples of HDL in 0.05 mol/l barbital buffer, pH 8.6, were incubated with “‘1-a Po A-I and labelled by in vitro exchange [7]. After labelling, HDL tracers were purified by ultracentrifugal flotation [8].
Preparation of 13’I-apo A-I-phospholipid complex L-a-dipalmitoyl- or dimyristoylphosphatidylcholine (DPPC or DMPC) purchased from Sigma Co. (St. Louis, MO, USA) was dissolved in Tris (10 mmol/l in 150 mmol/l NaCl, pH 8.6) (2-3 g/l). Single-lamellar vesicles were prepared by sonification for 30-60 min at 4°C in a thermostated cell using a Branson 50 sonifier equipped with a microtip [9]. Labelled apo A-I (in 0.01 mol/l Tris) was complexed to the liposomes in a ratio A-I/DPPC or DMPC 1.95 : 1 (w/w) [9]. The mixture was raised to a density of 1.225 g/ml by addition of solid KBr and unbound apoprotein separated from the complex by ultracentrifugation (20 h, at 260000 X g, 12’C). The 2 ml on the top of the tube were recovered, dialysed and sterilised by Millipore filtration (0.45 mn) before use in vivo. Study protocol Fifty microcuries of ‘25I-high density lipoproteins (5 mg of protein) were mixed with 50 CCi of ‘)‘I-apo A-I/DPPC or DMPC complex (1.2 mg of protein) or 10 PCi of delipidated free ‘3’I-apo A-I (0.5 mg of protein) and immediately administered intravenously. Thirty ml fasting blood samples were collected into disodium EDTA (1 g/I) after 40 min, 6 h, 24 h and then every second morning for up to 15 days (12-13 blood samples). Each plasma sample was assayed for radioactivity using a Philips autogamma counter. The plasma lipoproteins were fractionated by sequential ultracentrifugation and radioactivity in VLDL, LDL, HDL, VHDL and bottom (d ) 1.25) was determined in the Philips autogamma counter. Cholesterol, phospholipid and triglyceride were measured following the Lipid Research Clinics protocol [lo]. The protein content was determined by the method of Schacterle and PolIack [ 1I]. Delipidated apo HDL (see above) were separated by gel filtration on
204
Sephacryl S-200. Fractions III and IV corresponding to apo A-I and apo A-II respectively, were pooled, dialysed, concentrated. In each fraction, the protein content was determined [ 1I] and specific activities of ‘251-apo A-I and apo A-II were calculated. Apo A-I concentrations in Fraction I + II, III and IV of Sephacryl chromatography, in HDL and in total plasma were also determined by electroimmunoassay, using 2% agarose gels containing 1.5% apo A-I antiserum [ 121. 24-h urine specimens were collected in plastic jars containing 5 ml of a preservative [ 131. Urinary radioactivity excretion was measured in samples of each collection.
The plasma radioactivity decay curves of apo A-I/HDL, apo A-II/HDL and apo A-I/DPPC or DMPC were fitted to two exponentials. The initial space distribution of the tracer was estimated as the plasma volume. The proportion of apo A-I in the intravascular space, the biological half-life (7’4) and the fractional catabolic rate (FCR) were calculated using the matbematic~ procedures of Matthews 1141. The FCR was also calculated as the ratio urine acti~ty/plasma activity, i.e. &herate of excretion of radioactivity per 24 h divided by the mean plasma activity per each collection period. The metabolic clearance rate (MRC) = FCR X VP(plasma volume). The residence time (i) in the system or the average time that particles introduced into plasma spend in the system was calculated from the area under the curve [ 151. The synthetic rate of apo A-I was calculated as the products of FCR times plasma volume times apoprotein A-I concentration in HDL. Under steady-state conditions, the rate of synthesis is equal to its absolute catabolic rate. Results
~i~tribu?ion of rad~ooctivity in lipoprotein classes After injection of ‘251-apo A-I/HDL, 90% of the plasma radioactivity remained in the density range of HDL (1.063- 1.210 gjml); less than 1% was found in the density fraction x 1.063 (VLDL + LDL) and 9% in the d > 1.21. After injection of ‘311-A-I/DPPC or DMPC the percentages were for HDL, VLDL + LDL and bottom 72.0, 1.2 and 26.8% respectively. 20 min after injection of free apoprotein A-I, 60% of the plasma activity was recovered in HDL and the remaining in the d 1.21 fraction but after the first 24 h more than 80% remained in HDL. Plasma clearance curves Typical plasma clearance curves of ‘251-HDL/apo A-I, “‘1-HDL/apo A-II and ‘3’I-apo A-I/DPPC complexes for two normal volunteers are shown in Fig. 1, A and B. Apoprotein A-II decay curves differ only very slightly from those of apo A-I. Apo A-I injected as a complex labelled in vitro ( “‘1-A-I/HDL) is metabolically undistinguishable from apo A-I labelled in situ in the lipoprotein (Table I). Our results do not confirm those of Shepherd et al [16] who found a 20% faster metabolism for iodine-labelled A-I incorporated into the HDL than for that labelled in the whole HDL particle. When labelled apoprotein A-I is injected as a free protein (Fig. 1, C) the
205
radioactivity decay curves of apo A-I/HDL and of total plasma apo A-I differ in their early portion but become parallel to one another after 4-5 days, 90% of plasma activity then being recovered in apo A-I/HDL. U/P
activity curves
Fig. 1, D shows the difference in pattern of urine activity/plasma activity (U/P) when apo A-I is injected as an integral part of HDL or as free apoprotein. However, no difference in the U/P ratio is observed when apo A-I is complexed to HDL, to DPPC or DMPC.
Fig. I. (A), (B) and (C), HDL activity decay curves expressed as the fraction of initial activity remaining as a function of time (days), (A) after simultaneous injection of ‘251-HDL/A-I (x), ‘ZSI-HDL/A-II (0) and ‘3’I-A-I/DPPC (0) in a subject C.R.; (B) in subject H.G.; (C) after injection of free apoprotein A-I-“‘1 in subject D.J. (X). Total plasma activity curve (0) is given for comparison. (D) Urine activity/plasma activity after ‘3’I-HDL/A-I (x) and free ‘251-apo A-I (0) injection, calculated by dividing the rate of urinary excretion of radioactivity per 24 h by the mean plasma radioactivity for each collection period.
1.10
1.07
2974
3304
H‘G.
_
3.85 z?I
5.I5
3.9t G
t .a7
A-I/DPPC CR.
5.14
1.10
3304
5.363 b7??i
t .O?
-.
4.93 5,58
6.23
6.04 5.08 5.56
5.64 5,86
6.29 5.64
WY9
(days)
4.85 4.19
Residence time
Apo A-I half-life
1.04 1.06
2974
3098 26a2 2906
HDL-A-I wncentration (g/O
H.G.
*
*
Plasma volume (ml)
HDL,/A-I C.R.
N.Y+
A-I/HDL R.O. M.P.
Material injected
0.265 rt 0.020 0.285+0.019 0.275
0.257 0.288 0.272
0.270+0.015 0.265
0.260 rt 0.0 18
0.290 + 0.022 0.316+0.025 0.271 ltO.014 0.292
(b)U/P ratio ( f SD)
-
95(1 857
0.275 0.26f
0.255
0.260 0.293 0.236 izz
(a)calculated
Fractional catabolic
764
760 907 834
686 752
806 763
Metabolic rate per day clearance fate (ml/day)
Kinetic parameters of apoprotein A-l and A-II turnover in normal male subjects
TABLE I
12.70 izz
11.67
Il.61 12.13 11,87
9-66 iiz
11.17 12.84
Synthesis rate mgakg-‘-day-’
70.9 z?
62.5
71.4 ss.r
64.9
61.0 61.0
61.0 61.0
% of apo A-I intravascular
3304
H.G.
5.87 5.38
4.88 590 814
1037
0.32
0.29 6.09
4.17
6.65 5.53
5.08 4.46
620
685
749 0.218
0.227
0.208
0.73s
0.307
0.260 0.285
0.310
0.222
2.08
1.98 2.18
0.22s It 0.01s
10.92
10.74 11.33
11.91
0.219&0.018
0.318f0.019
0.243 f 0.0 11 0.254
0.26SrtO.014
65.8
71.9
-71.9
71.9
55.2
66.2 66.0
l
Iodin~labelled A-I/HDL, labelled apoprotein A-I incorporated in vitro into HDL [8]; iodine-labelled HDL/A-I, native HDL labelled by the iodine monochloride procedure [ 11; iodine-lab&d A-I/DPPC or DMPC, complexes formed between dipalmitoyl- or dimyristoylphosphatidylcholine vesicles and iodine-Welled apoprotein A-I [9]. ** These results were not used for cakulation of the means for apoprotein A-I. *** Apo A-I comxntration in total plasma.
Mean A-II
2974
HDL/A-II CR.
D.J.
3.13 4.30 4.02
radioactivity decay curve 4.63 3841 O.% 5.92 1181 caindated from totalphsma radioactivity curve 4.87 1.24 *** 5.55 2822 3841
1.47
1.15
cahdatedfrom apo - A- i/HDt
SM.
Free apo A-I ** D.J.
3342
2272
A-I/DMPC D.B.
208
Kinetic parameters of apolipoprotein A-I The kinetic parameters of apoprotein A-I metabolism are presented in Table I. The biological half-life of apo A-I/HDL, the mean residence time, the fractional catabolic rate (whether determined by calculation from the decay curves or from the U/P ratios) show no difference regardless of whether apo A-I was injected into one of the three forms: native HDL labelled with iodine, HDL incubated with labelled apo A-I or labelled apoprotein A-I/DPPC or DMPC complex. Values of biological half-life, metabolic clearance rate, fractional catabolic rate, and synthetic rate are comparable to those found in the literature [ 1,16,17]. After injection of free apo A-I, apoprotein A-I/HDL has a similar metabolic clearance rate and intravascular percentage of apo A-I, indicating it is likely that apo A-I is incorporated into HDL. The curve of total plasma radioactivity shows a higher metabolic clearance rate (2822 ml/day) but a similar half-life (4.87 days).
Discussion Under normal conditions apo A-I and A-II are catabolised together as shown already by Blum et al [l] and Shepherd et al [ 161. Moreover, the kinetics of apo A-I, whether injected as HDL or as a dipalmitoyl (or dimyristoyl) phosphatidylcholineapoprotein complex, are identical. In contrast to what happens to complexed A-I, free A-I is degraded at a much faster rate but the half-life of the terminal portion of apo A-I/HDL curve (free A-I incorporated in vivo into HDL) is the same. A similar observation was made before in the dog [18]. The U/P ratio after injection of apo A-I/HDL or apo A-I/DPPC or DMPC provides a direct measurement of the fractional catabolic rate of the apoprotein. The interpretation of the U/P ratio curve observed after free apo A-I administration is more subtle. The fractional catabolic rate calculated from the terminal portion of the U/P curve (6 values) reaches after 5 days a value similar to the fractional catabolic rate calculated from the plasma radioactivity curve of apo A-I/HDL and must be interpreted as representing the metabolism of apo A-I incorporated into HDL. The early portion of the U/P ratio curve shows higher values explaining the high fractional catabolic rate of total plasma apo A-I given in Table I and making evident an heterogeneity in the fate of the injected molecules. It seems indeed that part of injected A-I is quickly metabolised and directly excreted in urine and that part of apo A-I is incorporated into HDL in vivo, as was shown in vitro by Shepherd et al [7]. Although a dynamic equilibrium was shown in vitro [16], the release of apo A-I activity from HDL is slow, as suggested by the ratio between the two pools (free and bound apo A-I) in steady-state conditions, to maintain the high value of FCR; this part thus follows the fate of apo A-I/HDL, which is later degraded and excreted in urine. The early urine excretion of radioactivity after free apoprotein therefore expresses the quick degradation of the unbound apoprotein. The fact that apo A-I either as an integral part of HDL or as a DPPC or DMPC complex turns over at a similar rate is important. The decay curves of the A-I/DPPC complex obtained from the measurement of activity in total plasma and in ultra-
209
centrifugally separated HDL are parallel, and half-lives of the terminal portions of the decay curves are identical. As the number of subjects in each group is limited, no statistical calculation was done. More studies would be required for that purpose. The use of apoprotein A-I/DPPC or DMPC complexes, which are easy to obtain, offers some advantages: (a) it only requires the measurement of total plasma activity for biological half-life determination; (b) preparations of apo A-I/DPPC or DMPC complexes may be better standardised than HDL incubated with apoprotein A-I; (c) the amount of cold A-I injected is negligible and is less likely to give an immunological reaction than HDL proteins. Conclusion
A particle as simple as the apo A-I/DPPC or DMPC complex behaves like HDL-apo A-I, and its kinetics are hardly distinguishable from those of HDL. Therefore, this type of study may be proposed as a simple index for apo A-I turnover measurement without using sophisticated procedures. Before this kind of procedure can be used, however, to test the value of hypolipidaemic agents or dietary manipulations, it must be shown that the different preparations give the same results in hyperlipidaemia. Such studies are now in progress and will be reported separately. Acknowledgements
This investigation was supported in part by grants from the Ministhe de la Politique Scientifique within the framework of the Association Euratom-University of Brussels and by the Research Grant No. 3.4520.82 from the Belgian National Fund for Scientific Research (FRNS). References I Blum CB, Levy RI, Eisenberg S, Hall MIII, Goebel RH, Berman M. High density lipoprotein metabolism in man. J Clin Invest 1977; 60: 795-807. 2 Schaefer EJ, Blum CB, Levy RI et al. Metabolism of high-density lipoprotein apolipoproteins in Tangier disease. N Engl J Med 1978; 299: 905-910. 3 Rao SN, MagilI PJ, Miller NE, Lewis B. Plasma high-density lipoprotein metabolism in subjects with primary hypertriglyceridaemia: altered metabolism of apoproteins AI and AII. Clin Sci 1980; 59: 359-367. 4 Scanu A. Forms of human serum high density lipoprotein protein. J Lipid Res 1966; 7: 295-306. 5 Malmendier CL, Christophe J, Ameryckx JP. Separation and partial characterization of new apoproteins from human plasma high density lipoproteins. Clin Chim Acta 1979; 99: 167-176. 6 Schonfeld G, Pfleger B. The structure of human high density lipoprotein and the levels of apolipoprotein A-I in plasma as determined by radioimmunoassay. J Clin Invest 1974; 54: 236-246. 7 Shepherd J, Gotto AM, Taunton OD, Caslake MJ, Farish E. The in vitro interaction of human apolipoprotein A-I and high density lipoproteins. Biochim Biophys Acta 1977; 489: 486-501. 8 Shepherd J, Packard CJ, Gotto AM, Taunton OD. A comparison of two methods to investigate the metabolism of human apolipoproteins A-I and A-II. J Lipid Res 1978; 19: 656-661. 9 Middelhof G, Rosseneu M, Peeters H, Brown VW. Study of the lipid binding characteristics of the
210 apolipoproteins from human high density lipoprotein. Biochim Biophys Acta 1976; 441: 57-67. 10 Lipid Research Clinics Program. Manual of laboratory operations, Vol. 1. DHEW Publication (NIH) 1974: 75-628. 11 Schacterle GR, Pollack RL. A simplified method for the quantitative assay of small amounts of protein in biological material. Anal Biochem 1973; 51: 654-655. 12 Curry MD, Alaupovic P, Suenram CA. Determination of apolipoprotein A and its constitutive A-I and A-II polypeptides by separate electroimmunoassays. Clin Chem 1976; 23: 315-322. 13 Steinfeld JL, Paton RR, Flick AL, Milch RA, Beach FE. Distribution and degradation of human serum albumin labeled with 13’1by different techniques. Ann NY Acad Sci 1957; 70: 109-121. 14 Matthews CME. The theory of tracer experiments with ‘3’I-labelled plasma proteins. Phys Med Biol 1957; 2: 36-53. 15 Rescigno A, Gurpide E. Estimation of average times of residence, recycle, and interconversion of blood-borne compounds using tracer methods. J Clin Endocrinol Metab 1973; 36: 263-276. 16 Shepherd J, Packard CJ, Patsch JR, Gotto AM, Taunton OD. Metabolism of apolipoproteins A-I and A-II and its influence on the high density lipoprotein subfraction distribution in males and females. Eur J Clin Invest 1978; 8: 115-120. 17 Caslake MJ, Farish E, Shepherd J. Metabolism of apolipoprotein A-I in healthy young adults. Metabolism 1978; 27: 437-447. 18 Nakai T, Whayne TF. Catabolism of canine apolipoprotein A-I: purification, catabolic rate, organs of catabolism, and the liver subcellular catabolic site. J Lab Clin Med 1976; 88: 63-80.
201
CCA 2554
In vivo metabolism of human apoprotein A-I-phospholipid complexes. Comparison with human high density lipoprotein-apoprotein A-I metabolism Claude L. Malmendier *, Claude Delcroix and Jean P. Ameryckx Research Unit on Atherosclerosis, Laboratory of Chemical Parhology and Interdisciplinary Institute for Research on Human and Molecular Biology, Free University of Brussels, Brussels (Belgium) (Received September 7th, 1982; revision March 9th, 1983)
The metabolism of apolipoprotein A-I was investigated in healthy young adults. The apoprotein was injected intravenously in one of the four following forms: radiolabelled high density lipoproteins, HDL incubated with labelled apo A-I, labelled A-I-dipalmitoyl- or dimyristoylphosphatidylcholine complexes or free apoprotein A-I. The radioactivity-time curves of high density lipoproteins were followed for 15 days. The kinetic studies demonstrate similar fractional catabolic rates, half-lifes and rates of synthesis. Fractional catabolic rate calculated from plasma data was in reasonable agreement with the value derived from the urine/plasma activity ratio. When free apo A-I was injected, the unusual form of the U/P ratio curve is interpreted as the result of different fates of the injected molecules. This study indicates that the half-life of apo A-I seems not to be affected by the form in which the apoprotein is injected and that kinetics of apo A-I-phospholipid complexes may be used as a valid index of apoprotein A-I/HDL metabolism.
Introduction
The specific activity decay curves of two major apoproteins of high density lipoproteins (HDL), apoproteins A-I and A-II, are parallel to one another in normal * Correspondence.: Professor Claude L. Mahnendier, Research Unit on Atherosclerosis, Faculty of Medicine U.L.B., Bat D, 2 rue Evers, 1000 Brussels, Belgium. Abbreoiatiorw: ‘251-apo A-I, delipidated unbound apoprotein A-I labelled with lz51 [l]; ‘2sI-A-I/HDL, apoprotein A-I from ‘251-labelled apoprotein A-l incorporated in vitro into HDL [a]; ‘251-HDL/A-I, or DMPC, native HDL labelled with lz51 by the iodine monochloride method [l]; ‘3’I-A-I/DPPC “‘I-labelled apoprotein A-I incorporated into dipalmitoyl- or dimyristoylphosphatidylcholine vesicles.
0009-8981/83/SO3.00
0 1983 Elsevier Science Publishers B.V.
202
subjects studied either with a balanced diet or a carbohydrate-rich diet [I], as well as in Tangier heterozygotes [2]. But apolipoprotein A-I (apo A-I) is catabolised at a much greater fractional rate than apo A-II in Tangier homozygotes [2] whereas in h~ert~~yce~de~c subjects [3] a higher catabolic rate of apo A-II is associated with a lower rate of synthesis of apoprotein A-I. These studies used homologous or heterologous high density lipoproteins labelled in vitro by the iodine monochloride method of McFarlane. With this procedure of labelling all HDL apoproteins are iodinated and measurement of the specific activity of apo A-I and apo A-II requires therefore the previous isolation of HDL by preparative ultracentrifugation followed by either polyac~la~de gel electrophoresis ]2,3] or Sephadex c~omato~aphy [I]. In order to study the effect of a diet or a drug on apoprotein catabolism without using these time-consuming procedures, the turnover of ‘25I-high density lipoprotein, ‘3’I-A-I-dipalmitoylphosphatidylcholine complexes or free apoprotein A-I-‘25I was first tested in normal volunteers. Material and methods
The healthy volunteers participating in this study gave informed consent and the experimental procedure was approved by the Ethical Committee of the Faculty of Medicine and of the Medical Research Scientific Foundation. Each subject received for 3 days before and throu~out the study 400 mg of KI per day in two doses to prevent thyroidal uptake of radioiodide, Isolation of HDL Two hundred ml of venous blood were collected from normal donors after an ove~~t fast (14 h) into 0.1% disodium EDTA and plasma was isolated at 4°C and at 2500 X g in a J-6B Beckman refrigerated centrifuge. HDL (density 1.063- 1.210 g/ml) were isolated in a Beckman 60 Ti rotor at 12°C and at 260000 x g for 20 h at d 1.063, then for 20 h at 1.250 and finally for 20 h at d 1.210 g/ml [4]. They were then dialysed for 24 h at 4°C with several changes against 150 mmol/l NaCl, 0.01% EDTA, 0.01% thimerosal, pH 7.4. Purification of apoprotein A-I Isolated HDL were delipidated using ethanol/diethyl ether 3 : 2 (v/v). Apo A-I was fractionated on a Sephacryl S-200 column (2.6 x 100 cm) using Tris-HCl (200 mmol/l, pH 8.2) containing 6.0 mol/l urea, 2 mmol/l sodium dodecylsulphate and 0.02% sodium azide as elution buffer (rate of elution: 7.5 ml/m@. The Fraction III corresponding to apo A-I was dialysed for 24 h at 4°C against NH,HCO, (5 mmol/l, pH 8.0) and concentrated in an Amicon Cell (Diaflo UM2 membrane) prior to lyophilisation. Apo A-I was purified by ion-exchange chromatography on a DEAE-Sephacel(l.6 x 48 cm) colurnr~ as described previously [5]. Apo A-I was then dialysed against ammonium bicarbonate, concentrated and lyophilised as above and finally stored at -20% The purity of the isolated apo A-I was confirmed by polyacrylamide gel electrophoresis, amino acid composition and combined im-
munodiffusion/immunoelectrophoresis. Rabbit monospecific antisera were prepared as described by Schonfeld and Pfleger [6]. L.&ding
of HDL apoprotein A-I (1) Samples of HDL were radioiodinated with 125I by a modification of the iodine monochloride technique [7]. Unbound iodine was removed by dialysis at 4°C against 6 1 of 150 mmol/l NaCl, 0.01% EDTA and 0.01% thimerosal, pH 7.4. Ninety-nine percent of the radioactivity remaining after dialysis was attached to the lipoprotein. Before injection, HDL was filtered on a Millipore filter (0.45 nm, Millipore Corp., Bedford, MA, USA). The material was tested for the presence of bacteria and pyrogens prior to use. A maximum of 96 h elapsed before the taking of the initial blood sample and reinjection. (2) Lyophilised apo A-I was dissolved in a glycine-NaOH buffer (1 mol/l, pH 10) at a concentration of 3 g/l and radiolabelled with 13’1by the iodine monochloride method as adapted by Shepherd et al [8]. Free iodine was eliminated by dialysis for 48 h at 4OC against Tris (10 mmol/l) in NaCl, 150 mmol/l, pH 8.6. (3) Samples of HDL in 0.05 mol/l barbital buffer, pH 8.6, were incubated with “‘1-a Po A-I and labelled by in vitro exchange [7]. After labelling, HDL tracers were purified by ultracentrifugal flotation [8].
Preparation of 13’I-apo A-I-phospholipid complex L-a-dipalmitoyl- or dimyristoylphosphatidylcholine (DPPC or DMPC) purchased from Sigma Co. (St. Louis, MO, USA) was dissolved in Tris (10 mmol/l in 150 mmol/l NaCl, pH 8.6) (2-3 g/l). Single-lamellar vesicles were prepared by sonification for 30-60 min at 4°C in a thermostated cell using a Branson 50 sonifier equipped with a microtip [9]. Labelled apo A-I (in 0.01 mol/l Tris) was complexed to the liposomes in a ratio A-I/DPPC or DMPC 1.95 : 1 (w/w) [9]. The mixture was raised to a density of 1.225 g/ml by addition of solid KBr and unbound apoprotein separated from the complex by ultracentrifugation (20 h, at 260000 X g, 12’C). The 2 ml on the top of the tube were recovered, dialysed and sterilised by Millipore filtration (0.45 mn) before use in vivo. Study protocol Fifty microcuries of ‘25I-high density lipoproteins (5 mg of protein) were mixed with 50 CCi of ‘)‘I-apo A-I/DPPC or DMPC complex (1.2 mg of protein) or 10 PCi of delipidated free ‘3’I-apo A-I (0.5 mg of protein) and immediately administered intravenously. Thirty ml fasting blood samples were collected into disodium EDTA (1 g/I) after 40 min, 6 h, 24 h and then every second morning for up to 15 days (12-13 blood samples). Each plasma sample was assayed for radioactivity using a Philips autogamma counter. The plasma lipoproteins were fractionated by sequential ultracentrifugation and radioactivity in VLDL, LDL, HDL, VHDL and bottom (d ) 1.25) was determined in the Philips autogamma counter. Cholesterol, phospholipid and triglyceride were measured following the Lipid Research Clinics protocol [lo]. The protein content was determined by the method of Schacterle and PolIack [ 1I]. Delipidated apo HDL (see above) were separated by gel filtration on
204
Sephacryl S-200. Fractions III and IV corresponding to apo A-I and apo A-II respectively, were pooled, dialysed, concentrated. In each fraction, the protein content was determined [ 1I] and specific activities of ‘251-apo A-I and apo A-II were calculated. Apo A-I concentrations in Fraction I + II, III and IV of Sephacryl chromatography, in HDL and in total plasma were also determined by electroimmunoassay, using 2% agarose gels containing 1.5% apo A-I antiserum [ 121. 24-h urine specimens were collected in plastic jars containing 5 ml of a preservative [ 131. Urinary radioactivity excretion was measured in samples of each collection.
The plasma radioactivity decay curves of apo A-I/HDL, apo A-II/HDL and apo A-I/DPPC or DMPC were fitted to two exponentials. The initial space distribution of the tracer was estimated as the plasma volume. The proportion of apo A-I in the intravascular space, the biological half-life (7’4) and the fractional catabolic rate (FCR) were calculated using the matbematic~ procedures of Matthews 1141. The FCR was also calculated as the ratio urine acti~ty/plasma activity, i.e. &herate of excretion of radioactivity per 24 h divided by the mean plasma activity per each collection period. The metabolic clearance rate (MRC) = FCR X VP(plasma volume). The residence time (i) in the system or the average time that particles introduced into plasma spend in the system was calculated from the area under the curve [ 151. The synthetic rate of apo A-I was calculated as the products of FCR times plasma volume times apoprotein A-I concentration in HDL. Under steady-state conditions, the rate of synthesis is equal to its absolute catabolic rate. Results
~i~tribu?ion of rad~ooctivity in lipoprotein classes After injection of ‘251-apo A-I/HDL, 90% of the plasma radioactivity remained in the density range of HDL (1.063- 1.210 gjml); less than 1% was found in the density fraction x 1.063 (VLDL + LDL) and 9% in the d > 1.21. After injection of ‘311-A-I/DPPC or DMPC the percentages were for HDL, VLDL + LDL and bottom 72.0, 1.2 and 26.8% respectively. 20 min after injection of free apoprotein A-I, 60% of the plasma activity was recovered in HDL and the remaining in the d 1.21 fraction but after the first 24 h more than 80% remained in HDL. Plasma clearance curves Typical plasma clearance curves of ‘251-HDL/apo A-I, “‘1-HDL/apo A-II and ‘3’I-apo A-I/DPPC complexes for two normal volunteers are shown in Fig. 1, A and B. Apoprotein A-II decay curves differ only very slightly from those of apo A-I. Apo A-I injected as a complex labelled in vitro ( “‘1-A-I/HDL) is metabolically undistinguishable from apo A-I labelled in situ in the lipoprotein (Table I). Our results do not confirm those of Shepherd et al [16] who found a 20% faster metabolism for iodine-labelled A-I incorporated into the HDL than for that labelled in the whole HDL particle. When labelled apoprotein A-I is injected as a free protein (Fig. 1, C) the
205
radioactivity decay curves of apo A-I/HDL and of total plasma apo A-I differ in their early portion but become parallel to one another after 4-5 days, 90% of plasma activity then being recovered in apo A-I/HDL. U/P
activity curves
Fig. 1, D shows the difference in pattern of urine activity/plasma activity (U/P) when apo A-I is injected as an integral part of HDL or as free apoprotein. However, no difference in the U/P ratio is observed when apo A-I is complexed to HDL, to DPPC or DMPC.
Fig. I. (A), (B) and (C), HDL activity decay curves expressed as the fraction of initial activity remaining as a function of time (days), (A) after simultaneous injection of ‘251-HDL/A-I (x), ‘ZSI-HDL/A-II (0) and ‘3’I-A-I/DPPC (0) in a subject C.R.; (B) in subject H.G.; (C) after injection of free apoprotein A-I-“‘1 in subject D.J. (X). Total plasma activity curve (0) is given for comparison. (D) Urine activity/plasma activity after ‘3’I-HDL/A-I (x) and free ‘251-apo A-I (0) injection, calculated by dividing the rate of urinary excretion of radioactivity per 24 h by the mean plasma radioactivity for each collection period.
1.10
1.07
2974
3304
H‘G.
_
3.85 z?I
5.I5
3.9t G
t .a7
A-I/DPPC CR.
5.14
1.10
3304
5.363 b7??i
t .O?
-.
4.93 5,58
6.23
6.04 5.08 5.56
5.64 5,86
6.29 5.64
WY9
(days)
4.85 4.19
Residence time
Apo A-I half-life
1.04 1.06
2974
3098 26a2 2906
HDL-A-I wncentration (g/O
H.G.
*
*
Plasma volume (ml)
HDL,/A-I C.R.
N.Y+
A-I/HDL R.O. M.P.
Material injected
0.265 rt 0.020 0.285+0.019 0.275
0.257 0.288 0.272
0.270+0.015 0.265
0.260 rt 0.0 18
0.290 + 0.022 0.316+0.025 0.271 ltO.014 0.292
(b)U/P ratio ( f SD)
-
95(1 857
0.275 0.26f
0.255
0.260 0.293 0.236 izz
(a)calculated
Fractional catabolic
764
760 907 834
686 752
806 763
Metabolic rate per day clearance fate (ml/day)
Kinetic parameters of apoprotein A-l and A-II turnover in normal male subjects
TABLE I
12.70 izz
11.67
Il.61 12.13 11,87
9-66 iiz
11.17 12.84
Synthesis rate mgakg-‘-day-’
70.9 z?
62.5
71.4 ss.r
64.9
61.0 61.0
61.0 61.0
% of apo A-I intravascular
3304
H.G.
5.87 5.38
4.88 590 814
1037
0.32
0.29 6.09
4.17
6.65 5.53
5.08 4.46
620
685
749 0.218
0.227
0.208
0.73s
0.307
0.260 0.285
0.310
0.222
2.08
1.98 2.18
0.22s It 0.01s
10.92
10.74 11.33
11.91
0.219&0.018
0.318f0.019
0.243 f 0.0 11 0.254
0.26SrtO.014
65.8
71.9
-71.9
71.9
55.2
66.2 66.0
l
Iodin~labelled A-I/HDL, labelled apoprotein A-I incorporated in vitro into HDL [8]; iodine-labelled HDL/A-I, native HDL labelled by the iodine monochloride procedure [ 11; iodine-lab&d A-I/DPPC or DMPC, complexes formed between dipalmitoyl- or dimyristoylphosphatidylcholine vesicles and iodine-Welled apoprotein A-I [9]. ** These results were not used for cakulation of the means for apoprotein A-I. *** Apo A-I comxntration in total plasma.
Mean A-II
2974
HDL/A-II CR.
D.J.
3.13 4.30 4.02
radioactivity decay curve 4.63 3841 O.% 5.92 1181 caindated from totalphsma radioactivity curve 4.87 1.24 *** 5.55 2822 3841
1.47
1.15
cahdatedfrom apo - A- i/HDt
SM.
Free apo A-I ** D.J.
3342
2272
A-I/DMPC D.B.
208
Kinetic parameters of apolipoprotein A-I The kinetic parameters of apoprotein A-I metabolism are presented in Table I. The biological half-life of apo A-I/HDL, the mean residence time, the fractional catabolic rate (whether determined by calculation from the decay curves or from the U/P ratios) show no difference regardless of whether apo A-I was injected into one of the three forms: native HDL labelled with iodine, HDL incubated with labelled apo A-I or labelled apoprotein A-I/DPPC or DMPC complex. Values of biological half-life, metabolic clearance rate, fractional catabolic rate, and synthetic rate are comparable to those found in the literature [ 1,16,17]. After injection of free apo A-I, apoprotein A-I/HDL has a similar metabolic clearance rate and intravascular percentage of apo A-I, indicating it is likely that apo A-I is incorporated into HDL. The curve of total plasma radioactivity shows a higher metabolic clearance rate (2822 ml/day) but a similar half-life (4.87 days).
Discussion Under normal conditions apo A-I and A-II are catabolised together as shown already by Blum et al [l] and Shepherd et al [ 161. Moreover, the kinetics of apo A-I, whether injected as HDL or as a dipalmitoyl (or dimyristoyl) phosphatidylcholineapoprotein complex, are identical. In contrast to what happens to complexed A-I, free A-I is degraded at a much faster rate but the half-life of the terminal portion of apo A-I/HDL curve (free A-I incorporated in vivo into HDL) is the same. A similar observation was made before in the dog [18]. The U/P ratio after injection of apo A-I/HDL or apo A-I/DPPC or DMPC provides a direct measurement of the fractional catabolic rate of the apoprotein. The interpretation of the U/P ratio curve observed after free apo A-I administration is more subtle. The fractional catabolic rate calculated from the terminal portion of the U/P curve (6 values) reaches after 5 days a value similar to the fractional catabolic rate calculated from the plasma radioactivity curve of apo A-I/HDL and must be interpreted as representing the metabolism of apo A-I incorporated into HDL. The early portion of the U/P ratio curve shows higher values explaining the high fractional catabolic rate of total plasma apo A-I given in Table I and making evident an heterogeneity in the fate of the injected molecules. It seems indeed that part of injected A-I is quickly metabolised and directly excreted in urine and that part of apo A-I is incorporated into HDL in vivo, as was shown in vitro by Shepherd et al [7]. Although a dynamic equilibrium was shown in vitro [16], the release of apo A-I activity from HDL is slow, as suggested by the ratio between the two pools (free and bound apo A-I) in steady-state conditions, to maintain the high value of FCR; this part thus follows the fate of apo A-I/HDL, which is later degraded and excreted in urine. The early urine excretion of radioactivity after free apoprotein therefore expresses the quick degradation of the unbound apoprotein. The fact that apo A-I either as an integral part of HDL or as a DPPC or DMPC complex turns over at a similar rate is important. The decay curves of the A-I/DPPC complex obtained from the measurement of activity in total plasma and in ultra-
209
centrifugally separated HDL are parallel, and half-lives of the terminal portions of the decay curves are identical. As the number of subjects in each group is limited, no statistical calculation was done. More studies would be required for that purpose. The use of apoprotein A-I/DPPC or DMPC complexes, which are easy to obtain, offers some advantages: (a) it only requires the measurement of total plasma activity for biological half-life determination; (b) preparations of apo A-I/DPPC or DMPC complexes may be better standardised than HDL incubated with apoprotein A-I; (c) the amount of cold A-I injected is negligible and is less likely to give an immunological reaction than HDL proteins. Conclusion
A particle as simple as the apo A-I/DPPC or DMPC complex behaves like HDL-apo A-I, and its kinetics are hardly distinguishable from those of HDL. Therefore, this type of study may be proposed as a simple index for apo A-I turnover measurement without using sophisticated procedures. Before this kind of procedure can be used, however, to test the value of hypolipidaemic agents or dietary manipulations, it must be shown that the different preparations give the same results in hyperlipidaemia. Such studies are now in progress and will be reported separately. Acknowledgements
This investigation was supported in part by grants from the Ministhe de la Politique Scientifique within the framework of the Association Euratom-University of Brussels and by the Research Grant No. 3.4520.82 from the Belgian National Fund for Scientific Research (FRNS). References I Blum CB, Levy RI, Eisenberg S, Hall MIII, Goebel RH, Berman M. High density lipoprotein metabolism in man. J Clin Invest 1977; 60: 795-807. 2 Schaefer EJ, Blum CB, Levy RI et al. Metabolism of high-density lipoprotein apolipoproteins in Tangier disease. N Engl J Med 1978; 299: 905-910. 3 Rao SN, MagilI PJ, Miller NE, Lewis B. Plasma high-density lipoprotein metabolism in subjects with primary hypertriglyceridaemia: altered metabolism of apoproteins AI and AII. Clin Sci 1980; 59: 359-367. 4 Scanu A. Forms of human serum high density lipoprotein protein. J Lipid Res 1966; 7: 295-306. 5 Malmendier CL, Christophe J, Ameryckx JP. Separation and partial characterization of new apoproteins from human plasma high density lipoproteins. Clin Chim Acta 1979; 99: 167-176. 6 Schonfeld G, Pfleger B. The structure of human high density lipoprotein and the levels of apolipoprotein A-I in plasma as determined by radioimmunoassay. J Clin Invest 1974; 54: 236-246. 7 Shepherd J, Gotto AM, Taunton OD, Caslake MJ, Farish E. The in vitro interaction of human apolipoprotein A-I and high density lipoproteins. Biochim Biophys Acta 1977; 489: 486-501. 8 Shepherd J, Packard CJ, Gotto AM, Taunton OD. A comparison of two methods to investigate the metabolism of human apolipoproteins A-I and A-II. J Lipid Res 1978; 19: 656-661. 9 Middelhof G, Rosseneu M, Peeters H, Brown VW. Study of the lipid binding characteristics of the
210 apolipoproteins from human high density lipoprotein. Biochim Biophys Acta 1976; 441: 57-67. 10 Lipid Research Clinics Program. Manual of laboratory operations, Vol. 1. DHEW Publication (NIH) 1974: 75-628. 11 Schacterle GR, Pollack RL. A simplified method for the quantitative assay of small amounts of protein in biological material. Anal Biochem 1973; 51: 654-655. 12 Curry MD, Alaupovic P, Suenram CA. Determination of apolipoprotein A and its constitutive A-I and A-II polypeptides by separate electroimmunoassays. Clin Chem 1976; 23: 315-322. 13 Steinfeld JL, Paton RR, Flick AL, Milch RA, Beach FE. Distribution and degradation of human serum albumin labeled with 13’1by different techniques. Ann NY Acad Sci 1957; 70: 109-121. 14 Matthews CME. The theory of tracer experiments with ‘3’I-labelled plasma proteins. Phys Med Biol 1957; 2: 36-53. 15 Rescigno A, Gurpide E. Estimation of average times of residence, recycle, and interconversion of blood-borne compounds using tracer methods. J Clin Endocrinol Metab 1973; 36: 263-276. 16 Shepherd J, Packard CJ, Patsch JR, Gotto AM, Taunton OD. Metabolism of apolipoproteins A-I and A-II and its influence on the high density lipoprotein subfraction distribution in males and females. Eur J Clin Invest 1978; 8: 115-120. 17 Caslake MJ, Farish E, Shepherd J. Metabolism of apolipoprotein A-I in healthy young adults. Metabolism 1978; 27: 437-447. 18 Nakai T, Whayne TF. Catabolism of canine apolipoprotein A-I: purification, catabolic rate, organs of catabolism, and the liver subcellular catabolic site. J Lab Clin Med 1976; 88: 63-80.