The disappearance rate of human versus rat intermediate density lipoproteins from rat liver perfusion'

ROBERT DUPRAS, LOUISE BRISSETTE, 2'4 PAUL D. ROACH, SYLVAIN BEGIN, ANDRB TREMBLAY, AND SIMON-PIERRE NOEL^

De'partement de biochimie, Universite' de Montreal, Montr6al (Que'bec), Canada H3C 3J7

Received June 4, 1991

DUPRAS, R., BRISSETTE, L., ROACH, P. D., BEGIN, S., TREMBLAY, A., and NOEL, S.-P. 1991. The disappearance rate of human versus rat intermediate density lipoproteins from rat liver perfusion. Biochem. Cell Biol. 69: 537-543.

The aim of this work was to compare the disappearance rate of human and rat intermediate density lipoproteins (IDL) using the rat liver perfusion system. Human and rat IDL were produced in vitro by incubating human or rat very low density lipoproteins (VLDL) with either rat post-heparin plasma (method I) or a resolubilized isopropanol precipitate of rat post-heparin plasma (method 11). With both methods, the degree of triacylglycerol lipolysis was approximately 55%. The different preparations of IDL were labelled with 12'1 and added to perfusates of rat livers. The disappearance rates of 1Z5~-labelled IDL were monitored by measuring the radioactivity associated with apolipo- protein (apo) B in the perfusate during a 15-min period. Both human and rat IDL prepared with method I had an increased apoE to apoC ratio as compared with their native counterparts. Furthermore, human IDL had a significantly higher apoE to apoC ratio than rat IDL. However, when IDL were produced in the absence of exchangeable apolipo- proteins (method 11), no change in the apoE to apoC ratios was observed for the transformation of VLDL to IDL and the ratios were similar for human and rat IDL. Despite these differences, human IDL were always removed at a lower rate than rat IDL. The only striking difference between the two types of IDL made by method I1 was that the apoBlOO to apoB48 ratio was considerably higher in human than in rat IDL. These results suggest that the apoBlOO to apoB48 ratio is likely to be responsible for the observed differences in liver uptake between rat and human IDL.

Key words: very low density lipoproteins, intermediate density lipoproteins, low density lipoproteins, hepatic lipoprotein receptors, intermediate density lipoprotein uptake, in vitro lipolysis, very low density lipoprotein remnants, apolipoproteins.

DUPRAS, R., BRISSETTE, L., ROACH, P. D., BEGIN. S.. TREMBLAY, A., et NOEL, S.-P. 1991. The disappearance rate of human versus rat intermediate density lipoproteins from rat liver perfusion. Biochem. Cell Biol. 69 : 537-543.

La vitesse de captation des 1ipoprotBnes de densitt intermtdiaire (IDL) humaines ou de rat a t t t mesurCe par le systkme de perfusion de foie de rat. Les IDL ont Ctt produites in vitro en incubant des prtparations de 1ipoprotCines de trks faible densitt (VLDL) avec du plasma post-heparin6 de rat (mCthode I) ou une preparation de plasma post- hCparinC prkipitke a l'iso ropanol (mCthode 11). Dans chacun des cas, environ 55% des triacylglycbols ont kt6 hydrolysks. Ces IDL marquees I I"$ ont CtC ajoutkes skparkment au milieu de perfusion de foie et leur disparition a hC suivie pendant 15 min dans le milieu de perfusion, en dtterminant la radioactivitt associCe a I'apolipoprotCine (apo) B pen- dant 15 min. Les IDL humaines et de rat prtpartes par la mtthode I se sont caractCrisCes par un rapport apoE sur apoC plus ClevC que celui des VLDL de dipart. De plus, les IDL humaines ont un rapport significativement plus ClevC que les IDL de rat. Quelle que soit la mCthode de lipolyse utilide, les IDL humaines ont t t t capttes plus lentement que les IDL de rat, indiquant que le rapport apoE sur apoC ne peut &tre la seule cause de la difftrence de captation observte entre les IDL humaines et celles de rat. La seule diffkrence entre la composition des IDL humaines ou de rat prCparCes par la mtthode I1 est un rapport apoBlOO sur apoB48 plus ClevC pour les IDL humaines que pour les IDL de rat, suggtrant que la composition en apoBlOO et apoB48 est responsable de la captation plus rapide des IDL de rat que des IDL humaines.

Mots cle's : lipoprottines de t r b faible densitt, lipoprottines de densitt intermkdiaire, lipoprottines de faible densite, rkcepteurs hkpatiques des lipoprottines, captation des lipoprotkines de densitt intermtdiaire, lipolyse in vitro, rtsidus de 1ipoprotCines de trts faible densite, apolipoprotCines.

Introduction Low density lipoproteins (LDL) have been shown to

correlate positively with the incidence of coronary heart disease (Kannel1983). The understanding of the formation of LDL is, therefore, crucial to delineate what regulates the concentration of these atherogenic particles in the circula-

ABBREVIATIONS: EDTA, ethylenediaminetetraacetic acid (sodium salt); VLDL, very low density lipoproteins; IDL, inter- mediate density lipoproteins; LDL, low density lipoproteins; HDL, high density lipoproteins; apo, apolipoprotein; PHP, post-heparin plasma.

his paper is dedicated to the memory of the late Dr. Simon- Pierre Nod. R. Dupras and L. Brissette are joint senior authors.

'present address: Institut de recherches cliniques de Montrtal, 110, avenue des Pins ouest, MontrCal (QuCbec), Canada H2W 1R7.

3~eceased February 5, 1987.

tion. LDL come from the catabolism of intermediate den- sity lipoproteins (IDL) from which in turn come very low density lipoproteins (VLDL).

The fate of IDL differs between human and rat species. In human, a highly variable proportion of IDL (averaging 50%) is converted to LDL (Janus et al. 1980). In contrast, the vast majority of rat IDL is rapidly taken up by the liver (Faergerman and Have1 1975). While human VLDL contain the high molecular weight form of apolipoprotein B (apoB100) and only trace amount of the low molecular weight form (apoB48), rat VLDL contain both the high (apoBh) and the low (apoB1) molecular weight forms (Wu and Windmueller 1981). Rat apoBh has been shown to disappear more slowly from circulation than rat apoBl (Elovson et al. 1981 ; Sparks and Marsh 1981). Moreover, the VLDL catabolism cascade is different for the particles

Printed in Canada / Imprim6 au Canada

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y SA

VA

NN

AH

RIV

NA

TL

AB

BF

on 1

1/10

/14

For

pers

onal

use

onl

y.

538 BIOCHEM. CELL BIOL. VOL. 69, 1991

A. HUMAN B. R A T

FIG. 1. Apolipoprotein profiles of isopropanol-precipitated and isopropanol-soluble fractions of human and rat VLDL. One-half millilitre of (A) human and (B) rat VLDL (2 mg/mL of VLDL protein) was mixed with 0.5 mL of isopropanol. The precipitate was pelleted by centrifugation. The supernatant was removed and evaporated to dryness. Both the isopropanol precipitate and the dried supernatant were delipidated with ethanol-ether (3: 1) before solubilization with the electrophoresis sample buffer. Control VLDL was lyophilysed and delipidated as described (Brissette and Noel 1986) before solubilization of apolipoproteins. Lane 1,20 pg of isopropanol- soluble apolipoprotein; lane 2. 20 pg of isopropanol-precipitated apolipoprotein; lane 3, 40 pg of control VLDL apolipoprotein. The molecular mass (in kilodaltons) is given on the left of each set of gels. apoCs are proteins of the apoC family (EI. EII, and EIII).

containing either apoBh or apoB1, since only the high molecular weight apoB remains associated with rat LDL. Since Arbeeny et al. (1987) have shown that the hepatic uptake of rat IDL is mediated by apoE, it can be speculated that apoBh impedes or that apoBl favors the uptake of IDL via apoE. The former possibility is more likely, since apoB48 has been shown not to influence the binding of lipoproteins containing apoE to hepatic and extrahepatic receptors (Hui et al. 1984). Also, Windler et al. (1980) have shown that apoE promotes and apoC opposes the uptake of triacyl- glycerol-rich lipoproteins by perfused rat livers.

The aim of our work was to compare the disappearance rate of human and of rat IDL from rat liver perfusate. Since rat IDL are at a too low concentration in rat plasma to be isolated, rat and human IDL were prepared by in vitro lipolysis of plasma VLDL. The choice of a lipolysis protocol and particularly the enzyme source is crucial in regards to the final lipid and protein composition of IDL. Two prepa- rations of lipoprotein lipase were used that gave IDL par- ticles with the same degree of lipolysis, but with higher or similar apoE/apoC ratios than the native VLDL. Further- more, the disappearance of apoB was exclusively monitored to preclude any transfer of labelled phospholipids between labelled IDL and rat liver membranes or any exchange of apoE and apoC with freshly synthesized lipoproteins. We

found that the uptake by the rat liver of human IDL is slower than that of rat IDL and that a faster uptake is associated with a low apoBlOO/apoB48 (apoBh/apoBl) ratio. In addi- tion, we show that the uptake of either rat or human IDL is faster when the particles have higher apoE/apoC ratios.

Materials and methods Materials

Male Sprague-Dawley rats from Charles River Canada Inc., (St-Constant, Que.) were used. They were fed regular rat chow (Ralston Purina Co., St-Louis, MO), had free access to water, and were maintained in a 12-h light-dark cycle. Human blood (3-5 days old) was obtained from the Montreal Transfusion Center of the Canadian Red Cross. '"I (as sodium iodide, 100 mCi/mL; 1 Ci = 37 GBq) was bought from Amersham Corporation (Arlington Heights, IL). Fatty acid free or regular bovine serum albumin (fraction V), phenylmethylsulfonylfluoride, and heparin (sodium salt, grade I) were purchased from Sigma Chemical Co. (St-Louis, MO). Acrylarnide and bis-acrylamide were bought from Bio-Rad (Richmond, CA). Isopropanol(2-propanol), 99% pure, was obtained from Fisher Scientific Co. (Montreal, Que.). Sodium pentobarbital (as Somnotol) was bought from Centre de Distribu- tion de Medicaments Vettrinaires (St-Hyacinthe, Que).

Preparation of lipoproteins Rat blood from rats weighing 600-800 g and fed ad libitum was

obtained by heart puncture under ether anesthesia with the use of

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y SA

VA

NN

AH

RIV

NA

TL

AB

BF

on 1

1/10

/14

For

pers

onal

use

onl

y.

DUPRAS ET AL.

TABLE 1. Apolipoprotein distribution in human and rat VLDL and IDL

Percent of total apolipoprotein Ratios

Bl00 B48 B100/B48 (Bh) (Bl) apoE apoC apoE/apoC apoE/apoB, total (Bh/Bl)

Human Control VLDL 47 + 4a 6 +- l a 18 + 2 23 k 3a 0.6 + 0.1 0.4 k 0.1 9 k 2 a IDL (PHP) 51+5a 7 + l a 33+-4b 6 + l a b 6.4+1.8ab 0.6k0.1 8 + l a IDL (isopropanol

precipitate) 54+6a 3 k 2 a 10+3c 29flOac 0 .4kO. l~ 0.2k0.11 8 k 6ac Rat

ControlVLDL 1 7 k 3 15+-1 17k1 51k4 0.4k0.1 0.6k0.11 2 k 0.2 IDL (PHP) 23k3 24+5d 29+4d 24+5d 1.4-1-0.4 0.6k0.1 1.1k0.3 IDL (isopropanol

precipitate) 14k2e 21+3 15+2de 51+2e 0.3kO.le 0.7 k0.2 0.7 k0.2

NOTE: The lipoproteins were lyophilysed and delipidated. The apolipoproteins were separated on sodium dodecyl sulfate - polyacrylamide gel electrophoresis. The apolipoprotein distribution was determined by scanning laser densitometry and the percent content of each band was calculated using the software program Gelscan adapted for the Apple IIe microcomputer. The values are the mean+SEM of three determinations. a, statistically different from its rat counterpart; b, statistically different from human control VLDL and IDL (isopropanol precipitate); c, statistically different from human IDL (PHP); d, statistically different from rat control VLDL; e, statistically different from rat IDL (PHP).

a syringe containing an anticoagulant solution (1.32% sodium citrate, 0.44% citric acid, and 1.32% glucose). Cells of human and rat blood were removed by low-speed centrifugation, and 0.01% EDTA. 0.02% sodium azide, and 10 pM phenylmethylsulfonyl- fluoride were added to the plasma before the isolation of lipopro- teins which was achieved by ultracentrifugation as described by Hatch and Lees (1968). Chylomicrons were removed by centrifuga- tion at 8.5 x lo4 x g for 30 min. VLDL were isolated by apply- ing a total centrifugal force of 100 000 x g for 18 h and recen- trifuged once in a sodium chloride solution of density 1.006 g/mL.

IDL were produced by lipolysis of VLDL by the two following methods. Method I: Rat post-heparin plasma (PHP) was prepared from overnight fasted rats as described (Nod and Rubinstein 1981) and used within 1 h. VLDL was added to PHP at a concentration of 5 mg VLDL-triacylglycerols per mL of PHP in the presence of 2.5% (w/v) fatty acid free bovine serum albumin and incubated at 37°C for 1 h. Method 11: The isopropanol precipitate of rat PHP was prepared as follows: one volume of isopropanol was added to PHP and mixed thoroughly. Five minutes later, the mixture was centrifuged at 12 000 x g for 5 min at 4OC. This procedure was repeated with the pellet. Then, one volume of acetone was added to the pellet and mixed. The mixture was then filtered on Whatman paper no. 1. The precipitate was washed with several volumes of anhydrous diethyl ether and air-dried. The isopropanol precipitate (approximately 10 mg per mL of PHP) retained 70-75070 of the total lipase activity contained in PHP. The powder was stored at - 20°C in a dessicator and was stable for at least 1 week without any apparent loss in enzymatic activity. Just prior to the lipolysis reactions, the lipase was extracted by incubating the powder at a proportion of 37.5 mg per mL of 25 mM Tris-HC1. pH 8. at 4OC for 30 min. The amount of isopropanol precipitate equivalent to 0.25 mL PHP was used for each mg of VLDL-triacylglycerol in the presence of 2.5% (w/v) fatty acid free bovine serum albumin. The mixture was incubated at 37°C for 1 h.

For both lipolytic protocols, the reaction was stopped by putting the tubes on ice. Sodium chloride was added to raise the density to 1.019 g/mL and IDL were isolated and washed by applying a total centrifugal force of 100 000 x g for 20 h. In all cases, the degree of triacylglycerol hydrolysis was 55 k 2 (mean + SEM, N = 9).

Iodination of ZDL Rat and human IDL were iodinated by a modification (Langer

et al. 1972) of the iodine monochloride method of McFarlane (1948). One millicurie of Na '"I was used to iodinate 1 mg IDL- protein in the presence of 100 nmol of iodine monochloride in 1 M glycine-NaOH buffer, pH 10. The reaction was stopped by adding

NazSz05 and KI (Roach and Noel 1985). Free iodine was removed by gel filtration on Sephadex G-25 (Pharmacia). The specific radioactivity ranged from 100 to 200 cpm/ng protein. The iodine to protein molar ratio (per 100 000 Da) was 1.1 k 0.2 (mean +- SEM of 10 experiments). The radioactivity incorporated into the lipid moiety was 36 ? 4 and 25 k 3% (mean * SEM of 5 exper- iments) for rat and human IDL, respectively.

Liver perfusion Rats, weighing 175-200 g and fed ad libitum, were anesthetized

with sodium pentobarbital (60 mg/kg body weight). Livers were perfused in situ according to the method of Mortimore (1961) as fully described elsewhere (Noel and Dupras 1983), with the fol- lowing modifications. Livers were f is t flushed with approximately 20 mL of Krebs-Henseleit buffer, pH 7.4, containing 0.1 % glucose (w/v) and 25% (v/v) of human washed red blood cells at room temperature. The livers were transferred into the perfusion chamber at 40°C and perfused with 40 mL of Krebs-Henseleit buffer con- taining 0.1% glucose (w/v) and 25% red blood cells (v/v). After a 10-min "recovery" period, each liver was examined carefully and only livers perfused uniformly (as judged by an even coloration of the liver surface) were kept for further processing. At this time. using a three-way stopcock, a switch was made to another recir- culating perfusate (40 mL) containing radioactive rat or human IDL (100 pg proteins in a 300-pL volume). Aliquots of 1 mL of perfusate were taken at 0,0.5, 1,2,4, 8, and 15 min of perfusion and maintained over ice. The red cells were removed by low-speed centrifugation at 4°C and cell-free perfusate was processed as follows: 400 pL were mixed with an equal volume of isopropanol in the presence of 50 pg of unlabelled VLDL protein as carrier in order to selectively precipitate apoB (Holmquist and Carlson 1977). The pellet containing apoB was washed with diethyl ether and chloroform-methanol (2: 1, v/v) to remove any labelled lipids. The counting of '25~-labelled apoB radioactivity was carried out in a gamma counter (LKB) working at 80% efficiency.

Other methods The lipase activity in PHP or in its isopropanol precipitate was

determined with the use of a glycerol-stabilized emulsion of tritiated triolein (Nielson-Ehle and Schotz 1976) as fully described elsewhere (Brissette and Noel 1983). Protein was determined by the method of Lowry et al. (1951) with bovine serum albumin as standard. Triacylglycerols were determined by the method of Van Handel (1961) after extraction of the VLDL lipids (Folch et al. 1957) and adsorption of polar lipids onto silicic acid.

Sodium dodecyl sulfate - polyacrylamide gel electrophoresis of apolipoproteins was carried out as described elsewhere (Brissette

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y SA

VA

NN

AH

RIV

NA

TL

AB

BF

on 1

1/10

/14

For

pers

onal

use

onl

y.

540 BIOCHEM. CELL BIOL. VOL. 69. 1991

and Noel 1986). The apolipoprotein profiles were scanned with the use of an Ultra-Scan Laser densitometer (LKB) coupled to an Apple IIE microcomputer: The intensity of each band was calculated using the GELSCAN (LKB) software program and was expressed as percent of total intensity measured in a given apolipoprotein profile.

Results Experiments were designed to specifically monitor the

disappearance of IDL apoB from rat liver perfusate. Isopropanol was shown to precipitate human VLDL apoB in a slightly more quantitative fashion than tetramethylurea (Holmquist et al. 1978). Before extending the procedure to rat VLDL, we have verified that isopropanol precipitates rat apoB quantitatively and selectively. Figure 1 shows that isopropanol precipitates rat and human apoB, whereas apoE and apoC remain soluble. Furthermore, there is practically no contamination of apoE and apoC in isopropanol precipitate and no trace of apoB in the soluble extract. In addition, both rat apoB forms (Bh and B1) co-precipitate in isopropanol. Similarly, the small traces of apoB48 in human VLDL appear to be recovered in the isopropanol precipitate.

In the first series of perfusion experiments designed to compare the hepatic uptake of rat and human IDL, the remnant particles were prepared by incubation of VLDL with rat PHP. The apolipoprotein content has an enormous influence on the ability of a given lipoprotein to recognize hepatic receptors. We have, therefore, examined the apolipo- protein content of both human and rat IDL by polyacryl- amide gel electrophoresis. The relative content of each apolipoprotein band was determined by densitometry. The results summarized in Table 1 show that the original human and rat VLDL differ significantly in their apoB composition, since the apoBlOO/apoB48 (apoBh/apoBl) ratio is 7.5 times bigger for human IDL than rat IDL. However, the VLDL particles do not differ significantly in their apoE/apoC and apoE/apoB ratios. The production of IDL by incubation with rat PHP produced human and rat IDL particles strongly depleted of apoC and relatively enriched in apoE as compared with the original VLDL. The apoE/apoC ratio was considerably higher for IDL, the difference being par- ticularly striking for human IDL. Furthermore, the more important apoE contribution and the higher apoE/apoB ratio of human IDL reveals that the particles are enriched in apoE, presumably in rat PHP apoE.

The rat and human IDL produced by lipolysis were iodinated and added separately to liver perfusates. The radioactivity associated with apoB was measured in per- fusates as a function of time. The results shown in Fig. 2 indicate that both types of IDL are taken up rapidly by the liver and that human IDL disappear slower than rat IDL. It is worth noticing that for both curves there is a faster com- ponent of disappearance followed by a slower component and that the former is identical for rat and human IDL, while the later is slower for human than rat IDL. This may indicate that each preparation contains approximately 20% of IDL particles that are taken very rapidly by the liver and with the same apparent affinity. Furthermore, in accordance with the slower component of disappearance, increasing the perfusion time to 30 or 60 min did not generate much more IDL uptake than at 15 min (data not shown).

To avoid apoE transfer from rat PHP, we have carried out a second series of experiments in which VLDL were

S I , , , I 0 5 10 15

PERFUSION T I M E (min)

FIG. 2. The disappearance rate of iodinated apoB associated with human and rat IDL prepared with rat PHP. Human (0) and rat ( 0 ) IDL were prepared by incubation of the corresponding VLDL with rat PHP (method 3. IDL were isolated by ultracen- trifugation and labelled with ' 'I. One hundred micrograms of '25~-labelled protein was added to each liver perfusate. Aliquots were sampled from the perfusate at the indicated times. Cell-free perfusate (0.4 mL) was mixed with an equal volume of isopropanol in the presence of 50 g of unlabelled VLDL proteins. The precipitate containing ''$-labelled apoB was washed with organic solvents to remove labelled lipids before counting. Each point represents the mean & SEM of nine different perfused livers using three different batches of labelled IDL. The differences between rat and human IDL were significant (p < 0.05) at 4, 8, and 15 min.

incubated with partially purified rat lipoprotein lipase devoid of exchangeable apolipoproteins. To obtain an appreciable amount of such a lipase preparation, rat PHP was treated with isopropanol as described in Materials and methods. Under these conditions, most of the lipoprotein lipase was recovered in the isopropanol precipitate, whereas plasma apoE and apoC remained in the soluble fraction. Also, we found that isopropanol co-precipitates lipoprotein lipase and hepatic lipase, since 30% of the lipase activity is insensitive to NaCl(1 M). The apolipoprotein distribution on human and rat IDL produced with the isopropanol precipitate was quite different from that of IDL produced with PHP. Table 1 shows that the apoC depletion observed in IDL made with PHP did not occur with the isopropanol precipitate procedure in spite of the fact that the degree of lipolysis was similar for these two methods of lipolysis (approximately 55% triacylglycerols hydrolysis). Also, the apoE/apoC ratio of IDL made with the isopropanol precipitate is not different from that of the original VLDL. An apparent increase in apoB100/apoB48 is also observed with human IDL. This is possibly an artefact of the deter- mination of apolipoprotein levels where a small decrease in the estimation of apoB48 can have an important effect on the apoB100/apoB48 ratio.

The disappearance rate from the liver perfusate of human and rat 125~-labelled IDL produced with the isopropanol precipitate of rat PHP is shown in Fig. 3. These rat IDL were again showing a faster rate of disappearance than human IDL. Also, as with the IDL produced with PHP (Fig. 2), early after injection of the particles into the liver

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y SA

VA

NN

AH

RIV

NA

TL

AB

BF

on 1

1/10

/14

For

pers

onal

use

onl

y.

DUPRAS ET AL. 54 1

PERFUSION TIME (min)

FIG. 3. The disappearance rate of iodinated apoB associated with human and rat IDL prepared with partially purified rat lipoprotein lipase. Human (0) and rat ( 8 ) IDL were prepared by incubation of the corresponding VLDL with lipase-containing isopropanol precipitate of rat PHP (method 11). Preparation of labelled IDL and treatment of samples were as described in the legend of Fig. 2. Each point represents the mean k SEM of six different perfused livers using two different batches of labelled IDL. The differences between rat and human IDL were significant (p < 0.05)at0.5, 2.4, 8, and 15min.

perfusate, human and rat IDL disappeared approximately at the same rate (Fig. 3). It is also worth noticing that the disappearance rate of IDL was lower when particles were prepared with an isopropanol precipitate of PHP instead of PHP itself. Indeed, significant differences (p < 0.05) were found at perfusion time of 2,4, and 15 min for rat IDL and 0.5, 2, 4, 8, and 15 min for human IDL.

Discussion Differences in IDL uptake by livers may depend either

on the abundance of hepatic lipoprotein receptors or on the structural features of the IDL particles themselves. The experiments reported in this paper were designed to study the effect of the composition of IDL on the uptake by the liver; receptors were presumed constant in each perfused liver. However, to reduce normal biological variations between rat livers, each preparation of IDL was perfused through at least three different livers. As a result, the stan- dard error of the mean of each point on the disappearance curve was always less than 10% (Figs. 2 and 3). Thus we can reasonably assume that each perfused liver displayed a comparable ability to take up circulating IDL. The dif- ferences observed in the disappearance rates of IDL could therefore be attributable to the nature of the ligands.

When IDL are iodinated, the radioactive iodine incor- porates not only into the various apolipoproteins but also into the lipids and particularly the surface phospholipids. During liver perfusion, the labelled phospholipids can easily exchange with unlabelled phospholipids of both the liver and the erythrocyte membranes of the perfusate. Moreover, labelled apoE and peptides of the apoC family may exchange with their unlabelled counterparts of newly secreted VLDL and high density lipoproteins (HDL). To avoid these undesirable effects, we have chosen to follow the disap-

pearance rate of apoB which is known to remain associated with its original particle as an integral constituent.

Using the perfused rat liver system, we have shown that human and rat IDL prepared with rat PHP are catabolized differently, the later being removed at a faster rate. These particles are relatively enriched in apoE and depleted in apoC compared with both control VLDL and IDL made with the partially purified lipase. The increased apoE/apoC ratio of the particles produced with rat PHP should facilitate their uptake, since it was shown that apoE promotes and apoC reduces the uptake of triacylglycerol-rich lipoproteins by perfused rat livers (Windler et al. 1980). Rat PHP is a natural, efficient donor of apoE, since the concentration of this apolipoprotein in rat serum reaches 0.18 mg/mL (Fainaru et al. 1977), which is about five times more than apoE levels in human serum (Breslow 1984). In the normal physiological situation, remnants are always produced in the presence of transferable apoE, since the lipolysis occurs in the circulation. Rat plasma is likely a better donor of apoE than human plasma for the simple reason of apoE being more abundant in the rat than in the human circulation. Thus, the very high apoE/apoC ratio observed in human IDL produced by incubation with rat PHP may be a conse- quence of rat apoE association and, therefore, may not represent the physiological human IDL. A similar effect was also noted by Hui et al. (1984) using human PHP and dog chylomicrons. In their study, chylomicron remnants made by incubation with human PHP produced apoE-enriched particles that were efficient ligands for adult dog liver mem- branes. In contrast, chylomicron remnants made by incu- bation with purified bovine milk lipoprotein lipase produced apoE-depleted particles that bound very poorly to the hepatic membranes. Taken together, these experiments indicate that the method of preparation of triacylglycerol- rich lipoprotein remnants is of crucial importance to their ability to recognize hepatic lipoprotein receptors and that apoE is the determinant of the binding. This is in accordance with the demonstration by another group (Arbeeny et al. 1987) that the apoE of rat IDL mediates the uptake of this lipoprotein by the rat perfused liver and with a recent study by Friedman et al. (1990) showing that the binding to HepG2 cells of human plasma IDL is mediated by apoE. This accen- tuates the need of choosing the best conditions of lipolysis mimicking as much as possible the physiological conditions.

Nevertheless, our results indicate that a factor in addi- tion to the apoE/apoC ratio modulates hepatic IDL uptake. Rat and human IDL produced by incubation of VLDL with PHP results in an apoE/apoC ratio that is 4.5 times higher for human IDL than rat IDL, but, nevertheless, the former particles disappear more slowly than the later. We con- sidered the possibility that the uptake of human IDL, although slower than the uptake of rat IDL, could be a con- sequence of an artificial enrichment in rat apoE. To avoid this enrichment, the human and rat IDL were obtained using rat PHP that had been depleted in its apolipoproteins by selective precipitation with isopropanol. We found that these particles with a low apoE/apoC ratio are also taken up by the rat liver but at a lower rate than the corresponding IDL, with a high apoE/apoC ratio prepared with PHP. These results are in agreement with the demonstration that a high apoE/apoC ratio is a favorable determinant for lipoprotein uptake (Windler et al. 1980). Furthermore, we showed (Fig. 3) that rat IDL still disappear faster than that of human

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y SA

VA

NN

AH

RIV

NA

TL

AB

BF

on 1

1/10

/14

For

pers

onal

use

onl

y.

542 BIOCHEM. CELL BIOL. VOL. 69, 1991

IDL having the same apoE/apoC ratio and a similar apoE content (approximately 12%). On one hand, it is possible that human IDL are simply not as good ligands as rat IDL for the rat receptors. Indeed, the receptor binding domains of rat apoB and (or) apoE may have higher affinity for the rat receptors than the analogous domains in human apolipo- proteins. On the other hand, we found that rat and human IDL apoB100/apoB48 (apoBh/apoBl) ratios are several fold higher in human IDL than in rat IDL, suggesting that the contribution of the two forms of apoB may also be respon- sible for the faster disappearance of rat IDL than human IDL. Furthermore, this finding is in accordance with another series of experiments (Brissette and Noel 1988), in which we have compared the binding of rat IDL rich in apoBh or in apoBl to rat liver membranes. We found that both IDL bound specifically to the membranes with a high and a low affinity and that the binding of B1-rich IDL to the high affinity sites was significantly more avid than that of Bh-rich IDL. Since apoB48 does not influence the binding of dog chylomicron remnants or human 0-VLDL to hepatic lipoprotein receptors (Hui et al. 1984), taken together, the results we obtained with human IDL and the two different preparations of rat IDL substantiate the idea that apoB100 (apoBh) interferes with the binding of apoE to hepatic lipoprotein receptors as suggested by Brown and Goldstein (1983).

The disappearance curves also showed that at time shorter than 2 min, rat IDL and human IDL disappear at the same rate, while between 2 and 4 min there is a small increase in human IDL apoB in the perfusate, indicating that some human IDL are released into the perfusate. It can be speculated that there is, at first, a trapping of rat and human IDL into the space of Disse of the hepatocyte and that human IDL, not being as good ligands as rat IDL, are released back, instead of remaining associated or being endocytosed by the hepatocytes. Alternatively, it is possible that human IDL, while being associated with hepatic cells, lost part of their apoE load, became weaker ligands, and were released.

Differences in affinity can also reflect the lipid composi- tion of lipoproteins. Our preparations of human IDL were moderately poorer in triacylglycerols than of rat IDL. This is based on the observation that this lipid contribution was somewhat less important in human VLDL than its rat counterpart (data not shown) and that the same degree of lipolysis was achieved with the two types of VLDL to obtain IDL. If such a difference plays a role in the binding, it would be by increasing the expression of apoBl00 epitopes involved in the binding to the LDL receptor (Krul et al. 1985). This difference between the two preparations of IDL cannot explain the slower rate of disappearance of human IDL com- pared with that of rat IDL, since it would result in making human IDL a better ligand. A recent study by Sehayek and Eisenberg (1990) showed a positive correlation between cholesteryl ester / protein ratios of VLDL and their capacity to interact (via apoE) with the LDL receptors. Our prepa- rations of VLDL were similar in their cholesteryl esters / protein ratios (data not shown). Thus, lipid com- position cannot explain the difference in the disappearance rate of human and rat IDL in our rat liver perfusion system.

In conclusion, by using an in situ rat perfusion system and by specifically monitoring the disappearance of IDL- apoB, we showed that rat liver takes up rat IDL more avidly

than human IDL and that this difference is associated with the composition in apoBlOO (apoBh) and apoB48 (apoB1); a higher apoBlOO/apoB48 (apoBh/apoBl) ratio is unfavor- able to the uptake. These results could explain, at least in part, why IDL (precursors of LDL) accumulate more in human blood than in rat blood.

Acknowledgements We thank the Montreal Transfusion Center of the

Canadian Red Cross and Ms. M. Caisse for supplying us with fresh human blood. We gratefully acknowledge Ms. L. Charette and C. Lemire for typing this manuscript and Ms. C. Ostiguy for drawing the figures. We are indebted to Drs Y.L. Marcel, R.W. Milne, and Z. Zawadzki for their useful suggestions in the preparation of this manuscript. This work was supported by the Quebec and Canadian Heart Foundations. L.B. and P .D.R. were recipients of student- ships from the Canadian Medical Council and the Canadian Heart Foundation, respectively.

ARBEENY, C.M., RIFICI, V.A., HANDLEY, D.E., and EDER, H.A. 1987. Determinants of the uptake of very low density lipopro- tein remnants by the perfused rat liver. Metabolism, 36: 1106-1113.

BRESLOW, J.L. 1984. Molecular genetics of lipoprotein disorders. Circulation, 69: 1 190-1 194.

BRISSETTE, L., and NOEL, S.-P. 1983. Rapid preparation of a relatively stable, partially purified lipoprotein lipase from per- fused rat heart. Anal. Biochem. 135: 197-200. - 1986. The effects of human LDL and HDL on the bind-

ing of rat IDL to rat liver membranes. J. Biol. Chem. 261: 6847-6852. - 1988. Binding of intermediate density lipoproteins rich or

poor in high molecular weight apolipoprotein B to rat liver mem- branes. Biochem. Cell Biol. 66: 1238-1247.

BROWN, M.S., and GOLDSTEIN, J.L. 1983. Lipoprotein receptors in the liver. Control signals for plasma cholesterol traffic. J. Clin. Invest. 72: 743-747.

ELOVSON, J., HUANG, Y.O., BAKER, N., and KAUNAN, R. 1981. Apolipoprotein B is structurally and metabolically heterogeneous in the rat. Proc. Natl. Acad. Sci. U.S.A. 78: 157-161.

FAERGERMAN, O., and HAVEL, R.J. 1975. Metabolism of cholesteryl esters of rat VLDL. J. Clin. Invest. 55: 1210-1218.

FAINARU, M., HAVEL, R.J., and IMAIZUMI, K. 1977. Radio- immunoassay of arginine-rich apolipoprotein of rat serum. Biochim. Biophys. Acta, 490: 144-155.

FOLCH, J., LEES, J., and SLOANE-STANLEY, G.H. 1957. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226: 497-509.

FRIEDMAN, G., GAVISH, D., VOGEL, T., and EISENBERG, S. 1990. Cellular metabolism of human plasma intermediate density lipoprotein (IDL). Biochim. Biophys. Acta, 1044: 118-126.

HATCH, F.T., and LEES, R.S. 1968. Practical methods for plasma lipoprotein analysis. Adv. Lipid Res. 6: 1-68.

HOLMQUIST, L., and CARLSON, K. 1977. Selective extraction of human serum very low density apolipoproteins with organic solvent. Biochem. Biophys. Acta, 493: 400-409.

HOLMQUIST, L., CARLSON, K., and CARLSON, L.L. 1978. Com- parison between the use of isopropanol and tetramethylurea for the solubilization and quantitation of human serum very low den- sity lipoprotein. Anal. Biochem. 88: 457-460.

HUI, D.Y., INNERARITY, T.L., MILNE, R.W., MARCEL, Y.L., and MAHLEY, R.W. 1984. Binding of chylomicron remnants and VLDL to hepatic and extrahepatic lipoprotein receptors. J. Biol. Chem. 259: 15060-15068.

JANUS, E.D., NICOLL, A., N O O ~ N , R., TURNER, P.R., MAGILL, P.J., and LEWIS, L. 1980. Quantitative studies of VLDL con-

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y SA

VA

NN

AH

RIV

NA

TL

AB

BF

on 1

1/10

/14

For

pers

onal

use

onl

y.

DUPRAS ET AL. 543

version to LDL in normal controls and primary hyperlipideamic states and the role of direct secretion of LDL in heterozygous familial hypercholesterolaernia. Eur. J. Clin. Invest. 10: 149-159.

KANNEL, W.B. 1983. HDL: epidemiologic profile and risk of coronary artery disease. Am. J. Cardiol. 52: 9B-12B.

KRUL, E.S., TIKKANEN, M.J., COLE, T.G.. DAVIS, J.M., and SCHONFELD, G. 1985. Roles of apolipoproteins B and E in the cellular binding of very low density lipoproteins. J. Clin. Invest. 75: 361-369.

LANGER, T., STROBER, W., and LEVY, R.I. 1972. The metabolism of LDL in familial type I1 hyperlipoproteinemia. J. Clin. Invest. 51: 1528-1536.

LOWRY, O.H., ROSEBROUGH, N. J., FARR, A.L., and RANDALL, R. J. 195 1. Protein measurement with the Folin-phenol reagent. J. Biol. Chem. 193: 265-275.

MCFARLANE, A.S. 1948. Efficient trace-labeling of proteins with iodine. Nature (London), 182: 53-54.

MORTIMORE, G. 1961. Effect of insulin on potassium transfer in isolated rat liver. Am. J. Physiol. 200: 1315-1319.

NIELSON-EHLE, P., and SCHOTZ, M.C. 1976. A stable radioactive substrate emulsion for assay of lipoprotein lipase. J. Lipid Res. 17: 536-541.

NOEL, S.P., and DUPRAS, R. 1983. The kinetic parameters of the uptake of VLDL remnant cholesteryl esters by perfused rat livers. Biochem. Biophys. Acta, 754: 117-125.

NOEL, S.P., and RUBINSTEIN, D. 1981. Uptake of cholesterol from VLDL or its remnants by the perfused rat liver. Can. J. Biochem. 59: 447-453.

ROACH, P.D., and NoBL, S.P. 1985. Solubilization of the 17 6-ethinyl estradiol-stimulated LDL receptor of male rat liver. J. Lipid Res. 26: 713-720.

SEHAYEK, E., and EISENBERG, S. 1990. Abnormal composition of hypertriglyceridemic very low density lipoprotein determines abnormal cell metabolism. Arteriosclerosis, 10: 1088-1096.

SPARKS, C.E., and MARSH, J.B. 1981. Metabolic heterogeneity of apolipoprotein B in the rat. J. Lipid Res. 22: 519-527.

VAN HANDEL, E. 1961. Suggested modification of the micro- determination of triglycerides. Clin. Chem. 7: 249-251.

WINDLER, E., CHAO, Y.S., and HAVEL, R.J. 1980. Determinants of hepatic uptake of triglyceride-rich lipoproteins and their remnants in the rat. J. Biol. Chem. 255: 5475-5480.

Wu, A., and WINDMUELLER, H.G. 1981. Variant forms of plasma apolipoproteins B. Hepatic and intestinal biosynthesis and het- erogeneous metabolism in the rat. J. Biol. Chem. 256: 3615-3618.

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y SA

VA

NN

AH

RIV

NA

TL

AB

BF

on 1

1/10

/14

For

pers

onal

use

onl

y.