Effect of different neutral phospholipids on apolipoprotein binding by artificial lipid particles in vivo

MING-FOR TONG AND ARNIS KUKSIS'

Bunting and Best Department ofMedica1 Research, 112 College Street, University of Toronto, Toronto, Ont., Canada M5G 1M Received November 12. 1985

Tong, M-F. & Kuksis, A. (1986) Effect of different neutral phospholipids on apolipoprotein binding by artificial lipid particles in vivo. Biochem. Cell Biol. 64, 826-835

Soybean triacylglycerol particles, stabilized with egg yolk sphingomyelin (SPH), phosphatidylcholine (PC), phosphatidyl- ethanolamine (PE), or PC-PE mixtures, with diameters ranging from 170 to 550 nm were prepared by sonication and isolated by ultracentrifugation. Binding of apoproteins to the lipid particles was studied in vivo using the strategy of Connelly and Kuksis. The recoveries of the injected particles, which had decreased in size and undergone minimal changes in lipid composition, ranged from 70% and 57% for SPH- and PC-stabilized particles to 14% for particles stabilized with egg yolk PC-PE mixtures. The apoprotein (apo) composition of the recovered particles showed qualitative and quantitative differences, which were affected by the number of washes during isolation. After four washes, the SPH and PC particles contained apoE, apoC-11, and apoC-111 as major components and apoA-IV as minor components. In addition, all particles, except those stabilized with egg yolk PC, contained large amounts of albumin. In contrast to egg yolk PC, the dipalmitoyl PC particles bound albumin as a major compo- nent. The recovered PC-PE and PE particles were characterized by a relative decrease of apoC and greatly increased binding of albumin. The higher rate of clearance of the PE-containing particles was attributed to a relative absence of apoC-111, which is known to delay hepatic uptake of lipid particles containing it, and to a more rapid hydrolysis of PE by lipoprotein lipases. Since PE occurs as a minor and variable component of chylornicrons and plasma lipoproteins, the present observations are of physio- logical interest.

Tong, M-F. & Kuksis, A. (1986) Effect of different neutral phospholipids on apolipoprotein binding by artificial lipid particles in vivo. Biochem. Cell Biol. 64, 826-835

Nous avons prkpark par sonication des particules de triacylglyckrol de la e v e soya et nous les avons isolkes par ultracentri- fugation. Ces particules sont stabiliskes avec la sphingomykline (SPH), la phosphatidylcholine (PC), la phosphatidylkthanol- amine (PE) ou des mklanges de PC-PE du jaune d'oeuf et elles ont des diamktres allant de 170 ti 550 nm. Utilisant la technique de Connelly et Kuksis, nous avons ktudik in vivo la liaison des apoprotkines aces particules lipidiques. Les particules injectkes et recouvrkes sont plus petites et leur composition lipidique est 1kgBrement changke. La dcup6ration est de 70% pour les particules stabiliskes avec la SPH, de 57% pour les protkines stabiliskes avec la PC et de 14% pour celles stabiliskes avec des mklanges de PC-PE du jaune d'oeuf. La composition en apoprotkines des particules recouvrkes montre des diffkrences qualitatives et quantitatives qui sont influenckes par le nombre de lavages lors de I'isolation. AprBs quatre lavages, les principales apoprotkines (apo) des particules SPH et PC sont I'apoE, l'apoC-I1 et I'apoC-111 et I'apoA-IV y est un constituant mineur. @ plus, toutes les particules, sauf celles stabiliskes avec PC du jaune d'oeuf, contiennent d'importantes quantitks d'albumine. A l'encontre des particules avec PC du jaune d'oeuf, les particules avec phosphatidylcholine dipalmitique lient I'albumine comme principal constituant. Les particules PC-PE et PE recouvrkes sont caractkriskes par une diminution relative de I'apoC et la liaison de I'albumine y est considkrablement accrue. Le taux plus klevk d'klimination des particules contenant laPE est attribuk ii I'absence relative de I'apoC-111, reconnue pour retarder I'incorporation hkpatique des particules lipidiques la contenant, et a une hydrolyse plus rapide de la PE par les lipoprotkine lipases. Comme la PE est un constituant mineur et variable des chylomicrons et des lipoprotkines plasmatiques, les prksentes observations ont un intkrkt physiologique.

[Traduit par la revue]

Introduction It is well established that during the complex course of

plasma lipoprotein metabolism, various apolipoproteins

ABBREVIATIONS: SPH, sphingomyelin; PC, phosphatidyl- choline; PE, phosphatidylethanolamine; apo, apoprotein; LPC, lysoPC; DPPC, dipalmitoyl PC; TLC, thin-layer chro- matography; TG, triacylglycerol(s); Sf, Svedberg flotation rate; d , density; VLDL, very low density lipoproteins; GLC, gas-liquid chromatography; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; MW, molecular weight; LDL, low density lipoproteins; HDL, high density lipoproteins; DMPC, dimyristoyl PC; FC, free cholesterol; PL, phospholipid; TMS, trimethylsilyl.

'Author to whom all correspondence should be addressed.

are exchanged and transferred among the different lipoprotein classes (1, 2). The critical parameters of the lipid particles governing the redistribution of the apo- proteins, however, are not understood. Previous work has shown that apolipoproteins are also transferred between plasma lipoproteins and artificial emulsions during incubation in vitro (3) or following injection in vivo (4). Recent work has demonstrated characteristic apoprotein binding during incubation of purified apo- proteins with artificial lipid emulsions (5-7). The above studies have established that the size of lipid particles may be an important regulatory factor in the distribution of the apoproteins (3,4,6), while the composition of the core lipid may be of lesser importance (4). Other factors

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y Y

OR

K U

NIV

on

11/2

0/14

For

pers

onal

use

onl

y.

TONG AND KUKSIS: I 827

contributing to the apoprotein binding by the artificial lipid particles have been recognized in the content of free cholesterol and LPC in the surface monolayer (8) and of Ca2+ ions in the incubation medium (9).

In an extension of these studies we have examined the effect of substituting all or part of the surface PC with SPH or PE upon the stability, particle size, rate of plasma clearance, and apoprotein binding in rats using the strategy of Connelly and Kuksis (4). PE and SPH are components of the surface coating of chylomicrons (lo), but their role in the binding of apoproteins and chylo- micron clearance has not been investigated. Our results show that the phospholipid class composition of the particle surface affects both distribution and relative strength of binding of the apoproteins, as well as the rate of clearance of the particles. The physiological signifi- cance of these observations is discussed.

Materials and methods Materials

SPH, PC, and PE, all from egg yolk, were obtained through Sigma Chemical Co., St. Louis, MO; synthetic DPPC was also from Sigma. These preparations were found to be 98% or more pure (single lipid class) by TLC (10).

Lipid emulsions TG emulsions stabilized with pure SPH, PC or DPPC, and

PC-PE (4: 1) were prepared in the laboratory by dispersing 10% lipid in 90% saline (4). Of the lipid, 90% by weight was TG and 10% was phospholipid. The lipid was dissolved in 1 mL of diethyl ether in a 40-mL conical Pyrex tube (heavy wall, 28 x 133 mm) and the solvent was evaporated under a stream of nitrogen. After addition of saline the contents were agitated using a vortex mixer to effect an initial dispersion. The samples were then sonicated above the transition temperature of the phospholipids (45°C) and trio1eoylglycero1(21"C) using Bronson W-350 sonifier (Bronson Ultrasonics Corp., Dan- burt, CT) at a power setting of 3 (75 W) for 15 min with 10-s interruptions after each minute of sonication. The final emulsions appeared milky but were stable for several hours or days at room temperature. Alternatively, PC-PE (4:l) stabi- lized and pure PE-stabilized emulsions were made by a modification of the method of Stollery and Vail (1 1). The lipids were suspended in 0.005 M glycine-NaOH - 1 mM EDTA buffer at pH 9.2 and sonicated as above. After sonication the resulting milky emulsion was dialyzed against a lOmM Tris-HC1 buffer (pH 7.3) overnight. The final emul- sion had a pH 7.3 and showed no tendency to precipitate.

The emulsions were separated by ultracentrifugation into Sf 2 400, and d 5 1.006, 1.063, and 1.21 g/mL fractions as described by Connelly and Kuksis (4).

Animal preparations Male rats of the Wistar strain (400-450 g) were obtained

from Charles River (Montrkal (QuC.)) and were fed on laboratory chow. After fasting overnight, the tail vein was cannulated (12) and infused with 2 mL of emulsion. After 10 min the rat was exsanguinated via the abdominal aorta under diethyl ether anaesthesia. The blood was transferred to chilled graduated plastic centrifuge tubes and Na2EDTA was added to a final concentration of 1 mg/mL blood.

Lipoprotein isolation Plasma was obtained by centrifuging at 1400 x g for 20

min. It was overlaid with a salt solution of d = 1.006 g/mL and the chylomicron or Sf 2 400 fraction was isolated (15"C, 22 000 x g, 15 min). It was suspended in salt solution (d = 1.006 g/mL) and washed four times (three times 22 000 x g x 15 min and 65 000 x g x 18 h). Up to four washes were necessary to eliminate contamination with traces of VLDL, as indicated by the presence of apoB. The remaining plasma lipoproteins were isolated by flotation at 4OC after adjusting the sample to the appropriate density by means of solid KBr as previously described (4, 8, 9).

Lipid analysis Total lipid profiles of whole plasma and of each lipoprotein

and emulsion fraction were determined by direct GLC (13) using an automated Hewlett-Packard model 5700A gas chromatograph (Hewlett-Packard, Palo Alto, CA). An aliquot of plasma (0.2 mL) or an equivalent amount of lipid from a lipoprotein or emulsion fraction was dephosphorylated with phospholipase C (Bacillus cereus), which converts the diacyl- glycerophospholipids into diacylglycerols and the sphingomy- elins into ceramides. The different lipid classes are quantitated by high-temperature GLC (175-350°C), following total lipid extraction and trimethylsilylation in the presence of tridecan- oylglycerol used as internal standard. Total lipid extracts were obtained with chloroform-methanol (2:l) (14).

Electron microscopy and calculation of particle size Particle size of Sf 2 400 fractions of plasma lipoproteins and

of the synthetic emulsions was determined after fixing the samples in 1% Os04 (15). The samples were examined in a Phillips 300 electron microscope at 60keV accelerating voltage. The particle size was measured using negative film with S-45 Durst enlarger (Durst Laboratories, Balzano-Bozen, Italy). The particle size enlarged an appropriate number of times was characterized by two points representing the diameter and this length was stored in the computer (Zenith Data Systems, Zenith Corp., Chicago, IL) and used to measure at least 200 particles, from which an average diameter was calculated.

In addition, the average core radii of the lipid particles were obtained from the surface to volume ratio by assigning all free cholesterol and phospholipid to the surface monolayer and the triacylglycerols and cholesteryl esters to the lipid core. The total radii of the particles were obtained by adding the thickness of the surface monolayer, 2.0nm, as previously described (1 6).

Apolipoprotein analysis For this purpose the lipoprotein samples were dialyzed

against 25 mM Tris-HC1 - 0.1% EDTA buffer (pH 7.3). Protein concentration was determined by the method of Peterson (17) using bovine serum albumin as standard. Lipid turbidity was removed by extracting twice with petro1e;m ether after color development. For electrophoresis an aliquot of 50 pg of protein was lyophilized and delipidated with absolute diethyl ether - ethanol (1:3) (18). The protein residue was dissolved in 40 pL of 2% SDS - 50 mM Tris-HC1 (pH 8.0) plus 20 mg sucrose, by incubation for 2 min at 100OC. Ten microlitres of 0.02% Pyronin Y (Bio-Rad Laboratories, Richmond, CA) and 10 pL of 400 mM dithiothreitol were added prior to electrophoresis. SDS-PAGE was performed

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y Y

OR

K U

NIV

on

11/2

0/14

For

pers

onal

use

onl

y.

BIOCHEM. CELL BIOL. VOL. 64, 1986

TABLE 1 . Mass distribution of lipids in different density classes of synthetic emulsions and post- infusion plasma

Weight % Density

class PC DPPC PE* PC-PE* PC-PEP DPPC-PEP SPH

Synthetic emulsion Sf 2 400 96.1 96.8 96.7 98.0 99.5 97.9 97.6 d 1.006 2.1 3.2 2.8 1.6 0.4 1.2 1.3 d 5 1.063 1.8 0.0 0.5 0.4 0.1 0.9 1 . 1 d = 1.21 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Postinfusion plasma Sf r 400 86.854 65,624 73.754 72.422 51.027 76.95 1 70.323 d < 1.21 13.2 34.4 26.3 27.6 49.0 23.1 29.7

NOTE: Values for emulsions represent averages of two determinations. Values for Sf 2 400 fraction of postinfusion plasma represent mean ? SD from four to five determinations. All mixed emulsions were 80% PC - 20% PE.

*Emulsions prepared at neutral pH. tEmulsions prepared at pH 9 . 2 and then brought to pH 7 .3 by dialysis against buffer.

using a modification of the glycerol method described by Con- nelly and Kuksis (19). A 4% acrylamide - 18% glycerol gel was prepared and used in a model 155 tube gel apparatus (Bio- Rad, Richmond, CA) equipped with a Pharmacia (Pharmacia, Stockholm, Sweden) power supply at a constant current of 4rnA/gel for 4.5 h, until the tracking dye (Pyronin Y) was 1 cm from the end of each gel. Gels were stained in 9.2% acetic acid - 45.4% methanol - 0.025% Commassie brilliant blue (3-250 for 2 h and were destained in 7.5% acetic acid - 5% methanol. Protein standards for molecular weight determina- tion consisted of a lyophilized mixture of thyroglobulin (MW 670 000) gamma-globulin (MW 158 OW), ovalbumin (MW 44 000), and myoglobin (MW 17 000) (Bio-Rad) and were run along with each set of tubes. Following destaining, the gels were scanned by optical densitometry in a Joyce-Loebl Chromoscan Mark I1 densitometer (Joyce-Loebl Co., Gates- head, U.K.). The percentage of each apolipoprotein compo- nent was calculated on the basis of the distribution of the optical density.

Determination of relative strength of apoprotein binding The relative strength of apoprotein binding was assessed by

determining the apoprotein profile of the recovered emulsion particles after a different number of times of washing under standard conditions. The standard wash consisted of suspen- ding the particles in salt solution ( d = 1.006g/mL) and cenMuging at 22000 X g for 15 min. After each wash an aliquot of the sample was taken for SDS-PAGE and the apolipprotein profile was determined by scanning the gels that had been stained and destained undw standard conditions.

Results Characterization of lipid emulsions

The artificial lipid emulsions used in the apoprotein binding were characterized on the basis of their density distribution, chemical composition, and particle size. The mass distribution of the lipids in the different density classes of the emulsions and in the postinfusion plasma is given in Table 1 . It is seen that more than 96%

TABLE 2. Chemical composition of Sf 2 400 particles from synthetic emulsion and from postinfusion plasma

Weight % Source of particles FC PL TG

PC DPPC SPH PC-PE DPPC-PE PE* PC-PE*

PC DPPC SPH PC-PE DPPC-PE PE* PC-PE*

-.

Emulsion particles 0.1 5 .1 0.0 3.7 0.0 3.6 0.0 5.0 0.0 3.2 0.0 6.2 0.0 7.4

Postinfusion plasma 0.550.1 5.521 0.720.5 3.3k0.1 0.520.1 3.7+1 0.4k0.2 15.020.2

0.0420 4.1 20. 1 0.620.1 7.322.4 0.620.1 7.7k2.0

NOTE: Values for emulsion particles are averages of two determinations. Values for plasma are means 2 SD from four to five experiments.

'Emulsions prepared at pH 9 . 2 and then brought to pH 7 .3 before infusion.

of the total mass of each emulsion was recovered in the Sf 2 400 fraction, 1-395 in the VLDL range (d 5 1.006 g/rnL), and the remainder in the LDL range (d 5 1.063 g/mL), which was attributed to the mesophase of the phospholipid and presumably possessed a liposomal structure (20. 21). No material was recovered in the HDL (d 5 1.21 g/mL) range. In view of the minimal amount of the mesophase, these emulsions were injected into the animals without further fractionation.

The chemical composition of the Sf r 400 emulsion

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y Y

OR

K U

NIV

on

11/2

0/14

For

pers

onal

use

onl

y.

TONG AND KUKSIS: I 829

SPH

FIG. 1 . Electron micrographs of Sf r 400 particles from synthetic emulsion (a) and from postinfusion plasma (b ) . SPH, particles stabilized with egg yolk sphingomyelin; PE, particles stabilized with egg yolk phosphatidylethanolamine at pH 9.2 then brought to pH 7 .3 before infusion. Bar represents nanometre units.

particles and of postinfusion plasma is given in Table 2. The particles of Sf r 400 consisted primarily of TG (93-97%), with the phospholipids accounting for 3-7% by weight. The particles in the LDL density range contained over 88% phospholipid and maximum of 10% of TG (data not shown). Both free and esterified cholesterol were absent.

The electron micrographs of the SPH- and the PE-stabilized emulsion particles of Sf r 400 are shown in Fig. 1. The particles possess a spherical structure, as anticipated, but vary considerably in size. The size distribution of all the preparations is given in Fig. 2. The SPH particles are two to three times larger than the PC particles. The pure PE-stabilized particles prepared under alkaline pH had the smallest size. The addition of 20% PE had no significant effect upon the particle size formed using PC or DPPC alone. Table 3 compares the measured and calculated average diameters for the Sf s

400 lipid particles. The agreement is generally good, which attests further to the validity of the assumptions about their structure.

Effect of injected emulsion on plasma lipids Figure 3 compares the total lipid profiles of rat plasma

10 min after injection of the PE, SPH, and DPPC emulsions. The presence of the emulsion lipids is clearly seen. Before infusion, the TG accounted for 36% and phospholipids, represented as diacylglycerols and ceramides, accounted for 25% of the total lipids by mass. In the postinfusion plasma, the TG make up 50% or more of the total lipid. Furthermore, the postinfusion plasma TG is distinguished by the presence of increased proportions of the CS4 molecular species of soybean oil TG. Likewise, the plasma from animals receiving the DPPC- and SPH-stabilized particles shows the presence of increased proportions of the characteristic C34 diacyl-

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y Y

OR

K U

NIV

on

11/2

0/14

For

pers

onal

use

onl

y.

830 BIOCHEM. CELL BIOL. VOL. 64, 1986

1 DPPC a 1 I

20

10

0 200 400 600 800 200 400 600 800

Diameter, n m

FIG. 2. Size distribution of Sf 2 400 particles from synthetic emulsion ( a ) and from postinfusion plasma (b). PE, as in Fig. 1; PC, particles stabilized with egg yolk phospha- tidylcholine; PC-PE, particles stabilized with egg yolk phos- phatidylcholine and phosphatidylethanolamine in a 4: 1 ratio at neutral pH; DPPC, particles stabilized with dipalmitoyl

TABLE 3. Comparison of measured and calcu- lated a;erage diameters of the Sf 2 400 par- ticles from synthetic emulsion and postinfusion

plasma

Diameter, nm Source of particles Measured Calculated

Emulsion particles PC* 2582 86 230 DPPC* 3232 101 327 SPH* 546k 144 504 PC-PE* 255 2 85 240 DPPC-PE* 3552 127 379 PET 171266 177

Postinfusion plasma PC 204k74 196 DPPC 257 k 86 294 SPH 3722 142 396 PC-PE 103263 76 DPPC-PE 218k89 29 1 PE 161k49 138

NOTE: Experimental values are means 5 SD derived from counting a minimum of 200 particles. Calculated diameters (D) based on average lipid composition and the equation for calcu- lation is shown as follows:

D = ~(VTG[TGI + VCE[CEI)/(APL[PLI + AFC[FCI)

VTO and VCE are the molecular volumes of triacylglycerol and cholesteryl ester, respectively. An and AFC are the molecular surface areas of phospholipid and free cholesterol, respec- tively. [TG], [CE], [PL], and [FC] are the mole percentages of triacylglycerol, cholesteryl ester, phospholipid, and free cholesterol, respectively. *P < 0.001 compared with postinfusion particle size. tP < 0.05 compared with postinfusion particle size.

Table 5 gives the recovery of the injected Sf 2 400 emulsion particles from the plasma after 10 min of in vivo circulation. The recovery was calculated on the basis of the lipid content of the particles isolated in the Sf 2 400 range and the total blood volume. For the PC emulsion, the recovery was 57%, which is similar to that reported by Connelly and Kuksis (4). The recoveries of the PC-PE emulsions prepared under neutral pH were the lowest and varied from 1 1 to 17%.

phosphakdilcholine; DPPC-PE, particles stabilizd with di- Chnracreriiation ofrecovered lipid particles palmitoyl phosphatidylcholine and phosphatidylethanolamine in a 4:1 ratio; SPH, as in Fig. 1. Data ae divided into 50-nm The lipid paRicles were characterized

intervals. +, diameter. their density distribution, chemical composition, parti- cle size, and apoprotein makeup. The mass distribution

glycerol and C34 and C38 ceramide moieties, respec- of the postinfusion plasma lipoproteins is shown in tively. TLC analyses showed that the recovered PE Table 1. The Sf r 400 fraction, which was absent in the emulsions contained some PC that had been transferred fasting state, now accounts for 54-87% of the total to the emulsion from the plasma lipoproteins. The lipid lipoprotein mass. There were marginal increases in the class composition of the postinfusion and of the control absolute amounts of the plasma LDL and HDL. plasma is given in Table 4. Table 2 gives the lipid composition of the recovered

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y Y

OR

K U

NIV

on

11/2

0/14

For

pers

onal

use

onl

y.

TONG AND KLJKSIS: I

FIG. 3. Plasma total lipid profiles of rats following infusion of emulsions stabilized with egg yolk phosphatidylethanolamine stabilized at pH 9.2 and brought to pH 7.3 before infusion (PE), egg yolk sphingomyelin (SPH), dipalrnitoyl phosphatidylcholine (DPPC), or of fasting animals (control). Peaks: 16-18, TMS esters of free fatty acids of 16-18 acyl carbons; 20-24, TMS ethers of monoacylglycerols with 16-20 acyl carbons; 27, TMS ether of free cholesterol; 30, tridecanoylglycerol, internal standard; 34-42, TMS ethers of diacylglycerols with 32-40 acyl carbons (PC-PE emulsions) or TMS ethers of ceramides with 32-40 fatty chain carbons (SPH emulsions); 43-47, cholesteryl esters of fatty acids with 16-20 acyl carbons; 48-56, triacylglycerols with a total of 48-56 acyl carbons. GLC conditions and sample size as given elsewhere (12).

TABLE 4. Lipid composition of fasting and of postinfusion plasma

Weight % - Source of plasma FC PL CE TG FC/PL

Fasting plasma Control 6.820.5 24.923.7 31.822.1 36.525.1 0.3

Postinfusion plasma PC 1.420:3 14.02 1.9 9.5k1.9 75.1k3.2 0.1 DPPC 3.920.7 11.9k2.5 7.9k2.3 76.3k5.5 0.3 SPH 6.221.4 26.026.9 14.8a2.0 52.9k5.5 0.2 PC-PE 5.6k1.4 34.022.0 12.923.8 49.424.6 0.2 DPPC-PE 2.620.7 20.224.0 14.022.2 63.226.8 0.1 PE 1.820.2 19.121.4 9.220.5 69.9+ 1.7 0.1 PC-PE 3.1"0.9 17.423.4 10.6k0.5 68.9k4.8 0.2

NOTE: Postinfusion plasma values are means * SD of four to five experiments. Control plasma was from fasting rats and the values represent means t SD from four separate detenninations.

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y Y

OR

K U

NIV

on

11/2

0/14

For

pers

onal

use

onl

y.

832 BIOCHEM. CELL BIOL. VOL. 64, 1986

TABLE 5. Recovery of infused emulsion (Sf 2 400) particles from plasma after 10 min of in vivo circulation

- -

Emulsions

PC DPPC SPH PC-PE DPPC-PE PE* PC-PE*

No. of experiments 4 3 5 3 3 4 2

% recovery 56.9?11 50.9?8 70.0?7 14.2?3t$ 35.4?6 23.2?47 32.3?2

NOTE: Values represent means + SD for the number of experiments indicated. *Emulsions prepared at pH 9.2 and then brought to pH 7.3 before infusion. t P < 0.05 compared with PC group. $P < 0.05 compared with DPPC-PE group.

particles of Sf r 400. In comparison to those of the emulsion, the plasma particles in most instances contain somewhat more phospholipid and less triacylglycerol. Nevertheless, the triacylglycerol still accounts for 8.5- 96% of the total lipid. The recovered particles also contain small amounts of free cholesterol and choles- teryl esters. Only the free cholesterol has been tabulated (<I%).

The electron micrographs of the recovered particles of Sf 2 400 are shown in Fig. 1, while the particle size distribution is given in Fig. 2. The diameters of the recovered Sf r 400 particles are given in Table 3. The average size of the recovered particles in all instances is significantly smaller than that of the injected particles. There is a good agreement between the measured and the calculated values for the diameters of the recovered particles, except for the PC-PE particles, which were underestimated by calculation.

The SDS-PAGE profiles of the apoproteins of the recovered particles of Sf 2 400 are given in Fig. 4 and the quantitative evaluations of the results are tabulated in Table 6. The SDS-PAGE system, which was modified from the work of Connelly and Kuksis (19), permitted the resolution of apoC-11, apoC-111, apoA-I, apoE, apoA-IV, and apoB-48 and apoB-100, when present. In addition to variable amounts of the A, C, and E apoproteins, all the samples contained albumin and higher molecular weight proteins, of which an unknown protein of MW 117 000 was very prominent in some samples. There were minor differences in the quantita- tive apoprotein patterns of the particles recovered from different rats, as indicated by the standard deviations in Table 6. The recoveries of the apoproteins varied with the extent of the washing. After four washes there was no evidence of any remaining apoB, which would have indicated cross-contamination with VLDL. As a result of washing there were losses of apoC, which greatly exceeded those anticipated from the removal of the apoC contained in the VLDL.

a b c d e f g h - - -

- 670 000

117000- Alb - A-IV - g.r E -a A - I -- c-I l l - A

FIG. 4. SDS-polyacrylamide profiles of apoproteins asso- ciated with Sf r 400 emulsion particles following 10 min of circulation in the blood stream. (a) PE particles stabilized at pH 9.2; (b) PC-PE particles stabilized at pH 9.2; (c ) PC-PE particles stabilized at neutral pH; ( d ) PC particles: (t) DPPC particles; V) DPPC-PE particles; (g) SPH particle&; ( h ) molecuIar weight standard. SDS-PAGEconditions as given in text. Abbreviations of emulsions as in Figs. 1 and 2. Alb, albumin. A-I, A-IV, C-11, C-111, and E are all apolipoproteins.

Discussion In this study we have attempted to determine the

effect of different neutral and weakly acidic phospholip- ids as sole or major components of the surface mono- layer upon plasma protein binding and clearance of injected artificial emulsions. For this purpose we have prepared synthetic TG emulsions of a particle size and po1ar:neutral lipid ratio comparable with that of chylo- microns.

Stability of emulsions Exchanges of the surface phospholipids were made

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y Y

OR

K U

NIV

on

11/2

0/14

For

pers

onal

use

onl

y.

TONG AND KUKSIS: I 833

TABLE 6. Quantitative estimates of apoproteins associated with the recovered emulsion (Sf 2 400) particles

OD % - - - -

117000 Protein Emulsions C-II C-1n A-I E A-IV Albumin MW Other content*

PC 23.622 33.526 7.821 28.121 1.920 2.121 0.0 3.020 1.620.5 DPPC 20.521 14.321t 4.020f 19.521$ 13.222t 19.5+1$ 0.0 9.1+-1$ 1.010 SPH 12.023t 13.0+ 17 12.52 l t 18.4+5 8.020t 23.9257 6.32 1 6.02 l t 4.222 PC-PE 4.8*0$ 6.920t 8.620 10.8klf 2.821 34.7+9t 19.427 12.021$ 2.42 1 DPPC-PE 6.021$ 5.121t 8.622 5.820s 4.321 52.523f 5.9%2 12.0+-lj: 4.5?0 PE 19.222 13.5227 15.42lt 2.923 11.222t 6.022 6.122 4.221 4.2Ic-0 PC-PE 16.822 16.921 10.820 17,221 15.621 7.122 9.922 5.721 2.5+1

- - --- -

Nore: Values are means k SD from three replicates. C-11, C-111, A-1, E. and A-IV are apoliproteins as commonly defined ( 3 ) . *Rotein content is expressed as milligrams protein per milligram protein plus lipid. tP < 0.05 compared with PC. $ P < 0.01 compared with PC group. OP < 0.001 compared with PC group.

within the experimental limits compatible with both the mechanical and chemical stability of the particles, while the composition of the neutral core lipid remained the same. The prepared particles closely resembled those of Intralipid as verified by centrifugation, electron micro- graph~, and chemical analyses (4).

It was not possible to prepare stable emulsions using PE as the sole surfactant at physiological pH. However, stable emulsions could be made when the PE was incorporated into PC up to 20%. These findings are consistent with the earlier reported difficulty in obtain- ing pure PE or PC-PE liposomes above 0.7 mole fraction PE (22) at neutral pH. The instability of pure and mixed PE vesicles was thought to be a reflection of the formation of a hexagonal phase, which is incompati- ble with the lamellar structure of the vesicle wall. In mixtures with PC, PE micelles yielded vesicles of the same size as PC. There was no significant difference in the particle size prepared with 80% PC - 20% PE and 100% PC. Other experiments have also shown that vesicle size is not a parameter totally dependent upon the chemical composition of the particle (23). It varies with the sonication time. Furthermore, a study of the incor- poration of PE into PC vesicles has shown that ideal mixing occurs below an initial mole fraction of 0.2 (22). However, pure PE stabilized TG emulsions could be readily obtained by sonication at pH 9.2, as first reported by Stollery and Vail(11). These emulsions were brought to pH 7.3 by dialysis against a buffer and they remained stable during infusion and recovery from the animals. There was no difficulty in preparing emulsions stabi- lized with pure DPPC or pure SPH when the sonication was performed above the transition temperature of these lipids. This is consistent with the ready formation of pure DPPC and SPH liposomes (24) and may be related

to the ease of formation of the lamellar phase by these phospholipids in aqueous solutions.

Plasma clearance The various emulsions differed in the relative rates of

plasma clearance. Those stabilized with SPH or PC were cleared relatively slowly (30-43% in 10 min), while those stabilized with pure PE or PE-PC were cleared more rapidly (65-86% in 10 min). The SPH-stabilized emulsions were cleared slowest. The slower clearance of the SPH emulsions is consistent with the relatively lower rates of clearance of SPH liposomes, when compared with PC liposomes (25). The DPPC-stabilized emulsions were cleared at about the same rate as the PC-stabilized emulsions.

The most rapidly cleared were the emulsions stabi- lized with the PC-PE mixtures. Furthermore, the emulsions stabilized with PC-PE were cleared signifi- cantly faster than those stabilized with DPPC-PE. This rapid clearance of the PE-stabilized emulsions may be related to the rapid disappearance of chylomicron PE reported by Landin and Nilsson (26). These workers demonstrated that there is a mechanism for rapid removal of chylomicron PE, which maintains a low plasma PE level despite a significant transport of PE with the intestinal lipoproteins. Approximately 70% of the chylomicron PE was cleared during the first 10 min after chylomicron injection. However, Landin and Nilsson (25) failed to note any selective loss of the PE from the chylomicrons, which suggests that both PC and PE were cleared at the same time and probably together with the rest of the chylomicron particle. Interestingly, in the present study the particles stabilized with PE alone were cleared at about the same rate as the particles stabilized with 80% PC - 20% PE. The increased rate of

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y Y

OR

K U

NIV

on

11/2

0/14

For

pers

onal

use

onl

y.

a34 BIOCHEM. CELL BIOL. VOL. 64, 1986

clearance of the PE-containing particles may have been related to a more rapid hydrolysis of PE than PC by lipoprotein lipase (27) and to a relative absence of apoC-I11 in these particles (see below).

The synthetic emulsions stabilized with the glycero- phospholipids were presumably cleared via the action of lipoprotein lipase (27) and hepatic lipase (28), as already established for chylomicrons. The mechanism of clear- ance of the SPH-stabilized emulsions is not known. Neither lipoprotein lipase nor hepatic lipase is known to attack SPH, but various tissue extracts (29) possess sphingomyelinases, which act on liposomes and could facilitate the breakdown of the surface coating of these particles, making the TG cores accessible to lipolysis by lipoprotein lipase or other lipases. The SPH species of plasma lipoproteins are not known to equilibrate (30), but they could be transferred to HDL or other lipoprotein fractions upon degradation of the synthetic emulsions. The mechanism of clearance of SPH liposomes also is not known.

Nature of recovered particles Qualitatively, the PC-stabilized emulsions absorbed

those apoproteins generally associated with chylomi- crons, as already reported by Imaizumi et al. (1) and Connelly and Kuksis (4). In striking difference to the chylornicrons of mesenteric lymph of rats, the synthetic emulsions contained no apoB and much less apoA, with the apoC-11, apoC-111, and apoE being the major components of the extensively washed emulsions, accounting for 85% of the total protein. The SPH- stabilized emulsions contained qualitatively the same apoproteins as the PC-stabilized emulsions, but quanti- tatively there were significant differences. Thus, the SPH emulsions retained much more albumin than the PC emulsions, along with an unknown circulating protein of MW 117 000. The emulsions stabilized with DPPC bound less apoC and especially apoA-I than emulsions stabilized with PC. The addition of PE to PC emulsions made at neutral pH led to decreased binding of total apoprotein and of apoC-111. In contrast, emulsions stabilized with pure PE or PC-PE at alkaline pH bound the common apoprotein in amounts intermediate be- tween those seen for the PC and PC-PE emulsions made at neutral pH.

The absence of the B apoproteins from the emulsion surface is anticipated from previous work (31, 32), which has shown that the apoB species are not transfer- red among lipoproteins or between lipoproteins and liposomes. The small amounts of apoB detected in the unwashed particles were apparently due to cross- contamination with plasma VLDL. Both SPH- and DPPC-stabilized emulsions bound much more albumin than the emulsions stabilized with PC or PC-PE. This is consistent with the high affinity of DPPC and SPH

liposomes for surfactant proteins, e.g., albumin (33). The SPH-stabilized emulsion also bound the apoC-I1 and apoC-I11 species less strongly than did any of the PC-stabilized emulsions. After two washes the PC-stabilized emulsions possessed apoprotein profiles, which were similar to those reported by Connelly and Kuksis (4). With further washing there was extensive loss of the C apoproteins, resulting in a preferential retention of the E apoprotein. In contrast, the presence of PE in the stabilizing surfactant led to a loss of E apoproteins with prolonged washing. The relative absence of apoC may have contributed to the rapid clearance of these particles, because apoC-I11 is known to inhibit hepatic uptake of lipid particles containing it (34, 35). However, SPH-stabilized particles were also low in apoC, yet they were cleared much more slowly.

The major forces involved in the binding of apopro- teins to zwitterionic lipid surfaces are believed to be hydrophobic (36). However, the strength of the binding might depend on the intimacy of interaction, which might vary depending on the exact nature of the hydrophobic and polar moieties of the phospholipids. The slight negative charge of the particles stabilized with pure PE would have been expected to repel any negatively charged apoproteins, unless these proteins were able to rearrange their tertiary structures to present a positively charged surface to these particles. Boggs (37) has discussed the possibility that a decreased binding of proteins to PE-containing surfaces might be related to an increased intermolecular interaction obser- ved for these phospholipids. In the present study we observed that the PE-stabilized particles tended to self-associate more than the particles stabilized with PC and SPH, especially following recovery from the plasma. It is possible that this aggregation reduced access to the surface, resulting in a lower protein binding. The tendency of PE to form gross structures which do not permit effective reactivity of the amino and phosphate groups with chemical reagents and enzymes (22) may also minimize any potential charge effects, owing to the presence of the PE in the surface mono- layer.

The present experiments do not permit an evaluation of the effect of the nature of the molecular species of the phospholipid classes upon the apoprotein binding. Although both saturated (DPPC) and unsaturated (egg yolk PC) lipids were employed, the molecular species differed too much in their transition temperatures for direct comparisons. It has been shown that the saturated phosphatidylcholines with lower transition temperatures (e.g., DMPC) react more readily with apoA-I than those with higher transition temperatures (e.g., DPPC) and that the smaller DPPC vesicles were more reactive than the larger ones (38).

In conclusion, the present study suggests that the

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y Y

OR

K U

NIV

on

11/2

0/14

For

pers

onal

use

onl

y.

TONG AND KUKSIS: I 835

characteristic apoprotein composition and relative resis- tance to displacement in chylomicrons may be specifi- cally and largely related to the presence of PC along with the minor PE and SPH contributions. Introduction of excess SPH or PE results in an alteration of the composition and strength of binding of the apoproteins. Landin and Nilsson (26) found that PE can exist only at a relatively Iow concentration in rat plasma and is rapidly cleared upon entry into the circulation. Myher et al. (30) have shown that the molecular species of the SPH of plasma lipoproteins do not equilibrate, suggesting that these phospholipids do not undergo a rapid transfer. Other investigators have obtained evidence that SPH- containing liposomes are toxic to mice when adminis- tered chronically (39). Boggs (37) and Cullis et al. (40) have reviewed the possible function of different phos- pholipids in the cell membrane bilayer and have pointed out that many interesting phenomena arise as a direct consequence of changes in phospholipid composition. The effect of different molecular species of PC and of other phospholipids upon apoprotein binding by lipid particles remains to be systematically investigated.

Acknowledgments These studies were supported by funds from the

Ontario Heart Foundation, Toronto, Ont., and the Medical Research Council of Canada, Ottawa, Ont.

1. Imaizumi,K.,Fainaru,M. &Havel,R. J. (1978) J. Lipid Res. 19, 712-722

2. Nilsson, A., Ehnholrn, C. & Floren, C. H. (1981) Biochim. Biophys. Acta 663, 408-420

3. Havel, R. J., Kane, J. P. & Kashyap, M. L. (1973) J. Clin. Invest. 52, 32-38

4. Comelly, P. W. & Kuksis, A. (1981) Biochim. Biophys. Acta 666, 80-89

5. Enkelens, D. W., Chen, C., Mitchell, C. D. & Glomset, J. A. (1981) Biochim. Biophys. Acta 665, 221-233

6. Tajima, S., Yokoyarna, S. & Yamarnoto, A. (1983) J. Biol. Chem. 258, 10073 - 10082

7. Weinberg, R. B. & Spector, M. S. (1985) J. LipidRes. 26,26-37

8. Connelly, P. W. & Kuksis, A. (1983) Can. J. Biochem. Cell Biol. 61, 63-71

9. Connelly, P. W. & Kuksis, A. (1983) Biochim. Biophys. Acta 752, 371-382

10. Huang, T. C. & Kuksis, A. (1967) Lipids 2,443-460 11. Stollery, J. G. & Vail, W. J. (1977) Biochim. Biophys.

Acta 471, 372-390 12. O'Doherty, P. J. A , , Kakis, G. & Kuksis, A. (1973)

Lipids 8, 249-255

13, Kuksis, A., Myher, J. J., Ma&, L. & Geher. K. (1975) J. Chromatogr. Sci. 13. 423430

14. Folch, J., Lees, M. & Sloane Stanley, G. H. (1957) J . Biol. Chem. 226,497-504

15. Frazer, R. (1970) J. LipidRes. 11,60-65 16. Kuksis, A., Breckenridge, W. C., Myher. 3 . J. & Kakis,

G. (1978) Can. J. Biocli~m. 56,630-639 17. Peterson, G. L. (1977) Anal. Riochem. 83, 346-356 18. Scanu, A. M. & Edelstein, C. (1971) Anal. Biochcm. 44.

576-588 19. Connelly, P. W. L Kuksis, A. (1982) Biochim. Biophys,

Acta 71 1,245-25 1 20. Breckenridge, W. C., Kakis, G . & Kuksis, A. (1978)

Can. J. Biockm. 57, 72-82 21. Kakis, G. & Kuksis, A. (1984) Can. J. Biochem. C d

Biol. 62, 1-10 22. Litman, B. J . (1973) Binchcmishy 12. 2545-2554 23. luliano, R. L. & Stamp, D. (1975) Biochem. Biophys.

Rex. Commun. 63, 651-658 24. Gemitsen, W. J . , Verkely, A. J., Zwaal, R. F. A. & Van

Deenen, L. L. M. (1978) Eur. J . Biochem. 85,255-261 25. Senior, J . Br Gregoriadis, G. (1982) FEBS Lett. 145.

109-1 14 26. Landin, B. & NiLson, A. (1 984) Biochim. Biophys. Acta

793, 105-1 13 27. Groot, P. H. E., Oerlernans, M. C. & Scheek, L. M.

( 1978) Biochim. Biophys. Acta 530, 91-98 28, Groot, P. H. E. &Van Tol, A. (1978) Biochim. Biopbs.

Acta 530, 188-196 29. Poulos, A., Shankamn, P., Jones, C. S. & Callahan, J .

W . (1983) Biochim. Biophys. Acta 751. 428-431 30. Myher, J. J., Kuksis, A., Breckenridge, W. C. &Little,

J. A. (1981) Can. J . Biochem. 59,626-636 31. Faergeman, 0.. Sata, I., Kane. J. P. & Havel, R. J.

(1975) J . Clin. Invest. 56, 1396- 1403 32. Granot, E., Deckelbaum, R. J., Eisenberg, S., Oschry,

Y . & Bengtsson-Otivecrona, G. (1985) Biochim. Bio- phys. Acta 833, 308-315

33. Slotte, 1. P. & Lundberg, B. (1983) Biochim. Biopiys. Acta 750, 434-439

34. Shelbume, F., Hanks, J.. Meyers, W. & Quarfordt, S. (1 980) J. Clm. Invest. 65, 652-658

35. Windler, E. & HaveI, R. J. (5985) J . Lipid Res. 26, 556-565

36. Stoffel, W.. Ziwenberg, O., Tunggal, B. & Schreiber,E. (1974) Proc. Natl. Acad. Sci. U.S.A. 71,3696-3700

37. Boggs, J. M. (1980) Can. f . Biochem. 58,755-770 38. Wettereau, I. R. &Jaw. A. (1982) 3. Biol. Chem. 257,

10961 - 10966 39. Weemtne, E. A. H., Gregoriadis,G. &Crow, I. (1983)

Br. J . Pathol. 64, 670-676 40. CulIis, P. R., DeKruijff. B . , Hope, M. J.. Verkleij, A.

T., Nayar, R., Farren. S. B..Tilcock, C., Madden, T. D. & Bally, M. B. (1983) in Membrane Fluidity in Biology, vol. 1, pp. 39-8 1, Academic Ress, Inc., New York

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y Y

OR

K U

NIV

on

11/2

0/14

For

pers

onal

use

onl

y.