155

Atherosclerosis, 84 (1990) 155-163 Elsevier Scientific Publishers Ireland, Ltd. ATHERO 04531

Apolipoprotein

E-rich HDL in patients with homozygous familial hypercholesterolemia

Shlomo Keidar, Richard E. Ostlund, Jr. and Gustav Schonfeld Atherosclerosis, Lipid Research and Metabolism Divisions, Washington University School of Medicine, St. Louis. MO 631IO (U.S.A.) (Received 8 January, 1990) (Revised received 18 May, 1990) (Accepted 23 May, 1990)

Ordinarily, HDL,, a fraction of HDL enriched in apoE, is a minor fraction of plasma, but in human subjects and experimental animals eating diets high in fat and cholesterol and in patients with homozygous familial hypercholesterolemia (HFH) or CETP deficiency, HDL, (or HDL,) concentrations in plasma are increased. However, little is known about the structures, compositions and metabolic sources of HDL, in HFH patients. To obtain HDL, for the study, we surveyed several fractions in the HDL density range for apoE by SDS-PAGE. The ratio of apoE to apoA1 in the HDL (d = 1.063-1.21 g/ml) of 8 HFH patients was 0.14 f 0.03 compared to 0.03 f 0.005 in a control group of 8 normolipidemic subjects (P < 0.001) suggesting that an apoE-rich fraction indeed was present in increased amounts. ApoE/apoAI ratios of lipoproteins of the density range 1.050-1.090 were even higher at 1.5 and 2.0 in 2 patients compared to 0.4 & 0.1 in controls, indicating that this density fraction may be particularly enriched with apoE-rich lipoproteins. By contrast, d = 1.020-1.050 g/ml and d > 1.090 fractions contained very little apoE. Therefore, we further characterized the d = 1.050-1.090 g/ml lipoproteins of HFH patients and controls. Fractionation of an d = 1.050-1.090 fraction by concanavalin-A chromatography (CONA) yielded an unbound apoE-rich fraction that contained apoE, apoA1 and apoC but no apoB, and a bound LDL-like fraction that contained mostly apoB-100, as determined by SDS-PAGE and by solid phase immunoassays, containing monoclonal antibodies directed against apoB, apoE and apoA1. The apoE/apoAI ratio of the CONA unbound fraction of HFH patients was greater, and the fraction also contained more free cholesterol and phospholipids than the fraction of control subjects. The diameters of these HDL, particles, determined by nondenaturing gradient gel electrophoresis, ranged from 12.2 to 17 run. HDL, of HFH patients were slightly larger than HDL, of controls. HDL, particles associated with and were degraded by cultured normal human skin fibroblasts with higher affinity than the LDL-like particles that were bound to the CONA column. Thus, in the composition and metabolic behavior HDL, isolated from the plasma of a fasted HFH patient resembled the HDL, (or HDL,) seen in animals fed diets enriched in cholesterol and fats. Key words: Apo E-containing HDL; HDL,; Familial hypercholesterolemia Correspondence to: Gustav Schonfeld, M.D., Lipid Research Center, 4566 Scott Avenue, Box 8046, St. Louis, MO 63110, U.S.A. 0021-9150/90/$03.50

0 1990 Ekevier Scientific Publishers Ireland, Ltd.

156 Introduction Familial hypercholesterolemia (FH) is characterized by elevated low density lipoprotein (LDL)-cholesterol concentrations in plasma, premature atherosclerosis, autosomal co-dominant inheritance, and the deposition of cholesteryl esters in tissues such as skin, tendons, and arteries [l]. In addition to LDL, HDL levels also have some prognostic value for CHD even in the presence of hypercholesterolemia, and many FH patients have low plasma concentrations of high density lipoproteins (HDL), with the HDL, subfraction being particularly depressed [2-41. HDL can be divided into HDL,, HDL,, and HDL, subfractions. Whereas the HDL, fraction in normolipidemic plasma is a minor fraction, in patients with homozygous familial hypercholesterolemia, where HDL, and HDL, concentrations may be quite low, HDL, is frequently 5-lo-fold increased over normal [5]. HDL, also is elevated in subjects with CETP deficiency [6] and in humans and animals eating diets high in saturated fatty acids and cholesterol [7]. Under these dietary conditions, HDL, is called HDL,. HDL, contains much of the apolipoprotein E (apoE) found in HDL, and variable amounts of apoA1. Therefore HDL, has also been called apoE-rich HDL. Lipoproteins that contain apoE can bind to both hepatic and extra-hepatic LDL receptors and to hepatic apoE receptors. Thus apoE is important in mediating the catabolism of apoE-containing lipoproteins in plasma, including VLDL, IDL, chylomicron remnants and HDL, [7]. The metabolic source of HDL, or HDL, in plasma is unknown, but it has been proposed that these particles may arise in peripheral tissues and function as intermediates in the mobilization of cholesterol from peripheral cells and its transport to the liver [8]. However, a portion of plasma HDL also originates from ‘surface remnants’ during the catabolism in plasma of triglyceride-rich lipoproteins [9]. ApoE-rich HDL could represent the mature form of some of these surface remnants. Our aims were to determine whether the apoErich HDL fraction of homozygous FH patients resembled HDL, in structure, composition and function. Accordingly, we isolated apoE-rich HDL

particles from FH patients, determined their compositions, sizes and interactions with normal human skin fibroblasts. Materials and methods Patients

The diagnosis of HFH was based upon the following criteria: (a) total plasma cholesterol concentration in excess of 500 mg/dl, (b) clinical or biochemical evidence of the heterozygous state in both parents, and (c) xanthomatosis occurring before age 10. Seven patients were from Arabic families living in the northern part of Israel. Genetic studies of the LDL receptor were performed on fibroblasts of 3 patients in the laboratory of Brown and Goldstein in Dallas and found to have the ‘Lebanese allele’ [lO,ll]. Patients S.L. and R.L. are brothers who live in Illinois and are compound heterozygotes for FH and have been reported on before [12]. Patient S.L. is undergoing regular plasmapheresis, the other HFH patients are treated by diet and various cholesterol lowering drugs. Probucol was not taken by any patients. The controls were normal volunteers from St. Louis aged 24-42 years of both sexes whose mean lipid levels are near the 50th percentile of the Lipid Research Clinics Population Study [ 131. Laboratory studies

Plasmas were separated immediately after blood collection by centrifugation (2500 rpm, for 25 min at 4” C). Aliquots were taken for lipid and apoprotein determinations. To prevent degradation of apoproteins, gentamicin (0.2 mg/ml), D-phenylalanyl-L-propyl-Larginine chloromethyl ketone (PPACK) (20 PM) and D-phenylalanyl-L-arginine-CMK (20 @l) were added to the plasma and ultracentrifugation was begun on the same day that blood was drawn. VLDL (d -c 1006 g/ml) was isolated at plasma density in a Beckman L8-55 ultracentrifuge (Beckman Instruments, Inc., Palo Alto, CA) using a 50.2 Ti rotor (45000 X g for 18 h at 10°C) and washed by recentrifugation in EDTA saline under the same conditions. Lipoproteins in the density ranges were separated by ultracentrifugation and re-centrifuged at their upper density limits (g/ml): d = 1.006-1.019, IDL; d = 1.020-1.050, LDL; d

157 = 1.063-1.124, HDL,; d = 1.090-1.21, “HDL,“; and d = 1.063-1.21, total HDL. Some apoproteins may have been ‘stripped off the lipoproteins during the centrifugations. The quantitative importance of this is unknown. To the extent that it happened, the compositions we report may not accurately reflect compositions in vivo. Quantitation of lipoproteins, lipids, and apoproteins

Cholesterol and triglycerides in total plasma and in lipoproteins were measured according to standard Lipid Research Clinic techniques in the Core Laboratory of the Washington University Lipid Research Center. Triglyceride, cholesterol ester, free cholesterol, and phospholipid contents of isolated lipoproteins were determined by commercially available enzymatic calorimetric methods (Wako Pure Chemical Industries, Ltd., Osaka, Japan). Phospholipids were also determined by the Bartlett procedure [14]. Protein was measured by the method of Lowry et al. [15]. Apoproteins AI, AII, and apoB were quantified by established radioimmunoassays [16-181. Affinity chromatography harose

on concanavalin A sep-

To remove any apoB from the d = 1.050-1.090 fraction, a concanavalin-A (CONA) sepharose column was used [19] (Pharmacia Fine Chemicals, Inc., Piscataway, NJ). Three to 5 mg of lipoprotein proteins (of density 1.050-1.090) were applied to the column and the unbound fraction was collected at a flow rate of 15 ml per h at 4” C. The equilibration buffer contained 0.05 M Tris, 1.0 M NaCl, 1 mM CaCl,, 1 mM MgCl,, and 1 mM MnCl,, pH 7.0. After the unbound fraction was eluted and the absorbance (at 280 nm) of the eluate had fallen to baseline, the retained fraction was eluted with 0.2 M methyl-&glucopyranoside (Sigma, St. Louis, MO). Column fractions were monitored (at 280 nm) to detect the appearance of retained fractions. The tubes containing the unbound and retained peaks were pooled separately into two aliquots, concentrated by dry dialysis (Aquacide II, Behring Diagnostics, La Jolla, CA) and dialyzed against EDTA-saline for 18 h at 4O C. Recoveries of protein in the bound plus unbound fractions were - 90% of the protein applied.

Electrophoresis

Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) of the apoproteins was performed as previously described [20,21]. Before electrophoresis, the samples were dialyzed against 5 mM ammonium bicarbonate, lyophilized and the dry residue incubated with ethanol/ diethyl ether (3 : 1, v/v) for 18 h at 4°C followed by a wash with anhydrous ether. Proteins were dissolved in loading buffer containing /?mercaptoethanol heated at 100” C for 4 min; at the end of electrophoresis, protein bands were identified by Coomassie blue staining. Stained gels were scanned with an LKB 2202 Ultrascan laser densitometer (LKB Instrument, Inc., Broma, Sweden) and the areas under the curves corresponding to the various apoproteins were measured with a Planix Tamaya 7 digital planimeter (Tamaya and Company, Ltd., Tokyo, Japan). Protein loads were kept within the linear range (pg protein vs. stained area) for gel scanning. Non-denaturing gradient gel electrophoresis (GGE)

To determine relative diameters of lipoproteins, electrophoresis was performed in non-denaturing gradient polyacrylamide gels (3-258) as described by Nichols et al. (22). For immunoblotting lipoproteins were electrophoretically transferred to nitrocellulose in a Tram-Blot cell (Bio-Rad, Richmond, CA). After drying, nitrocellulose transfers were blocked with 5% (w/v) non-fat dry milk in PBS and incubated overnight with the monoclonal antibodies. Transfers were then washed with PBS containing 0.1% Tween-80 and 0.1% Triton X-100, using an Omniblot System (American Bionetics, Emeryville, CA) and incubated with ‘251 affinity purified goat anti-mouse IgG (7.5 X 10’ cpm/ml) in 3% BSA-PBS. The nitrocellulose was washed again with the detergents, dried, and exposed to Kodak X-Omat AR x-ray film at - 70 ’ C. Solid phase antigen assay

The apoprotein contents of lipoproteins in density range 1.050-1.090 g/ml (both the retained and unbound fractions from the CONA column) were studied in a direct assay on microtiter plates (Dynatech Laboratories, Inc., Alexandria, VA). The wells were coated with 150 ~1 of the lipopro-

158 tein (10 pg/ml in PBS) overnight and then the wells were blocked with 3% BSA-PBS. The various monoclonal antibodies 1 pg/ml in 1% BSA-PBS were added followed by a constant amount of goat antimouse IgG radiolabeled with **‘I (1000 cpm/l_rl) in 1% BSA-PBS. After incubation for 4 h at 23’ C the wells were washed and counted. Monoclonal antihuman antibodies Antihuman monoclonal antibodies (MAbs) were produced in mice by using intact VLDL, LDL or HDL as immunogens as previously described 123,241. In these studies we used MAb BlB6 directed against apoB, MAb WU-E4 against apoE and MAb A5.4 against apoA1. Human fibroblasts cultures Monolayer cultures of normal human fibroblasts were grown and maintained as previously described [25]. Cells were seeded into 35 mm dishes 7.5-10 X lo4 cells per well with Eagle’s Modified Essential Medium containing 15% newborn calf serum. After 5 days the cells were washed and the medium was replaced with medium containing 10% lipoprotein-deficient serum for 48 h. The assay contained lZSI-LDL (5 pg/m.l) isolated from a normolipidemic subject as the labeled ligand, and the indicated doses of competitor lipoproteins at 37°C for 4 h in triplicate. At ther::d of the incubations, specific degradation of ILDL by the fibroblasts was determined by radia-

tion counting of TCA-soluble material in the spent medium after removal of free iodide with chloroform. After removal of the medium, cells were washed as described [26], dissolved in 0.1 M NaOH, and aliquots were taken for determination of cell protein and cell-associated radioactivity. Non-specific association and degradation were determined in the presence of 50-fold excesses of non-labeled LDL. A cell-free degradation blank was included in all the assays. Results were calculated as nanograms of ‘2sI-labeled lipoprotein associated or degraded per mg of cell protein after subtracting the non-specific cell association and degradation. Statistical analysis Significant differences between HFH patients and controls were evaluated by Student’s t-test. ReSUltS Clinical characteristics of patients and the concentrations in their plasmas of lipoprotein lipids and apoproteins are presented in Table 1. By definition, the levels of LDL cholesterol and apoB in the homozygous patients were markedly higher than in the control group of normolipidemic subjects. HDL cholesterol, apoA1, and apoAI1 concentrations were significantly reduced to - 50% of controls (P < 0.001). The mass ratio of apoB to apoA1 was almost 10 times higher in the HFH

TABLE 1 CLINICAL

FEATURES

AND PLASMA LIPIDS AND APOPROTEINS

Values are mg/dI. Patient

Age

f&X

Chol

TG

HDL-C

apoA1

apoAI1

apoB

M M F M F M M M M

825 701 804 652 608 622 638 670 634 684&26 200*10

90 185 131 119 128 73 108 119 174 125 f 12 89&11

21 16 22 28 29 30 23 25 22 25k1.5 56&7

88 57 90 87 64 67 92 62 53 73* 5* 177*11

25 18 17 27 15 17 20 16 19 19*1.4 39f3

360 330 542 413 413 321 338 367 365 383f23 * 98* 7

(Yrs) S.L.

10

R.L. 15 A.S. 7 N.M. 23 M.A. 15 M.V. 20 M.H. 17 M.aa. 12 H.V. 26 Mean f SEM Control (n =13)

*

*

*

* P < 0.001 compared to the normolipidemic control. Chol = cholesterol; TG = triglycerides; HDL-C = HDL-cholesterol.

159 patients compared to normals (5.2 vs. 0.55). The ratio of apoA1 relative to apoAI1, was 4.5 in the Icontrol subjects and 3.8 in HFH patients. Since the particles we sought are rich in apoE [5], we estimated the apoE/apoAI ratios of several HDL density fractions in order to identify the appropriate fraction. The highest apoE/apoAI ratios were found in the d = 1.050-1.090 g/ml fractions for both HFH and control subjects, but ratios were higher in the HFH patients than in controls in all HDL and HDL, fractions (Fig. 1, Table 2). LDL (d = 1.020-1.050 g/ml) and HDL (d = 1.090-1.21 g/ml) contained only traces of apoE (not shown). Therefore, maximum effort was concentrated on the d = 1.050-1.090 g/ml fraction. To separate apoE-rich HDL away from any apoB-containing lipoproteins, the lipoproteins of

2

3

4

5

Control 0

b

c

d

RATIO OF APGE TO APO AI IN HDL FRACTIONS Study subjects

Lipoprotein fractions d = 1.050

-1.090

HDL

HDL-2

(d =1.063

(d =1.063

- 1.21) HFH

S.L. 1.5 R.L. 2.0

0.14f0.03

NormaIs

0.4 & 0.1 (5)

0.03 f 0.005 (8)

- 1.124)

* (8)

S.L. 0.4 0.07 f0.02 (5)

The apoE to apoA1 ratio was determined from SDS-PAGE of the indicated lipoproteins, and scanned with laser densitometer as described in the Methods. d = density in g/ml HFH = homozygous familial hypercholesterolemia; Normals = normolipidemic control. * P -c 0.001; number of subjects in parentheses.

the density range 1.050-1.090 were fractionated by CONA chromatography. The fraction that bound to CONA when analyzed by SDS gel electrophoresis, contained only apoB, while the fraction that did not bind contained apoE, apoA1 and apoC (Fig. 2). The same apoproteins were detected when these 2 lipoprotein fractions were reacted with antibodies against apoE, apoA1 and apoB on microtiter plates (not shown). Thus, apoE-rich HDL particles were clearly separated from the apoB-containing lipoproteins by CONA chromatography. The apoE-rich HDL isolated from subject S.L. after 12 h of fasting contained more unesterified cholesterol and phospholipids than the compara-

HFH 1

TABLE 2

e

f

TABLE 3 A-l -

COMPOSITIONS OF CONCANAVALIN APOE RICH HDL, PARTICLES

UNBOUND

Iofmass Fig. 1. Apoproteins in HDL subfractions separated by ultracentrifugation and 3-251 gradient SDS-PAGE from HFH patients and normal controls. Upper panel: Homozygous familial hypercholesterolemia. Lane 1: apoproteins of density 1.050-1.063 g/ml from S.L. Lane 2: HDL, (d = 1.063-1.124 g/ml) from S.L. Lanes 3 and 4: apoproteins of density 1.0501.090 g/ml from patients R.L. and S.L., respectively. Lane 5: apoproteins of density 1.090-1.21 g/ml of S.L. Lower panel: Normolipidemic control. Lane a: d =1.050-1.063; lanes b and c. HDL-2 (d = 1.063-1.124). Lanes d and e: d = 1.050-1.090. Lane f: d =1090-1.21. E = apoE, A-I = apoAI.

Protein CE FC TG PL

S.L. Fasted a

controls (n=5)

42rfr2 13*1 13*1 2*0.1 3Of3

44&-2.9 22 + 2.0 Sf0.9 4*0.4 25*1.8

CE = cholesteryl ester, FC = free cholesterol, TG = triglycerides, PL = phosphohpids. a Mean f SD of 3 preparations.

ble control fractions (Table 3). The apoE/apoAI ratio of this particle from FH patients was greater than the ratio in control particles. The relative diameters of the d= 1.050-1.090 lipoproteins were analyzed by non-denaturing polyacrylamide gradient gel electrophoresis (GGE). ApoB-containing CONA retained particles were similar to LDL in size, while sizes of the apoE-rich HDL particles ranged from 12.2 to 17 nm (Fig. 3A). Immunoblotting of the unbound particle with a.ntiapoE monoclonal antibody confirmed the presence of apoE in the CONA non-retained fractions (Fig. 3B). The HDL in the d = 1.090-1.21 range showed no reaction with anti-

1

c

4

5

TH

FR

“?yqi

~~p’“~“-‘,

$8

11.1

.^^__

d

T

,,__

‘,

Fig. 3. Relative diameters of d =1.050-1.090 g/ml lipoproteins as determined by non-denaturing gradient gel electrophoresis (GGE, 3-25X). Panel A: Coomassie blue staining of the gel. Lanes 1 and 2: lipoproteins in density 1.050-1.090 g/ml from patients R.L. and S.L., respectively. Lane 3: the CONA unbound fraction from S.L. Lane 4: the CONA unbound fraction of a normal control. Lane 5: lipoprotein in density 1.050-1.090 g/ml from normal control, same as in lane 4. Panel B: Immunoblotting of d =1.050-1.090 lipoproteins. The lipoproteins in panel A were electrotransferred to nitrocellulose paper and immunoblotted with antihuman apoE MAb WU-E4, as described in the Methods. The presence of apoE in LDL and HDL sired fractions is shown. TH = thyroglobulin (sire 17.0 ma), FR = ferritin (sire 12.2 nm).

B-

E-

A-l

3

HDL-

m-F.

b

2

LDL -

-

c-

Fig. 2. Apoprotein compositions of density 1.050-1.090 g/ml lipoprotein (3-258 gradient SDS-PAGE, Coomassie blue stain). Lane a: apoproteins in density 1.050-1.090 of patient RL. Lane b: d-1.050-1.090 from normolipidemic control. Lane c: unbound fraction from CONA column of density 1.050-1.090 lipoproteins from patient S.L. Lane d: the retained fraction of density 1.050-1.090 of patient S.L. B = apoB, E = apoE, A-I = apoAI,C = apoC.

apoE but did react with antiapoA1 antibody (data not shown). The results of cultured fibroblast studies with the CONA column bound and unbound d = 1.050-1.090 fractions and density 1090-1.21 HDL are demonstrated in Fig. 4. Both CONA fractions competed with ‘251-LDL for association and degradation, but the unbound apoE-rich HDL competed more effectively than the retained apoBcontaining lipoprotein (Fig. 4, Table 4). The lipoproteins in the density 1.090-1.21 containing mainly apoA1 and almost no apoE did not react with the cells. Discussion

It has been known for some time that patients with HFH have increased levels of HDL, (or

161 HDL,) in their plasmas [5,7] but the compositions and sizes of the particles have not been characterized, nor is much known about their metabolic origins or roles. The availability of a compound heterozygote with FH undergoing regular plasma exchange made it possible to obtain sufficient lipoproteins to perform a limited number of measurements. The findings of this patient were confirmed in his brother R.L. and in other patients with HFH. Cell

assoclatlon

100 -*-A_A

logo1.21

A

6

\

+ .I< In

a UNB-CON

25

Oi

2.5 5.0

10

.-.-

logo1.21

.

BON-CON 0 UNB-CON -I>

to-

0

25

N

t 2.5 5.0 Unlabeled

10 lipoprotein

REACTIVITY OF d =1.050-l.O!Xt HUMAN SKIN FIBROBLASTS Competitor lipoprotein CONA unbound a (apoE-rich HDL) CONA bound a (” LDL”) CONA unbound b (apoE-rich HDL) LDL control (n = 5) a

LIPOPROTEINS

WITH

IC-50 @g/ml) Association 6.1

Degradation 7.5

25

20

3.3 14*4

4.0 16*2

IC-50 is the concentration (in kg/ml) of competitor lipopro teins that reduced degradation or binding of tZsI-LDL to fibroblasts by 50%. The concentration of competitors can be expressed in terms of total protein (apoE, apoA1, apoC) in the fraction or in terms of the concentration only of apoE (determined by scanning of SDS gels). The CONA fractions are from patient S.L., see Fig. 4 and Methods for details of the cell competition assay. b IC-50 expressed relative to total protein in the fraction. IC-50 expressed relative to the apoE contents of the fraction.

25

Degradation loo

TABLE 4

25 (fig protein/

ml)

Fig. 4. Competition assay of ‘%-LDL with association and degradation by normal human fibroblasts. Density 1.050-1.090 g/ml lipoproteins separated by concanavalin A chromatography, and lipoproteins of density 1.090-1.21 were used as competitors. The cells were preincubated in 10% lipoprotein deficient medium for 48 h at 37 a C to induce maximal expression of LDL receptor activity. Then cells were incubated with 5 rg/ml tzsI-LDL in the presence and absence of indicated amounts of the various competitor lipoproteins for 4 h at 37 o C. The amount of cell associated (top panel) and degraded (bottom panel) 12’I-LDL was determined as described in the Methods. Coefficients of variation of triplicate measurements were < 8%. BON-CON = concanavalin A, bound fraction of density 1.050-1.090 lipoproteins, UNB-CON = unbound fraction.

Based on the descriptions of other workers [7] and our own data [27,28] on HDL, or HDL,, we suspected that particles in the density ranges between LDL and HDL, i.e. between approx. d= 1.050 and 1.090 g/ml, would contain relatively large amounts of apoE. Indeed, when several fractions encompassing LDL through HDL density ranges were surveyed, d = 1.020-1.050 and d = 1.090-1.21 fractions contained very little apoE, the d = 1.063-1.124 fraction contained more, but the d = 1.050-1.090 fraction had the highest amounts relative to other apoproteins (Fig. 1, Table 2) and the apoE/apoAI ratios of the HFH patients were considerably and consistently higher than the comparable fractions of controls. Thus, while the most detailed analyses were carried out in S.L., the high apoE/apoAI ratios were seen also in the d = 1.050-1.090 fractions of R.L. and in the d = 1.063-1.21 fractions of all FH patients compared with controls (Table 2) suggesting that all had increased amounts of an apoE-rich fraction in their plasma. The d = 1.050-1.090 fraction in addition to containing apoE and apoA1 also contained apoB100 (Fig. 2). Two populations of particles were successfully separated by CONA column chro-

162 matography into LDL-like apoB-containing lipoproteins and apoE-rich HDL, (Figs. 2 and 3). The apoE-rich HDL, of FH patients was larger (Fig. 3), and contained more apoE relative to apoA1 (Fig. 2) and free cholesterol relative to esterified cholesterol than the controls. Thus the apoE/ apoA1 results obtained with the d = 1.063-1.21 HDL of all FH subjects (Table 2) probably reflected the presence of an apoE-rich HDL, in their plasmas. HDL, resembled the HDL, of cholesterol-fed animals in composition [5,7]. To ascertain whether apoBrich HDL also had the biologic activities of HDL,, fibroblast binding and degradation assays were performed. ApoE-rich HDL did react with the LDL receptor, with greater affinity than the LDL-like particle and LDL itself (Fig. 4, Table 4). This was probably due to the presence of apoE [7]. ApoE-rich HDL in HFH patients could arise from at least 2 sources: (a) In cholesterol-fed animals HDL, are thought to arise in peripheral cells and to transport excess dietary cholesterol from the periphery back to the liver for excretion [9]. The structural-compositional and functional resemblances between S.L.‘s HDL, (in the fasted state) to HDL, suggest similar metabolic origin and role for the apoE-rich particles of HFH patients. However, it is also possible that in FH, apoBrich HDL represents particles that originate from nascent HDL generated as ‘surface remnants’ during VLDL and chylomicron metabolism or as a result of lipid transfer proteins acting on other HDLs [29] or a combination of the above. Based on our data it is not possible to definitively distinguish between possibilities. Acknowledgements The authors wish to thank Dr. M. Aviram and G.J. Brook for providing the plasma samples of their HFH patients in Israel. This research was supported by NIH Grants HL3200, HL 15308, AGO5562 and HL42460. We acknowledge the expert secretarial assistance of Cheryl Doyon. References 1 Goldstein, J.L., and M.S. Brown, Familial hypercholesterolemia. In: The Metabolic Basis of Inherited Disease, 5th

ed., Stanbury, J.B., J.B. Wyngaarden, S.D. Fredrickson, J.L. Goldstein, and M.S. Brown (Eds.), McGraw-Hill Book Co., New York, 1983, p. 672. 2 Goldberg, R.B., G.M. Fless, S.G. Baker, B.I. Joffee, G.S. Getz, A.M. Scanu and H.C. Seftel, Abnormalities of high density lipoproteins in homozygous familial hypercholesterolemia, Arteriosclerosis, 4 (1984) 472. 3 Gibson, J.C., R.B. Goldberg, A. Rubinstein, H.N. Ginsburg, M.V. Brown, S. Baker, B.I. Joffe, and H.C. Seftel, Plasma lipoprotein distribution of apolipoprotein E in familial hypercholesterolemia, Arteriosclerosis, 7 (1987) 401. 4 Anderson, D.W., A.V. Nichols, S.S. Pan, and F.T. Lindgren, High density lipoprotein distribution. Resolution and determination of three major components in a normal population sample, Atherosclerosis, 29 (1978) 161. 5 Schmitz, G., and G. Assmamr, Isolation of human serum HDL 1 by zonal ultracentrifugation, J. Lipid Res., 23 (1982) 903. 6 Kobayashi, J., T. Nishide, M. Shinomiya, N. Sasaki, K. Shirai, Y. Saito, and S. Yoshida, A familial hyperalphalipoproteinemia with low uptake of high density lipoproteins into peripheral lymphocytes, Atherosclerosis 73 (1988) 105. 7 Mahley, R.W., and T.C. Innerarity, Lipoprotein receptors and cholesterol hemostasis, Biochim. Biophys. Acta, 737 (1983) 197. 8 Innerarity, T.C., and R.W. Mahley, Enhanced binding by cultured human fibroblasts of apoE-containing lipoproteins as compared with low density lipoproteins, Biochemistry, 17 (1978) 1440. 9 Small, D.M., HDL system: A short review of structure and metabolism, Atheroscler. Rev., 16 (1987) 1. 10 Lehrman, M.A., W.J. Schneider, M.S. Brown, and J.L. Goldstein, The Lebanese allele at the LDL receptor locus: Nonsense mutation produces truncated receptor that is retained in endoplasmic reticulum, J. Biol. Chem., 262 (1987)401. 11 Brook, G.J., S. Keidar, M. Boulos, E. Grenadier, A. Wiener, N. Shehada, W. Markiewica, A. Bender& and M. Aviram, Familial homozygous hypercholesterolemia: Clinical cardiovascular features in 18 patients, Clin. Cardiol., 12 (1989) 353. 12 Osthmd, R.E., R.A. Levy, J.L. Witztum, and G. Schonfeld, Familial hypercholesterolemia. Evidence for a newly recognized mutation determinin g increased fibroblasts receptor affinity but decreased capacity for low density lipoprotein in two siblings, J. Clin. Invest., 70 (1982) 823. 13 The Lipid Research Clinics Population Study Data Book, Vol. 1, The Prevalence Study. United States Government Printing Office, 1980 (NIH Publication No. 80-1527), Washington, DC. 14 Bartlett, G.R., Phosphorus assay in column chromatography, J. Biol. Chem., 234 (1958) 466. 15 Lowry, O.H., N.J. Rosebrough, A.L. Far-r, and R.J. Randall, Protein measurement with folin phenol reagent, J. Biol. Chem., 193 (1951) 265. 16 Schonfeld, G., R.S. Lees, P.K. George, and B. Pfleger, Assay of total plasma apolipoprotein B concentration in human subjects, J. Clin. Invest., 53 (1974) 1458.

‘LIOI (P86I) SZ “sax p!drl ‘1 ‘msqoqelaru matoIdodg &suap q8q~ “s ‘%aquasg 6Z ‘9817 (5861) vi ‘wwww WuaIowsaIoqa~acL(qq ppm3 IIF su!aloJdodg musaId uo larp 1~3$&I IoJalsaIoq:, $%I 30 s13e332 ‘pIa3uovs ‘f) Pm ‘JaeI3d ‘8 “f)‘.L ‘aIo3 8Z ‘8011 (E861) 9S “Iouu3oPW “VI3 ‘1 ‘ww 30 eumqd aql m su!aloJdodq %u~mmg u!aloldode 30 sIaAaI asyJ w3 pm2 IolalsaIoq3 hreja!p u! saseamq ‘pIa -UOqs

‘f)

pW2 ‘UaUOf)

‘8

‘?S!ll)I

‘SL (6~60

‘I ‘~3SlEd

‘M

c9 ‘waauI

“f)‘L

w3

‘3103

LZ

‘r ‘sn=

paml[n3 6q %uyssamoJdu!aloJdodq dl!suap MOI tq saiqnl -0~3~ 30 aIoa ‘pIa3uoqs ‘9 pm! ‘Ja%aUd‘8 “3.x ‘punI~s0 9Z ‘ESIS (PL6I) 6PZ ““aq3 ‘lo!8 ‘f ‘sl=?q -oJq!3 pamIn 1(q su!ajordodg &suap MOI 30 uoypefiap pw %urpp”~ ‘uhola .s’M pm ‘ema ‘3’s “~‘1 ‘FaapIof) SZ ‘60EI (8861) 6Z “sax p!d!T ‘r ‘uo~1~m3 3 u?loJdodgodv 103 mwodm! su!aloJdodg munq 30 uo!ssaJdxa ado@ 3 uyoldoda 30 QauaSoJaiaH ‘pIa3uoqq ‘f) put? ‘uauemJ ‘ryv ‘.~.a ‘F-~)I pZ ‘LE6 (8861) 6Z ‘-sax p!d!? ‘r sa!poqrw 8 uyoJdodqodB ueumq-gut? p~1oI3ouom 30 sagg!3~~ds Iwo@?3 ‘pIa3uoq* ‘0 pa ‘asead 23 ‘tws ‘r ‘ME-I ‘\I’ ‘*PKJ ‘3 ‘asaI3d ‘8 ‘wsov~ ‘kv ‘m=cay>I ‘A ‘5ra ‘~nr)~ Ez

3x13

‘(‘~3)

7

‘Lp-6Z dd ‘~861 ‘TJ ‘uovz?\dema “XII ‘spw

‘(sapvs

wum~

pue

‘SSald

Bow0qwq

FaloJdod!?) 111 ‘IOA %!saJoqdowaI-, 30 3ooqpuvH :ul .Faloldodg @nap @q auwqd umunq 30 s!saroqdowaIa Ia% lua!pelf) ‘%Iof) ‘7.3 puE ‘yXn2~g ‘T’d “A’V ‘SIOqXN zz ‘m (6961) trpZ ‘waq~ ‘Io!a ‘1 ‘s!saJoqdowaIa Ia% apymIh%?~Iod-alIns Ik~pop kq suoyXyuIalap I@!aM nqn3BIom 30 &qenal aq,j_ ‘uroqsg ‘g,q pue ‘y ‘JaqaM IZ ‘0% (EL60 ES ‘.marlw8 ‘@uv ‘sural -oIdodq unuas p su!aloldodv 30 sa!cads lpnqns 30 uog -w3yxap! 103 anbqw cyaroqdowaIa p!del v “&I ‘atrex OZ ‘6ZI (PL61) IP “~~a7

‘St@(LL6I) 81 ‘.sax p!d!T ‘1 ‘Qsslroun~o!peJ Lq pamst?atu suyoId -odg Qsuap q@q emseId 30 wawo:, 11-v malordodqodv ‘8uay ‘1 pm2 ‘lauuoaqq ‘d’~ ‘uaq3 ‘s.1 “f) ‘pIa3uoq* 81 ‘9EZ (PL6I) PS “Isa* -III ‘ug3 ‘1 ‘kssrrounurm!o!pe~ dq paguualap se museId II! 1-v u!ajoJdodgode 30 sIai\aI aqi pun uyoldodg D!suap q8g umunq 30 amwuls au ‘1aSaUd ‘8 pm2 “f) ‘pIa3uoq3s LI