In vitro remodelling of plasma lipoproteins in whole plasma by lipoprotein lipase in primary and secondary hypertriglyceridaemia

European Journal of Clinical Investigation (1 990) 20,422-43 1

ln vitro remodelling of plasma lipoproteins in whole plasma by lipoprotein lipase in primary and secondary hypertriglyceridaemia E. LEVY, R. J. DECKELBAUM*, R. L. THIBAULT, E. SEIDMAN, T. OLIVECRONAt & C. C. ROY, Departments of Gastroenterology and Nutrition, University of Montreal and Research Center, Sainte-Justine Hospital, Montreal, Quebec, Canada, *Department of Pediatrics, College of Physicians and Surgeons of Columbia University, New York, USA, and ?Department of Physiological Chemistry, University of Umea, Umea, Sweden

Received 1 August 1989 and in revised form 9 January 1990

Abstract. In patients with familial lipoprotein lipase deficiency (FLPL-d) and glycogen storage disease type I (GSD-I), hypertriglyceridaemia (1445 f 247 and 1082 3 12 mg dl- ‘, n = 5 per group) was associated primarily with reduced extrahepatic lipoprotein lipase (LPL) activity (0.33 f0.33 and 1.69kO.38 pmol FFA ml-’ h-I) when compared with controls (4.83f0.90). Hypercholesterolaemia was characterized by elevated LDL cholesterol (I91 f 30 and 344 f 34 vs. 1 15 5 mg dl-’ in controls P<O.OI) and low HDL cholesterol (12+2and22&2vs. 56f3incontrols, P<O.OOI). In order to ascertain the role of LPL in the interconversion and remodelling of lipoproteins in these disorders, we analysed lipid and lipoprotein profiles before and following in uitro incubation of patient plasma with purified milk LPL (EC 3. I . 1.34) for 6 h at 37°C. The efficiency of exogenous LPL in uitro was demonstrated by the extent of hydrolysis of chylomicrons and of VLDL-TG in both groups. Concomitant with the disappearance of TG-rich lipoprotein particles, a consistent per cent increment of IDL (99.2 f 30.8 and 43.9 k 70.5), LDL (152.8 f 36.2 and 137.0 f 36- 1) and of HDLz (144.8 f 29.4 and 99.8 & 18.7) was observed in both groups of patients. The enhancement of the latter fractions contrasted with the decline of HDL3 mass concentration (25.4 f 7.7 and 5 1.4f 5.8%), suggesting that a major shift of HDL3+HDL2 occurs following in uitro lipolysis by LDL. Simultaneous compositional and morphological changes of individual lipoprotein particles were noted, confirming the dynamic movement and exchange of neutral lipids and proteins. Specificity of LPL results was demonstrated by experiments in which incubation of the whole plasma at 37°C without exogenous lipolytic enzyme did not cause any substantial changes. The present study, therefore, demonstrates a correction of the major lipoprotein abnormalities associated with FLPL-d and GSD-I by exogenous LPL. No substantial difference was noted

Correspondence: E. Levy, Gastroenterology and Nutrition Unit, Hhpital Sainte-Justine, 3175 Sainte-Catherine Road, Montreal, Quebec, H3T 1C5, Canada.

between primary (FLPL-d) and secondary (GSD-I) hyperlipidaemias. These studies allow us to conclude that a simple in uitro system, utilizing an exogenous source of LPL and plasma from patients, may serve as a suitable model for the study of the metabolic relationships of lipoproteins. However, in view of the fact that the extent of lipolysis achieved in uitro did not differ between FLPL-d and GSD-I, it may not be able to separate primary from secondary hyperlipaemias.

Keywords. Glycogen storage disease type I; familiar lipoprotein lipase deficiency; chylomicron; VLDL; IDL; LDL; HDL,; apoprotein.

Introduction Lipoprotein lipases have an important role in the regulation of serum lipoprotein metabolism, composition and structure [I-31. Chylomicrons (CM) and very- low-density lipoproteins (VLDL) are hydrolysed by LPL bound to the capillary endothelium of adipose tissue, heart, skeletal muscle, lung and milk [ 1,4]. In the catabolic process, CMs are rapidly degraded to form remnants and VLDLs that are sequentially lipolysed to produce IDL and LDL [ 5 ] . The constituents originat- ing from the surface coat (phospholipids, free cholesterol and apoproteins) of hydrolysed TG-rich lipoproteins are thought to undergo transfer to HDL3, thereby promoting its conversion to HDL2 [5-81. LPL activity is, therefore, related to the levels of plasma HDL, particularly the HDL2 subfraction [7,8]. Based on these observations, it would appear that plasma lipolysis constitutes not only a delivery source of fatty acids to peripheral tissues, but also an important route for the modulation of the structure and concentration of lipoproteins.

Prior reports suggesting a regulatory role for LPL in the concentration and spectrum of HDL are primarily based on population studies and on the results of various treatment regimens [9- 121. However, few

422

REMODELLING OF PLASMA LIPOPROTEINS IN HYPERTRIGLYCERIDAEM IA 423

experimental studies have been performed to investi- gate the reciprocal changes between the levels of HDL and TG-rich lipoproteins (Sf values of 20-100 and 100-400) [8,13-161. In addition to the role of LPL in the conversion of HDL subfractions, other reports emphasize the regulatory function of lecithin : cholesterol acyltransferase (LCAT) [ 171, cholesteryl ester transfer protein [ I 81 and specific 'conversion factors' [ 191. Finally, some investigators even minimized the role of LPL in the process of HDL transformation while concluding that LCAT action is the major action of the various changes in HDL structure [20,21]. These wide variations in opinion might be due to differing experimental conditions, often consisting of incubation of isolated HDL with either microemulsion such as intralipid, or with variable amounts of separated VLDL in a system containing LPL and LCAT. Furthermore. most of these in "itro studies included a methodological approach using a buffer system, often at unphysiological pH and temperature, while other experiments were carried out with post-heparin lipolytic activity containing both extralipoprotein lipase and hepatic lipase. Moreover, in tiizw studies have yielded contradictory results concerning lipolytic activity deficiency disorder [22-241.

The purpose of the present investigation was to examine the behaviour of the different lipoprotein classes in two disorders: ( I ) classical type I hyperlipoproteinaemia, which is characterized by absolute deficiency of lipoprotein lipase resulting in the most severe form of hyperchylomicronaemia and hypertriglyceridaemia, and (2) glycogen storage disease type I with a relative lipoprotein lipase deficiency [25] leading to similar, albeit usually less severe lipid abnormalities. We analysed the effect of purified milk LPL on lipolysis, lipoprotein conversion and modification, and apoprotein redistribution in these patients. Given that these disorders are characterized by a partial or total LPL deficiency, an excess of triacylglycerols carried by CM and VLDL, and that subnormal amounts of HDL cholesterol are present [2], we opted to utilize the patient's own unmodified plasma for our experiments. The studies described herein utilized an in oitro system, consisting of the addition of purified milk LPL to the patient's plasma. We have demonstrated that exogenous LPL can modify the various lipoprotein profile derangements towards the normal state. Further we define lipid and apoprotein changes during lipolysis in whole plasma which provide further insight on the role of this enzyme in normal lipoprotein interconversion and metabolism.

Patients and methods

Patients

Five patients with a diagnosis of type I glycogen storage disease documented by abnormal storage of liver glycogen and by glucose-6-phosphatase deficiency participated in this study. They were treated

with standard frequent daytime feedings and nocturnal nasogastric gavage. The other five patients who were known to have familial lipoprotein lipase deficiency, were treated with a diet low in fat and high in carbohydrate. All patients gave their fully informed consent to the investigation, the protocol of which had been approved by the ethics committee of Sainte- Justine Hospital.

Incubation

Blood samples were obtained from FLPL-d patients after overnight fasting. Venous blood samples from GSD-I subjects were collected after 4 h of fasting following nocturnal intragastric feeding. Samples were collected in 0.01 % disodium EDTA and 0.05% sodium azide. Plasma was separated immediately after blood collection at 3000 r.p.m. x 30 min at 4°C. Incubations of 5 ml plasma in the presence of additional 3% albumin (Sigma, FRG) which contained no detectable phospholipids, cholesterol or apoproteins, 250 PI heparin with or without purified milk lipoprotein lipase was performed at 37°C for 6 h using a shaking water bath. The enzyme preparation used for these studies released 0.1 1 nmol of FFA h k ' ml-' enzyme from a 14C-triolein emulsion as described below. The amount of LPL added to the mixture was a function of the T G concentration (35 pI LPL for each 50 mg triglyceride contained in 5-ml plasma incubation samples).

Isolation of lipoproteins

At the end of the incubation period, CMs were removed from plasma by preparative ultracentrifugation at 25000 r.p.m. x 30 min, VLDL (1.006 g ml-I) and LDL (1.063 g ml- ') at 40 000 r.p.m. x 18 h, HDLz (1.125 g ml-') and HDL3 (1.21 g ml-I) at 40000 r.p.m. x 48 h. All ultracentrifugations were carried out at 4°C in a Beckman L8-80 using a Ti-50 rotor. The lipoprotein fractions were washed by recentrifugation under identical conditions and dialysed exhaustively against 0.15 mol I - ' NaCI, 0.001 mol I - ' EDTA, pH 7.0 at 4°C.

Chemical analyses

Lipoprotein protein was measured according to Lowry et al. [26] with bovine serum albumin as a standard. Phospholipids (PL) were determined by the Bartlett method [27] and triglycerides (TG) by commercial kit (Boehringer Mannheim, Montreal, Canada). Total (TC) and free (FC) cholesterol were quantitated by the oxidase-esterase method [28] using the Boehringer kit. Cholesteryl ester (CE) was calculated as the difference between total and unesterified cholesterol. HDL cholesterol (HDL-C) was measured after precipitation of very-low-density and low-density lipoproteins with phosphotungstic acid [29]. LDL cholesterol (LDL-C) assay was performed using polyvinylsulphate (Boehr-

424 E. LEVY et al.

inger Mannheim) [30]. The apolipoprotein content of lipoproteins was examined by electrophoresis in SDS- polyacrylamide gel (SDS-PAGE) [25]. The densitometric distribution of apoproteins was assayed as described previously [25]. Plasma was also subjected to agarose gel electrophoresis as reported previously [25]. Negative stain microscopy was performed on a Phillips EM 300 and the diameter of 400-600 particles was then determined.

Lipolytic activity measurement Post-heparin plasma was taken by venous puncture 10 min after an intravenous injection of heparin (10 units kg- ' body weight). Lipolytic activity was measured with an emulsion of glycerol tri [1-I4C] oleate as substrate [3 11. Hepatic triglyceride lipase (HL) activity was assayed in the presence of protamine sulphate which was verified to completely inhibit peripheral lipoprotein lipase (LPL). Extraction of FFA was performed by a modification of the procedure of Belfrage & Vaughan [32] as previously described [25].

Statistical analysis All values were expressed as the mean & standard error (SEM). Statistical differences were assessed by the paired t-test and Wilcoxon signed rank test.

Results All subjects in both groups had marked elevations of plasma triacyl-glycerols, with higher levels in FLPL-d than in GSD-I (Table 1). On the other hand, choles-

terol levels were increased more in the latter than in the former, while levels of LDL-C were similar in both groups. HDL-C was markedly reduced in both cat- egories of patients but more so in FLPL-d, leading to decreased HDL-C/TC and HDL-C/LDL-C ratios.

However, it is probable that because all patients were being treated by high carbohydrate diet at the

Table 2. Post-heparin lipolytic activity

(pmol FFA ml-' h r - '

Patients Total PHLA LPL HL

GSD-I I 2 3 4 5

FLPL-d 6 7 8 9

10

Normal

2.75 2.74 2.74 2.79 3.89

2.98 f 0.23' x f SE

5.52 5.1 1 3.60 5.42 5.14

4.95 f 0.35' x f SE

zfSE 14.15f0.50

2.08 I .40 1.27 2.10 I .62

1.69 f0.17*

N D N D N D 1.67 N D

0.33 +0.33*

4.83 f 0.90

0.67 I .34 I .41 0.69 2.27

I .29 0.29.

5.52 5.1 1 3.60 3.75 5.14

4.62 f 0.39*

9.32 f I .30

PHLA, post-heparin lipolytic activity; LPL, extra-hepatic lipo-

* vs. normal, P < 0.01. protein lipase; HL, hepatic lipase; ND, non-detectable.

Table 1. Characteristics of the patients

Plasma concentrations (mg dl - I ) HDL-C HDL-C Age Weight Relative body

Patients Sex (years-months) (kg) weight (%) TG TC LDL-C HDL-C TC LDL-C

GSD-I I f 18-8 40.7 90 752 275 181 77 0.098 0.15 2 f 16-7 47, I 75 433 263 192 24 0.09 I 0.12 3 f 24-6 41.0 50 I404 448 307 19 0.042 0.06 4 m 23-4 49.3 50 2153 489 182 16 0,033 0.09 5 m 21-3 60.1 75 672 244 182 22 0,090 0.12

x f S E 1083f312t 344+51* 209k24t 2 2 k 2 t 0~071f0~014t 0.11+0~01t

FLPL-d 6 f 7-5 23.7 50 1 I55 202 I67 14 0.069 0.08 7 f 4-8 18.8 60 963 182 I42 9 0.049 0.06 8 m 7-8 21.2 80 1752 348 30 I 12 0.034 0.04 9 m 10-8 35.0 10 2212 247 209 I I 0.044 0.05

10 f 17-4 43.1 5 1085 I88 135 12 0.064 0.09

x f S E 1445f247t 233+3I* 191 f 3 0 t 12+ 1' 0.052+0407t 0.06+0,02t

Controls n = 16 x*SE 6 0 f 4 172f5 114f4 4 7 f 1 0.28k0.01 0.42f0.01

* vs. controls, P<0 .05 . t vs. controls, P<O.OI.

REMODELLING OF PLASMA LIPOPROTEINS IN HYPERTRIGLYCERIDAEMIA 425

Chvlomicron VLDL I

10 1 - 100

80 -

- 60

- 40

-20 2 > - P - J VLDL

pre post

pre post I

6;l I_ ct

4 IK 4

2

3 7

100

80

60

40

20

-8 4 %;; -7

80

60

40

20 ln 1 1 8 -

9 -

Figure 1. Extent oftriglyceride hydrolysis and lipoprotein mass reduction of CM and VLDL in plasma of GSD-I (upper panel) and FLPL-d patients (lower panel). Individual values pre- and post-lipolysis are depicted.

GSD-I L D L GSD-I HDL3

Pre post

1.0 Pre post

5 2 -9 \: I

- 3.0 L E o, 0.8 E

-

v

In In

0.5 - 0 + I-"

1.0 - L E

i? v

In In

0.5 - 0 t

4

0.5

I 5 3

0.5

4

F L P L - d L D L

~ ~~

F L P L - d H D L j F L P L - d I D L F L P L - d HDL2 Pre post

1.0 Pre post

10

6

8 7

Pre post 2

6 7 0.5 9 10 8

0.5 / 6 4

Figure 2. Effect of LPL treatment on plasma lipoproteins. Values are given in mg/ml plasma of total mass.

426 E. LEVY et al.

140

120

w 100 E

80 + 2 2 6 0 - + 0 ar

0 L 0

40

8 20

time of study, VLDL made up the majority of TG-rich lipoprotein in both disorders. Moreover, a relatively small proportion of HDL concomitantly with an increase of LDL were observed in agarose gel electrophoresis (gel not shown).

To document the contribution of LPL deficiency to the higher TG values, post-heparin lipolytic activity was measured in the plasma of patients in both groups (Table 2). The extra-hepatic lipoprotein lipase activity was lower in GSD-I than in controls. As anticipated, it was non-detectable in FLPL-d except in one patient. In both patient groups, hepatic lipase activity was depressed, but to a greater extent in GSD-I.

In order to explore the role of LPL as a major variable in the composition and metabolism of the various lipoprotein classes, whole plasma was incubated in the presence and absence of exogenous LPL. Figure 1 shows the degree of lipolysis achieved in uitro in both of these hypertriglyceridaemic disorders. Fol- lowing incubation with LPL, there was a mean hydrolysis of 64.4 k 3.7 and 86.8 & 5.1 % of CM and VLDL, respectively, in GSD-I. However, a different pattern characterized the FLPL-d plasma: more hydrolysis was observed for CM (78.9+ 3.5%) than for VLDL (63.1 *9.3%). The extent of the in uitro hydrolysis of VLDL was larger (P< 0.05) for GSD-I than

-

-

-

-

-

-

0 -

I -

- 2 0

-40

-60

-

-

I DL

I , I 1

LDL

0 GSD-I ISS FLPL-d

Figure 3. Averaged change in post-hpolysis lipoprotein mass following the incubation with purified milk LPL. Data are expressed as mean+SE in YO of the initial values.

for the other group. As expected, lipolysis also led to decreases in total lipoprotein mass in CM and VLDL: 59.1 f 8.6 and 87.3 +4.6% in GSD-I, and 72.6 & 7.8 and 61.2+4.9% in FLPL-d. The magnitude of the decrease in VLDL mass was again higher (P < 0.05) in GSD-I than in FLPL-d. Because of their differences, we chose to analyse the results of the GSD-I separ- ately, free of FLPL-d. Changes in the mass of the lipoproteins other than CM and VLDL were moni- tored after lipolysis of the TG-rich particles. As shown in Fig. 2, alterations in IDL, LDL and HDL of both disorders were essentially similar. There was an increase of the total mass of IDL, LDL and HDL2 particles and a decrease of that of the HDL, fraction. Figure 3 compares the average change in the post- lipolysis lipoprotein mass with the pre-lipolysis values in IDL, LDL, and HDL. Incubation with LPL in GSD-I and FLPL-d, resulted in increased masses of IDL by 43.9k70.5 VS. 99.2&30.8%, of LDL by 137.0f36-1 vs. 152.8+36.2%, and of HDL2 by 99.8 f 18.7 vs. 144.8 k 29.4'%, respectively. The mass of HDL3 particles was reduced post-lipolysis more so in FLPL-d than in GSD-I (27 .0f8 .4 vs. 51.4k5.8 , respectively).

Changes in lipoprotein composition occurred as a result of the in uitro incubation of plasma with lipoprotein lipase (Table 3). VLDL particles of both groups showed significant decreases of TG at the end of the 6 h incubation period. There was an increase (P<0.05) in FC but it was limited to patients with FLPL-d. These large changes in TG, with relatively small increments in other VLDL constituents, led to decreases in the TG:PL, CE:PR and TG+CE: FC + PL + PR (core: surface) ratios. It should be noted

92 000 - 66 200 - 45 000 -

31 000 -

21 500 -

I4 400 -

P

]Albumin

A - E

1 E A1

] t;rs=

1 2 3 4 5 6 7

Figure 4. SDS-PAGE (12.5%) ekctrophoretogrdms of VLDL- apoproteins following the incubation at 37°C with (t) or without ( - ) LPL. ( 1 ) protein standards (phosphorylase B, BSA, ovalbumin. carbonic anhydrase, soybean trypsin inhibitor and lysozyme); (2) rat HDL; (3) rat VLDL; (4) GSD-I; -LPL; (5) GSD-I: +LPL (6) FLPL-d: - LPL. (7) FLPL-d: + LPL. After the incubation. VLDL were isolated by flotation at d = 1.006 g/ml and after both their dialysis and delipidation. 30 jig of apoproteins were applied on each gel, and staining was performed with Coomassie blue.

REMODELLING OF PLASMA LIPOPROTEINS IN HYPERTRIGLYCERIDAEMIA 427

Table 3. Effect of LPL on the relative composition of plasma lipoproteins in GSD-I and FLPL-d

mg 100 mg- ' lipoprotein mass

Subject TG CE

Very-lowdensit,. lipoproleins GSD-I prior 61.1 f4 .4 4.7 f 0.8 GSD-I post 41.9f 13.1' 6 .6f0 .5 FLPL-d prior 59.0 f 2.4 4.9 f 2.7 FLPL-d post 50.3 f 9.1 * 6.8 f 3.0

Inrerniediure-rlensiry lipoproleins GSD-I prior 4 5 0 f 2.8 7.2 f 0.8 FSD-I post 3 1.3 f 3.05 8.3 f 1.5

FLPL-d prior 50.6f9.5 7.7f 1.3 FLPL-d post 35.1 f7 .0 10.5 f 5.6

Lon,-densiiy lipoproieins GSD-I prior 23.8 f 7.0 14.0 f 3.2 GSD-I post 21.0f8.4 14'2f2.2 FLPL-d prior 27.4f6.9 25.8 f 10.2 FLPL-d post 19.5 f 6,4* 25.9 f 1 I .7

High-densi?,, lipoproieins ( HDL:) GSD-I prior 16.9f6.2 14.4 f 3 .3

FLPL-d prior 19.8 f 5.7 16.6 f 3.1 FLPL-d post I5.0f 4.9 12.8 f 2.6*

GSD-I post 16.5f8.5 7.7 f 1.7*

High-density lipoproteins ( H D L j ) GSD-I prior 9.0f4.5 4.3 f I .4

FLPL-d prior I 1.5 f 3.0 5.9 i 3.7 GSD-I post 10.8f4.9* 3.9f 1.5

FLPL-d post 8.7f4.2 10.5f2.9'

FC

5.0 f 0.7 6.1 f 0.8 8.9 f 0.8

10.2 f 1.6*

6.2 f 0.4 7.7f 1.1

3.5f 1.7 6.3f2.1

7.0f 1.0 7.2 f I .7 5.3f0.5 6.8 k 2.0

5.6 f 0.6 4.6 f 1.3 3.9 f 1.9 4,4+ 1.5

2.0 * 0.1 2.5f 1 . 1 2.5f 1 . 1 3.2 f 1.5.

PL PR TG + CE

~ ~

TG C E PL m F C + PL+ PR

18.5f5.2 31.7f 10.7 17. I f 2.0 20.0 f 4.6

27.2 f I .3 35.0 f I .O 18.9 f 5.0 25.3 f 5 . I

29.6 f I .7 32.2f2.9 21.4f 3.2 24.5 f 4.9

29.7 f 2.6 30.9 f 7.6 24.9 3.4 23.1 f 2.9'

42.4 f 1 I .4 22.8 f 6.4* 28.7 f 3.3 24.3 f 7.1

9.2f 1.7 13.7 f 2.9 1 1 . 1 f 4 . 2 12.0 f 3.6

14.6 f 0.9 17.7 f 2.5 19.3 f 5.1 22.9 f 2.9

24.8 f 1.5 25.0 f 2.7 20.1 f 3.4 22.8 f 5.9

33.8 f 3.0 41.2+ 5.6' 35.1 f 1.8 44.8 * 3.7'

42.7 f 8.6 60.2 +4.2* 513f 6.3 53.1 f 8.5

3.74f 0.88 0.89 f 0.44' 3.49 f 1.08 2.75f0.58

I .6 I f 0.08 0.69 f 0.2 I 3. I 1 f 0.90 1.55 f0.28

0.82 f 0.29 0.78 f 0.39 1.28 f 0.28 0.80f0.25*

0.58 fO.24 0.57 f 0.49 0.82 k 0.32 0.68 k 0.32

0.24 f 0. I8 0.54f0.34 0.41f0.14 0.39 f 0.24

0.51 f0 .13 2.04f0.17 0.30k0.28 0.61 f0.29 0.62 f 0.20 1.8 I f 0.03 0.61f0.21 1.23f0.13

0.48 f 0.08 0.80 f 0.24 0.40f0.25* 0.49f0.15 0.43 f0 .17 1.54f0.37 0.48 k 0.28 0.84 f 0.05

0.56f0.12 0.61 f0.I 1 0.51 f0.13 0.84f0.77 I .33 f0 .60 1.1 5 1 0 2 6 1.20f0.61* 0.85f0.24

0.43 f 0.08 0.19 f0.06' 0,32+0.15* 0.47 k 0.09 0.58 f 0. I6* 0.29 k 0.07' 0.39 f 0. I 2

0.46 f 0.10

0.10 k 0.03 0.06 f 0.03 0. I2 f 0.09 0.2 I f 0.08

0. I5 f 0.05 0. I7 f 0,05* 0.2 1 f 0.06 0.24 f 0.03*

The results are given as per cent of total lipoprotein mass and represent mean values f SEM of samples obtained before (prior) and after (post) exposure to LPL. Differences between samples in the presence and absence of LPL were tested for significance by the Student's i-test for paired samples.

* P < 0.05.

Table 4. Densitometric distribution of VLDL apoproteins pre- and post-lipolysis

Apoproteins ( ' W )

Subjects

GSD-I Case 4

Case 3

FLPL-d Case 7

Case 8

Pre post

Pre post

A-IV E A-I C

0.4 29 0 10 37 7

I 46 0 12 40 8

3 33 0.8 10 48 13 0.4 30 I

13 32 13

70 46 53 41

63 28 68 41

The densitometric evaluation does not include apo B and albumin fractions.

that the mass ratio of core constituents (TG+CE) to surface constituents (FC + PL + PR) can be used to make inference on the size of spherical lipoprotein particles. In general, lighter and larger populations are relatively enriched with core components compared with denser and smaller particles. Therefore, given the relation of these ratios to particle size, the VLDL of plasma treated with LPL were of smaller diameter. The 12.5% SDS-PAGE (Fig. 4) revealed that during the course of lipolysis, apoprotein C was deleted from the VLDL while there was an enrichment of apo A-IV, apo E and apo A-I. These alterations were present in both GSD-I and FLPL-d patients (Table 4). Furthermore, analysis of VLDL proteins on TMU-gel showed the disappearance of apo C , confirming the SDS-PAGE findings (Fig. 5).

A similar tendency was recorded in compositions of IDL and LDL. Whereas the relative amount of TG decreased, especially in the FLPL-d group, the per cent

428 E. LEVY et a/.

P 92 000 - 66 200 -

45 000 -

] ALB

A - E

E 31 000 -

A- I - CII - cmo - C r n l

- C I u 2 - cm3

I 2 3

- CII - cmo - cm, - cmp

- cm3

I 2 3

Figure 5. Disc electrophoresis of VLDL in 10% polyacrylamide gels containing 8 M urea of FLPL-d (upper panel) and of GSD-I patient (lower panel). Lane I represents the VLDL-apo C following the incubation at 37°C for 8 h without the addition of LPL; lane 2 shows the electrophoretogram of VLDL incubated in the presence of LPL; and lane 3 VLDL (non incubated) from normal subject.

content of the other components was scarcely modi- fied. This in turn resulted in the drop of the TG: PL, and TG + CE: FC + PL+ PR ratios, compatible with a reduced lipid core. Negatively stained electron micro- graphs of IDL and LDL preparations showed that post-lipolysis particles were smaller (data not shown) consistent with the ratios tabulated (Table 3). Changes in apoproteins were observed in the IDL and LDL fractions of both groups. Figure 6 demonstrates enrichment of these particles by A-IV, E and A-I following the in uitro incubation with LPL. Also, small amounts of apo B-48 were visualized by 4% SDS- PAGE (data not shown) in post-lipolysis samples in

21 500 -

14 400 -

1 2 3 4 5 6 7 8 9 1 0 Figure 6. Electrophoresis of IDL ( I ,019 g/ml) and LDL ( I ,063 g/rnl) in 12.5% SDS-polyacrylamide gels. ( I ) protein standards as described in Figure 4; (2) apoprotein A-1 standard; (3) FLPL-d: IDL ( - LPL); (4) FLPL-d: IDL (+ LPL); (5) GSD-I: IDL ( - LPL); (6) GSD-I: IDL( + LPL); (7) FLPL-d: LDL( - LPL); (8) FLPL-d: LDL (+ LPL); (9) GSD-I: LDL ( - LPL); (10) GSD-I: LDL ( + LPL).

the LDL fraction suggesting that some intestinally derived TG-rich particles were converted directly to particles at similar density to LDL.

The absolute increase of HDLz mass was accompanied by a relative decrease of HDL2-CE and elevated levels of HDL2-PR in each group of patients (Table 3). PL did not change in GSD-I, whereas a significant decrease was noticed in FLPL-d.

In contrast to HDL2, the HDL3 mass decreased after

Table 5. A-1:A-ll ratio in HDL2 and HDLj fractions

Subjects HDL2 HDLi

GSD-I Case 4 pre 0.65 0.35

post 2.53 0.77 Case 1 pre 2.85 2.61

post 3.92 1.38

post 6.94 2.34 Case 2 pre 2.67 3.81

post 4.29 2.37

x SE pre 2.7 f 0.82 2.24 f 0.72

Case 3 pre 4.70 2.2 I

post 4.42f0.92. 1.71 f0.39

FLPL-d Case 7 pre 4.71 2.56

post 6.87 3.96

post 4.95 3.16 Case 6 pre 2.56 1.77

Case 8 pre 3.81 3.35

post 2.84 2.04

z+SE pre 3.69f0.54 2.56k0.39 post 4.89 k 1 .OO* 3.05 f 0.48

vs. pre-lipolysis, P < 0.05.

REMODELLING O F PLASMA LIPOPROTEINS IN HYPERTRIGLYCERIDAEMIA 429

Figure 7. SDS-polyacrylarnide gel ( 15%) elcctrophoretograms of HDLz and HDL? from FLPL-d (upper panel) and GSD-I patient (lower panel). Lanes I and 2. apo A-I and apo A-11 standard respectively; Lanes 3 and 5 . HDLz and HDLl respectively without LPL; Lanes 4 and 6. HDLz and HDL3 respectively with LPL.

lipase incubation. This was accompanied by a significant increase in per cent level ofTG and PR as well as a pronounced decrease of PL (46'X)) in HDLx in GSD-I plasma. Similar experiments with LPL-treated plasma of FLPL-d led to a different pattern of HDLx fraction, in that both free and esterified cholesterol were higher than those from non-treated plasma.

Apoprotein characterization in the HDL subpopu- lations is shown in Fig. 7. The addition of LPL caused a significant elevation of the apo A-I : apo A-I1 molar ratio in the HDLz of both forms of hypertriglyceridaemia (Table 5). This ratio exhibited wide variations in HDL1; the apo A-I: apo A-11 ratio was lower in GSD-I patients and higher in FLPL-d patients following incubation with LPL.

It should be noted that in order to separate the influence of both LCAT and lipid transfer protein from LPL, incubations of freshly isolated plasma were performed at 4°C and 37°C without LPL supplemen- tation. There was no alteration in the composition, size and concentration of all the lipoprotein fractions.

Discussion The present study contributes to an understanding of the relevant role of LPL in lipoprotein abnormalities in severe hypertriglyceridaemia and shows that some of these alterations are reversible upon addition of LPL. Moreover, our data again point to a major role of lipolysis of triglyceride-rich particles in formation and remodelling of IDL, LDL, HDLl and HDLx. Absence of lipoprotein lipase is responsible for the most severe form of clinical hypertriglyceridaemia, primary hyperlipoproteinaemia type I [ 2 ] . This disorder is also known to be associated with very low concentrations of HDL cholesterol [33]. In addition, alterations are found in the chemical composition of the lipoproteins, including TG enrichment and cholesteryl ester depletion of LDL and HDL [2]. Similar findings have recently been described in glycogen storage disease type I. with secondary hyperlipoproteinaemia and low activity of lipoprotein lipase [25].

Our data suggest that the raised TG levels are attributable to impaired catabolism due to decreased activity of extrahepatic lipoprotein lipase in both disorders. The deficient breakdown led to an overpro- duction of chylomicron and VLDL particles, as documented in this study with two different systems: agarose gel electrophoresis and ultracentrifugation.

Our results indicate that LPL has an important role in the metabolism of all lipoprotein classes. With hydrolysis, the principal carriers of TG, CM and VLDL diminished, giving rise to increased levels of IDL, LDL and HDL?. On the other hand, HDL3 levels generally showed a reciprocal fall. During the conversion of VLDL to LDL, IDL particles accumulated in the plasma, as described by Eisenberg c't a/. [34]. Whether CMscontribute to IDL and LDL by a similar mechanism and to the same extent is uncertain. I t is interesting to note that apo B-48 was present in the LDL fraction visualized by 4% SDS-gel electrophoresis suggesting that some CM may be converted to particles in the LDL density range. In normal indi- viduals, if B-48-containing LDL-like particles are formed in plasma, they may be rapidly cleaved. Of interest, B-48 has been described in LDL of subjects with sluggish clearance of TG-rich particles and remnants (type 111).

LDL-like particles are formed in iiitro by incubating human VLDL with bovine milk LPL [8]. Conversely, lipolysis blocked by anti-LPL serum in zizw caused a decrease in LDL levels while VLDL accumulated [ 151. In the current study, as TG-rich lipoproteins were hydrolysed by LPL, smaller, cholesterol-enriched intermediate-density lipoproteins were formed. The

430 E. LEVY et al.

latter were probably converted to LDL, because it is believed that nearly all LDL derive from VLDL through IDL [35]. However, independent production of LDL has been reported in patients with familial hypercholesterolaemia [36,37] and in some species [38]. In our study, the formation of IDL and LDL was accompanied by specific changes in their composition and structure, while lipoprotein precursors (VLDL) were reduced in both size as well as total mass, mainly because of depletion of TG. Post-lipolysis triglyceride- rich particles then contributed to the increase in total mass of IDL and LDL. Quantitation of other VLDL constituents also showed substantial depletion. Taking into account the small size and the mass increment of the IDL and LDL after the incubation period, our data suggest that VLDL particles convert to numerous small IDL and LDL particles in whole plasma in oitro if LPL activity is present.

Evidence obtained from our studies stress the role of lipolysis in the rapid conversion of sub-populations of HDL. The incubation of whole plasma in the presence of active LPL led to a pronounced decrease of HDL3 accompanied by an increase of HDLz. The variation of the two HDL subclasses support previous reports indicating a conversion of HDL3 to HDLz during degradation of TG-rich lipoproteins by lipoprotein lipase. Patsch et al. have shown that the degradation of VLDL by LPL is accompanied by a partial conversion of HDL3 towards HDL2 [8]. A similar shift was observed in the ultracentrifugal Schlieren pattern during heparin-induced intravascular lipolysis [20]. The transformation of HDL is probably the result of the release of surface components from TG-rich lipoproteins and their accumulation by HDL3. This interpretation is supported by the replenishment of cholesterol and PL of HDL3 which thus became less dense (i.e. HDL2). Changes in the levels of HDL are of special interest because of the negative association of this lipoprotein with atherosclerotic cardiovascular disease [ I 1,391. This inverse relationship between plasma HDL concentration and coronary atherosclerosis is mediated by HDL2 rather than by HDL3 [40]. Investigation into the cause of the changes in the properties of HDL have yielded conflicting conclusions. Studies of pig and rabbit serum by Ha et al. [41] and by Rye & Barter [42] implicated a serum factor in this process, but tended to rule out the involvement of LCAT or the lipid transfer factor; similar conclusions were reached by Gambert et al. [43] in studies of rabbit, human and rat serum. Recently, a protein factor distinct from LCAT has been reported as solely responsible for most, if not all, of the conversion of smaller to large HDL species [19]. Contrary to these results were the studies of pig and human plasma by Knipping et al. [44] in which no evidence was found supporting the existence of such HDL transformation factors apart from LCAT.

The processes by which lipid and protein molecules move from one lipoprotein to another are not entirely understood. Possibilities include the shuttling of lipids

by transfer proteins, movement of free soluble apolipoproteins in the plasma, and the transient formation of physical complexes between lipoprotein particles. A recent investigation, reported that increasing amounts of HDL diminished and eventually completely inhi- bited VLDL-LDL complex formation [45]. It is still unknown whether the small amounts of HDL charac- terizing FLPL-d and GSD-I patients might favour such a physical association. This possibility merits additional study in order to better understand remodelling of lipoproteins in the course of their intravascular metabolism. The results of our study document myriad consequences of human plasma incubation, which appear to be largely attributable to the action of LPL. The major effects of LPL resulted in changes of size, protein content, lipid composition and apoprotein distribution in HDL fraction. As mentioned above, a transformation of HDL3 occurred. These alterations were not observed when the plasma (containing all its physiological constituents such as LCAT and lipid transfer protein) was deprived of LPL. We interpret our data to mean that the initiation of the process HDL3+HDL2 needs first the lipolytic action of TG-rich particles by LPL and subsequently, the other factors might give their contribution. This obligatory step is excluded in patients with LPL deficiency.

Consequently, the cascade of reactions including the hydrolysis of phospholipids at the surface of TG-rich lipoproteins, the breakdown of T G in the core, the formation of cholesteryl esters from free cholesterol to produce spherical high-density lipoproteins, and the conversion of HDL3 to HDLz is blocked.

Acknowledgments

This work was supported in part by the Medical Research Council of Canada (MT-10584), the US National Institutes of Health (HL 21006) and by the Swedish Medical Research Council grant ( B 13X-727). Emile Levy is the recipient of a scholarship from Le Fonds de Recherche en Sante du Qukbec.

We appreciate D. St-Cyr Huot’s assistance in manu- script preparation.

References I Nilsson-Ehle P, Garfinkel AS, Schotz MC. Lipolytic enzymes

and plasma lipoprotein metabolism. Ann Rev Riochem

2 Breckenridge WC. Deficiencies of plasma lipolytic activities. Am Heart J 1987;113:567-73.

3 Nikkili EA, Taskinen MR. Plasma high-density lipoprotein concentration and subfraction distribution in relation to triglyceride metabolism. Am Heart J 1987;113:543-8.

4 Egerlud T, Olivecrona T. The purification of a lipoprotein lipase from bovine skim milk. J Biol Chem 1972;247:6212-17.

5 Eisenberg S . Plasma lipoprotein conversions. Methods in Enzy- mology 1986;129:347-66.

6 Tall AR, Small DM. Plasma high-density lipoproteins. N Engl J Med 1978;299: 1232-6.

7 Nikkila EA. Metabolic and endocrine control of plasma high

1980;49:667-93.

REMODELLING OF PLASMA LIPOPROTEINS IN HYPERTRIGLYCERIDAEMIA 43 1

density lipoprotein concentration. In: Gotto AM Jr, Miller N E & Oliver MF. eds. High Density Lipoproteins and Atherosclerosis. Amsterdam: ElsevieriNorth Holland Biomedical Press. 1978: 177-92.

8 Patsch JF, Gotto AM Jr, Olivecrona T, Eisenberg S. Formation of high density lipoproteinz-like particles during lipolysis of very low density lipoproteins in riifro. Proc Natl Acad Sci USA 1978;75:45 19-23,

9 Tall AR. Blum CB. Forester GP. Nelson CA. Changes in the distribution and composition of plasma high density lipoproteins after ingestion of fat. J Biol Chem 1982;257:198-207.

10 Patsch JR. Karlin JB, Scott LW. Smith LC. Gotto AM Jr. Inverse relationship between blood levels of high d t 4 ty lipoprotein subfraction 2 and magnitude of postprandial lipema. Proc Natl Acad Sci USA 1983;80:1449 53.

I I Tyroler HA. Epidemiology of plasma high density lipoprotein cholesterol levels. Circulation 1980;62(suppl IV): I 3.

12 Nikkill EA, Taskincn MR. Kekki M. Relation of plasma high- density lipoprotein cholesterol to lipoprotein-lipase activity in adipose tissue and skeletal muscle of man. Atherosclerosis 1978:29:497- 501.

13 Taskinen MR, Kashyap ML, Srivastava LS. Ashraf M, Johnson JD, Perisutti G . Brady D, Glueck CJ. Jackson RL. In ritro catabolism of human plasma very low density lipoproteins. Effects of VLDL concentration on the interconversion of high density lipoprotein subfractions. Atheroscl~rosis 1982;41:381 94.

14 Anderson DW, Nichols AV. Pan SS, Lindgrcn FT. High density lipoprotein distribution. Resolution and determination of three major components in a normal population sample. Athero- sclerosis I978;29: I6 1-79,

IS Behr SR. Patsch JR. Forte T, Bensadoun A. Plasma lipoprotein changes resulting from immunologically blocked lipolysis. J Lipid Res I98 I :22:443 5 I .

16 Tall AR. Green PHR. Incorporation of phosphatidylcholine into spherical and discoidal lipoproteins during incubation of egg phosphaditylcholine vesicles with isolated high density lipoproteins o r with plasma. J Biol Chem 1981;256:2035 44.

17 Schmitz G . Assmann G . Melnik B. The role of lecithin: cholesterol acyl transferase in HDLjIHDL2 interconversion. Clin Chim Acta 1981;119:225 36.

18 Hopkins GJ. Chang LBF, Barter PJ. Role of lipid transfers in the formation of a subpopulation of small high density lipoproteins. J Lipid Res 1985;26:218-29.

19 Rye KA. Barter PJ. Changes in the size and density of human high-density lipoproteins promoted by a plasma-conversion factor. Biochim Biophys Acta 1986;875:429-38.

20 Dieplinger H. Zechner R, Kostner GM. The in isitro formation of HDLz during the action of LCAT: the role of triglyceride-rich lipoproteins. J Lipid Res 1985;26:273-82.

21 Swaney JB. Orishimo MW. Girard A. Enzymatically induced alterations in the structure of rat serum lipoprotein. J Lipid Res 1987;28:982-92.

22 Brunzell JD, Miller NE. Alanpovic P. St-Hilaire RJ, Wang CJ. Sarson DL. Bloom SR, Lewis B. Familial chylomicronemia due to a circulating inhibitor of lipoprotein lipase activity. J Lipid Res 1983;24:12 00.

23 Fielding CJ, Have1 RJ. Lipoprotein lipase. Arch Pathol Lab Med 1977;101:225 9.

24 Miller NE. Rao SN. Alanpovic P, Noble N, Slack J, Brunzell JD. Lewis B. Familial apolipoprotein C-11 deficiency: plasma lipoproteins and apolipoproteins in heterozygous and homozygous subjects and the cffccts of plasma infusion. Eur J Clin Invest 1981;11:69-76.

25 Levy E, Thibault L, Roy CC, Lepage G . Letarte J. Circulating lipids and lipoproteins in glycogen storage disease type I following nocturnal intragastric feeding. J Lipid Res l988;29:215~ 26.

26 Lowry OH, Rosebrough NJ, Farr AL, Randall RT. Protein measurement with the Folin phenol reagent. J Biol Chem I95 I ; I93:265-75.

27 Bartlett GR. Phosphorus assay in column chromatography. J Biol Chem 1959;234:466-8.

28 Patsch W, Sailer S, Braunsteiner H . An enzymaticmethod for the determination of the initial rate of cholesterol esterification in human plasma. J Lipid Res 1976;17:182-5.

29 Lopez-Virella MF, Stone P, Ellis S, Colwell JA. Cholesterol determination in high density lipoprotein separated by three different methods. Clin Chem 1977;23:882-91.

30 Schriewer B, Nolte W, Assmann G . VLDL apolipoprotein B determination in blood serum following precipitation of LDL with polyvinylsulphate. J Clin Chem Clin Biochem 1985;23:349 53.

31 Krauss RM, Levy R1, Frederickson DS. Selective measurement of two lipase activities in post-heparin plasma from normal subjects and patients with hyperlipoproteinemia. J Clin Invest

32 Belfrage P, Vaughan M. Simple liquid-liquid partition system from isolation of labelled oleic acid from mixtures with glycer- ides. J Lipid Res 1969;10:341-4.

33 Frederickson DS, Levy RI. Lindgren FT. A comparison of heritable abnormal lipoprotein patterns as defined by two different techniques. J Clin Invest 1968;47:2446-57.

34 Eisenberg S, Bilheimer DW, Lindgren FT. Levy R1. On the metabolic conversion of human plasma very low density lipoprotein to low density lipoprotein. Biochim Biophys Acta I973;326:36 1-71,

35 Berman M, Eisenberg S, Hall MH, Levy RI. Bilheimer DW, Phair RD. Goebel RH. Metabolism of apo B and apo C lipoproteins in man: kinetic studies in normal and hyperlipopro- teinemic subjects. J Lipid Res 1978: 19:38-56.

36 Soutar AK, Myant NB, Thompson G R . Simultaneous measurement ofapolipoprotein B turnover in very density lipoproteins in familial hypercholesterolemia. Atherosclerosis I977;28:247 -56.

37 Soutar AK, Myant NB, Thompson G R . Metabolism ofapolipoprotein B-containing lipoproteins in familial hypercholesterolemia. Atherosclerosis 1979;32:315-25.

38 Hamilton RL. Hepatic secretion of nascent plasma lipoproteins. In: Glaumann H, Paters T J r & Redman C, ed. Plusniu Protein Secretion by the Lirier. London: Academic Press. 1983:357 ~ 7 4 .

39 Miller RJ, Miller NE. Plasma high density lipoprotein concentration and development of ischaemic heart disease. Lancet 1975;i: 16- 19.

40 Ballantyne FC. Clark FS, Simpson HS, Ballantyne D. High density and low density lipoprotein subfractions in survivors of myocardial infarction and in control subjects. Metabolism

41 Ha YC. Gorjatschko L, Barter PJ. Changes in the density distribution of pig high density lipoproteins during incubation in ritro. Atherosclerosis 1983;48:253-63.

42 Rye KA, Barter PJ. Evidence of the existence of a high-density lipoprotein transformation factor in pig and rabbit plasma. Biochim Biophys Acta 1984;795:230- 7.

43 Gambert P, Lallemant C, Athias A. Padieu P. Alterations of HDL cholesterol distribution induced by incubation of human serum. Biochim Biophys Acta 1982;713: 1--9.

44 Knipping G , Zechner R, Kostner G M , Holasek A. Factors affecting the conversion of high-density lipoproteins: experiments with pig and human plasma. Biochim Biophys Acta

45 Musliner TA, Michenfelder HJ, Krauss RM. Interactions of high density lipoproteins with very low and low density lipoproteins during lipolysis. J Lipid Res 1988;29:349-61.

1974;54: 1107-24.

1982;31:433-7.

1985;835:24452.

Documents

In vitro remodelling of plasma lipoproteins in whole plasma by lipoprotein lipase in primary and secondary hypertriglyceridaemia