Ethanolamine and choline phospholipids in nascent very-low-density lipoprotein particles

Clin Invest Med 1996; 19 (4): 243-250

(Original manuscript submitted July 6, 1995; received in revised form Feb. 26, 1996; accepted Feb. 29, 1996)

Copyright 1996, Canadian Medical Association

• Abstract
• Résumé
• Introduction
• Materials and methods
• Results
• Discussion
• Conclusions
• Acknowledgements
• References

Design: Analysis of PE and PC fractions in blood samples taken from the subjects while fasting and at four 2-h intervals after eating a 4749-kJ breakfast.

Participants: Five healthy subjects.

Outcome measures: Levels of PE and PC fractions isolated from blood samples by ultracentrifugation (to isolate the d<1.006 fraction of plasma) and reversed-phase chromatography.

Results: The concentration of triglyceride in particles of density less than 1.006 g/cm² in the samples increased rapidly, peaking 4 h after the meal at 0.51 mmol/L above the fasting level. Apolipoprotein B-48 was maximal in the sample taken 2 h after eating and below 1%, as a percentage of apolipoproteins B-48 and B-100, in all other samples. The concentration of PC exceeded that of PE by an order of magnitude, but there was a proportionately greater increase in PE postprandially. PE contained a rapidly cleared alkenylacyl fraction, which was maximal (at a threefold increase over the baseline level) in the sample taken 4 h after eating.

Conclusions: As blood triglyceride levels rise and fall postprandially, the polar lipid composition of the TRLs also changes; these changes are more marked in the PE than in the PC fraction. The fact that the level of the PE fraction peaked later (at 4 h after eating) than that of B-48 apolipoprotein (which peaked at 2 h after eating) suggests that the peak in the PE fraction had a hepatic (very-low-density lipoprotein) origin. Alkenylacyl (plasmalogen) phospholipid accompanied both PE and PC fractions but was about 10 times greater in PE. This result is significant if alkenylacyl PE does indeed function as an antioxidant.

[Table of Contents]

Résumé

Méthodes : Cinq sujets en bonne santé ont ingéré à 8 heures un petit déjeûner de 4749 kJ. Des prises de sang eurent lieu à jeûn puis aux 2 heures pendant 8 heures. La fraction de plasma de d<1,006 a été isolée par ultracentrifugation; les fractions PE et PC ont été isolées par chromatographie à polarité de phase inversée.

Résultats : La concentration de triglycérides dans les particules de d<1,006 a augmenté rapidement et atteint son niveau maximal 4 heures après le repas, à 0,51 mmol/L au-dessus du niveau à jeûn. L'apoliprotéine B-48 a atteint son niveau maximal à 2 heures; le niveau était moins de 1 % dans les autres prélèvements. Dans les LRT, PC était supérieur à PE, mais une augmentation proportionnellement plus grande de PE s'est produite durant la période post-prandiale. PE contenait une fraction d'alkénylacyl rapidement éliminée et qui a atteint son niveau maximal (du triple de la valeur de base) dans le prélèvement effectué 4 heures après le repas.

Conclusions : Pendant que les triglycérides augmentent puis diminuent durant la période post-prandiale, la composition des lipides polaires des LRT est modifiée; ces changements sont plus marqués dans la fraction PE que dans la fraction PC. Le fait que la fraction PE atteigne son niveau maximal plus tard que l'apoliprotéine (B-48) suggère une origine hépatique (LTFD) à la fraction PE. La présence de phospholipide alkénylacyl a été observée dans les fractions PE et PC, mais sa concentration était 10 fois supérieure dans la fraction PE. Ceci serait très significatif s'il se trouve que l'alkénylacyl PE a véritablement une fonction antioxydante.

[Table of Contents]

Landin and Nilsson[3] studied the clearance of phospholipids from postprandial triglyceride-rich particles in rats, and they drew attention to the more rapid removal of the glycerophosphocholine (PC, also known by the older term phosphatidylcholine) fraction than of the glycerophosphoethanolamine (PE, also known by the older term phosphatidylethanolamine) fraction. No similar studies in
humans have been reported. Landin and Nilsson's observation is of interest because of the importance ascribed to oxidation of plasma lipids, particularly the LDL particles, in atherogenesis.[4] The PE fraction, although the smaller of the two, is relatively enriched in alkenylacyl glycerophospholipids (plasmalogens).[5] Several recent articles have ascribed an antioxidant function to this species of polar lipid.[6-8] This antioxidant function may be important if some part of the lipid-oxidation process is initiated in triglyceride-rich lipoprotein (TRL) particles as these are processed in the circulation.

Postprandial studies of TRL particles have generally involved a test meal consisting predominantly of fat, with flavoured dairy cream a common choice of food.[2,9] This type of meal increases chylomicrons, but may also increase VLDL, either because of competition for clearance by the influx of chylomicrons or because delivery of free fatty acid from peripheral tissues to the liver (where chylomicrons have been cleared) stimulates VLDL synthesis. We argue that a mixed meal is a more effective stimulus to triglyceride-rich VLDL particle secretion. The carbohydrate load, together with elevated postprandial insulin levels, stimulates triglyceride synthesis in the liver.

We wished to determine whether there was a greater relative increase in the PE fraction than in the PC fraction postprandially and whether the PE fraction contains a rapidly cleared component of alkenylacyl-PE.

[Table of Contents]

Materials and methods

Materials

We obtained 2-¹⁴C-arachidonyl-sn-glycerol-phosphoethanolamine, specific activity 2.04 × 10⁹) Bq/mmol, from Amersham Canada Ltd.(Oakville, Ont.), phosphatidylethanolamine (cat. no. A-351) and phosphatidylcholine (cat. no. A-29) from Serdary Research Laboratories (London, Ont.), phospholipase C, grade II from Bacillus cereus, from Boehringer Mannheim (Montreal), reagents for polyacrylamide gel electrophoresis from Biorad Laboratories (Mississauga, Ont.), and solvents for extraction and chromatography from Caledon Scientific (Georgetown, Ont.). Chromatographic-grade solvents were purchased and were not redistilled before use. Other laboratory reagents were of the best available quality from either Sigma or Fisher Scientific. The aminopropyl reversed-phase preparatory columns were from J.T. Baker (Toronto; cat. no. 7088-03).

Subjects

The five subjects (four men and a woman) were healthy members of the staff of the McMaster University Health Sciences Centre. Their fasting total cholesterol levels were less than 6.2 mmol/L and their fasting triglyceride levels were less than 1.7 mmol/L (Table 1). The male subjects regularly engaged in physical activity; the female was sedentary. Subjects were instructed to fast overnight; they attended the clinical investigation unit at 0800 h, when the first blood sample was drawn. They then ate the meal provided, and further blood samples were drawn every 2 h, with the last taken at 1600 h.

Meal

The meal comprised cereal (508 kJ, 2 g fat), 2% milk (945 kJ, 2.5 g fat), six pieces of side bacon (676 kJ, 18 g fat), two eggs (697 kJ, 8 g fat), two containers of jam (462 kJ, no fat), two slices of whole-wheat toast (550 kJ, 2 g fat) and a muffin (1016 kJ, 10 g fat) for a total of 4749 kJ and 54.5 g fat. (Nutritional analysis by Ms. Shannon Gadowski, RDA, based on the Canadian Nutrient File and the USDA Handbook #8.) This meal may be compared with a typical fat load of 1.5 g/kg, which is used to study the postprandial response[2] that provides a 70-kg person with 3970 kJ.

Preparation of TRL particles

Blood samples of 20 mL were collected into EDTA, and plasma was prepared. The density was brought to 1.006 g/cm³, and aliquots of 4 mL were centrifuged at 37 500 g for 18 h. Tubes were sliced to harvest the VLDL and any lighter particles present. Total triglyceride levels were determined, an aliquot was set aside for protein electrophoresis and the remaining sample taken for analysis of polar lipids.

Analysis of polar lipids

2-¹⁴C-PE (918 Bq) was added to 4 g of VLDL-enriched supernatant after centrifuging, and a dry chloroform extract was prepared according to Bligh and Dyer.[10] This extract was taken up in a minimum volume of chloroform and methanol, in a 1:1 ratio, containing 0.01% (weight per volume) of butylated hydroxytoluene, and fractionated on a reversed-phase aminopropyl column according to Pietsch and Lorenz;[11] only neutral lipids, free fatty acids (if any) and the PC and PE fractions were eluted. Using a mixture of PC, PE and olive oil (as a neutral lipid), we performed thin-layer chromatographic analysis[12] of the eluted fractions. This analysis confirmed that this method gives clean separation into these same fractions. One tenth of the PE and PC fractions eluted from the aminopropyl column were taken for phosphorus determination according to Chalvardjian and Rudnicki.[13]

The balance of the choline and ethanolamine glycerolipids were treated with phospholipase-C according to Myher, Kuksis and Pind.[14] After meticulous drying, the 1,2-substituted glycerides were redissolved in acetonitrile (dried by passage through anhydrous sodium sulfate) with 10 mg of benzoic anhydride and 4 mg of dimethylaminopyridine. This reaction was stopped with ammonium hydroxide after 1 h and the diradyl glycerobenzoates extracted into n-hexane, dried under nitrogen, taken up in chloroform and passed through silica gel to remove the dimethylaminopyridine. The chloroform eluate was dried off and the residue taken up in a minimum volume of chloroform with 1 mg/mL of benzoic anhydride as internal standard.

High performance liquid chromatography (HPLC) was performed according to Blank, Cress and Snyder[15] with the use of a 5 µ, normal-phase, Absorbosphere silica column in a Waters 625 LC system. The mobile phase was cyclohexane:methyl-t-butyl-ether:acetic acid (in a ratio of 97:3:0.035). Benzoates were detected at 230 nm using a Waters 486 Tunable Absorbance Detector. Fractions were collected using a Gilson #203 microfraction collector and effluent collected into 20-mL glass scintillation vials containing 14 mL of scintillation fluid. Radioactivity was determined in a Beckman LS 5801 Liquid Scintillation System. Peak areas were determined using Waters' Baseline 810 Chromatography Workstation Software (version 3.3, Waters Canada, Toronto) and a calibration curve. The internal standard permitted correction for evaporation of the solvent in which the benzoates were redissolved and for column performance. Overall recovery of the PE fraction was calculated from the recovered radioactivity. Separation of PE and PC was confirmed by the absence of radioisotope from the eluted PC fractions; it was assumed that overall recovery of the PE and PC fractions was similar.

Sodium dodecyl sulfate (SDS) gel electrophoresis

The Lowry method was used to determine the protein content of the fractions with a density of less than 1.006 g/cm³ isolated from plasma by ultracentrifugation. Aliquots of 50 *g of protein were delipidated and examined by SDS-gel electrophoresis according to Zilversmit and Shea.[16] After staining and drying, gels were scanned by laser densitometry (Howtek Laser Scanner with Whole-band Analysis Software from Bioimage, Ann Arbor, Mich.) to determine the relative amounts of material with molecular weights of 514 kd and 264 kd; the gels were calibrated using a mixture of proteins of known molecular weight (high range) obtained from Biorad (cat. no. 71605) and consisting of myosin (205 kd), beta-galactosidase (118 kd), bovine serum albumin (85 kd) and ovalbumin (47 kd).

Statistical methods

Standard methods were used to calculate means and standard deviations; Student's t-test was used to test the significance of the difference between sample pairs.

[Table of Contents]

Results

The mean triglyceride response in the fraction with a density of less than 1.006 g/cm³ of the five subjects after eating a standard meal is shown in Fig. 1. In every subject, the triglyceride response peaked 4 h after eating and started to decline by 6 h after the meal. Considerable variation among individuals is to be expected in this response; although the response in the four men was homogeneous, the woman had a higher fasting triglyceride level of 1.53 mmol/L. Although the postprandial increase in the triglyceride level in the woman was similar to that in the four men, the higher fasting level in the woman increased the variance of the pooled data. For this reason, Fig. 1 shows the triglyceride levels relative to that at baseline.

We conducted SDS polyacrylamide gel electrophoresis of each of the samples collected at each time point. Except in the 2-h postprandial sample, the level of apolipoprotein B-48 did not exceed 1% of the sum of the combined apolipoprotein B-48 and apolipoprotein B-100. The percentage of apolipoprotein B-48 in this combined level is shown in Fig. 1; at 2 h, the apolipoprotein B-48 level reached 2.8% of the combined level (standard deviation [SD] 1.41%), a level significantly different (p <0.05) from that found in both the baseline and 4-h samples.

The mean fasting level of PE was 22.16 (SD 14.20) nmol per sample, as determined by phosphorus assay of the eluate from the aminopropyl columns. The level of PE peaked in the 2-h sample (taken at 1000 h), at 69.70 (SD 22.05) nmol per sample, then returned close to baseline, to 29.3 (SD 16.5) nmol per sample in the 8-h sample. The mean fasting level of PC was some 40 times greater than that of PE, at 834 (SD 594) nmol per sample. The mean PC level also peaked in the 2-h sample, at 1469 (SD 852) nmol per sample, but returned more slowly to the baseline, at 910 (SD 637) nmol per sample after 8 h. Much of the variance was contributed by the higher baseline levels in the female subject.

The results of HPLC analysis are shown in Fig. 2. The PE fractions reached maximal levels after 4 h (the 1200-h sample); these fractions showed larger proportional increases than did the PC fractions, and alkenylacyl-PE increased more, in relative terms, than the diacyl component of PE. There was a threefold average increase in alkenylacyl-PE between baseline and 4 h, and a twofold increase in diacyl-PE during this period. The increase in the PC fractions was maximal in the first sample, taken at 2 h; diacyl-PC increased 1.67 times and alkenylacyl-PC increased 1.73 times. The amounts of alkenylacyl-PE and PC phospholipids were comparable (6 to 30 nmol per sample), whereas the level of diacyl-PC exceeded that of diacyl-PE by an order of magnitude. Thus, the time at which a maximum level was reached after the test meal differed between the two phospholipid fractions.

The fractions' postprandial enrichment with an alkenylacyl component also differed (Fig. 3). The ratio of alkenylacyl- to diacyl-PE increased from 0.17 (SD 0.069) at 0800 hours to 0.31 (SD 0.07) at 1200 hours (p <0.01), whereas the ratio of alkenylacyl- to diacyl-PC showed no significant change during the 8 h of the study.

[Table of Contents]

Discussion

Postprandial lipids have been studied extensively. Common practice[2] has been to provide a meal predominantly composed of fat. We chose to provide a mixed meal, in the expectation that the non-fat energy load (2100 kJ) would stimulate VLDL secretion by the liver. Others have also reported results after various types of mixed meals. Karpe and associates[17] provided subjects with a meal in which about 60% of energy was in the form of fat; they studied the composition of TRL particles at 3, 6 and 12 h after the meal and analysed proteins by SDS-PAGE. Our results regarding apoliproteins B-48 and B-100 are similar to theirs, with an increase in apolipoprotein B-48 shown only in the first postprandial sample. This is consistent with the rapid clearance of chylomicrons, which have a half-life of about 5 min in people with normal triglyceride levels.[2]

A previous study reported that postprandial triglyceride levels peak 5 to 6 h after a meal of 1.5 g fat/kg;[2] however, in this study, with an similar energy load but less fat, the level peaked 4 h after the meal. We determined the relative amounts of B-100 and B-48 apolipoprotein in the particles with a density of less than 1.006 g/cm³ recovered by ultracentrifugation and showed that the gut-derived particles contributed less than 1% of the total protein level, except in the 2-h sample, in which the proportion of apolipoprotein B-48 reached 2.6%. The test meal we used reflects what people eat more realistically and, therefore, may provide a more normal view of postprandial events. We could not have used the ultracentrifuge to separate gut- and liver-derived particles, because of the overlapping sizes of chylomicrons, chylomicron remnants and triglyceride-rich VLDL particles. Only affinity chromatography can separate particles on the basis of their apolipoprotein-B protein,[18] but it cannot do so easily or on a scale that would permit the phospholipid analysis reported here.

The time-course of events shown in Figs. 1 and 2 provides evidence for a hepatic origin for the postprandial increase in PE, and particularly of alkenylacyl-PE. First, apolipoprotein B-48 appears only briefly in the 2-h sample, suggesting that chylomicron appearance and clearance has been largely completed before the 4-h sample. Second, both levels of triglycerides and PE continue to increase between the 2-h and 4-h samples, and, third, this increase is more marked for the alklenylacyl-PE fraction.

Fig. 2 also shows that the increase in alkenylacyl- and diacyl-PE phospholipids occurred later (it was maximal at 4 h) than the increase in PC phospholipids. An explanation for this, consistent with the foregoing, is that liver-derived TRL membrane is more highly enriched in PE phospholipid. Fig. 3 emphasizes that not only does the PE fraction of phospholipid become relatively enriched during the postprandial clearance of TRLs, but the alkenylacyl fraction of PE also increases relative to the diacyl fraction during the same period.

Few studies of the polar lipids of postprandial TRL-particles have been reported. During lipolysis of TRLs, apolipoproteins contribute to the growth of HDL-particles, as does the excess lipoprotein membrane. Thus, following these polar lipids into the HDL-particle family is of particular interest. One recent study[19] demonstrated the preferential labelling of erythrocyte membrane by acyl moieties from the sn-2 position of PE in the LDL lipoprotein fraction. The results presented here also suggest that the labile phospholipid membrane of TRL particles could selectively donate components of PE-plasmalogen to the tissues. Compared with other biological membranes, and with the content of choline phosphoglycerides, the content of plasmenyl ethanolamine in VLDL is small, but it does appear to be particularly labile. Vance,[8] in a study of PE in cultured rat hepatocytes and in the surrounding medium, found that the medium (into which VLDL particles are secreted in this system) was enriched about fivefold in PE relative to cell membrane.

The biological function of plasmenyl phospholipids is largely unknown, but the fact that these phospholipids tend to enrich PE fractions, and the association of PE with the inner leaflet of the phospholipid bilayer, is well established.[20] Glaser and Gross[21] have suggested that plasmenyl phosphoethanolamines may play a special role in membrane fusion; this process is presumed to be important in both the secretion and clearance of VLDL particles. Morand, Zoeller and Raetz[6] put forward the interesting idea that the 1-O-alk-1'-enyl substituent of the plasmalogens has antioxidant properties, and they showed that protection against ultraviolet irradiation was lost in Chinese hamster ovary cells when a peroxisomal defect prevented plasmalogen biosynthesis. Engelmann, Brautigam and Thiery[22] compared the loss of tocopherol and plasmalogen from LDL particles exposed to oxidative challenge, and they also advanced evidence that plasmalogen is an antioxidant.

The 2-position of the glycerol skeleton of plasmalogens is preferentially occupied by a polyunsaturated fatty acid,[23] such as arachidonic acid. It has been postulated that this position is "protected" from oxidation by the 1-O-alkenyl group. If this theory is true, the phospholipid membrane of VLDL could be acting as a "courier" to deliver essential polyunsaturates from the liver to the peripheral tissues, while protecting them from peroxidation. This view is supported by the findings of Engelmann's group.[19]

This discussion would be incomplete without mentioning the evidence concerning the oxidation of LDL particles and atherogenesis. Most of the work in this area has dealt with the postabsorptive state, but, as argued by Zilversmit,[1] more attention should be paid to postprandial events. If products of free-radical attack (e.g., lipid peroxy radicals) begin to accumulate in apolipoprotein B-100 particles early in their life history, while they are being rapidly depleted of triglycerides, and if plasmenyl-PE does indeed possess antioxidant properties, the persistence of these particles after depletion of their antioxidant plasmalogen could have a special significance.

[Table of Contents]

Conclusions

[Table of Contents]

Acknowledgements

[Table of Contents]

References

1. Zilversmit DB. Atherogenesis: a postprandial phenomenon. Circulation 1979;60:473-85.
2. Tall AR. Metabolism of postprandial lipoproteins. Meth Enzymol 1986;129:469-82.
3. Landin B, Nilsson A. Metabolism of chylomicron phosphatidylethanolamine in the rat. Biochim Biophys Act 1984;793:105-13.
4. Steinberg D, Parthasarathy S, Carew TE, Khoo JC, Witztum JL. Beyond cholesterol: modifications of low-density lipoprotein that increase its atherogenicity. N Engl J Med 1989;320:915.
5. Lazarow PB, Moser HW. Disorders of peroxisome biogenesis. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic basis of inherited disease. New York: McGraw-Hill, 1989:1479-509.
6. Morand OH, Zoeller RA, Raetz CR. Disappearance of plasmalogens from membranes of animal cells subjected to photosensitized oxidation. J Biol Chem 1988;263:11597-606.
7. Zoeller RA, Rangaswamy S, Herscovitz H, et al. Mutants in a macrophage-like cell line are defective in plasmalogen biosynthesis, but contain functional peroxisomes. J Biol Chem 1992;267:8299-306.
8. Vance JE. Lipoproteins secreted by cultured rat hepatocytes contain the antioxidant 1-alk-1-enyl-2-acylglycerophosphoethanolamine. Biochim Biophys Acta 1990;1045:128-34.
9. Patsch JR, Karlin JB, Scott LW, Smith LC, Gotto AM. Inverse relationship between blood levels of high density lipoprotein subfraction 2 and magnitude of postprandial lipemia. Proc Natl Acad Sci U S A 1983;80:1449-53.
10. Bligh EG, Dyer WJ. A rapid method of total lipid extraction and purification. Can J Biochem Physiol 1959;37:911-17.
11. Pietsch A, Lorenz RL. Rapid separation of the major phospholipid classes on a single aminopropyl cartridge. Lipids 1993;28:945-7.
12. Bitman J, Wood DL, Mehta NR, et al. Comparison of the phospholipid composition of breast milk from mothers of term and preterm infants during lactation. Am J Clin Nutr 1984;40:1103-19.
13. Chalvardjian A, Rudnicki E. Determination of lipid phosphorus in the nanomolar range. Anal Biochem 1970;36:225.
14. Myher JJ, Kuksis A, Pind S. Molecular species of glycerophospholipids and sphingomyelins of human erythrocytes: improved methods of analysis. Lipids 1989;24:396-407.
15. Blank ML, Cress EA, Snyder F. Separation of phospholipid subclasses as their diradylglycerobenzoate derivatives by normal-phase high performance chromatography. J Chromatogr 1987;392:421-5.
16. Zilversmit DB, Shea TM. Quantitation of apoB-48 and apoB-100 by gel scanning or radio-iodination. J Lipid Res 1989;30:1639-46.
17. Karpe F, Steiner G, Olivecrona T, Carlson LA, Hamsten A. Metabolism of triglyceride-rich lipoproteins during alimentary lipemia. J Clin Invest 1993;91:748-58.
18. Cohn JS, Johnson EJ, Millar JS, et al. Contribution of apo-B48 and apo-B-100 triglyceride-rich lipoproteins (TRL) to postprandial increases in the plasma concentration of TRL triglycerides and retinyl esters. J Lipid Res 1993;34:2033-40.
19. Engelmann B, Brautigam C, Kulschar R, et al. Reversible reduction of phospholipid bound arachidonic acid after low density lipoprotein apheresis. Evidence for rapid incorporation of plasmalogen phosphatidylethanolamine into red blood cell membrane. Biochim Biophys Acta 1994;1196:154-64.
20. Zachowski A. Phospholipids in animal eucaryotic membranes: transverse asymmetry and movement. Biochem J 1993;294:179-92.
21. Glaser PE, Gross RW. Plasmenylethanolamine facilitates rapid membrane fusion: a stopped flow kinetic investigation correlating the propensity of a major plasma membrane constituent to adopt an H-II phase with its ability to promote membrane fusion. Biochemistry 1994;33:5805-12.
22. Engelmann B, Brautigam C, Thiery J. Plasmalogen phospholipids as potential protectors against lipid peroxidation of low density lipoproteins. Biochem Biophys Res Commun 1994;204:1235-42.
23. Mead JF, Alfin-Slater RB, Howton DR, Popjak G. The amphipathic lipids. In: Lipids: Chemistry, Biochemistry and Nutrition. New York: Plenum, 1986:369-404.

The full text may also be ordered from the Canada Institute for Scientific and Technical Information (CISTI) or Institute for Scientific Information (ISI).