Effects of n-3 fatty acids on the NMR profile of plasma lipoproteins

Effects of n-3 fatty acids on the NMR profile of plasma lipoproteins

J. D. Bell,',* M. L. Barnard,* H. G. Parkes,§ E. L. Thomas,* C. H. Brennan,f S. C. Cunnane,** and P. C. Dagnelie* NMR Unit* and Department of Molecular Medicine,t Hammersmith Hospital, Du Cane Road, London; Chemistry Department,§ Birkbeck College, Gordon Square, London, United Kingdom; and Department of Nutritional Sciences,** University of Toronto, Toronto, Ontario, Canada

Abstract The effects of fish oil supplementation (14.5 g n-3 fatty acids/day) on plasma lipoprotein particles in healthy volunteers were assessed by high resolution "C and 'H nuclear magnetic resonance (NMR) spectroscopy. Resonances not previously observed in the 13C and 'H spectra of plasma and isolated lipoproteins were detected after fish oil ingestion. The 13C resonances, centered at 14.3, 127.1, and 131.6 ppm, have been assigned to specific carbon groups

CH2-IIZH = CH-CHp, respectively) in eicosapentaenoic acid (C20:5n-3) and docosahexaenoic (C22:6n-3) DHA. The new lipid resonance observed in the 'H spectra of plasma (0.941 ppm) is consistent with the incorporation of these n-3 fatty acids into lipoprotein particles. The presence of increased EPA and DHA in plasma lipids was confirmed by gas-liquid chromatography. A marked reduction in the intensity of the methylene signal from very low density lipoproteins (VLDL) was also observed with fish oil. This reduction arises from a decrease in plasma triglyceride concentration (ca. 18%) and a reduction in the number of VLDL particles. Transverse relaxation studies of isolated VLDL and low density lipoprotein (LDL) showed significant elevation in the T2 of the -(C&)"- and C&- signals from non-n-3 fatty acids. The relaxation characteristics and signal intensity of the novel 'H peak (0.941 ppm) point to the existence of n-3 enriched microenvironments within lipoprotein particles. I These findings suggest that incorporation of EPA and DHA into VLDL and LDL, after fish oil ingestion, leads to significant alteration in the molecular architecture of lipoprotein particles.-Bell, J. D., M. L. Barnard, H. G. Parkes, E. L. Thomas, C. H. Brennan, S. C. Cunnane, and P. C. Dagnelie. Effects of n-3 fatty acids on the NMR profile of plasma 1ipoproteins.J. Lipid

(GHs-CHpCH = CH-, CHs-CH2-CH = GH-CHz-, CH3-

Res. 1996.37: 1664-1674.

Supplementary key words microenvironments

transverse relaxation lipid mobility

In recent years, there has been increasing interest in the effects of dietary fish oil on plasma lipid metabolism (1-3). The main effect of dietary fish oil appears to be a lowering of total plasma triglyceride and VLDL concentrations (I). This is due to a combination of factors, including inhibition of triglyceride secretion and an

apparent enhancement of triglyceride removal (1,4,5). Several reports have shown a significant rise in the cholesterol/triglyceride ratio contained in VLDL and LDL, suggesting smaller lipoprotein particles (4). These studies have concentrated mainly on determining the concentration of eicosapentaenoic acid (EPA: C20:5n-3) and docosahexaenoic (DHA C22:6n-3) in different lipoprotein fractions and/or their effects on relative levels of plasma lipids and lipoprotein fractions. The effects that n-3 fatty acids may have on the structure and function of plasma lipoproteins have not been fully investigated. Parks and Rude1 ( 5 ) and Nenseter et ai. (6) have shown that dietary fish oil can induce changes in the thermal transition of LDL particles. However, due to the methodology used in their studies, little could be ascertained as to the general organization of the lipid components within the intact lipoprotein particles. To address this question we have in this study used nuclear magnetic resonance spectroscopy.

High resolution nuclear magnetic resonance (NMR) spectroscopy is a non-destructive technique which is finding increasing favor in metabolic studies. NMR spectroscopy has been applied to the study of a range of metabolic questions in a variety of biological systems, including body fluids (7,s) and tissue extracts (9). NMR spectroscopy has also been applied to the study of isolated lipoproteins, yielding valuable information re- garding their composition and structure (10). Recent improvements in NMR techniques and data processing have opened the possibility of studying different lipoprotein fractions in intact plasma (11-13). These meth-

Abbreviations: NMR, nuclear magnetic resonance; VLDL, very low density lipoproteins; LDL, low density lipoproteins; HDL, high density lipoproteins; EPA, eicosapentaenoic acid; DIU, docosahexaenoic acid: FA, fatty acids; Tz, transverse relaxation time; TI, longitudinal relaxation time; n-3, omega-3.

'To whom correspondence should be addressed.

1664 Journal of Lipid Research Volume 37, 1996

by guest, on February 13, 2016

ww

w.jlr.org

Dow

nloaded from

http://www.jlr.org/

ods enable complete qualitative and quantitative measurements of different lipoprotein fractions directly from the ‘H NMR spectra of intact plasma.

The purpose of the present study was to examine the effects of dietary fish oil on plasma lipoproteins, principally VLDL and LDL, by ’H and 13C NMR spectroscopy. A combination of chemical shift and spinecho measurements suggests that fish oil induces marked changes in the physical properties, as well as the composition, of lipoprotein particles. The use of NMR spectroscopy to assess the effects of dietary supplementation on the physicochemical status of plasma lipoprotein particles is discussed.

SUBJECTS AND METHODS

Subjects

Nine healthy volunteers (29-40 years old) took part in this study. All subjects had normal liver function and normal fasting plasma triglyceride concentrations (0.62 * 0.17 mmol/l). From 7 days before and throughout the study, subjects excluded alcohol and fish from their normal diet.

Protocol and blood samples

All volunteers consumed 50 ml/day MaxEPA fish oil (EPA: 17.0%, DHA 11.9% and DPA (docosapentaenoic acid): 2.7%, expressed as g / l O O g of fish oil) for 7 days. The fish oil was kindly supplied by Seven Seas Ltd, Hull UK. This amount of fish oil corresponds to 15.8 g of n-3 fatty acids per day. All blood samples were drawn after a 12-h fast, immediately prior to starting fish oil (baseline samples) and again after 7 days of fish oil supplementation. Venous blood samples for lipid and NMR analyses were collected in tubes containing sodium EDTA and phenylmethylsulfonylfluoride (PMSF). PMSF was added to prevent enzymatic lipoprotein degradation ( 14). Plasma (ca. 20 ml/subject) was immediately separated by centrifugation at 700 g for 20 min at 4°C.

Lipoprotein fractionation

Lipoprotein fractions were isolated by sequential flotation ultracentrifugation as described by Lindgren et al., with small modifications (15). Basically, the density of whole plasma was adjusted to 1.25 g/ml with KBr and the solution was spun at 150,000 g (Beckman 70.1 Ti rotor, 45,000 rpm) at 15°C for 20 h. The top of each tube was sliced off and the lipoproteins were collected using a Pasteur pipette and the density was increased with solid KBr. The lipoproteins were then loaded into Sorvall03988 polyallomer ultracentrifuge tubes and the

samples were carefully overlaid with a solution of NaCl 0.15 M, EDTA 1 mM. Tubes were spun at 162,000 g (Sorvall TV8 65B rotor, 45,000 rpm) for 1 h at 15’C. Tubes were then allowed to stand at room temperature for 30 min to allow the VLDL fraction to move up the walls of each tube. Each lipoprotein fraction could be seen as a distinct band within the gradient. The top of each tube was sliced off and the various fractions were collected with a Pasteur pipette. The density of each fraction was then sequentially adjusted to isolate VLDL ((1.006 g/ml), LDL (1.006-1.06.3 g/ml) and high density lipoprotein particles (HDL: 1.063-1.21 g/ml). Den- sity was adjusted by adding solid KBr and all centrifugation was at 150,000 g (Beckman 70.1 Ti rotor) for 20 h at 15°C. Each tube was sliced and the top 3 ml was collected. These volumes of each lipoprotein fractions were more than sufficient for all subsequent NMR and GC measurements. After separation, lipoprotein fractions (1-4 ml) were dialyzed for 24 h against NaCl 0.15 M, EDTA 1 mM, CaC12 0.2 mM, NaN3 15 mM. Fractions were stored at 4”C, for a period of less than 12 h, until NMR analysis. Minimal cross contamination and effec- tive separation of the lipoprotein fractions was confirmed by determining apolipoprotein content of each fraction, after delipidation, by polyacrylamide gel electrophoresis. Briefly, the Laemmli SDS polyacrylamide gel electrophoresis method was used (Bio-Rad Labora- tories, Richmond, CA), with a 3-18% resolving gel (16).

Plasma lipids and glucose

Total plasma triglyceride concentration was measured using standard enzymatic test kits (Boehringer, Mannheim, Germany) (17). Total plasma cholesterol, plasma LDL, and plasma HDL cholesterol concentrations were determined enzymatically (18-20). HDL cholesterol was determined using the method described by Lopes-Virella et al. (20) where VLDL and LDL are precipitated by adding phosphotungstic acid and mag- nesium ions to the sample. LDL cholesterol concentration was measured using polyvinyl sulfate precipitation (19). Total plasma eicosapentaenoic acid (EPA; C20:5n-3) and docosahexaenoic acid (DHA; C22:6n-3) concentrations were determined by capillary gas-liquid chromatography (GLC), as described by Sanders,

TABLE 1. Plasma cholesterol and triglyceride concentrations

Subjecu Total Plasma Ratio LDYHDL Triglyceride Cholesterol Cholesterol

“ d / l

Control 0.62 It 0.17 5.21 f 0.69 2.71 f 0.73

Fish oil, 7 days 0.51 f 0.1lu 2.18 f 0.90 4.87 f 0.30

Mean f SEM. “P < 0.03.

Bell et al. Fish oil effect on lipoproteins 1665


ww

w.jlr.org

Dow

nloaded from

http://www.jlr.org/

TABLE 2. Plasma and LDL EPA and DHA concentrations determined bv GLC

Subjects Total Plasma Total LDLa

EPA DHA EPA DHA

Control 1.06 f 0.46 3.63 f 1.18 0.73 f 0.37 2.64 f 0.34

Fish oil, 7 days 13.02 f 2.49 6.52 f 1.44 18.15 f 2.44 5.93 f 0.64

change Significance of P < 0.0001 P < 0.001 P < 0.001 P < 0.001

Values are given as means f SEM. EPA and DHA (n = 9) are expressed as mass % of total fatty acids. "Ultracentrifugally isolated LDL.

Hinds, and Pereira (21) and Chen, Menard, and Cun- nane (22) and expressed as mass 5% of total plasma fatty acids. Plasma glucose levels were measured using standard enzymatic methods (23). All analyses were performed in duplicate.

NMR measurements

l3C and 'H NMR measurements were carried out on a JEOL GSX500 spectrometer operating in quadrature detection mode at 125.7 and 500 MHZ, respectively. All spectra were recorded at constant temperature (37°C).

A

olefdc n

I3C NMR

Protondecoupled 13C NMR of plasma and lipoprotein fractions were obtained essentially as described by Peeling et al. (24), with small modifications. A 30" pulse angle with a 1 sec relaxation delay was used to acquire fully relaxed spectra. Due to the low sensitivity of 13C, all spectra required more than 10,000 scans for good signal to noise ratio, leading to a total scan time of 12 h. This long acquisition time meant that determination of absolute TZ and TI values for the 13C resonances were not feasible. Free-induction-decays (FID) were zero- filled to 64K and an exponential line broadening of 2 Hz was applied before Fourier transformation.

Fig. 1. assignments: a-f, lipids (see Table 3); G, glucose (96.2 ppm); P, protein related peak (40.3 ppm).

13C protondecoupled spectra of plasma: (A) befox; (B) after fish oil supplementation. Peak



ww

w.jlr.org

Dow

nloaded from

http://www.jlr.org/

'H NMR

A sample (0.5 ml) of intact plasma or isolated lipoprotein fractions, VLDL, LDL and HDL, was placed in a 5-mm diameter NMR tube (Fluorochem Limited, UK) and 0.05 ml of D20 was added as a field-frequency lock. Samples were allowed to equilibrate at 37°C in the spectrometer for 10 min prior to acquisition. The in- tense water peak and the broad protein resonances were suppressed by the application of a gated secondary irradiation at the water resonance. The Hahn spinecho (90-z-180-z-collect) spectra of plasma were obtained with a z = 60 ms and a repetition time of 10 sec (25,26). Single pulse spectra were obtained with a pulse angle of 35" and a total relaxation delay of 5 sec to allow full relaxa-

i

C I

b

m 1 " " I " " I " " I " " I " " I " " I " " I " "

15.0 14.5 14.0 13.5 13.0 12.5 12.0 11.5

h

tion. Transverse (spin-spin) relaxation times ( 7 ' 2 ) of the methylene (-(CH&-) and methyl (CH3-) signals of VLDL and LDL were obtained by the use of CPMG spin-echo sequence with a 1 ms delay and increasing from 2 to 600 ms (7). The FIDs were zero-filled to 32K prior to Fourier transform. Plasma spectra were cross- referenced to the methyl group resonances of lactate at 1.33 ppm and N-acetyl resonances at 2.02 ppm for chemical shift.

Data analysis

1H NMR spectra of whole plasma and isolated lipoproteins were analyzed using standard JEOLspectrome- ter software, allowing direct measurements of chemical

b

m 1 " ~ ' 1 " ' ' I ' ~ ' ~ 1 ~ ~ " I ' ' " I ' ' " I ' ' " I ' ' " I

15.0 14.5 14.0 13.5 13.0 12.5 12.0 11.5 11.0

h

D

Fig. 2. assignments see Table 3.

13C protondecoupled spectra of plasma (A,C) before; (B,D) after fish oil supplementation. Peak



ww

w.jlr.org

Dow

nloaded from

http://www.jlr.org/

TABLE 3. Assignment, chemical shift, and peak height ratios of the spectra of intact plasma before and after 7 davs fish oil suDDlementation

Resonances

(a) Cholesterol C18

(b) CHs-CHn-CHp-

(b’) CHs-CH2-CH = CH- (n-3)

( c ) CHs-CH2-CHz-

(d) CH = CH-CHZCH = CH

(e) Fatty acyl -(CHZ)n-

(f) Choline -N+ (CHs)s

(9) CHs-CHZ-CH = CH-CHZ (n-3)

(vi) Polyunsaturation Index

(i) CHy-CH2-CH = CH- (n-3)

Chemical Shift Peak Ratio Significance Fish oil/Control of Change

P P l

11.9 1.20 f 0.31

14.1 1.16

14.3 U P < 0.01

22.6 0.91 & 0.04

25.6 1.67 k 0.31 P < 0.02

29.5 0.72 f 0.17 P < 0.01

54.1 1.14

127.1 U P < 0.01

127.7/129.9 1.74 f 0.46 P < 0.01

131.6 U P 0.01

Values are given as means f SEM. “Unique resonances from n-3 fatty acids (b’, g, and j) were present only in spectra of plasma samples

taken after fish oil supplementation. Therefore, no fish oil/control ratios are available

shift, peak areas, and T2 measurements. 13C resonances were measured after computer-based baseline correc- tion of the areas of interest. Spectra were internally scaled relative to the glucose resonance at 96.2 ppm and the protein-related peak at 40.3 ppm (24), both of which were unaffected by fish oil supplementation. Peak areas of the resonances in the baseline plasma spectrum were compared to those obtained from the plasma spectra after 7 days of fish oil supplementation to give the peak height ratios for each subject. Each subject served as his own control.

Carbon and proton NMR signals were assigned on the bases of chemical shifts from previously published data and chemical shifts obtained from spectra of isolated lipoproteins and pure lipids (7, 8, 10,24-27).

Statistical analysis

All values presented here are expressed as mean k SEM. Statistical analysis was performed by Student’s paired t-test and changes were considered significant when P < 0.05.

RESULTS

Changes in plasma lipid concentrations after 7 days fish oil supplementation are shown in Table 1. Plasma triglycerides concentrations decreased by 18% (P < 0.03), while total plasma cholesterol concentration and the LDL/HDL cholesterol ratio showed no significant changes. Furthermore, the proportions of EPA and DHA in total plasma were raised ca. 10-fold (P < 0.001) and 2-fold (P < 0.001) respectively (Table 2). Similarly, isolated LDL particles showed marked increases in rela-

tive levels of EPA (20-fold; P < 0.001) and DHA (2-fold; P < 0.01). No changes in plasma glucose concentration were observed (pre: 4.92 k 0.60 vs. post: 5.21 k 0.45 mmol/l).

Typical 13C NMR spectra of intact plasma and isolated LDL before and after fish oil supplementation are shown in Fig. 1 and Fig. 2. All the spectra were acquired at constant temperature (37°C). The spectrum from control plasma consists of resonances from low molecular weight metabolites, including glucose, glutamine, lactate, alanine, and lipids from lipoprotein particles superimposed on a broad protein envelope (24). Using 13C NMR, distinctive chemical groups can be distin- guished by their chemical shift, including specific sites in the fatty acid chains and cholesterol (Table 3) (10,24). In particular, as shown in Fig. lA, resonance f (54.1 ppm) has been assigned to the choline groups in phospholipids (-N+(CH3)3) from HDL and LDL particles; resonance e (29.5 ppm) to the fatty acyl methylene (-(G&)n-3) carbons; resonance b (14.1 ppm) to the terminal fatty acid methyl (GHs-) carbon; resonance a (11.9 ppm) to the G-18 carbon of cholesterol. Two resonances dominate the olefinic region of the spectrum (Fig. 2C) and can be assigned to monounsaturated fatty acids (i, 129.7 ppm: -CHz-GH = GH-CHp) and polyunsaturated fatty acids (i, 129.7 ppm:-GH =

CH-CH2-CH = CH- and h, 127.4 ppm:-CH =

GH-CH2-CH = CH). The extent of polyunsaturation in the plasma fatty acids can be further assessed by a sharp resonance at 25.6 ppm (d), arising from -CH =

CH-CHz-CH = CH- carbons. After fish oil supplementation, as shown in Figs. lB, 2B, and 2D, three new resonances were detected in the plasma spectra from all volunteers: at 14.3 ppm (b’), 121.1 ppm (g), and 131.6 ppm 0). These resonances have been assigned, based on



ww

w.jlr.org

Dow

nloaded from

http://www.jlr.org/

8-

-8

-B

-8

-8

-5

Y -E

m -5

a

64

71 9 - M k c .-



ww

w.jlr.org

Dow

nloaded from

http://www.jlr.org/

* B

Fig. 4. 500 MHZ spinecho lH NMR spectra of plasma: (A) base line plasma; (B) plasma from the same subject after 7 days fish oil supplementation. Assignments: Peaks 1 + 2 and 3 to HDL + LDL and VLDL terminal CH3- lipid group, respectively; peaks 5 and 6 to -(CH2)n- resonances of LDL and VLDL, respectively (10); Val. valine CH3-; Lac, lactate CH3-; Ala, alanine CH3-. The resonance at 0.941 ppm (*) arises from n-3 fatty acids in lipoprotein particles.

3

U

Ala U J/ Val Lac

chemical shift and literature values, to L;Hs-CH~-CH =

CH- (b': 14.3 ppm), CHs-CHz-CH = CH- (g: 127.1 ppm), and CHs-CHyCH = CH- (i: 131.6 ppm) of n-3 fatty acids, respectively (27). Fish oil also led to significant changes in the intensity of other lipid resonances (Table 3). Peak e (-(CH*)n-) at 29.5ppm was consistently reduced (ca. 15%; P < O.Ol), while resonance d (25.6 ppm), arising from -CH = CH-CSI2-CH = CH- carbons, was markedly increased (ca. 72%, P < 0.01). Further- more, the polyunsaturation index, defined as the peak height ratio of signals h/i (127.'7/129.7 ppm), was significantly increased (ca. 83%; P < 0.01) in all plasma samples after fish oil supplementation. No significant changes in the relative intensities of the resonances arising from cholesterol G-18 (a: 11.9 ppm), fatty acid terminal methyls (b: 14.1 ppm), and choline -N' (CH3)3 (f: 54.2 ppm) were observed. The 13C NMR spectra of isolated LDL is shown in Fig. 3. The spectra consist principally of signals from lipids, including triglycerides, phospholipids, cholesterol, and cholesreryl esters ( 10). As with intact plasma, supplementation of fish oil led to an increase in the polyunsaturation index and the ap- pearance of resonances (b', g, and j) not previously observed. These peaks were assigned, based on their chemical shift and lipid extract analysis, to specific chemical groups of n-3 fatty acids. Similar resonances were observed in the spectra of isolated VLDL and HDL particles (results not shown).

The resonances at 14.3 and 14.1 ppm can be used to determine, by I3C NMR, the relative plasma levels of EPA + DHA after fish oil ingestion. The results were compared to those obtained by GLC from the same samples. Good agreement between the two methods was observed (GLC: 19.8 ? 3.5% vs. NMR: 17.6 k 1.77%). Similar agreement was observed for the LDL levels of EPA + DHA as measured by GLC and NMR (GLC: 24.1 f 4.2% vs. NMR. 20.0 +- 1.6%).

'H spectra of blood plasma have been extensively reported in the literature (7, 8). Resonances from lipoproteins arise mainly from their lipid components (fatty acids from triglycerides, cholesteryl esters, and phospholipids) as contributions from apolipoproteins are too broad to be observed under standard 'H NMR conditions (10). The signals arising from different lipoprotein fractions in the 'H NMR spectra of plasma are assigned in Fig. 4A: peaks 1 (0.858 ppm) and 2 (0.863 ppm) which overlap, assigned to the C&- groups of HDL and LDL respectively; peak 3 (0.886 ppm) to the C&- groups of VLDL; peaks 5 (1.25 ppm), and peak 6 (1.279 ppm) to -(C&)n- groups of LDL and VLDL, respectively (1 1,25). After fish oil supplementation, the plasma spectra were significantly altered, Fig. 4B. Reso- nances from VLDL (peaks 3 and 6) are completely absent (>99% reduction) and a new lipid resonance, centred at 0.941 ppm (*) and of similar intensity as peak 2, can be observed. The spin-echo 'H spectra of isolated

1670 Journal of Lipid Research Volume 37,1996


ww

w.jlr.org

Dow

nloaded from

http://www.jlr.org/

LDL and VLDL, before fish oil, are shown in Fig. 5A and Fig. 6B, respectively. The methyl and methylene resonances shown in these spectra arise from non-n-3 fatty acids (27, 28). After fish oil supplementation, the spectra from both lipoprotein fractions show the presence of a new lipid resonance centre at 0.941 ppm. A similar resonance was detected in the 'H NMR spectra of isolated HDL (results not shown). The new peak (*) has been assigned, on basis of its chemical shift and the 13C NMR findings, to the methyl group of EPA and DHA within the lipoprotein particles.

The effects of fish oil supplementation on the transverse relaxation (T2) of the LDL and VLDL methyl, methylene and (*) lipid signals are shown in Table 4. Spin-echo spectra were recorded at constant temperature (37°C) to minimize TZ changes known to arise with temperature variation. The T2 of the methyl (CfIs-) signals from LDL (1 12 -t 28 ms vs. 162 f 37 ms; P < 0.05) and VLDL (182 k 24 ms vs. 268 f 19 ms; P < 0.01) were significantly elevated after fish oil supplementation. Similar changes, although of lesser magnitude, were observed for the Tz of the methylene (-(C&)"-) signals from LDL (52 f 12 ms vs. 78 f 21 ms; P < 0.02) and VLDL (88f 17msvs. 106k24ms;P<O.Ol).The Tzofthelipid resonance (*) arising from EPA and DHA were relatively longer and for both VLDL (326 f 42 ms) and LDL (227 f 33 ms) compared to the other methyl lipid signals within the same particles.

I

DISCUSSION

.(CHI ,". .CHI A , I

In the present study we have used high resolution NMR to examine the effects of fish oil supplementation on the plasma lipoprotein molecular architecture. The NMR properties of lipoprotein particles in isolation and in intact plasma have been extensively analyzed (10-13, 29). Each lipoprotein fraction (chylomicrons, VLDL, LDL, HDL) has its own spectral characteristics including chemical shift, lipid signal intensities, longitudinal (TI) and transverse (Tz) relaxation times (7, 8, 10-13, 29). These features allow the identification of different lipoprotein fractions in the lH NMR spectra of intact plasma or serum (11-13, 29). Furthermore, the 'H and *3C NMR properties of the lipoprotein lipid signals can shed light on the physico-chemical state of lipoprotein particles, including chemical composition, structural organization and molecular motion (7, 10, 11, 24-26,29).

The fish oil supplement used in this study (MaxEPA) consists mainly of triglycerides, in which 28.9% (mass) are represented by EPA and DHA. These n-3 fatty acids are thought to be the active components of fish oil and are consistently incorporated into plasma lipids (1, 29-31). In this study, and as previously shown by others, fish oil supplementation led to a significant decrease in plasma triglyceride concentration and an increase in the concentration of EPA and DHA. No changes in choles-

Fig. 5. 500 MHZ spin-echo lH NMR spectra of isolated VLDL (A) before, (B) after fish oil supplementation. The new lipid resonance at 0.941 ppm (*) has been assigned to methyl proton (Cb-CHz-CH = CH-) from EPA and DHA. The resonance at 1.284 ppm has been identified as a contaminant arising from the lipoprotein isolation pro- cedure.



ww

w.jlr.org

Dow

nloaded from

http://www.jlr.org/

- CA,

Fig. 6. 500 MHZ spin-echo 'H NMR spectra of isolated LDL (A) before, (B) after fish oil supplementation. The new lipid resonance at 0.941 ppm (*) has been assigned to methyl proton (W-CHz-CH =

CH-) from EPA and DHA.

A

R*( I " " I " " I " " I " " I " " I " " I " " I " " I " " I " " I

1.6 1.5 1.4 1.3 1.2 1.1 1.0 0.9 0.e 0.7 0.6

terol levels were observed, even at this relatively high concentration of fish oil.

The lipid changes induced by fish oil had noticeable effects on the NMR profile of intact plasma and isolated lipoproteins. In the absence of fish oil supplementation, the 'H and I3C lipid resonances arise principally from the fatty acids of triglycerides and phospholipids including linoleic, oleic, and palmitic acids (10, 1 1, 24). After fish oil supplementation, new 13C (14.3,127.1, and 131.6 ppm) and 'H (0.941 ppm) resonances were observed in the spectra of intact plasma and isolated lipoproteins. Gunstone (27) has shown that a number of chemical

groups from EPA and DHA have very distinctive chemical shifts and can be differentiated from those of other fatty acids by NMR. The new I3C and 'H resonances have been assigned, based on their chemical shift and NMR characteristics, to specific carbon groups of the EPA and DHA moiety in the lipoprotein particles. Furthermore, we have shown that these specific resonances can be utilized to determine the relative levels of these fatty acids in intact plasma and isolated lipoprotein fractions.

The disappearance of the VLDL resonances from the 'H NMR spectra of intact plasma after fish oil supplementation is consistent with the decrease in triglyceride

1672 Journal of Lipid Research Volume 37,1996


ww

w.jlr.org

Dow

nloaded from

http://www.jlr.org/

TABLE 4. Effects of fish oil on the transverse relaxation (Tz) of niques. Another more plausible explanation is the existence of "microenvironments" rich in EPA and DHA within the lipoprotein particles. The lipid constituents

the methylene (-(CHs).-) and methyl (CHs-) resonances of isolated plasma lipoproteins

-(CH2)n- CHs- Peak(*)

ms

VLDL

Pre-FO, 88f 17 182 f 24

Post-FO 106 f 24 268 f 19 326 f 42

LDL

Pre-FO 52f6 112 f 28 Post-FO 78 f 21 162 f 37 227 f 33

Mean f SEM (ms). Pre-FO. pre-fish oil supplementation; Post-FO. post-fish oil supplementation.

concentration (1, 30-32). Moreover, the fact that the methylene (-(CH&-) resonance from VLDL particles is absent in the 'H NMR spectra of intact plasma after fish supplementation suggests that the n-3 resonance observed in the spectra of intact plasma samples arises predominantly from LDL particles.

The transverse relaxation, chemical shift, and intensity of the signals of plasma lipoproteins are determined by a number of factors including composition and packing. Furthermore, these NMR parameters can shed light on the structural organization of lipoprotein particles (10). In this study, the Tz relaxation times of the VLDL and LDL lipoprotein methyl and methylene resonances from non-n-3 fatty acids were significantly increased after fish oil supplementation. This implies an increase in relative mobility of the acyl chain within the lipoprotein particles. The incorporation of n-3 fatty acids, therefore, appears to disrupt the hydrophobic domains of VLDL and LDL, resulting in alteration in the packing and increased isotropic mobility of the lipid components within lipoprotein particles (10,ll). This supports previous findings suggesting alteration in core fluidity being induced by increased amounts of n-3 polyunsaturated fatty acids in LDL (5, 6).

Interestingly, the intensity of the new peak at 0.941 ppm, in the 'H spinecho spectra of plasma and isolated lipoproteins, was similar to those from non-n-3 fatty acids, although the relative level of EPA and DHA was (20% of the total lipid concentration. Furthermore, its T2 was notably longer than those from non-n-3 lipids. These findings are surprising as one would expect the T2 of all lipid components within a given type of lipoprotein to be similar. There are several possible expla- nations for these findings, including the existence of a subfraction of LDL and VLDL particles that contains most of the n-3 fatty acid components detected by NMR and GLC. This would give rise to lipoprotein particles with NMR characteristics distinct from normal LDL and VLDL. N o such lipoprotein subfraction has been observed during separation by ultracentrifugation tech-

within these microenvironments will thus give rise to 'H signals with NMR relaxation characteristics very different from those of other lipids within the same lipoprotein particle. Furthermore, as there appears to be an overall alteration in the architecture of the lipoproteins, reflected in this study as changes in lipid mobility, the n-3 microenvironments must be distributed throughout the lipoprotein particles.

The existence of lipid microenvironments has been previously postulated from work carried out with phospholipid vesicles and intact cells (33-35). Gordon et al. (33) used electron spin resonance (ESR) spectroscopy to measure membrane fluidity in platelet and hepato- cytes and suggested the existence of cholesterol-rich and cholesterol-poor lipid domains in the membranes. Simi- larly, the use of the fluorescent probe merocyanine (MC540) suggests that the incorporation of DHA into cell membranes may produce DHA-rich, cholesterol- poor domains (34, 35). This would, in turn, lead to a variable distribution of plasma membrane proteins and possible local variations in biochemical activity. Our results appear to suggest that similar n-3-rich domains may be present in plasma lipoprotein particles. This is the first time that such a phenomenon has been observed in plasma lipoproteins. The possible effects of these microenvironments on the function and uptake of plasma lipoproteins are unclear and warrant further investigation.

In conclusion, our results show for the first time that fish oil supplementation leads to significant alterations in plasma lipoprotein architecture. These alterations arise from changes in the lipid composition and physical characteristics of the lipoprotein particles, particularly the possible existence of n-3-rich micro-domains. Our findings suggest that multinuclear NMR may be a pow- erful tool in the study of the effects of n-3 fatty acids on intact lipoprotein partic1es.l

The authors wish to thank the Medical Research Council, Ministry of Agriculture, Fisheries and Food (MAFF) and the Natural Sciences & Engineering Research Council (Canada) for financial support. The University of London Intercolle- giate Research Service for NMR facilities at Birkbeck College (JEOL 500) and Dr. T. Sanders, M. A. Ryan, and G. Frost for their invaluable help. Manuscript received 7 December 1995 and in revisedfwm 14 May 1996.

REFERENCES

1. Harris, W. S. 1989. Fish oils and plasma lipid and lipoprotein metabolism in humans: a critical review.]. Lipid Res. 3 0 785-807,



ww

w.jlr.org

Dow

nloaded from

http://www.jlr.org/

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

Nestel, P. J. 1990. Effects of n-3 fatty acids on lipid metabolism. Annu. Rev. Nutr. 10: 149-167. Sanders, T. A. B., D. R. Sullivan, J. Reeve, and G. R. Thompson. 1985. Triglyceride-lowering effect of marine polyunsaturates in patients with hypertriglyceridemia. Ar- teriosclerosis. 5: 459-465. Harris, W. S., W. E. Connor, D. R. Illingworth, D. W. Rothrock, and D. M. Foster. 1990. Effects of fish oil on VLDL triglyceride kinetic in humans. J. Lipid Res. 31:

Parks, J. S., and L. L. Rudel. 1990. Effects of fish oil on atherosclerosis and lipoprotein metabolism.

Nenseter, M. S., A. C. Rustan, S. Lund-Katz, E. Soyland, G. Maelandsmo, M. C. Phillips, and C. A. Drevon. 1991. Effect of dietary supplementation with n-3 polyunsaturated fatty acids on physical properties and metabolism of low density lipoprotein in humans. Atheroscler. Thromb.

Bell, J. D., J. C. C. Brown, and P. J. Sadler. 1989. NMR studies of body fluids. NMR Biomed. 2: 246-256. Nicholson, J. K., and I. D. Wilson. 1989. High resolution 'H NMRspectroscopy of biological fluids. Prog. NMR Spec.

Petroff, 0. A. C. 1988. Biological 'H NMR spectroscopy. Comp. Biochem. Physiol. 903: 249-260. Hamilton, J. A., and J. D. Morrisett. 1986. Nuclear magnetic resonance studies of lipoproteins. Methods Enzymol.

Bell, J. D., P. J. Sadler, A. F. Macleod, P. R. Turner, and A. La Ville. 1987. 'H NMR studies of human blood plasma. FEBS Lett. 219: 239-243. Otvos, J. D., E. J. Jeyaraja, and D. W. Bennett. 1991. Quantification of plasma lipoproteins by proton nuclear magnetic resonance spectroscopy. Clin. Chem. 37:

Hiltunen, Y., M. Ala-Korpela, J. Jokisaari, S. Eskelinen, K. Kivinitty, M. Savolainen, and Y. A. Kesaniemi. 1991. A lineshape fitting model for 'H NMR spectra of human blood plasma. M a p . Reson. Med. 21: 222-232. Cardin, A. D., K. R. Witt, J. Chao, H. S. Margolius, V. H. Donaldson, and R. L. Jackson. 1984. Degradation of apoprotein B-100 of human plasma. J. Biol. Chem. 259:

Lindgren, F. T., L. C. Jensen, R. D. Wills, and N. K. Freeman. 1969. Flotation rates, molecular weights and hydrated densities of the low-density lipoproteins. Lipids. 4: 337-44. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 227: 680-685. Bucolo, G., and H. David. 1973. Quantitative determination of serum triglycerides by the use of enzymes. Clin. Chem. 19: 476-482. Siedel, J., E. 0. Hagele, J. Ziegenhorn, and A. W. Wahle- feld. 1983. Reagent for the enzymatic determination of serum total cholesterol with improved lipolytic efficiency. Clin. Chem. 29: 1075-1080. Assman, G. 1979. Lipiddiagnostik heute. In Lipoproteine und Herzinfarkt H. Gneten, editor. Witzstrock-Verlag, Baden-Baden/Germany. 29 ff. Lopes-Virella, M. F., P. Stone, S. Ellis, and J. A. Colwell. 1977. Cholesterol determination in high-density lipopro-

1549-1558.

Ath~osclerOSis. 84: 83-94.

12: 369-379.

21: 449-501.

128A 472-515.

377-386.

8522-8528.

teins separated by three different methods. Clin. Chem.

21. Sanders, T. A. B., A. Hinds, and C. C. Pereira. 1989. Influence of n-3 fatty acids on blood lipids in normal subjects. J. Int. Med. 225 Suppl. 1: 99-104.

22. Chen, Z. Y., C. R. Menard, and S. C. Cunnane. 1995. Moderate selective depletion of linoleate and a-linolenate in weight-cycled rats. Am. J. Physiol. 268: R498-505.

23. Werner, W., H. G. Reg, and H. Wielinger. 1970. Uber die eigenschaften eins neven cromogens fur die blutzucker- bestimmung nach der POD/GOD methods. Z. Analyt. Chem. 252: 224-228.

24. Peeling, J., S. Garnette, K. Marat, E. Tomchuk, and E. Bock. 1988. 'H and 13C nuclear magnetic resonance studies of plasma from patients with primary intracranial neoplasms. J. Neurosurg. 68: 931-937.

25. Nicholson, J. R., N. J. Buckingham, and P. J. Sadler. 1983. High resolution 'H NMR studies of vertebrate blood and p1asma.J. Biochem. 211: 605-615.

26. Bell, J. D., J. C. C. Brown, R. E. Norman, P. J. Sadler, and D. R. Newell. 1988. Factors affecting 'H NMR spectra of blood plasma: cancer, diet and freezing. NMR Biomed. 1:

27. Gunstone, F. D. 1991. High resolution NMR studies of fish oils. Chem. Phys. Lipids. 5 8 83-89.

28. Arus, C., W. M. Westler, M. Barany, and J. L. Markley. 1986. Observation of the terminal methyl group in fatty acids of the linolenic series by a new 'H NMR pulse sequence providing spectral editing and solvent suppres- sion. Application to excised frog muscle and rat brain. Biochemistry. 25: 3346-3351.

29. Lounila, J., M. Ala-Korpela, J. Jokisaari, M. J. Savolainen, and Y. A. Kesaniemi. 1994. Effects of orientational order and particle size on the NMR line positions of lipoproteins. Phys. Rev. Lett. 72: 4049-4052.

30. Bronsgeest-Schoute, H. C., C. M. van Gent, J. B. Luten, and A. Ruiter. 1981. The effects of various intake of n-3 fatty acids on the blood lipid composition in healthy subjects. Am. J. Clin. Nutr. 34: 1752-1757.

31. Harris, W. S., W. E. Connor, and M. P. McMurray. 1983. The comparative reductions in the plasma lipids and lipoproteins by dietary polyunsaturated fatty acids: salmon oil versus vegetable oils. Metabolism. 32: 179-184.

32. Huff, M. W., and D. E. Telford. 1989. Dietary fish oil increases conversion of very low density lipoprotein apoprotein B to low density lipoprotein. Atherosclerosis. 9

33. Gordon, L. M., P. W. Mobley, J. A. Esgate, G. Hofman, A. D. Whetton, and M. D. Houslay. 1983. Thermotropic lipid phase separations in human platelet and rat liver plasma membrane.J. Membr. Biol. 76: 139-149.

34. Stillwell, W., S. R. Wassall, A. C. Dumaual, W. D. Ehringer, C. W. Browning, and L. J. Jenski. 1993. Use of merocyanine (MC540) in quantifying lipid domains and packing in phospholipid vesicles and tumor cells. Biochim. Biophys.

35. Zerouga, M., L. J , Jenski, and W. Stillwell. 1995. Compari- son of phosphatidylcholine containing one or two docosahexaenoic acyl chains on properties of phospholipid monolayers and bilayers. Biochim. Bwphys. Acta. 1236:

23: 882-886.

90-94.

58-66.

Acta. 1146 136-144.

266-272.



ww

w.jlr.org

Dow

nloaded from

http://www.jlr.org/

Effects of n-3 fatty acids on the NMR profile of plasma lipoproteins

Documents

Transcript of Effects of n-3 fatty acids on the NMR profile of plasma lipoproteins