Purchased by U. S. Dept. of Agric. for Official Use Only. 6468 l\1etabolism ill vivo .... acyltransferase for 18:2.w6-d.j' The TAG plot suggests that. 18: 3. w3-d2 was ...
Biochemical Society Transactions Volume 18 1990
6468
Purchased by U. S. Dept. of Agric. for Official Use Only
l\1etabolism ill vivo of deuterium-labelled linolenic and linoleic acids in humans E. A. EMKEN,* R. O. ADLOF,* H. RAKOFP' W. K. ROHWEDDER* and R. M. GULLEYt * Vegetable Oil Research, Northern Regional Research Center, Agricillwral Research Service, USDA, 1815 N. University Street, Peoria, lL 61604, U.S.A. and t St. Francis Medical Center, Peoria, lL 61604, U.s.A. Abbreviations used: pc, phosphatidyJcholine; PC-2. 2-acy)phosphatidyJcholine; TAG. triacylglycerol; CEo cholesterol ester.
Introduction
Over 60 years ago, linoleic acid (18: 2, (6) was recognized as an essential fatty acid with important physiological functions. Many earlier studies failed to identify a physiological role for linolenic acid (18:3,w3) and it was 40 years later before 18: 3, w3 was accepted as an essential fatty acid. The observations by Sinclair [1] deserve credit for biochemists becoming aware of the importance of w3 fatty acids. However, the metabolism and biochcmistry of w3 fatty acids have 1990
767
FATTY ACID i'vlETABOLISM not been as intensivelv investigated as those of the w6 fatty acids. ,~ Results from a variety of studies indicate that w3 and w6 fatty acids have separate physiological roles [1-3]. Also, there is considerable evidence indicating that the fattv acids in the w3 and w6 families are competitors for th~ same enzvmes and that their eicosanoid metabolites are antipathetic [1-4]. However. biochemists still do not understand the metabolism of linoleic acid well enough to develop a biochemical explanation for the physiological effects produced bv linolenic acid. . Our development of methods for the preparation of deuterium-labelled linoleic and linolenic acids provided the opportunity to investigate the concurrent metabolism of both of these essential fatty acids in man. The direct comparison of metabolic events in the same subject at the same time reduces some of the problems normally encountered in interpretation of data obtained from separate subjects. In addition. other deuterium-labelled fatty acids ( 16: 0, IS: 0, IS: I) can be included in the experimental design in order to provide a more complete comparison of interactions between specific fatty acids. In this brief paper. absorption. incorporation. desaturation and elongation of deuterium-labelled polyunsaturated. monounsaturated. and saturated fatty acids are compared in subjects prefed diets containing two different levels of linoleic acid. Experimellwl Four healthy. young-adult male subjects (two in each group) were fed a low-fat (24% fat calories) diet containing either 1.6 or 6.6% by energy (7% or 2S% of total fat) linoleic acid. The polyunsaturate to saturate (·P/S') ratio was 0.13 or O.S. After 12 days. a mixture of triacylglycerols containing deuterium-labelled fattv acids was blended with caseinate (30 g), sucrose (1S g). b-glucose (30 g) and water (1S0 ml), and administered as a single oral dose. The triacvlglycerol mixture contained 3.0-3.5 g each of the follo\~i;}g' fatty acids: cis-9.cis-l2.cis-lS-[ 15.IS-zHzJoctadecatrienoi~ acid ( IS :3. w3-d z I. cis-9. cis-I 2-[ 15 .15 .16.16-ZH~]octadecadi-
enoic acid (IS:2.w6-d~). cis-9.[14.14.IS.IS.17,IS- 2H(']octadecanoic acid (18: L w9-d(,), [9,9.1 0.1 O-zH~Joctadecanoic acid (IS:O-d~). [9,10- zH 2Jhexadecanoic acid (16:0-d zJ. These deuterium-labelled fatty acids were synthesized by procedures similar to those previously published [S-S]. A total of nine blood samples were drawn from each subject over a 48 h period. Chylomicron fractions were isolated from a portion of the plasma samples by ultracentrifugation. The plasma and chylomicron lipids were extracted and separated by l.l.c. Portions of the phosphatidylcholine (PC) isolated from the plasma samples were treated with phospholipase A 2 and the I-acyl- and 2-acyl-phosphatidylcholine (PC-2) fractions were isolated by procedures used previously [9]. An internal standard (17 :0) was added to all fractions. Methyl esters were prepared and analysed by g.c.-m.s. using chemical ionization conditions with isobutane ~s the reagentgas and selective-ion monitoring [9]. Results Chylomicron triacylglycerol (TAG) data was used to follow and compare the absorption of the deuterium-labelled fatty acids. The maximum incorporation of the deuteriumlabelled fatty acids in the chylomicron TAG samples occurred 4-6 h after the mixtures of deuterium-labelled TAG were fed. The ranges for maximum incorporation (in mg/ml of plasma) of the various fatty acids were: IS- 2S. IS:3.w3-d z: 16-21, lS:2-d 4 : 20-29. lS:1-d,,: 7-24.18:0d 4 : 14-2S. 16:0-d z. The per cent absorption was calculated by assuming 9S% absorption for 18: I-d" and correcting for differences in the amount and isotopic purity of the deuterium-labelled fats in the fed mixtures. The ranges of per cent absorption of the deuterium-labelled fats were: 64-SS%, IS:3.w3-d 2: 79-S2%, 18:2,w6-d~: 27-94%. IS:0-d 4 : 76-79%. 16:0-d 2. Plasma lipid data were used to compare the relative selectivitv of the acvltransferase enzymes for the deuteriumlabe'lled fatty ;cids. The area selectivity values are summarized in Table 1. They were calculated by dividing the deuterium-labelled fatty acid/ 18: I-d" ratio in the plasma
Table 1. Incolporlllion of deuterilllll-labelled fatty acids. relative plasma lipids
fO
18: I-do> into human
The SAT diet contained 1.6% (by energy) linoleic acid and the PUFA diet contained 6.6% (by energy) linoleic acid. Plasma lipid TAG*
CE
Subject
Diet
I 4 2 3
PUFA PUFA SAT SAT
I
PUFA PUFA SAT SAT
4 2 3 PC
PC-2
4 2 3
PUFA PUFA SAT SAT
I 4 2 3
PUFA PUFA SAT SAT
I
16:01
-, 1.68 1.04 1.14 -1.01 4.61 3.01
> -50 > -50
18:2.w6/ 18:1
18:3.w3/ 18:1
18:2.w6/ 18:3.w3
-5.04 -1.29 -3.76 -1.61
- 1.17 -1.24 -1.05 - 1.13
- 1.29 -2.05 - !.II -1.62
1.32 1.64 1.06 1.42
-10.27 - 2.08 -4.79 - 1.91
5.35 4.36 4.91 5.25
-1.44 -1.81 - 1.42 -1.51
7.71 7.83 6.96 7.85
1.54 5.27 2.89 4.71
5.81 8.20 3.93 8.21
-2.47 -2.15 -4.85 -2.32
14.37 17.46 10.87 18.89
- 5.46 - 3.53 -16.34 -6.17
11.24 10.49 11.26 11.21
1.19 -3.10 -1.30 - 1.98
23.14 32.23 14.64 22.01
18:01
*TAG selectivity values reflect absorption. Values must be different from TAG values to indicate difference in incorporation. t 16 :O-d, was not included in mixture fed to subjects I and 2. Vol. 18
768
BIOCHErvllCAL SOCIETY TRANSi\CTIOr-:.)
lipid fraction. obtained from the areas under the time course plots (Fig. I i. by the same ratio that was in the mixture of deuterium-labelled fatty acids fed. These calculations produce a baseline selectivity value of 1.0 for 18: I-d". If the area selectivity value was less than 1.0. then the inverse value was used and it was given a negative sign. Thus. negative values indicate lower in~corporati;n or di;crimination ~and positive values inclicate higher incorporation or selective acylation relative to 18: I-cl". The area selectivitv values (Table 1) inclicate that 18: 2. w6 is the preferred substrate for both lecithin: cholesterol acyltransferase and phosphatidylcholine acyltransferase. The selectivity values for TAG and free fatty acids are influenced
bv the removal of the deuterium-labelled fattv acids by o~idation and their transfer to other tissue lipid~. The time course plots in Fig. I for cholesterol ester iCE). TAG. PC and PC-2 illustrate graphically the high selectivity of lecithin: cholesterol acyltransferase and phosphatidylcholine acyltransferase for 18:2.w6-d.j' The TAG plot suggests that 18: 3. w3-d 2 was removed slightly more rapidly relative to 18: 2. w6-d.j' This observation agrees with data from parenchymal rat liver cells indicating that 18: 3. w3 is more rapidly oxidized than 18:2.w6[1O]. Isotope enrichment data are summarized in Table 2 for 18: 3, w3-d 2 and 18: 2, w6-d.j and their desaturated and elongated products. These data show a high isotopic enrichment
Table 2. .\1aximlim per cellI isolOpe cnrichlllclIlS ill plasma lipid w3 alld w6 filiI)' acids The SAT diet contained 1.6% (bv energvi linoleic acid and the PUFA diet contained 6.6% (by energy! linoleic acid. 18:4.o;3-d> !·S:3.w6-d.j. 22:4.w6-cl.j and 22:5.w6-d.j were not detected. Maximum isotope enrichment i%) Plasma lipid ... Timerhi ... Diet.,.
TAG
CE
PC
12 SAT
12 PUFA
24 SAT
24 PUFA
12 SAT
12 PUFA
39.4
35.0 0.0 0.0 8.9 0.0 4.5 2.7 Cl.Cl 0.0
37.1 Cl.O Cl.Cl 4.9 0.0 0.0
62.8 0.0 0.0 19.0 7.0 1.0 7.6 Cl.O 0.3
55.9 0.0
Fallyacid
18:2.w6-d.j
61.0 42.9 45.8 38.3 2Cl.6 7.Cl 3.8
20:3.w6-d.
1.7
2Cl:4.w6-d.j
Cl.2
18:3.w3-d. 20:3.w3-d: 20:4.w3-d:
2Cl:5.w3-d:
22 :5.lU3-d~ 22:6.w3-d;
33.3
37.5 10.5 72.2 17.1 4.1 4.0 0.4
~
,
"'..) Cl.O 0.0
Cl.D
11.1 8.6 4.1 5.6 0.0 0.4
*Time at which maximum incorporation of deuterium-labelled 18:3.w3 and 18:2.w6 occurred.
--------------;1::
::'(-a-)-/.-,
j( b ) /
)
18: 2,w6-d•./
I
i
,/
15'
.-/ I ./ 18: 1·d 6 10jl/,/ 18:3,w3'd 2 i. .' 16.0-d 2 • 5 / 18:0-d 2 I r
._~ ?-':::,/ ,.-:-_'-'~'-::::I
I
I~
f-------.;:..----"'~2""4'-""i480 0
r-l;:; 1.5
,
'-
/ .' ,
/' 0.00
2
18:0-d 2 ...- -
,,:
/ - 16:0-d
..). ...-
6
_ 2
1
'-.
01
2.
,i
,
1 1 .51
\.
\
·,ll.0j
..
'-'_.-.,...::.
0.5'/
\',
18:3,w3-d 2 16:0-d 2 / 18:1-d 6 !18:0-d 2
0
2
I j 1
, _ . ..:....,t._. "i/.':-::----r-'-'~z;,j
,/
0.0>«-'-- - - -
8 10 12 14 16 24 48
, "
I
j
.............,,---i---.lS . 1-d,
'
/ 6d' " '.' 18: 2, w -.
,
i
...........
-,=,3
8 10 12 14 16 24 48 _'
,'
1S:3,w3.d,.
-
4
'
6
1
,_., "', 18 : 2 ,w 6.• d '.\.
/
4
3Oj . ' (d) 2.5,
_''',
./
2
4
6
8 10 12 14 16 24 48
Time (h) Fig. 1. IlIeOlporation of dcwcrilim-labclled fatty acids into plasma (a) TAG, (b) CE, (e) PC and (d) PC-2
Subject was fed a diet containing 1,6% (by energy) linoleic acid. 1990
FATTY ACID METABOLlSivl for the w3 fattv acids relative to the w6 fattv acids. The total amount (in m~/l1ll) of IS:3,w3-d, metab~)lites formed bv desaturation a~d elon~ation was over 100 times greater tha~l the total metabolites ~mlled from 18: 2. w6-d 4 \~'hen a lowIS: 2. w6 (SAT diet was fed. i\leasurable levels of 22: 6. w3de were present in many fractions. whereas 20: 4. w(1-d 4 was present at above the detection limit of the g.c.-m.s. method in only a few samples.
DiSCltSsiOll Variability between subjects for absorption of 18: O-d~ was much higher than for the other fattv acids. There was no correlati~n between the percentage ~bsorption of i 8:0 and dietary fatty acid composition. although this relationship has been observed in animal studies [4]. Conversion of 18:0-d 4 to 9c-18: 1-9.1 O-d e by t>9 desaturase was very low. High conversion has been suggested as an explanation for the lack of a cholesterol-increasing effect \vhen diets high in 18:0 arc ~ ~ fed[11]. Studies ill "itl"O indicate that 18: 2. w6 competes with 18:3.w3 for the 2-acyl position of PC [12J and for conversion to long-chain metabolites [12, 13]. However, the selectivity data in Table 1 and the plot for the PC data in Fig. 1 clearly show that 18: 3. w3 is a relatively poor competitor of 18: 2, (1)6 for phosphatidylcholine acyltransferase. This poor affinity for phosphatidylcholine acyltransferase illustrates the biological importance of an additional double bond at the w3 position. The variability of the plasma lipid area selectivity values and the absolute amount (in mg/ml plasma) of deuteriumlabelled fatty acids incorporated in to the plasma lipids from different subjects was generally small. The variability in the 18: O-d~/ 18: I-d 6 values is related to the variability of 18: 0d 4 absorption because poor absorption reduces the amount of 18: 0-d 4 available for incorporation. The between-subjects consistency for incorporation of the other deuterium-labelled fatty acids into plasma lipids was better than expected considering the difference in the 18: 2, w6 content of the t\VO diets. This consistency is evidence that fatty acid composition of tissue lipids (s tightly controlled. This tight control of tissue lipid composition has been observed in animal studies [14]. From a nutritional viewpoint, it means that man can tolerate a rather wide fluctuation in the composition of dietary fatty acids without experiencing major problems due to changes in cell membrane lipid composition. In contrast to the consistency of the selectivity values, the isotopic enrichment data for the C zo w3 metabolites of 18:3,w3-d z from subjects fed the low-18:2,w6 (SAT) diet were 10-30% higher than w3 metabolites from subjects fed the high-18:2,w6 (PUFA) diet. However, the C zz w3 metabolites from subjects fed the SAT diet were two to three times lower than the w3 metabolites from subjects fed the PUFA diet (Table 2). This relationship between the isotopic enrichment data for w3 fatty acid metabolites and dietary 18:2,w6 content indicates that the amount of 18:2,w6 in the diet had different effects on the activities of the desaturases and elongases involved in the conversion of 18:3,w3 to 20:5,w3 and 22:6,w3. The isotope enrichment data in Table 2 is evidence for a much lower conversion rate for 18: 2. w6 than for 18: 3, w3.
Vol. 18
7h6desaturase. as the first step in the conversion pathway [12, 171. Thus. it is difficult to understand why 18: 3, w3-d e convcrsion to 20: 5. w3-d. and 22: 6. w3-d, is much more extensive than 18:2,w6 coiwersion to 20:4,w6. The implication is that 18: 3. w3 is a much better substrate for t>6-desaturase than 18 :2.w6. but this is not consistent with rat [121 and human [18] liver microsome data ill vitro that has utilized 14C-labelled 18 :2.w6 and 18 :3,w3 substrates. However. fetal rat data ill FiFO has shown similar differences in conversion [19]. Isotopic dilution due to the large endogenous pool 01 18 :2. (1)6 docs nol explain the absence of delectable amounls of 20 :4, w6-d •. The 20: 4. (1)6 in liver lipids from mice fed the same 18: 2. w6-d 4 as used in the human studies was found 10 contain 15- 20% 20: 4. (I)6-d~ after 24 h. These results raise several questions about the metabolism of (1)3 and w6 fatty acids in man and point out thaI we need to learn mllch more before a rationale can be developed to explain the physiological effects produced by w3 falty acids. I. Sinclair. H. M. (198 I) Prog. Lipid Res. 20,897-899 Lands, W. E. M. i.ed.1 (1987i PolVlInsawrated FallV Acids and Eicosanoids. American Oil Chemists' Society, Chanlpaign, IL 3. Galli. C. & Simopoulos, A. P. (eds.) (1989) Dieul/Y (1)3 alld (1)6 Fau)' Acids, Plenum Press, New York 4. Garg, M. L.. Wierzbicki, A. A., Thomson, A. B. R. & Clandinin. M. T. (1989) Lipids 24. 334-339 5. Rakoff, H. ( 1988) Lipids 23, 280- 285 6. Adlof. R. O. & Emken, E. A (1978) J. Labelled Compd. Radiopharm. 25,97-104 7. Adlof, R. O. & Emken. E. A (198 I) Chem. Ph}'s. Lipids 29, J
3-9 8. Adlof, R. O. & Emken. E. A. (1980) 1. Labelled Compd. Radiopharm. 18, 419-426 9. Emken, E. A., Rohwedder. W. K.. Adlof, R. 0 .. Rakoff. H. & Gulley, R. M. (1987) Lipids 22, 495-504 10. Hagve, T. A & Christophersen (1988) Scand. J. Clin. Lab. Invest. 48.8 I 3-816 11. Bonanome, A & Grundy, S. M. (1988) N. Engl. J. Med. 318. 1244-1248 12. Emken, E. A. & Dutton, H. J. (eds.) (1979) Geometrical and Positional Fatl\' Acid Isomers. American Oil Chemists' Society. . Champaign, IL 13. Brenner, R. R. & Peluffo. R. (I 966) J. BioI. Chem. 241. 5213-5219 14. Gibson, R. A, Mcmurchie, E. J., Charnock, J. S. & Kneebone, G. M. (1984) Lipids 19.942-951 15. Nichaman, M. Z .• Olson, R. E. & Sweeley, C. C. (1967) Am. J. Clin. Niar. 20, 1070-1083 16. Boustani. S. E .. Causse, J. E., Descomps, B., Monnier. L.. Mendy. F. & de Paulet. A. C. ( 1989) Metabolism 38. 315-321 17. Sprecher, H. (1989) in Dietmy (1)3 and (:)6 Fau)' Acids (Galli, C. & Simopoulos, A. P.. eds.), pp. 69-79, Plenum Press, New York 18. Irma, N. T., Dumm. D. G. & Brenner, R. R. (1975) Lipids 10. 315-317 19. Sanders, T. A. B. & Rana, S. K. (1987) Ann. Nil/I'. Metab. 31. 349-353 Received 25 April 1990
