THE PROSTACYCLIN/THROMBOXANE BALANCE IS

SHIFTED IN GREENLAND ESKIMOS

Sven Fischer and Peter C. Weber

PROSTAGLANDINS

FAVOURABLY

Medizinische Klinik Innenstadt der Universitat Mtinchen, Ziemssenstrasse 1, 8000 MUnchen 2, F.R.G.

Jdrn Dyerberg

Department of Clinical Chemistry, Aalborg Regional Hospital, 9000 Aalborg, Denmark

ABSTRACT

The rare incidence of cardiovascular disease in Eskimos has been ascribed to their diet rich in eicosapentaenoic acid (EPA, C20:5n-3) and hence a possible formation of trienoic prostanoids. In this study we compare endogenous formation of prostacyclin (PGI), which is formed by the endothelial cell, and thromboxane (TXA), which is formed by platelets in 20 Eskimos and 20 age and sex matched Danish controls by measurement of the main urinary metabolites. Considerable formation of bioactive PGI from dietary EPA was shown in Eskimos, which was barely detectabls in the controls. Furthermore synthesis of PGI was significantly higher in Eskimos in spite of a markedly lower ar chi- $ donate content in membrane lipids. In contrast formation of TXA was lower in Eskimos as compared to the Danish controls. We con?lade, that the balance between PGI and TXA, which may regulate the inter- action of platelet and vessel wall, is favourably shifted in Green- land Eskimos to an antithrombotic state.

INTRODUCTION

The low incidence of myocardial infarction in Greenland Eskimos (1) can be only partly explained by their favourable plasma lipid and lipoprotein concentrations (2). Differences in haemostatic function like a prolonged bleeding time which is shortened by aspirin and an impaired platelet aggregability (3) point to a difference in prosta- glandin production. Especially the balanced effect of PGI and TXA in the interaction between vessel wall and platelet (4) should be shifted by utilization of EPA toward an antiaggregatory condition (5). Recently, in vivo formation of bioactive PGI after dietary EPA has been reported (6,7) whereas inactive TXA was fo?med from ex vivo stimulated plate- lets at a reduced synthesi .s3 of aggregatory TXA logically less active leukotriene B5 was gener ted from ex vivo stimu- $

(8,9). Similarly, bio-

lated neutrophils after dietary EPA (10,ll). In this study we report on an enhanced prostacyclin activity and a reduced thromboxane forma- tion in Greenland Eskimos living on their traditional diet rich in EPA.

AUGUST 1986 VOL. 32 NO. 2 235

PROSTAGLANDINS

SUBJECTS, METHODS AND MATERIALS

Subjects: Twenty Eskimos, eleven women and nine men, living on their traditional diet rich in EPA in the Upernavik district in North West Greenland, were recruited for the study in 1982. Their food was composed essentially as has beep described in de)ail (12). They were aged between 22-55 years (42.7-7.5 years, mean - S.D.). Results of the urinary analyses were compared with those obtained by the same methods in twenty healthy age and sex matched Danes li- ving on a western diet in Denmark. Both groups had not taken any medication for two weeks, and refrained from alcohol for two days before examination.

Collection of urine: 24 hour urine samples were collected for ana- lvsis of the main urinarv metabolites of dienoic and trienoic orosta- cyclin and thromboxane. Within two hours after sampling the urine volumes were measured and frozen at -2DY. The urines were kept at this temperature for 2-3 months while stored on location and transported to Denmark. They were then stored at -6O'C until transportation to Munich in dry ice. The control urines were treated likewise except that they were stored at -6OY irmnediately after sampling.

Analysis of urinary prostacyclin metabolites. The metabolites PGI 72 3-dinor-6-keto-PGt ) d PGI M (Al7 2 3 dinor-6-keto-PGF ) o’f

-M

dienoic and trienoic ~ro~~acycl?n were a&zed by combined b&MS. They were extracted according to the specific method described in detail Breviously (13) from 50 ml urine after addition of 25 ng 19,19', 20,20'- H -2,3-dinor-6-keto-PGF as internal standard. The methoxime, pentafluo$obenzylester, trimethjfsilylether-derivatives were prepared as described (14) and fragments m/z 590, 586 and 584 (M--181(C H F )) of deuterated internal standard, PGI -M and PGI -M were monito e ir 6 ?n the negative ion chemical ionization2mode (isobatane). These fragments represent the intact molecule minus the pentafluorobenzyl-moiety and are ideally suited for simultaneous monitoring of the dienoic and tri- enoic PGI-metabolites, which differ in two mass units due to the addi- tional double bond in PGI -M. PGI -M and PGI -M derivatives were se- parated on the GC-capillara column3by about 16 sec. Detection limit of PGI -M was 3 % of PGI -M. The GC-MS system was a Finnigan MAT 44s (Brsmen, F.R.G.) equi$ped with a SE 30 fused silica capillary column (25 m, 0.25 mm i.d.). Operating conditions of the GC-MS were: injection port 290°C, interface 28O"C, ion source 2OO"C, electron impact energy 150 eV, current of emission 1.2 mA, electron multiplier voltage 2.5 kV. Quantification of PGI-metabolites relied on a standard curve, which was linear in the observed range.

Analysis of thromboxane metabolites in urine. The metabolites TXB -M

(2 3-d'nor-TXB?) an

thbomboxane we e measure together by radi immunoassay. The lipids of d TxBa-M (2 3 -dinor-TXBa) of dienoic and triengic

the urine were extracted and thromboxane B /B were separated from TXB2-M and TXB

i -M, which cochromatographed? O?I reversed-phase HPLC (15).

-M were TXB$ 1 ssayed with a TXB -antiserum which showed a 48 % cross- rea t vity to authentic 2,3-dine?-TXB2.

236 AUGUST 1986 VOL. 32 NO. 2

PROSTAGLANDINS

Analysis of platelet fatty acids. Platelet rich plasma was harvested from cltratea olooa centrltugea for 5 minutes at 200 g. The plasma was recentrifuged at 1500 g for 10 minutes and the platelet pellet was washed twice in saline. The platelets were then resuspended in 0.5 ml saline and stored like the urine samples until analysis. The fatty acid pattern in platelet lipids was analyzed after lipid extrac- tion and methylation by gas-liquid chromatography as described earlier (3).

Materials. Oeuterated PGI -M was a gift of Or. J. Pike, The Upjohn t Kalamazoo, Michl ,$ an. TXB -antiserum was a gift of Dr. J. L~:%~'Brandeis-University, Waltiam, USA.

RESULTS

%&in metabolite of PGI -M in Eskimos versus Danes. Fig. 1 shows the urinary levels of

,, and PGI, in 20 Eskimos and in 20 age and sex matched Danes. In Eskimos considerable amounts of PGI -M were excreted with a median at 49 rig/g creatinine, whereas in a anes PGI -M was not detectable (n = 12) or just over the detection limit (n = 3). Excretion of PGI -M in Eskimos was 146 rig/g creatinine and was signi- ficantly higher ? han the excretion in Danes, which was 109 rig/g crea- tinine (median, p<O.O5, Wilcoxon rank sum test). The sum of equally potent prostacyclins of the dienoic and trienoic series as measured by the main urinary metabolites is nearly twice as high in Eskimos (195 rig/g creatinine) as compared to Danes (109 rig/g creatinine).

GO0 -

.E 300 -

.-

0 0

; 200- WI

F 100 -

E

0 LZB

.

:*

: -

. 0

0

D

:

i . -cl . . :*

:’

Fig. 1: Urinary excretion of PG12-M (2,3-dinor-6-keto-PGFa)(O) and m(Al7-2,3-dinor-6-keto-PGF, )(O), the main urinary metabolites of dieaoic and trienoic PGI, in 20 'Eskimos (E) and 20 age and sex matched Danes (D). The median and the 25 to 75 percentile are indicated. Excretion of PGI -M is significantly higher in Eskimos as compared to Danes (pa.05; Wilcox& rank sum test).

AUGUST 1986 VOL. 32 NO. 2 237

PROSTAGLANDINS

Platelet fatty acids. Platelet lipids (Table I) in Eskimos show a decrease in n-6-polyunsaturated fatty acids as compared to Danes, whereas the content of n-3-polyunsaturated fatty acids is strongly increased. The ratios of EPA to AA were 0.72 in Eskimos and 0.03 in Danes.

Table 1:

16:0 18:O 18:1n-9 18:2n-6 20:ln-11 20:4n-6 20:5n-3 22:6n-3

E 20.9tl.9 11.7+1.1 20.6c1.6 5.7+2.1 4.8+1.6 9.0+1.3 6.4t1.3 4.020.8

D 19.0+2.0 17.2+3.5 17.2+0.9 8.2+1.5 l.OkO.7 22.122.1 0.5to.9 1.5to.4

Distribution of the main fatty acids (%) in platelet lipids of 20 Eskimos (E) and 20 Danes (D). Values are mean + S.D. Number of carbon atoms, number of double bonds and location of the last double bond in the fatty-acid molecule are indi- cated.

TXB2 -M in Eskimos versus Danes. In fig. 2 the excretion rates of s &in urinary metabolites 2,3-dinor-TXB

f and 2,3-dinor TXB

of dienoic and trienoic thromboxane in Esk mos and Danes are 2 (TXB2 3-M) hown.'

-M in Eskimos were significantly lower than -M excretion is 465 rig/g creatinine in

in Danes (~(0.05, Wilcoxon rank sum test

E

1500- .

.c 1000 - . c . ._ . 2 . Q, L

: 500 - i . c”

+I i:

D

.

:* 1 : -1 . . . . . : i

0-

Fig. 2: Excretion of immunoreactive TXB main urinary metabolites of dienoic and 2 FM (2*3-d'

inor-TXBg,3in ;;e t ienoic thromboxan

Eskimos (8) and 20 age and sex matched Danes (D). The median'and the 25 to 75 percentile are indicated. Excretion of TXB lower in Eskimos as compared to Danes (pxO.05;

-M is significantly ?iI?coxon rank sum test).

.).

238 AUGUST 1986 VOL. 32 NO. 2

PROSTAGLANDINS

Prostacyclin/thromboxane ratios. Table 2 shows the ratios of prosta- cyclins to thromboxanes formed in Eskimos and Danes as measured by the main urinary metabolites. In Eskimos the ratio was three times greater than in Danes indicating a considerable shift of the PGI/TXB balance towards an antiaggregatory state.

Table 2:

ESKIMOS DANES

PG12-MtPGI3-M = 0.42

PG12-M+PGI3-M* = 0.14

TXB2,3-M TXB 2,3-M

*which was formed in Danes only in negligible amounts.

Ratio of endogenous prostacyclin to thromboxane formation in Eskimos and Danes as measured by median urinary excretion of the main meta- bolite of dienoic and trienoic prostacyclin and thromboxane per g creatinine.

DISCUSSION

In this study we compared endogenous formation of prostacyclin and thromboxane in native Greenland Eskimos living on a diet rich in EPA and in Danish controls living on a normal western diet. We found the main urinary metabolite PG13-M of trienoic prostacyclin in levels of 30 % of PGI

B -M-formation in Eskimos, whereas in the control group it

was barely etectable. Furthermore PGI -M-excretion was significantly higher in Eskimos as compared to the Dgnish controls. This is surpri- sing, since the concentrations of AA in plasma phospholipids and in platelet lipids are twice as high in Danes as compared to Eskimos (2,3). However, Eskimos have a much higher intake not only of EPA but also of AA from their food which consists mainly of meat and blubber (12). This dietary AA might serve as a metabolic pool of free precursor fatty acid (16) which is rapidly available for the cyclooxygenase and PGI-synthase of the endothelial cell (6).

The results confirm our previous finding of PGI -formation and un- changed or even increased PGI supplementation of their west g

-release in healt y volunteers after ;i rn diet with fish oil or fish rich in

EPA and DHA (6,7). The scattering in formation of PGI -M in Eskimos reflects biological variation and may be induced by t&jr special nutrition, which varies widely in fat consumption (from 4-600 g/day) depending on accessibility of food (12). We have shown that this is associated with rapid changes in the excretion of PG12 3-M as already one day after ingestion of 300-600 g of mackerel PG13-PI is drastically elevated (6). Our analyses in urine do certainly not allow conclusion about the origin of the PGI-metabolites, they have, however, the ad- vantage to reflect true in vivo formation.

Polyunsaturated fatty acids in the platelet lipids of Eskimos showed a typically "marine" pattern with a strikingly decreased content of AA and an increased concentration of both EPA and DHA as compared to the Danish controls. As the metabolites of TXB2 and TXB were measured to_ gether by radioimmunoassay, formation of 2,3-dinor- 3 XB3 which presumably

AUGUST 1986 VOL. 32 NO. 2 239

PROSTAGLANDINS

cross-reacts with the TXB -antiserum could not be monitored separately in Eskimos. However, the $ xcretion of immunoreactive 2,3-dinor-TXB was significantly lower in Eskimos as compared to the Danish contr%l$. This confirms ex vivo studies with stimulated platelets after dietary EPA where formation of TXB was strongly reduced (9) and only small amounts of TXB were for-me with our findi .il

6 (8). The results are also in aggreement g that chronic intake of fish oil reduces the excretion

of elevated control leve!s of 2,3-dinor-TXB2 (7).

The balance between PGI and TXA has been postulated to regulate plate- let aggregation in vivo and haemostatic plug formation (4) and may eventually influence atherogenesis. Additional formation of PGI at reduced synthesis of TXA leading to a favourable spectrum of bjolo- gically active eicosanoi& from nutritional EPA may persistently shift this balance to a more antithrombotic state (5).

ACKNOWLEDGEMENTS

We thank E. Schmidt, U. Katzner and E. Stoffersen for expert technical assistance. Supported by Deutsche Forschungsgemeinschaft and the Danish Heart Association (Hjerteforeningen).

1.

2.

3.

4.

5.

6.

7.

8.

9.

REFERENCES

Kromann, N. and Green, A. 1980. Epidemiological studies in the Upernavik district, Greenland. Acta Med. Stand. 208:401-406 - Bang, H.O., Dyerberg, J. and Nielsen, A. 1971. Plasma lipid and lipoprotein pattern in Greenlandic west-coast Eskimos. The Lancet I: 1143-1146.

uyerberg, J. and Bang, H.O. 1979. Haemostatic function and platelet polyunsaturated fatty acids in Eskimos. The Lancetg: 433-435.

Moncada, S. and Vane, J.R. 1978. Unstable metabolites of arachidonic acid and their role in haemostasis and thrombosis. Br. Med. Bull. 34: 129-138.

Dyerberg, J., Bang, H.O., Stoffersen, E., Moncada, S. and Vane, J-R. 1978. Eicosapentaenoic acid and prevention of thrombosis and athe- rosclerosis. The Lancet II: 117-119. - Fischer, S. and Weber, P.C. 1984. Prostaglandin I is formed in vivo in man after dietary eicosapentaenoic acid. Natur2 307. 165-168. -*

Schacky v., C., Fischer, S. and Weber, P.C. 1985. Long-term effects of dietary marine w-3 fatty acids upon plasma and cellular lipids, platelet function, and eicosanoid formation in humans. J. Clin. Invest. 3, 1626.

Fischer, S. and Weber, P.C. 1983. Thromboxane A (TXA ) is formed in human platelets after dietary eicosapentaenojc acia (C20:5w-31. Biochem. Biophys. Res. Comm. 116: 1091-1099. - Siess, W., Roth, P., Scherer, B., Kurzmann, I., Bohlig, B and Weber, P.C. 1980. Platelet-membrane fatty acids, platelet aggregation, and thromboxane formation during a mackerel diet. The Lancet I: 441-444. -

240 AUGUST 1986 VOL. 32 NO. 2

PROSTAGLANDINS

10. Strasser, T., Fischer, S. and Weber, P.C. 1985. Leukotriene B is formed in human neutrophils after dietary supplementation wit8 eicosapentaenoic acid. Proc. Natl. Acad. Sci. USA 82: 1540-1543.

11. Lee, Th., Hoover, R.L., Williams, J.D., Sperling, R-I., Ravalese, J., Spur, B.W., Robinson, D.R., Corey, E.J., Lewis, R.A. and Austen, K.F. 1985. Effect of dietary enrichment with eicosapentaenoic and doco- sahexaenoic acids on in vitro neutrophil and monocyte leukotriene generation and neutrophil function. New Engl. J. Med. 312: 19, 1217-1224. -

12. Bang, H.O., Dyerberg, J. and Sinclair, H.M. 1980. The composition of the Eskimo food in north western Greenland. Am. J. Clin. Nutr. 33: 2657- 2661.

-

13. Falardeau, P., Oates, J.A. and Brash, A.R. 1981. Quantitative analysis of two dinor urinary metabolites of prostaglandin 12. Anal. Biochemistry 115: 359-367.

14. Waddell, K.A., Barrow, S.E., Robinson, C., Orchard, M-A., Dollery, C.T. and Blair, I.A. 1984. Quantitative analysis of prostanoids in biologi- cal fluids by combined capillary column gas chromatography negative ion $emical ionization mass spectrometry. Biomed. Mass Spectrometry 11: 68-

15. Fischer, S., Struppler, M., Bohlig, B., Wober, W. and Weber, P.C. 1983. The influence of selective thromboxane synthetase inhibition with a novel imidazole derivative, UK-38, 485, on prostanoid formation in man. Circulation 68: 821-826. -

16. Crawford, M.A. 1983. Background to essential fatty acids and their prostanoid derivatives. Brit. Med. Bull. 39: 210-213. -

Editor: E. Anggard Received: 4-4-86 Accepted: 7-15-86

AUGUST 1986 VOL. 32 NO. 2 241