Isolation of a new lipoxygenase metabolite of arachidonic acid, 8,11,12-trihydroxy-5,9,14-eicosatrienoic acid from human platelets

PROSTAGLANDINS

ISOLATIONOF A NEW LIPOXYGENASEMETABOLITE OF ARACHIDONICACID, 8,11,12-TRIHYDROXY5,9,14-EICOSATRIENOIC ACID FROM HUMAN PLATELETS*

Robert W. Bryant and .I.Martyn Bailey

Departmentof Biochemistry George WashingtonUniversity School of Medicine and Health Sciences Washington,D.C. 20037

Abstract Washed human platelets incubatedwith l-14C -arachidonicacid (1mM) produced a new metabolitewhich migrated on thin layer chromatographyclose to thromboxaneB2, but which was identifiedby mass spectrometryas a trihydroxy fatty acid. The mass spectrumwas consistentwith the assigned structure,8,11,12-trihydroxy-5,9,14-eicosatrlenolc acid (THETE). Platelet THETE synthesis from arachidonatewas not inhibitedby preincubationwith aspirin or lndomethacinbut was blocked by 5,8,11,14-eicosatetraynoic acid. Therefore,THETE appears to arise via the platelet lipoxygenasepathway rather than via the prostaglandincyclooxygenase. Two proposed structures, includinga novel dihydro-hydroxy-pyran cyclic intermediate,which could give rise to THETE are presented.

Introduction It was reported previouslyby us that platelets incubatedwith high concentrationsof arachidonicacid produce a mixture of isomeric trihydroxy eicosatrienoicacids (1). While 12-hydroxy-5,8,10,14-eicosatetraenoic acid (12-HETE)and its 12-hydroperoxyprecursor (12-HPETE)have been the only identifiedplatelet lipoxygenaseproducts of arachidonicacid (2-41, 8,11, 14-eicosatrienoicacid was reported by Farlardeauet al. (4) to yield several platelet lipoxygenaseproducts in addition to 12-hydroxy-8,10,14-eicosatrienolc acid. 8,9,12-trihydroxy-10,14-eicosadienoic acid and 8,11,12-trihydroxy-9,14-eicosadienoic acid were tentativelyidentifiedamong these (4). We now report the isolationof additionalplatelet lipoxygenaseproducts from platelets incubatedwith arachidonicacid and the identificationof one of these as 8,11,12-trihydroxy-5,9,14-eicosatrienoic acid (THETE).

* A report of this work was presented at the American Society of Biological Chemists Meeting in Atlanta June 4-8, 1978.

JANUARY 1979 VOL. 17 NO. 1

9

PROSTAGLANDINS

Materials and Methods Prostaglandin and thromboxane B2 standards were obtai$$C]_ through the courtesy of Dr. John Pike, Upjohn Company. arachidonic acid (Amersham-Searle, 60 uCi/mole) was diluted with purified, unlabelled arachidonic acid (AA) as previously described (5). Indomethacin was obtained from Sigma Chemical Co,, while 5, 8,11,14-eicosatetraynoic acid was supplied by Hoffman-LaRoche. Outdated human platelet concentrate was provided by the Red Cross of Washington, D.C. For washed platelet incubations, one unit of concentrate (used within 72 hours of expiration) was centrifuged at 1,100 x g for 15 min. The platelet pellet was resuspended in Tris-EDTA-NaCl buffer, pH 7.5 (6), recentrifuged (1,100 x g, 15 min.) and finally resuspended in ten volumes of Krebs-Henseleit buffer, pH 7.5, minus calcium 016 glucose (6). These platelet suspensions (approxima&ely 2 x 10 /ul) contained 1 to 2 percent red blood cells. [l- Clarachidonic acid was added in ethanol (the final concentration of ethanol was less than 0.5% v/v> to the suspension which was stirred at 37'C. The incubations were stopped after 10 minutes with addition of l/10 volume of 10% formic acid and the suspension was extracted 3 times with 2 volumes of ethyl acetate. The combined ethyl acetate extracts were backwashed with water, dried over anhydrous sodium sulfate and evaporated under vacuum or a nitrogen stream. The anhydrous extracts were fractioned and analyzed by radioscanning TLC as previously described (5) using solvent systems; C3 (chloroform:methanol:acetic acid, 90:5:1, v:v;v), Bl (benzene; dioxane:acetic acid, 40:20:1), and I (iso-octane:ethyl acetate: acetic acid:water, 5:11:2:10, upper phase). Separation of radioactive derivatives by Gas Liquid Chromatography-Radioactivity Monitoring (GLC-RAM) was carried out as previously described (5). GC-MS data was obtained on a HewlettPackard 59928 GC/MS system fitted with a jet separator and a 2 x 910 mm glass column packed with 3% SP-2250. The helium carrier gas flow was 20 mlfmin., the injection port temperature was '24O'C and the oven was operated at 200°C. All spectra were obtained at 70 eV. Samples (l-5 ug) of TXB2 or THETE, methylated with diazov methane (5), were silylated by either addition of I.0ul of bis (trimethylsilyl)-trifluoracetamide (BSTFA)/trimethylchlorosilane (TMCS) (Sylon BFT, Supelco) followed-by heating at 60°C for 30 mfn, or by addition of 10 ul of trimethyls~lylimidazole (TSM) in benzene (l:lO, v:v). The latter method effects immediate silylation. Deuterated trimethyl-Dg-silyl ether (D9-TMS> derivatives were prepared with N-(trimethyl-D9-silyl)-imidazole (Merck, Sharp, and Dohme).

10

JANUARY

1979 VOL. 17 NO. 1

PROSTAGLANDJNS

Results Washed platelet suspensionswere incubated in paral el with two different concentrations (0.1 mM and 1.0 mM) of l-[lbCl-arachidonic acid. The radioactive "TXB -like" products were isolated and purified by TLC in two successzve solvent systems and then analyzed by GLC-RAM as described in Figure 1. While the "thromboxane B2" fraction from the incubationswith a lower concentration of arachidonate gave only one radioactive peak (Figure lA), those incubationswith higher concentration;r consistently gave two peaks (Figure 1B). One peak had the same retention time, 6.1 min., equivalent chain length (24.1) as the tris, trimethylsilylmethyl ester of thromboxane B2 (TXB2-ME-(TMS)3)and was positively identified as such by GC-MS (7). The other more rapidly eluting peak (compound I) at 4.5 min. (C23.0), also eluted slightly before PGF2,ME-(TMS)3 (4.7 min. C23.1) on this column. The main component of compound I was tentatively identified by GC-MS as 8,11,12-trihydroxy-5,9,14_eicosatrienoic acid (THETE). The mass spectrum of the methylated, silylated compound I (Figure 2) showed prominent ions at m/e 569 (M-15, loss of 'CH3), 479 (M(15+90), loss of 'CH3 and (CH3)3SiOH),463 (M-(31+90),loss of 'OCH3 and (CH3)3SiOH),444 (M-140, loss of CH3(CH2)4CH=CH-CH2CH0 by rearrangement),443 (M-141, loss of 'CH2-CH=CH-(CH2)3-COOCH ), 371 (M-213, loss of *[(CH3)3Si-0-CH-CH2-CH=CH-(CH2) -CH3]), 353 (M-(141+90)),281 (M-(213+90)),230 (+[(CH3)3SI-O=&-CH=CH-CH-OSi(CH3)3]), 213 (+[(CH3)3Si=O-CH-CH2-CH-CH-(CH2)4-CH3]). These cleavage assignments were supported by the mass ion shifts seen in the mass spectrum of the methyl ester, trimethylDg-silyl ether derivative of I, which showed prominent ions (with ion shifts given in parentheses) at 471 (444 + 27), 470 (443 + 27), 389 (371 + 18), 371 (353 + 18), 290 (281 + 9), 248 (230 + 18), and 222 ( 213 + 9). While THETE and TXB2 were not well separated by the solvent systems C3 and Bl, they could by separated routinely in System I (see Materials and Methods for composition). Typical RF values in System I were THETE, 0.28-0.32; TXB2, 0.24; PGE 0.26; PGF2o, 0.17; 6-keto-PGFlo,0.10; PGD2, 0.36; HHT, 0.!Y 3; HETE, 0.69; and AA, 0.79. This TLC system was used to assay the effect of c clooxygenase and lipoxygenase inhibitors on THETE synthesis from [114 Cl-AA (Figure 3). Neither aspirin (Figure 3B) nor indomethacin (Figure 3C) inhibited THETE synthesis, while eicosatetraynoicacid (Figure 3D) totally blocked THETE formation. In addition, THETE formation was not detected in control incubationswith boiled platelets or red blood cells. Furthermore, significant THETE synthesis took place in platelet incubations containing less than 0.1% red blood cells.

JANUARY 1979VOL. 17 NO. 1

11

PROSTAGLANDINS

Figure 1 Gas-Liquid Chromatography-Radioactivity Monitoring (GLC-RAM) Analysis of TXB2 TLC Fractions from Platelets Incubated with [l-14C]-arachidonate.

Platelet suspensions (10 ml) were incubated in parallel with either 300 ug (5 uCi> or 3,000 pg (5 pCi) of [l-14C]-arachidonic acid. The acidic ethyl acetate extract from both incubationswere purified by TLC in system C3. The [14C]-TXB2fractions (Rf O.lO0.23) were rechromatographedin system Bl. Aliquots of these purified TXB2 fractions (Rf 0.33 - 0.46) were then analyzed by GLC-RAM on 3% OV-17 at 2400 C after methylation and silylation. (A): TXB2 fraction from the incubationwhere [AA] was 0.1 mM; (B): TXB2 fraction from the incubationwhere [AA] was 1.0 mM.

12

JANUARY 1979 VOL. 17 NO. 1

Figure 2 Mass Spectrum of Compound I: Identified as the Methyl Ester Tris-trimethylsilyl Ether of 8,11,12-trihydroxy5,9,14-eicosatrienoicAcid (THETE)

“Y

?

cn,-lcn.l.-cw-cn-cn,-pn-~M-cn=cM-~n-CH.-CH-cn-ccn,,._coocn.

FM+a,y ala

+ 444=

c&cn=cn-cH-cH,cH=cH-lcn,l~-c-ocn,

*

[

JANUARY

1979 VOL. 17 NO. 1

0

II

b

TM8

TM8

I

&MS

13

PROSTAGLANDINS

Figure 3 Effect of Inhibitors on TXB2 and THETE Synthesis from [l-14C]-arachidonic Acid

n

Platelet suspensions (0.5 ml), containing 0.5 mgfml hemoth inhibitors (added globin, were preincubated 5 to 6 minutes Yi in 1 ul ethanol) prior to addition of [l- Cl-arachidonate (40 ug, 0.2 nCi). After a 10 minute incubation, the metabolites were extracted and analyzed by radioscanning thin-layer chromatography (System I). The spots represent the location on the plate of carrier TXB2 and arachidonic acid which were added to the incubation extracts and detected on the plate by phosphomolybdic acid spray (5). (A): Control, no inhibitors: (B): Aspirin, Indomethacin, 20 ug/ml; (D): 5,8,11,14-eico100 pg/ml; (0: satetraynoic acid, 3 ug/ml.

14

JANUARY 1979 VOL. 17 NO. 1

PROSTAGLANDINS

Discussion An unidentified radioactive peak (compound I) was observed by GLC-EAM in experiments designed to prepare large amounts of TXB from platelets incubated with high concentrations of [14C]AA.* This product had co-purified with TXB2 in two TLC solvent systems and was formed in amounts about equivalent to TXB2. The major component of compound I was tentatively identified by GC-MS as 8,11,12-trihydroxy-5,9,14-eicosatrienoic acid (THETE). The mass spectrum of its methyl ester trimethylsilyl ether derivative and its corresponding deuterated trimethyl-Dg-silylether derivative clearly support this proposed structure. Additional support is gained by comparing its mass spectrum (Figure 2) with the published spectrum (8) for 9,12,13-trihydroxy-10,15-octadecadienoic acid(THODE), an incubation product of linolenic acid with a wheat lipoxygenase preparation. For instance, both spectra showed very weakhigh mass ions, e.g. M-15, but intense ions for the loss of an unsaturated aldehyde, i.e. 444 (M-140, loss of CH3-(CH2)4-CH= CH-CH2CHO) for THETE and 460 (M-98, loss of CH3-CH2-CH=CH-CH2CHO) for THODE. This rearrangement, caused by the migration of a vicinal trimethylsilyl group, has also been observed in the spectrum of 9,10-di-trimethylsilylmethyl stearate (9). In addition to intense ion fragments due to cleavage between vicinal 0-TMS groups (213, 371 for THETE and 171, 387 for THODE) both gave ions at 230 for a di-1, 4-trimethylsilyl-2,3-butenefragment. The latter fragment ion (m/e 230) establishes the position oflZhe A9 double bond in THETE, while the positions of the A and A double bonds are tentatively assigned. The presence of a fairly intense ion at 243 in the spectrum of compound I indicates that the sample may also contain the isomerit 8,9,12-trihydroxy-5,10,14-eicosatrienoic acid, which would give rise to the 243 ion via vicinal 0-TMS cleavage. Further structural work on the hydrogenated derivatives has confirmed this identification. THETE appears to arise via the platelet lipoxygenase pathway as its synthesis is not blocked by the cyclooxygenase inhibitors aspirin or indomethacin, but is totally blocked by 5,8,11,14eicosatetraynoic acid, a potent lipoxygenase inhibitor (10). Two possible routes of THETE formation from 12-HPETE are shown in Figure 4. Pathway A involves an epoxy-hydroxy intermediate, several of which have been proposed as intermediates in the synthesis of trihydroxy fatty acids in plant lipoxygenase systems (11,12). Pathway B involves an unusual 2H-3,6-dihydro-3-hydroxypyranintermediate. While the chemical stability of this structure is uncertain, dihydropyrans in general appear to be quite reactive (13). Both intermediates would yield THETE by hydrolytic opening of either the epoxy or dihyropyran ring.

JANUARY 1979VOL. 17 NO. 1

15

PROSTAGLANDINS

Figure 4

Possible Pathways for THETE Synthesis From 12-HPETE

The diagram illustrates the structures of two possible alternative intermediates in the synthesis of THETE from the 12-hydroperoxy precursor (12-HPETE) and also indicates the importance of the platelet glutathione (GSH) peroxidase system in determining the relative proportions of the mono and trihydroxy fatty acid products.

16

JANUARY 1979VOL. 17 NO. 1

PROSTAGLANDINS

We speculatethat THETE is synthesizedduring high rates of 12HPETE synthesis,e.g. at high [AA], when the platelet glutathione (GSH) peroxidasesystem might not be able to reduce all of the 12HPETE formed to 12-HETE. The possibilityof a limited supply of reduced glutathionefor 12-HPETE reductionis supportedby our finding that THETE synthesisfrom high concentrationsof AA is totally eliminatedby addition of glucose (2 mg/ml) to the platelet incubationbuffer (data to be publishedelsewhere). Glucose is known in red cells to be the main source of reducing equivalentsto maintain reduced glutathionelevels via the pentose phosphatepathway and glutathionereductase (14). Key enzymes in this pathway, glucose-6-phosphate dehydrogenose(15) and glutathioneperoxidase (16) have been detected in platelets. The coupling in plateletsof glucose oxidationand peroxide reduction has been demonstratedby the observationthat ,Jdition of either hydrogen peroxide or tert-butylhydro eroxide to antitreated plateletscaused a burst of [1%Cl-CO2 release from mycft 1-I Cl-glucosepresent in the incubation (17). The observationthat trihydroxy-eicosadienoic acids were formed in platelets incubatedwith 8.11,14-eicosatrienoic acid (4) at lower concentrations(0.1 mM) than from arachidonicacid may be related to the limitationsof the glutathioneperoxidase system. While 8,11,14-eicosatrienoic acid is a good substratefor the platelet lipoxygenase(3). it is a much poorer substratefor the platelet cyclooxygenase(4) than arachidonate. Thus relatively more substratewould be available for hydroperoxyfatty acid synthesis from eicosatrienoatethan from arachidonate. A number of factors have been found which influenceTHETE production, in addition to arachidonateconcentration. In particular, glutathioneand, as noted above, glucose were both inhibitory.Further work is necessary to evaluate the physiologicalor patho-physiological conditionswhich regulate the formationof these new lipoxygenaseproducts during platelet aggregation.

Acknowledgements The authors wish to acknowledgethe skilled assistanceof Dr. Amar N. Makheja and Mr. Shelby J. Fleischer. This work was supported by NIH grants 5 SO7HH539-17,HL 05062 and CA 15356 and by NSF grant PCM 07147. Note Added in Proof Subsequentto the submissionof this manuscript,the mass spectrometric identificationhas been reported (Jones,J.L, et&., ProstaglandinsI&, 583-589, 1978) for the 8,11,12- and the 8,9,12-trihydroxy-eicosatrienoic acids.

JANUARY

1979 VOL. 17 NO. 1

17

PROSTAGLANDINS

1.

Bryant,

R.W. and Bailey, J.M. Fed. Proc. 37:

1317 (1978).

2.

Hamberg, M and B. Samuelsson: Prostaglandin Endoperoxides. Novel Transformationof Arachidonic Acid in Human Platelets. Proc. Nat. Acad. Sci. USA 71: 3400-3404, 1974.

3.

Nugteren, D.H.: Arachidonate Lipoxygenase in Blood Platelets. Biochem. et Biophys. Acta 380: 299-3404, 1974.

4.

Falardeau, P., M. Hamberg and B. Samuelsson: Metabolism of 8,11,14-EicosatrienoicAcid in Human Platelets. Blochem. et Biophys. Acta 441: 193-200, 1976.

5.

Bailey, J.M., R.W. Bryant, S.J. Feinmsrk and A.N. Makheja: Differential Separation of Thromboxanes and Prostaglandins by One and Two-DimensionalThin Layer Chromatography. Prostaglandin13: 479-492, 1977.

6.

Samuelsson, B., M.Hamberg, .I.Svensson, and T. Waksbayashi: Isolation and Stru&ture of Two Prostaglandin Endoperoxides That Cause Platelet Aggregation. Proc. Nat. Acad. Sci. USA 2: 345-349, 1974.

7.

Bryant, R.W., S.J. Feinmark, A.N. Makheja, and J.M. Bailey: Lipid Metabolism in Cultured Cells XVII. Synthesis of Thromboxane A2 from [14C1-ArachidonicAcid by Cultured kung Fibroblasts. J. Biol. Chem., in press.

a.

Graveland, A.: Enzymatic Oxidation of Linolenic Acid in Aqueous Wheat Flour Suspensions. Lipids 8: 606-611, 1973.

9.

Odham, G. and E. Stenhagen.: In: Biochemical Applications of Mass Spectrometry (Wailer, G.R., Ed.) Wiley-Interscience, New Yrok, 1972, p. 225.

10.

Downing, D.T.,D.G. Ahern and M. Bachta: Enzyme Inhibition by Acetylenic Compounds. Biochem. Biophys. Res. Common. 60: 218-223, 1970.

11.

Veldink, G.A., J.F.G. Vliegenthart, and J. Balding: Plant Lipoxygenases. Prog. Chem. Fats Other Lipids 15: 131-166, 1977.

12.

Gardner, H.W.: Decomposition of Linoleic Acid Hydroperoxides. Enzymatic Reactions Compared with Nonenzymatic. J. Agric. and Food Chem. 2: 129-136, 1975.

13. Fried, J: In: Heterocyclic Compounds, Vol. I (Elderfield,R.C., Ed.) John Wiley and Sons, New York, pp. 349-350, 1950. 14. Beutler, E.: In: The Metabolic Basis of Inherited Disease (Stanbury,J.B.,J.B. Wyngaarden and D.S. Fredrickson, Eds.),N.Y., McGraw Hill, p. 1369, 1972. 15. Gross, R.. G.W. Lohr, and H.D. Waller: Biochemistry of Thrombocytes. Proc. Internatl. Congr. Biochem 10: 92-96, 1959. 16. Karpathkin, S. and H.J. Weiss. Deficiency of Glutathione Peroxidase Associated with High Levels of Reduced Glutathione in Glanzmann's Thrombasthenia. New Eng. J. Med. 287: 10621066, 1972. 17. Holmsen, H. Received

18

(personal communication).

10/2/78

- Approved

U/21/78

JANUARY

1979 VOL. 17 NO. 1