Characterization of a novel metabolic pathway of arachidonate in coronary arteries which generates a potent endogenous coronary vasodilator.

Bovine coronary artery strips were incubated with [1-14C]arachidonic acid and the chemical properties of the various prostaglandins (PG) formed were studied. Arachidonate was converted to two major prostaglandin products, PGE2 and a novel prostaglandin having chemical (i.e. base hydrolysis and borohydride reduction) and chromatographic properties identical with 6-keto-PGF1alpha. This final compound was inactive on coronary artery strips. The endoperoxide intermediates, PGG2 or PGH2, previously shown to induce coronary relaxation, were not released into the medium from isolated bovine coronaries. The arachidonic acid-induced dilation may have been due to an intracellular action of PGH2 (or PGG2) or to the action of another, yet unidentified, labile intermediate formed in the enzymatic conversion of endoperoxides to 6-keto PGF1alpha. When PGH2 was incubated with bovine coronary microsomes, the PGH2 was completely metabolized (i.e. loss of rabbit aorta contraction) but a compound was generated which was a much more potent coronary relaxant. We suggest that this major novel metabolic pathway of arachidonate generates a substance, intermediate between PGH2 and the final 6-keto PGF1alpha-like product, which is a potent coronary vasodilator.


SUMMARY
Bovine coronary artery strips were incubated with [l-%larachidonic acid and the chemical properties of the various prostaglandins (PG) formed were studied. Arachidonate was converted to two major prostaglandin products, PGE, and a novel prostaglandin having chemical (i.e. base hydrolysis and borohydride reduction) and chromatographic properties identical with f%keto-PGF,,. This final compound was inactive on coronary artery strips. The endoperoxide intermediates, PGG, or PGH,, previously shown to induce coronary relaxation, were not released into the medium from the isolated bovine coronaries.
The arachidonic acidinduced dilation may have been due to an intracellular action of PGH, (or PGG,) or to the action of another, yet unidentified, labile intermediate formed in the enzymatic conversion of endoperoxides to 6-keto PGF,,. When PGH, was incubated with bovine coronary microsomes, the PGH, was completely metabolized (i.e. loss of rabbit aorta contraction) but a compound was generated which was a much more potent coronary relaxant. We suggest that this major novel metabolic pathway of arachidonate generates a substance, intermediate between PGH, and the final 6-keto PGF,,-like product, which is a potent coronary vasodilator. Bovine coronary arteries maintained in vitro continuously release a PGE'-like substance (determined by biological assay) (1). Treatment with PG-cyclooxygenasc inhibitors (indomethacin, meclofenamate, or aspirin) increased basal coronary artery tone (1, 2). Low oxygen tension increased the rate of prostaglandin release and simultaneously reduced coronary vascular tone; these responses were also inhibited by indomethacin (3).
Exogenous PGE, or PGF,, contracted bovine and human coronary arteries but the precursor arachidonic acid caused relaxation; the latter effect was abolished by prostaglandin synthetase inhibitors (2). These results suggest that arachidonate was converted by coronary cyclooxygenase to a relaxing substance arachidonate and that its subsequent enzymatic conversion to PGE, (by PGH, --?r PGE, isomerase) was slow, thus allowing the endoperoxide to exert its relaxant effect, or (b) that a novel dilating substance was produced by the isolated coronaries.
In the present investigation we studied the metabolism of [f-*4C]arachidonic acid by bovine coronary artery strips. We identify a novel pathway of arachidonate metabolism which has as its final product a compound similar to 6-keto-PGF,,. In addition we find that bovine coronary microsomes convert PGH, to a substance which is a potent coronary relaxant. We suggest that arachidonic acid is converted to a labile intermediate in the biosynthesis of 6-keto-PGF,. and that this intermediate is responsible for the coronary relaxation induced by both arachidonate and PGH2.

MATERIALS AND METHODS
Bovine coronary arteries were excised from freshly removed hearts and cut into spiral strips (2). The coronary spirals were suspended in lo-ml chambers (37") containing Krebs-Henseleit medium continuously bubbled with O,:CO, (95:5%). [l-YlArachidonic acid (2 FCi, 10 pg, prepared as the sodium salt) was incubated with bovine coronary artery strips (1.5 to 2 g) for 1 h. The medium was removed and the coronaries were washed twice with Krebs-Henseleit medium, and then incubated for 4 to 6 h in fresh medium. Aliquots of the medium were removed periodically and tested for biological activity over a cascade of prostaglandin-sensitive smooth muscles (rat stomach strips, chick rectum) and compared to PGE, standards (5). At the end of the 4-to 6-h incubation, the medium was acidified with 0.5 M citric acid to pH 3.5 and extracted twice with 15 ml of ethyl acetate. The combined extract was dried over anhydrous Na,SO, and aliquots were analyzed by thin layer chromatography. The thin layer zones corresponding to PGE, were subjected to alkali treatment (6) and to NaBH, reduction (7). The tissues were weighed and extracted and analyzed for their total lipid content. The coronaries were homogenized in 20 volumes of chloroform:methanol (2:1)/g of tissue followed by two washings with 0.1 M KCl. The extracts were evaporated to dryness and subjected to thin layer chromatography. The radioactivity of the lipid extract (93 ? 2% of the original arachidonate was incorporated) consisted of: free arachidonate (55 + 6%), phospholipids (29 -t 4%), triglycerides (8 ? 4%), and cholesterol esters (5 f 3%) (n = 6). Among the major radioactive phospholipids were phosphatidylethanolamine (18% of total radioactivity) and phosphatidylcholine (10% of total). The bioassay experiments were performed with superfused bovine coronary artery spirals as previously described (4). The bovine coronary artery microsomes were prepared from freshly dissected arteries. The arteries were homogenized (Polytron) and centrifuged for 10 min at 10,000 x g. The supernatant was centrifuged for 40 min at 100,000 x g to obtain the microsome pellet.  ID). When PGH, was incubated in Krebs-Henseleit medium in the absence of tissue, it spontaneously decayed to PGE, and PGD, (Fig. lC), as previously described (9); similarly, incubating bovine coronaries with exogenous ['QPGH, resulted in the appearance of PGE, (Fig. LY). Thus, PGH, was not released from the coronaries. We previously observed that isolated perfused rabbit hearts convert arachidonic acid into a novel prostaglandin (10). A similar situation appears to exist with bovine coronary arteries (Fig. 2). When the extract of the coronary medium was chromatographed in a solvent system of benzene:dioxane: acetic acid (60:30:3; System B/D/A) the "PGE" peak separated into two compounds, one with chromatographic mobility identical to PGE, and the other with chromatographic mobility similar to PGF2, (Fig. 2, Panel 1). The lack of a radioactive peak co-migrating with PGF,, ( (2); the novel prostaglandin produced by the rabbit heart (10) and that produced by bovine coronary arteries were inactive on coronaries at the doses tested (up to 10 pg) (Fig. 3). These results suggest that either the endoperoxides directly relax coronary arteries or are converted to a labile precursor of 6-keto-PGF,,. Therefore we incubated microsomes from bovine coronary arteries with PGH, for short times (2 min) and tested the products formed on bovine coronary arteries to determine if the endoperoxides are converted to a coronary relaxant. As shown in Fig. 4 which is the potent coronary relaxing substance (Fig. 4).
pearance in the eflluent of a prostaglandin-like substance (10). We studied the products of arachidonate metabolism by prelabeling the cardiac phospholipids with [lJ4Clarachidonate (14). Surprisingly, we found that the major product formed by the heart was not PGE, but was, in fact, a novel prostaglandin (10). This substance, like the coronary compound described here, shared all the chemical and chromatographic properties of 6-keto-PGF,,.
Thus, a major possibility which arises from this work is that the biosynthesis of prostaglandin in the heart is largely restricted to the coronary vascular smooth muscle. It has been suggested that the coronary dilation produced by 6 arachidonic acid or hormone stimulation with bradykinin or angiotensin in the rabbit heart was due to PGE,; however, it has not been possible to match this dilation with exogenous PGE,, even at very high concentrations (15). Two explanations for these findings were given: (a) the dilation was due to the endoperoxide intermediates, which decomposed to PGE, and A Prostaglandin Vasodilator from Arachidonate by a Novel Pathway (b) exogenous PGE, cannot match the high, local concentration of endogenously produced PGE,, which may be synthesized very near the site of action. Our observation that the major product synthesized by both the isolated perfused heart and the coronary artery strips is not PGEZ, suggests another alternative, namely that this novel prostaglandin intermediate (Fig. 3) is responsible for the coronary dilation which occurs with arachidonate or bradykinin administration to an intact heart. Indeed, there are numerous demonstrations that isolated or cultured vascular smooth muscles intrinsically synthesize prostaglandins and therefore possess the potential for the local regulation of blood vessel tone (16)(17)(18)(19).
The interaction between the arachidonate metabolites synthesized in blood vessels and platelets is of considerable pathophysiological significance. The two synthetic pathways appear to be physiologically antagonistic and the net response may reflect the algebraic sum of the two systems. Thus, stimulated platelets release thromboxane A2 which induces aggregation and vasoconstriction, whereas blood vessels synthesize a vasodilator that inhibits aggregation. Both substances are highly labile and therefore would be much more critical for local vascular regulation rather than systemic responses.