COX‐1 and Vascular Disease

Since the discovery of the pivotal role of cyclooxygenase (COX) in the metabolism of arachidonic acid, vascular biologists have been confronted with the duality of this system. Indeed, one substrate (arachidonic acid) transformed by one enzyme (COX) yields end products (endoperoxides) that exist only very briefly before being metabolized to more stable prostanoids by a set of specific downstream synthases that were initially believed to be tissue specific. For instance, platelets contain mainly the synthase that produces thromboxane A 2 (a potent proaggregatory and vasoconstrictor substance), whereas endothelial cells contain mainly the enzyme that generates prostacyclin (an equally potent antiaggregatory and vasodilator substance). The overproduction of thromboxane A 2 by platelets leads to thrombosis; endothelial cells resist vascular occlusion by producing prostacyclin. This duality of the metabolism of arachidonic acid has dominated our thinking about atherothrombosis for decades, and rightfully still does. As scientific understanding progressed, it became evident that two isoforms of COX exist: COX‐1 and COX‐2. COX‐1 was initially considered to be the “good,” constitutive isoform, whereas COX‐2 appeared to be mainly a “bad” inducible enzyme involved in inflammatory responses. However, more recently, the unexpected events resulting from the widespread use of selective COX‐2 inhibitors has suggested that, from a cardiovascular point of view, the products of COX‐2 exert a protective role and that this isoform cannot necessarily be regarded as “bad.” Likewise, evidence has emerged that initiation of the metabolism of arachidonic acid by COX‐1 is not necessarily a “good” thing in terms of vascular protection. This brief review focuses on the potential contribution of endothelial COX‐1 to vascular dysfunction. It is based on a number of review articles, to which the reader will be referred in order to identify the original references to the statements made; these references are not cited here because of space limitations. Clinical Pharmacology & Therapeutics (2009); 86 , 2, 212–215 doi: 10.1038/clpt.2009.108