Inhibition of cyclooxygenase (COX)-2 increases cardiovascular deaths. Identifying a biomarker of COX-2 is desirable but difficult, since COX-1 and COX-2 ordinarily catalyze formation of an identical product, prostaglandin H2. When acetylated by aspirin, however, COX-2 (but not COX-1) can form 15(R)-HETE, which is metabolized to aspirin-triggered lipoxin (ATL), 15-epi-lipoxin A4. Here we have used COX-1- and COX-2-knockout mice to establish whether plasma ATL could be used as a biomarker of vascular COX-2 in vivo. Vascular COX-2 was low but increased by LPS (10 mg/kg; i.p). Aspirin (10 mg/kg; i.v.) inhibited COX-1, measured as blood thromboxane and COX-2, measured as lung PGE2. Aspirin also increased the levels of ATL in the lungs of LPS-treated wild-type C57Bl6 mice (vehicle: 25.5±9.3 ng/ml; 100 mg/kg: 112.0±7.4 ng/ml; P<0.05). Despite this, ATL was unchanged in plasma after LPS and aspirin. This was true in wild-type as well as COX-1(-/-) and COX-2(-/-) mice. Thus, in mice in which COX-2 has been induced by LPS treatment, aspirin triggers detectable 15-epi-lipoxin A4 in lung tissue, but not in plasma. This important study is the first to demonstrate that while ATL can be measured in tissue, plasma ATL is not a biomarker of vascular COX-2 expression.
Keywords: 15-HETE; COX-2 biomarker; nonsteroidal anti-inflammatory drugs; vascular inflammation.