Background and objective Cyclooxygenase-2 (COX-2) is involved in the generation of prostanoids such as prostaglandin (PG)E₂, prostacyclin (PGI₂) and thromboxane (TXA₂); PGE₂is involved in inflammation/pain, whereas the second two are involved in cardiovascular homeostasis. Nonsteroidal anti-inflammatory drugs (NSAIDs) inhibit COX-2 and are used to treat inflammation, however they cause cardiovascular side effects. The mechanisms for this include reduced vascular and/or renal PGI₂ and, by effects on the renal medulla, increased dimethylarginine (ADMA), an endogenous inhibitor of endothelial nitric oxide synthase (eNOS). Microsomal PGE₂ synthase 1 (mPGES-1) has been intensively investigated as a safer, alternative anti-inflammatory drug-target. We aimed to investigate how mPGES-1 versus COX-2 influences pathways associated with PGI₂, ADMAand eNOS in order to assess the potential cardiovascular safety of mPGES-1 as a drug-target.

Materials and Methods Experiments were performed on mPGES-1^+/+ (WT) and mPGES-1^-/- (KO) mice as well as PGI₂ synthase (PGIS^-/-) KO mice. In some experiments WT mice were treated orally with the selective COX-2 inhibitor, parecoxib (100mg/kg; 5 days). Plasma was analysed by immunoassay for ADMA and PGI₂, or by biochemical assay for creatinine. Renal medulla was stimulated with Ca²⁺ionophore and PGE₂ release measured by immunoassay. Gene expression levels were determined using TaqMan assays. Vascular function (eNOS response) was assessed in isolated aorta stimulated with acetylcholine by wire myography.

Results WT mice treated with parecoxib displayed no change in plasma PGI₂ levels but, in contrast, mPGES-1 KO mice had significantly higher plasma levels of PGI₂. In the kidney, mPGES-1 gene expression was lower in the renal medullathan the cortex, and in agreement; mPGES-1 deletion did not alter renal medulla PGE₂ production. mPGES-1deletion had no effect on genes responsible for the production/breakdown of ADMA in the kidney (PRMT1, AGXT2, DDAH1/2). Plasma creatinine/ADMA was elevated in mice treated with parecoxib or PGIS KO mice but unaffected in mPGES-1 KO mice. Furthermore, mPGES-1 deletion significantly improved the eNOS-driven dilator response to acetylcholine in aorta.

Conclusions Targeting mPGES-1 KO avoids the effects of COX-2 inhibition on renal PGI₂/ADMA pathway and therefore spares vascular eNOS responses. This supports the development of selective inhibitors of mPGES-1 to more safely treat inflammation.