Introduction: Urinary measurement of PGI-M, a metabolite of the cardioprotective hormone prostacyclin, has become a gold-standard approach to assessing its systemic production. Urinary PGI-M levels are dependent on the cyclooxygenase (COX)-2 enzyme, which has been taken to mean that systemic vascular prostacyclin is produced by COX-2 and that its inhibition in blood vessels by nonsteroidal anti-inflammatory drugs (NSAIDs) explains the cardiovascular side effects of these drugs. However, where tested directly, evidence shows that systemic vascular prostacyclin is synthesised via the COX-1 isoform. Importantly, however, COX-2 is expressed in the renal system. In order to consolidate these anomalies we have tested the idea that renal tissues have the ability to generate PGI-M locally and directly from endogenous and exogenous prostacyclin.
Methods: COX-2 expression was analysed by bioluminescent imaging of Cox2fLuc/+ reporter mouse tissue. PGI-M generation (30 mins) was measured by ELISA in tissues stimulated with Ca2+ ionophore to activate COX or in tissues treated with exogenous prostacyclin in which endogenous COX activity was blocked with diclofenac.
Results: COX-2 was constitutively expressed in the renal medulla and bladder but was essentially absent in the liver or aorta. All tested tissues were able to convert exogenous prostacyclin to PGI-M in the following rank order: renal medulla (81±15%) > liver (33±5%) = bladder (29±6%) > aorta (8±1%). In agreement, when endogenous COX activity was stimulated, all tissues were able to synthesise PGI-M de novo with the largest production seen in the renal medulla.
Conclusions: These data demonstrate that the kidney and uro-renal tract contain all the machinery required for local COX-2-dependent production of PGI-M, independent of any hepatic metabolism. This supports data we presented at this meeting in 2016 showing in a patient unable to synthesise prostaglandins, transplant of a normal kidney is sufficient to normalise urinary PGI-M levels. Together, these findings mean urinary PGI-M can no longer be used as evidence of general systemic vascular COX-2 and that we must now consider alternative mechanisms, such as blocking COX-2 in the kidney, to explain NSAID cardiovascular side effects.