Although nonsteroidal anti-inflammatory drugs (NSAIDs) are used as cancer chemopreventative agents, their mechanism is unclear because NSAIDs have cyclooxygenase-independent actions. We investigated an alternative target for NSAIDs, peroxisome proliferator-activated receptor-gamma (PPARgamma), activation of which decreases cancer cell proliferation. NSAIDs have been shown to activate this receptor, but only at high concentrations. Here, we have examined binding of diclofenac to PPARgamma using a cis-parinaric acid displacement assay and studied the effect of diclofenac effect on PPARgamma trans-activation in a COS-1 cell reporter assay. Unexpectedly, diclofenac bound PPARgamma at therapeutic concentrations (K-i = 700 nM) but induced only 2-fold activation of PPARgamma at a concentration of 25 muM and antagonized PPARgamma trans-activation by rosiglitazone. This antagonism was overcome with increasing rosiglitazone concentrations, indicating that diclofenac: is a partial agonist. No effect of diclofenac: was seen without exogenous receptor, confirming that it was working through a PPARgamma-specific mechanism. This is the first description of an NSAID that can antagonize PPARgamma. In addition, this is the first time that an NSAID has been shown to bind this receptor at clinically meaningful concentrations. The physiological relevance of these findings was tested using adipocyte differentiation and cancer cell proliferation assays. Diclofenac decreased PPARgamma-mediated adipose cell differentiation by 60% and inhibited the action of rosiglitazone on the prostate cancer cell line, DU-145, allowing a 3-fold increase in proliferation. This work shows that standard doses of diclofenac may have pharmacodynamic interactions with rosiglitazone and this has therapeutic implications, both in the management of type 2 diabetes and during cancer treatment.
