CB1-induced side effects of specific COX-2 inhibitors: a feature, not a bug.

Abstract

In the current issue of Pain, Telleria-Diaz and colleagues [6] shed novel light on the mechanism of action of cyclooxygenase-2 (COX-2) showing that a “side-effect” of specific COX-2 inhibitors is the significant reduction in the metabolism/degradation of the endogenous cannabinoid (eCB), 2-arachidonoyl glycerol (2-AG). The late Michael Walker's group first showed that the eCB, anandamide, was released in the periaqueductal grey (PAG) during noxious peripheral stimulation and that eCB levels correlated with analgesia [7]. Subsequently, Andrea Hohmann and colleagues [2] demonstrated another and more abundant eCB, 2-AG, is the most likely candidate for the non-opioid component of stress-induced analgesia in the PAG. These data placed the endogenous cannabinoid system as a primary player in the regulation of pain at the CNS level. Biochemical relationships between the eCB system and over-the-counter pain medication are an emerging field of interest. In 2005, Hogestatt and colleagues [1] proposed that an analgesic property of acetaminophen (also called paracetamol) arose from metabolic production of N-arachidonoyl-4-aminophenol (AM404) by conjugation of arachidonic acid to exogenously administered p-acetamidophenol. AM404 works as an eCB hydrolytic enzyme inhibitor that elevates or blocks the degradation of eCBs. Some of my colleagues and I who were studying both endogenous cannabinoids and pain joked at the time that while these results were fascinating and encouraging, we almost hoped that no one in the media would pick up on this fact. We feared that a media report that acetaminophen was, in fact, a cannabimimetic drug would lead to its removal from the shelves. Funnily enough, the popular media did not seem to notice and the interactions of NSAIDS and the eCB system on pain remained firmly in the minds of a relatively small group of scientists. In a separate but related area of research, Marnett's group demonstrated that 2-AG is readily oxidized by COX-2 in vitro and introduced a potential metabolite of 2-AG, PGE2-G, that may act more like a prostaglandin than an eCB [4]. We later showed that PGE2-G is produced endogenously, is pro-nociceptive, and acts through an as yet unidentified receptor [3]. Data presented here by Telleria-Diaz suggest that it is primarily the activation of 2-AG at CB1 receptors that inhibits the central sensitization driven by the continuous sensory input of sustained inflammation. More recently, Marnett's group [5] added another caveat to this beautifully complex system. They used in vitro recombinant COX-2 and showed that COX-2 inhibitors, at low doses that still allow arachidonic acid metabolism into PGE2, will block 2-AG metabolism into PGE2-G. That observation perhaps provides a molecular mechanism of action for what is shown in the in vivo studies of Telleria-Diaz and colleagues. Therefore, in addition to the lack of 2-AG degradation as a mechanism of action, suppression of PGE2-G production by COX-2 inhibitors may also contribute to their role in reducing post-inflammation central sensitization. The data in this new report, of course, challenge the traditional desired pharmacological profile of therapeutics, which is to have very narrow and selective activity. Generally, “off-target” actions increase the likelihood of deleterious side effects. In this case, we can delight in the fact that there may be a category of COX-2 inhibitors that block 2-AG degradation during post-injury pain as well as reducing prostaglandin levels. In other words, this “bug” may actually be a “feature”.