Structure of COX-2 Active Site Solved
Vanderbilt University (Nashville, TN) and Searle/Monsanto (St. Louis) researchers have caught a molecular snapshot of the first step in the biochemical pathway to prostaglandins, the COX-2 reaction. COX-2, otherwise known as cyclooxygenase-2 or the bad COX, has been a popular target for the pharmaceutical industry because of its role in pain, inflammation, and even cancer. (Click here for a recent story on COX-2 inhibitors.) Their findings, described in the May 4 issue of Nature, may help guide future drug development.Until now, work to develop drugs that inhibit the action of COX-2 has been based on a theoretical structure of the active site for arachidonic acid. No one knew for certain where it bound onto COX-2 and the precise shape of the pocket into which it fits. But now, X-ray crystallography has revealed the three-dimensional structure at 2.4 Angstrom resolution of the enzyme cyclooxygenase-2 (COX-2) with its substrate arachidonic acid bound to it.
"This results from a long line of research to understand how COX-2 interacts with its substrates and inhibitors," said Lawrence J. Marnett, Mary Geddes Stahlman Professor of Cancer Research, Professor of Biochemistry, and Associate Director of Basic Research Programs for the Vanderbilt-Ingram Cancer Center. "It helps us to understand in very specific terms how arachidonic acid is bound on the enzyme. This kind of information helps in identifying and developing new inhibitors."
The snapshot also caught the initial product of the reaction, prostaglandin G2, still bound to the enzyme and showed that it bound in the way predicted from earlier experiments. In the COX-2 reaction, this product would then detach from the enzyme to be picked up by the next enzyme in the pathway to prostaglandin synthesis.
Cyclooxygenase exists in two forms: COX-1 (a.k.a., the good COX), present in the stomach, whose prostaglandin products protect the stomach lining from irritation; and COX-2, produced in response to various stimuli such as inflammatory cytokines and growth factors, whose prostaglandin products result in pain and inflammation.
Recently developed drugs, like Celebrex, that are specific for COX-2 bind at the same site on COX-2 where the arachidonic acid is bound, according to Marnett, although they do not fit the pocket in precisely the same way.
Marnett and his colleagues have developed another COX-2 selective molecule, called APHS, that differs from other COX-2 inhibitors in that it permanently inactivates COX-2 like aspirin does. Other COX-2 inhibitors and non-steroidal anti-inflammatory drugs only temporarily block its action. Marnett noted that their research suggests APHS binds instead in the same pocket where prostaglandin PG2 was attached to COX-2.
"So there's something inside that pocket that appears to be important for the COX-2 selectivity of APHS," Marnett said.
In addition to Marnett, collaborators in the Nature article include James Kiefer, Jennifer Pawlitz, Kirby Moreland, Roderick Stegeman, William Hood, James Gierse, Anna Stevens, Williams Stallings, and Ravi Kurumball of Searle Discovery Research, Monsanto Co.; and Douglas Goodwin and Scott Rowlinson of Vanderbilt's Department of Biochemistry.
For more information: Ravi Kurumbail, Searle Discovery Research, Monsanto Co., 700 Chesterfield Parkway North, St. Louis, MO 63198. Email: ravi.g.kurumbail@monsanto.com.