Paradox suggests reasons why COX-2 inhibitors hurt and help

Laboratory studies at Johns Hopkins have revealed that certain products of the enzymes COX-1 and COX-2 can both protect and damage the brain. The findings, published in the February 2005 issue of the Journal of Neurochemistry, offer tantalizing clues to why drugs like Vioxx and Celebrex, which block COX-2, can ease arthritis but potentially harm the heart and brain.

Katrin Andreasson, M.D., an assistant professor in the neurology and neuroscience departments at Hopkins, and lead author of the study, explains that the recent discoveries of cardiovascular complications with long-term use of some COX-2 inhibitors are thought to be due to blocking effects of "good" prostaglandins, which are the downstream products of COX activity, potentially leading to heart attacks and strokes. "Defining which prostaglandin pathways are good and which promote disease would help to design more specific therapeutics," she says.

In their latest laboratory studies, the Hopkins scientists discovered that the prostaglandin PGD2 has a protective or harmful effect in the brain depending on where it docks on a brain cell's surface. After brain cells experience the laboratory equivalent of a stroke, PDG2 can protect them from being killed if it binds to one docking point, or receptor, on the cells' surface, but causes them to die in greater numbers if it binds to a second receptor instead, the researchers report. Prostaglandins are involved in a wide variety of bodily activities including relaxation and contraction of muscles and blood vessels, control of blood pressure and inflammation.

"PGD2 is the most-produced prostaglandin in the brain," says Andreasson. "It trumps all of the rest. So we theorized that high levels are protective. But it was a surprise that it was so effective at protecting neurons."

Because the Hopkins team found that PGD2's positive effects generally outweigh its negative ones, the group speculates that PGD2 may provide a potential target for medicines to combat conditions involving brain damage, including stroke, Parkinson's disease and Alzheimer's disease.

In these neurologic diseases, nerve cell death is thought to be carried out in part by a huge release of glutamate, an important signaling molecule in the brain. In their experiments with brain cells and brain tissue from rats, the Hopkins researchers used glutamate to simulate the aftereffects of a stroke. After strokes and other injuries to the brain, levels of glutamate rise, triggering a number of chemical reactions, including an increase in COX-2 production and prostaglandin production. Increased COX-2 activity then leads to further neuron death.

The new findings come in the wake of two previous studies Andreasson has worked on, each finding unexpected protective roles for another prostaglandin produced by COX-2. In one, a different prostaglandin (PGE2) prevented brain cells from dying after a stroke. In the other, mice lacking a docking point for the PGE2 experienced strokes far more severe than normal animals.

Andreasson is now trying to determine if these prostaglandins have a similar protective effect in mouse models of Lou Gehrig's disease, in which excessive glutamate is believed to damage neurons, and will begin work to see if the beneficial side of PGD2 activity can outweigh its toxic activity.

The Hopkins studies were funded by the National Institute of Neurological Disorders and Stroke (NINDS) and the American Federation for Aging Research. Other authors of the study were Xibin Liang, Liejun Wu and Tracey Hand, all from the Johns Hopkins School of Medicine.