cetaminophen poses a pain-relief puzzle. Despite sharing some properties with conventional nonsteroidal anti-inflammatory drugs, such as aspirin, it doesn't inhibit cyclooxygenase (Cox) -1 or -2. Three years ago, researchers postulated that acetaminophen might inhibit Cox-3--a Cox-1 splice variant.¹ Since then, however, research into Cox-3 failed to generate much consensus. "The scientific community was very enthusiastic at the beginning," says Bela Kis of Wake Forest University. But that excitement has faded. Some continue to probe Cox-3, however, hoping the molecule may provide clues for conditions including Alzheimer disease and some cancers as well as offer some new physiological insights.

Even the name is controversial. Several researchers believe that Cox-3 should be reserved for the product of a third independent COX gene. Kis prefers the term Cox-1b, even though, he says, it doesn't seem to show Cox activity. Indeed, Kis' group found that Cox-1b shows an entirely different amino acid sequence to other Coxs.² More recently, Kis' group found that prostaglandin production in cerebral endothelial cells, a crucial step in fever development, is very sensitive to inhibition by acetaminophen and suggested that the drug inhibits Cox-2.³ "The evidence seems to indicate, at least in humans and rodents, that [Cox-1b] has no relevance to prostaglandin production or the effect of acetaminophen," he says. Regardless, Cox-1b's physiological role isn't clear.

Other groups question whether humans express Cox-3. But Francis Berenbaum, from the Saint-Antoine Hospital, Paris, comments that last December researchers reported during the Osteoarthritis Research Society International Congress that human chondrocytes express Cox-3. Berenbaum recently confirmed this suggestion and plans to present his research later this year.

The research's implications go beyond nociception. Neal Davies, from Washington State University, says that Cox-3 might also be involved in Alzheimer disease and some cancers. More fundamentally, Kis says that studies of Cox variants may illuminate the regulation of gene transcription and splicing. "It is an interesting question--why and how an organism sometimes transcribes Cox-1 and synthesizes a protein with Cox activity and sometimes transcribes a Cox-1b and produces a completely different protein with a completely different function."

Resolving the outstanding issues requires research on several fronts. Berenbaum, for example, called for investigations into Cox-3 expression and function compared with Cox-1 and Cox-2 in different organs and characterizations of the prostaglandins produced by the enzymes in different situations. Davies adds that researchers need specific inhibitors and inducers. "There are many unanswered pharmaceutical hypotheses. Many laboratories and companies, including my own, are answering these questions," he says. Nevertheless, he notes, our "understanding of Cox splice variants is still in its infancy."