Representatives of the drug's manufacturer, Isis Pharmaceuticals Inc. of Carlsbad, Calif., say fomivirsen (Vitravene) excels at binding to mRNA. However, some antisense researchers outside the company caution that, in the past, antisense molecules thought to be highly specific have latched onto other proteins. A U.S. Food and Drug Administration advisory committee, which recently recommended that the drug be approved for marketing, only weighs in on the safety and efficacy of the drug, not the mechanism by which it works.

If that mechanism is, indeed, truly antisense--that is, works primarily through a complementary RNA-like molecule that binds to the portion of mRNA that produces the targeted protein, thus blocking the production of that protein--the product's arrival into the marketplace could signal a new wave in molecular medicine, Stanley T. Crooke, Isis founder and CEO told The Scientist.

"We could make antisense drugs for virtually every gene product," Crooke says, noting that the drug's navigation through clinical trials sparks optimism because, during the trials, it demonstrated high binding specificity along with low toxicity.

Researchers at the company selected a gene sequence that made a protein essential to cytomegalovirus (CMV) retinitis. "We picked a site on that messenger RNA that we've empirically demonstrated was a very sensitive site for antisense, and we administered [the drug] to the cells that are infectable by the virus. It binds to the messenger RNA that's made by the virus and causes the degradation of the messenger RNA and then prevents the production of the protein that the virus needs to live," Crooke explains. "It's a classic antiviral [strategy] through a totally novel mechanism."

Alan M. Gewirtz, professor of internal medicine, pathology, and laboratory medicine at the University of Pennsylvania, agrees with Crooke that a successful antisense therapy could signal a sea change in medicine. "It has enormous ramifications if it can be made to work." In interviews with The Scientist, Gewirtz was careful not to elaborate on fomivirsen's mechanism, noting that he hadn't seen any clinical data on the drug. (The company has not yet published results of its clinical trials.)

However, he notes that phosphorothioates-- the molecule on which Isis's oligonucleotide is based--have many nonsequence-dependent effects. "These molecules are very highly charged," Gewirtz comments. "They interact with lots of proteins." Gewirtz defines a true antisense drug as something that inhibits activity in a sequence-specific manner and binds only to messenger RNA. In a recent antisense review article, Gewirtz and co-authors write that many nonsequence-specific interactions occur between the phosphorothioate and proteins in the extracellular environment, on the cell surface, and intracellularly (Gewirtz et al., Blood, 92:712-36, August 1, 1998).

Sudhir Agrawal, of Hybridon Inc., a Cambridge-based biotech firm, agrees that the activity of antisense molecules aren't as simple as thought when the approach was first tried in vitro (P.C. Zamecnik, M.L. Stephenson, Proceedings of the National Academy of Sciences, 75:280, 1975). "These compounds do have other mechanisms of action." Hybridon has its own CMV retinitis antisense drug in Phase I trials, as well as an antisense drug for cancer.

In a diagram accompanying a review article on antisense approaches that is in press for Current Opinion in Chemical Biology, Agrawal and his co-author splits the effects of oligonucleotides into antisense and nonantisense activity. Antisense activity includes nuclease stability, cellular uptake, specificity for target RNA, and hybridization by Watson-Crick base pairing. Nonantisense activity includes antiviral activity, antitumor activity, and antimicrobial activity.

Agrawal, like Gewirtz, was careful not to comment directly on Isis's new drug. He notes that some of the nonantisense activity can be minimized. "If one is careful in designing, ... one can come up with a very specific oligo."

Arthur Krieg, professor of internal medicine at the University of Iowa, is convinced that fomivirsen acts through antisense, not another mechanism. "They've provided some pretty good evidence that it can have antisense activity," Krieg comments. "They've shown that if you change the sequence so it's not complimentary to the target, then it doesn't work." Krieg adds that Isis researchers also modified the sequence so that the immune-response-boosting part of it doesn't play a role.

Crooke agrees that it is difficult to say for certain that fomivirsen acts solely through antisense activity: "Proving the mechanism of action of drugs is challenging." Still, he feels confident that the drug's primary mechanism is antisense. "In human eyes, the virus is thought to be spread by cell-to-cell transfer, making it highly unlikely that this nonantisense mechanism plays any meaningful role in humans. Nor do the proinflammatory effects of the drug account for the therapeutic benefit," Crooke

writes in an in-press guest editorial for Antisense and Nucleic Acid Drug Development. The company has five other antisense drugs in human clinical trials.

Crooke notes that whatever the mechanism, the study demonstrated that the drug, injected directly into patients' eyeballs, slowed the progression of the disease. Researchers measured disease progression by taking masked photographs of patients' eyeballs, which were then read by scientists independent of the trial. It also caused few side effects and was well tolerated by most subjects, Crooke says.

Still, until the mechanisms are more clearly shown, Gewirtz thinks that simply calling the drug an oligonucleotide, rather than an antisense drug, may be the best prescription for the company.