For two solid years, Dr. Perry Molinoff and his colleagues methodically tinkered with a couple of molecules they had identified the old-fashioned way - by testing tens of thousands of compounds on cultures of growing cells. Among other things, they adjusted certain structural features to reduce the chance that the molecules would cause gastric distress and to increase the likelihood that they would cross over the blood-brain barrier. And all the while, they checked and rechecked to make sure that during this biochemical shuffling they did not lose the key property that made these particular molecules potentially so valuable: their ability to block the activity of gamma secretase, an enzyme thought to play a critical role in the development of Alzheimer's disease.
Out of more than 100,000 candidates in the initial screen, says Molinoff, head of neuroscience drug discovery at Bristol-Myers Squibb, only two were deemed promising enough to continue working on. That was in 1996. Now, thanks to the efforts of nearly 40 scientists - some at Bristol-Myers' Wallingford, Conn., research institute, others at SIBIA Neurosciences, based in La Jolla, Calif. - the number of compounds derived from these two templates has expanded to more than a thousand.
Last year one of these highly refined derivatives became the first so-called secretase inhibitor to enter clinical trials with Alzheimer's patients, and others seem sure to follow. In fact, not just Bristol-Myers Squibb but also Amgen, Elan Pharmaceuticals, Eli Lilly, Merck and SmithKline Beecham are racing to develop similar compounds. The reason? Over the past five years, an explosion of insights into the genetics of Alzheimer's has bolstered confidence that gamma secretase and a related enzyme called beta secretase are not innocent bystanders but rather are intimately involved in the disease process.
The rapidity with which this new class of drugs has emerged is nothing short of stunning. Of course, until the clinical trials now under way render a verdict, no one knows whether any of these novel compounds will turn into a pharmaceutical Cinderella, but at least the possibility is there. "Compared with what (Alzheimer's researchers) had coming down the pipeline a couple of years ago, it's night and day," says Dale Schenk, vice president of neurobiology at Elan Pharmaceuticals in South San Francisco. "Finally we have moved out of the laboratory and into the clinic."
Alzheimer's disease, many experts are all but convinced, starts with the abnormal buildup of a protein known as beta amyloid. The chief constituent of the scarlike plaques found in the brain of Alzheimer's patients, beta amyloid is made by nerve cells when beta and gamma secretase execute a one-two snip that cuts a larger precursor protein into a shorter fragment. Sometimes the fragment is 40 units long, sometimes 42. The slightly longer variant, scientists have found, is directly toxic to nerve cells. Among other things, it appears to stimulate the release of oxygen free radicals, thereby setting off in the brain a destructive biochemical cascade.
Gamma secretase emerged as an attractive target shortly after 1995, when scientists working with DNA donated by families prone to early-onset Alzheimer's disease finally succeeded in cloning two genes known as presenilin 1 and presenilin 2. In a series of experiments, researchers established that these genes exercised tight control over the activity of gamma secretase. They found that the particular mutations in the Alzheimer's-prone families not only increased the rate at which gamma secretase produces beta amyloid but also enhanced its penchant for making the more toxic version.
Beta secretase, the gene for which was cloned only in 1999, also provides drug designers with a target. Inside nerve cells it competes with a third enzyme known as alpha secretase, whose activity, some think, may help protect the brain against Alzheimer's disease. When alpha makes the first cut in the precursor protein, gamma secretase makes a second cut that produces not beta amyloid but an innocuous protein fragment known as p3. Elan and Pharmacia, based in Peapack, N.J., among others, are actively working to develop beta secretase inhibitors.
Logically it would seem that compounds that block the activity of either beta or gamma secretase should slow the progression of Alzheimer's disease. But there are reasons to remain cautious. For one thing, it may turn out that both these secretases play vital roles in other aspects of cellular metabolism, so that interfering with them will come at the price of serious side effects. For another, it is still far from proven that beta amyloid is as central to Alzheimer's disease as, say, cholesterol is to heart disease. Says molecular neurologist Dr. Peter St. George-Hyslop of the University of Toronto: "We have a theory and experimental data that support that theory, but we won't know the theory is right until we have a drug that actually prevents Alzheimer's disease."
A suggestion that drug designers might be on the right track comes from studies of mice that have been genetically engineered to develop Alzheimer's-like plaques. These mice exhibit at least some symptoms of memory loss, performing less well on tests that measure how long it takes them to get back to the one dry platform researchers have positioned out of sight in a watery maze. There are now strong hints that retarding the development of plaques helps preserve intellectual performance, at least in rodents. And that raises an intriguing question: Might getting rid of plaques once they have developed do more than slow a patient's decline?
"I used to think plaques were like bits of concrete scattered throughout the brain," says Bristol-Myers' Molinoff. "But there's now intriguing evidence that suggests you can get plaque regression." Some of the most striking evidence comes from studies of a vaccine against beta amyloid that Schenk and his co-workers at Elan have developed. In 1999 they administered their vaccine to mice whose brains were filled with plaques. A short time later, the plaques shrank. Currently the Elan vaccine, like the Bristol-Myers' secretase inhibitor, is in early-phase clinical trials, in which the primary objective is to test for safety as opposed to effectiveness.
At present, the likelihood that this particular vaccine or this particular secretase inhibitor will end up in the therapeutic arsenal has to be considered slim. Of all the compounds his team brings forward, says Molinoff, only 10% to 15% manage to pass the rigorous series of tests that lead to approval by the U.S. Food and Drug Administration. And even when they do, Harvard University neurologist Dr. Kenneth Kosik emphasizes, precious few new drugs prove to be anything close to magic bullets. Indeed, Kosik, along with many others, thinks it is quite likely that controlling Alzheimer's disease will require more than one type of drug. In addition to compounds that inhibit plaques, for example, patients may need drugs that prevent the formation of tangles that disrupt nerve cells from within.
For this reason, people whose advancing years place them at high risk for Alzheimer's disease would be wise to place their hopes not in any particular treatment strategy but in the broad range of options that the genetic toolbox is so rapidly opening up. Right now, it is encouraging to note, researchers are homing in on at least three more genes involved in Alzheimer's, each potentially a fat new target for drug developers.