(FORTUNE Magazine) – WE'D ALL LIKE a better memory, but for most of us the occasional lapse is only a fleeting embarrassment. Not so for the more than 25 million Americans over 65, whose ranks are growing three times faster than the population as a whole. An estimated 85% of the otherwise healthy elderly suffer from memory impairment. For the more than two million victims of Alzheimer's disease, most of them over 65, memory loss can lead to helplessness and despair. There's no real treatment for Alzheimer's, but an effective memory pill could drastically improve victims' lives. Since even in healthy people memory begins to deteriorate when they're still in their 30s, some scientists say, a memory pill would be a boon to untold millions of the middle-aged as well. The good news from university and industry research centers is that the memory-boosting job can be done -- not yet perfectly, but well enough to have ! set the world's drug companies off on a race for a workable memory pill. Pioneering research is unraveling the mysteries of how memories are formed and stored and how memory can be improved. That understanding has produced an approach to designing drugs that aims at the mechanisms of memory. A number of promising drugs are in the final stages of clinical testing on hundreds of volunteers. ''We're halfway there,'' says a research scientist at Warner- Lambert, believed by many analysts to be the leader in the hunt for the memory pill. The payoff, to be sure, won't come next week. Because of the special complexities of designing a drug to work safely in the brain, and the time it takes to navigate the labyrinth of testing required to win Food and Drug Administration approval, it will be five years or so before memory pills reach the market. Next to new cancer therapies (FORTUNE, November 25), the quest for memory- improving drugs is the hottest area of medical research. Right behind Warner-Lambert is a pack of drug companies: American Cyanamid's Lederle Laboratories, Hoffmann- LaRoche, American Hoechst, Ciba-Geigy, and such newcomers as Du Pont, all testing memory-enhancing compounds. European companies like Sandoz in Switzerland, UCB in Belgium, Organon in Holland, Fidea in Italy, and Japanese firms such as Takeda Chemical and Toyama Chemical are in the race too. Much of the work is zeroing in on Alzheimer's, since medications must be effective against a disease in order to be approved by the FDA and its counterparts abroad. Treating Alzheimer's memory loss would satisfy the regulatory requirement. Although ordinary memory slips aren't symptoms of disease, a drug could then be used by healthy people too. ''The first company that comes out with something that is even modestly effective is going to be very successful,'' says one industry consultant. Analysts foresee a $1- billion-plus annual market worldwide for memory drugs just for starters. New knowledge about the brain's remarkable plasticity -- the ability of brain cells, or neurons, to change in response to experience -- is what makes possible the search for drugs to improve failing memories. Under an electron microscope, the brain looks like a tropical rain forest. Neurons resembling clumps of uprooted trees intermingle with thick vinelike axons, through which the neurons transmit electrical signals, to form a seemingly chaotic, junglelike growth. Like a rain forest, the brain is a noisy place; the neurons fire electrical pulses constantly. It's also wet, drenched with chemicals. And like a forest, it's living, constantly changing. For example, dendrites, the rootlike receivers for electrical impulses sent to the neurons, can grow new protrusions in minutes in response to new experiences. Sheets of glial cells sweep through the brain like squalls, apparently to nourish the sprouting neurons. A new impression -- a sight, a thought, a sound, even a smell -- generates electrical pulses that race through the neurons and their connecting networks of dendrites and axons. Where neurons meet lie narrow slits known as synapses. The electrical signal doesn't cross a synapse. Instead, it causes the release of a chemical signal substance called a neurotransmitter. Transmitter molecules speed across the synapses to generate a duplicate of the electrical impulse in the adjoining neuron. According to the latest findings, when a memory is recorded the electric pulses representing the event lead to the creation of new synapses dotted with small sacs that release neurotransmitters. A SINGLE NEURON can have as many as 100,000 synapses connecting with its neighboring neurons. The exact number of neurons in the brain is unknown; it may be as high as 100 billion. Multiplying the number of neurons by the number of synapses suggests why the human brain has the power to make intuitive connections, which a computer does not, and an almost unlimited capacity to record new memories. Old memories don't have to be erased for new ones to be stored. Unlike computer memory disks, which store information in a particular place, the neuronal junctions in the brain that contain specific memories aren't rigidly confined. They're distributed throughout large areas of the neocortex, the great furrowed ''thinking cap'' on top of the brain. This redundant distribution of memories in the brain defeated an ambitious 35-year research effort by the Harvard neurophysiologist Karl S. Lashley. In the 1920s Lashley launched a search for the location of a specific memory, which he called an engram. Lashley trained thousands of rats to run mazes and then carefully cut out pieces of their cortexes, hoping to excise the memory of what they had learned. But like a shadow, the engram seemed to keep a step ahead of Lashley's surgery. The rats kept merrily running the mazes, although sometimes their ability to execute the tasks became blurred. Lashley's experiments suggested not only that memories of an event are distributed throughout the brain, but also that it contains localized centers that control a vast communications network of so-called cholinergic pathways. What's more, as has been shown in recent years, memory neurons are heavy producers and users of the neurotransmitter acetylcholine. Conveniently for memory pill designers, it appears that what creates memory impairments -- mild ones in healthy people, devastating ones in victims of Alzheimer's disease -- is deterioration of the cholinergic pathways because of a shortage of acetylcholine. THE DISCOVERY in the 1970s of cholinergic deterioration in both normal people and Alzheimer's victims -- as well as in aging rats, mice, and monkeys -- was a godsend to drug researchers. They saw two possible ways to increase the amount of acetylcholine in the cholinergic neurons. They could administer lecithin and choline, the sources of acetylcholine, both of which occur in such common foods as eggs, fish, and meat. And they could try to correct the problem with drugs. University researchers tried the first and seemingly most logical approach, giving the building blocks of acetylcholine. To their puzzlement, however, initial tests over two to four weeks failed to improve patients with impaired memories. ''We all thought it would be like taking L-dopa for Parkinson's disease,'' says Richard J. Wurtman, a neuroscientist who directs the clinical research center at MIT. ''Take it today, by Sunday you'll see something. It didn't work that way.'' (L-dopa is a synthetic building block of the neurotransmitter dopamine, in short supply in victims of Parkinson's disease.) A recent study in England shows that lecithin may have to be given to Alzheimer's patients for at least six months to have any effect. (Pure lecithin differs from the capsules sold in health food stores, which contain an unpurified form that would have to be consumed in enormous quantities to have any effect on memory.) Even then the memories of only the oldest patients improved, suggesting that their acetylcholine shortages were the most severe. To Wurtman, the fact that lecithin must be administered for relatively long periods indicates that it takes time to build up the membranes of damaged cholinergic neurons. As Wurtman and his colleague Jan Krzysztof Blusztajn have shown in laboratory tests, neurons short of acetylcholine may manufacture it by a process that Wurtman calls autocannibalism: they consume their own membranes, the ''overcoats'' of the neurons, which are made of lecithin. While Wurtman's hypothesis isn't universally accepted, he thinks the neuronal suicide means that any acetylcholine-boosting drugs will have to be given with large doses of lecithin to keep those membranes in good repair. There's also what could be called the Jewish mother approach. On a desk at the FDA is an application from Thomas J. Lipton Inc., for approval of a lecithin-enriched chicken noodle soup to treat symptoms that some schizophrenics suffer. This could be a convenient way to take lecithin for memory, partly because the daily dosage is so high -- up to three-quarters of an ounce. ''Once the lecithin soup is available,'' says neurologist John Growdon, director of the memory disorders unit at Massachusetts General Hospital in Boston, who with Wurtman helped Lipton incorporate lecithin into the noodle soup, ''we'll be able to test it in age-related forgetfulness and in Alzheimer's patients.'' So far drug companies have chosen the alternative approach -- trying to stimulate the cholinergic pathways with drugs. They find they get the best results with drugs that inhibit the action of the acetylcholinesterase enzyme, which destroys acetylcholine, or that mimic the action of the acetylcholine receptors, the ports that the transmitter molecules fit into. Under tightly controlled conditions, the drug company researchers have measured consistent if modest improvement in such objective measures of memory as recognition of photographs of famous people by Alzheimer's patients. Correcting cholinergic deficiencies hasn't proved as simple as it first looked. Most scientists think the results have not been more spectacular because the experimental cholinergic drugs have a number of drawbacks. Most have only a fleeting effect; physostigmine, for instance, which inhibits the breakdown of acetylcholine, has a half-life -- the time during which half of the drug is consumed -- of only about 15 minutes. And many aren't very effective at penetrating the blood-brain barrier, the sheath of microscopic blood vessels that protects the brain from foreign substances. How much difference getting such drugs into the brain in sufficient quantities can make was recently demonstrated by Dartmouth neurosurgeon Robert Harbaugh. He put a tiny infusion pump under the skin of the abdomen of four patients with advanced cases of Alzheimer's. Through an implanted catheter leading to the base of the skull, the pumps transmitted the drug bethanechol, which mimics the action of acetylcholine, directly into their brains. Three of the patients improved. To the delight of their families, they have shown no further deterioration since they went on Harbaugh's continuous treatment more than two years ago -- an uncommon stability in Alzheimer's victims. By March, Harbaugh hopes to have 50 patients in a follow-up study. Meanwhile, he has joined forces with Lederle to implant the drug-dispensing pumps in monkeys to see if other drugs work even better. The FDA has already okayed bethanechol, which Merck markets for treating bladder problems; it is the only clinically approved medication among the many that researchers would like to use to treat Alzheimer's. Bethanechol works in the bladder because cholinergic receptors, which play a role in muscle contraction, are scattered throughout the body. FOR EASE of administration, the ideal memory-boosting drug would have to be a pill. Accordingly, work is in progress in pharmaceutical houses on redesigning cholinergic molecules to make them act on the brain rather than elsewhere in the body. In addition, drug molecules are being redesigned to get through the blood-brain barrier with relative ease. Because several neurotransmitters are involved in memory impairment, the drugs may have to be given as a multichemical cocktail. Paradoxically, it may not be this logical and rigorous process that produces the first memory pill to reach the market. That drug may be one of another type of pharmaceutical that has been in development longer than acetylcholine restorers -- a class called nootropics (from the Greek noos, for ''mind,'' and tropos, for ''turning''). Piracetam, the prototype drug in this series of mind rectifiers, as they are also called, was not specifically designed to improve memory. Originally developed by the Belgian company UCB as a treatment for motion sickness, it is being sold in 85 countries as an antigeriatric drug, and Syntex hopes to introduce it in the U.S. soon. Since it's similar in structure to a neurotransmitter called GABA (gamma amino butyric acid), it's not surprising that piracetam increases the firing rate of those cholinergic neurons. It has been used to enhance memories in patients with mild to moderate dementia -- diminished cognitive abilities, including memory -- and has been shown to improve memory in healthy students. Apart from Syntex, U.S. drug companies haven't been completely persuaded of piracetam's efficacy. But they have found it a good base to work from: Warner-Lambert has been particularly successful, synthesizing a piracetam- based memory-enhancing drug, pramiracetam, and an unrelated nootropic called CI-911. Pramiracetam has been shown to sharpen memory in both monkeys and rats and in normal human volunteers; in mild cases of Alzheimer's, it has improved memory scores on tests. CI-911 enhances short-term memory and could be the prototype of a memory pill for normal people as well as those with mild memory impairments. With CI-911 Warner-Lambert scientists have observed a 10% increase in short-term memory in rhesus monkeys. The animals play a kind of a computerized game. Yellow squares on a CRT screen in front of them light up at random and then go off; the object is to remember which squares were lighted. Warner-Lambert and other companies are pursuing additional routes to memory pills. Raymond T. Bartus, who leads the hunt at Lederle, says his company's most promising experimental drugs are based on the premise that ''if you can improve the ability of the brain to utilize oxygen and glucose, you'd have improved neuronal functions.'' He reports that aged mice, monkeys, and rats all show sharper memory of events when given Lederle's new drugs. Bartus also hopes to stimulate electrical impulses in the brain by manipulating various chemicals inside the neuron. Other companies are pursuing still other leads. Academic researchers are hard at work too. For instance, Gary Lynch, a noted neurophysiologist at the Center for the Neurobiology of Learning and Memory at the University of California's Irvine campus, recently demonstrated after a ten-year effort that new synapses form when new memories are stored. Lynch thinks that an enzyme called calpain runs rampant in Alzheimer's patients, destroying neurons. ''The possibilities are just wide open,'' says James L. McGaugh, director of the Irvine center. ''There will be at least moderately effective compounds in the foreseeable future.'' None of these medications would cure Alzheimer's disease. Those cholinergic neurons in Alzheimer's victims are slowly dying. And as Kenneth L. Davis, a psychiatrist at Mount Sinai School of Medicine in New York City, puts it, ''You can't ask a dying horse to run at Belmont.'' But there's hope for slowing the progress of Alzheimer's and helping people with other types of serious memory impairment. Some of those neuronal ''horses'' may be not dead but simply underfed. Researchers at the University of California's San Diego campus have found that the acetylcholine deficiency in Alzheimer's disease can run as high as 80%, even though most of the neurons that use the transmitter are still functioning. The scientists conclude that some of those cholinergic neurons are still living but not producing the neurotransmitter. Furthermore, even an aging brain is still capable of sprouting new dendritic spines, forming new connections that may compensate for the loss of neurons in Alzheimer's disease. Neuron loss also occurs in healthy people, but it's much less extensive. Even partially successful treatment of Alzheimer's symptoms would be a great gift to victims and their relatives -- by prolonging normal functioning and keeping the victims out of nursing homes, and by alleviating the estimated $60-billion annual financial burden on their families and the government. For Alzheimer's sufferers, as for anyone with a normally decaying memory, a pill could help maintain what Shakespeare eloquently called ''the warder of the brain.''