By David N. Leff

Rational drug development is getting more and more sophisticated, what with combinatorial chemistry, X-ray crystallography and mass spectrometry, to name a few.

But irrational drug discovery also is turning up new pharmaceuticals, by sheer serendipity.

The latest example of stumbling upon a therapeutic compound is Viagra, the failed heart-disease medicine that unexpectedly enhances penile erectility in men suffering from minimal sexual potency.

The FDA has just approved Viagra for this indication, and Pfzer Inc., of New York, is poised to bring it to what looks like a humongous market.

Then there's Pharmacia & Upjohn's minoxidil, a drug for treating hypertension which turned out to reverse loss of scalp hair as well. Pharmacia & Upjohn, of London, purveys minoxidil as a remedy for male-pattern baldness.

Now comes estrogen, a reproductive system hormone that dwindles in women past reproductive age and is widely prescribed to combat this symptom of menopause.

Like so many therapeutic fixes, estrogen cuts two ways. While fending off the osteoporosis that can lead to life-threatening bone fractures, the hormonal supplement threatens post-menopausal women with endometrial (uterine) cancer.

Most recently, evidence has been found that estrogen slows down or wards off Alzheimer's disease (AD).

An article in the April 1998 issue of Nature Medicine reports this unsuspected spinoff finding under the title: "Estrogen reduces neuronal degeneration of Alzheimer amyloid-beta peptides."

The paper's senior author is biochemist and cell biologist Sam Gandy, a professor at New York University and a senior investigator in the dementia research program at the Nathan Kline Institute, in Orangeburg, N.Y. His post-doctoral fellow, molecular and cell biologist Suzana Petanceska told BioWorld Today of "the clues to the clue" that led Gandy and his co-authors to explore the estrogen-Alzheimer connection.

"In the first place," she recounted, "[there] was an etiological study by Gandy's close colleague Richard Mayeux at Columbia University some years ago. It showed very nicely that the decrease in relative risk of AD correlated with dose and duration of estrogen treatment."

The distinguishing hallmark of AD is accumulation of senile neuritic plaques around degenerating brain nerve cells, especially cholinergic neurons. Those plaques consist largely of amyloid-beta. This active ingredient of the disease is cleaved off from a larger parent molecule, the 700-amino-acid amyloid precursor protein (APP).

APP's True Function Still Elusive

"The true function of APP," Petanceska explained, "is still elusive. It's metabolized in several ways. In one, APP is processed to give rise to that small amyloid-beta peptide. This exists in several different isoforms, depending on its length, typically 40 or 42 amino acids. Both have the ability to aggregate as the major constituent of senile plaques.

"The 42 form," Petanceska pointed out, "is known to be far more aggregable than 40. It has a greater propensity to come out of solution, and seems to have a very important role in establishing the AD pathology.

"What Gandy's group wanted to know," she continued, "is how exposure of neuronal cells to estrogen affects the production of these potentially noxious peptides.

"It was essential to test this in a neuronal-cell milieu, as close as possible to the situation we have in vivo."

Initially, the multicenter team of co-authors assembled by Gandy tested a neuroblastoma cell line engineered at The Johns Hopkins Medical Institutions, in Baltimore, Md., to express the human version of APP.

"The reason why they first chose neuroblastoma cells," Petanceska pointed out, "was because this cell line produces quite a lot of amyloid-beta peptide, so that its detection was no problem."

What the investigators detected was that "the effect of estrogen on APP metabolism led to a decrease in the production and secretion of the amyloid-beta peptide. This was encouraging," she observed, "but not the ideal system for looking at AD, because neuroblastomas are cancer cells."

To get closer to the real situation, Gandy's group generated primary neuronal cultures. They obtained these embryonic cells from fetal rodent brain tissues of 17-gestational-day rat pups and 14-day mouse fetuses. In these too, estrogen decreased amyloid-beta output.

Following these confirmatory results, Petanceska recalled, "everybody's interest was, 'Yes, but mice are not men; how about people?' So the study was extended to human neurons, made possible through collaboration with Sweden's Karolinska Institute, in Stockholm."

These collaborators provided human fetal cerebrocortical neurons, which again confirmed the estrogen influence on APP. As compared to untreated neurons, amyloid-b generation was reduced by some 45 percent.

At the metabolic level, the co-authors found that the hormonal treatment diminished production of both the 40- and 42-amino-acid versions of amyloid-beta by half.

Gandy told BioWorld Today, "A number of laboratories — ours among them — both academic and pharmaceutical, are currently trying to elucidate which one of the neural activities of estrogen is responsible for its protective effect in Alzheimer's disease."

Petanceska went on, "If all goes well, if all our hypotheses hold true, it will mean that estrogen might be recognized as a lead compound for rational drug design.

"It means we can try to develop estrogen-like compounds that have desirable effects — but not the undesirable side effects —in treating or preventing AD."

This implies, she added, "someday developing estrogen analogs that can be used by men as well as women. It's hard to imagine," Petanceska concluded, "giving estrogen to men — that is, treating them with a hormone that has feminizing effects." *