By David N. Leff

In Africa¿s heart of darkness, 10 years ago, a safari of native hunters bagged another specimen. It was a leaf mold on an unidentified plant.

The party gathered this fungal sample in a rain forest of the Democratic Republic of Congo and sent it to the Basic Research Center of Merck, Sharp and Dohme in Madrid, Spain. ¿Our researchers there,¿ recalled natural products chemist Gino Salituro, ¿examined the plant samples and isolated the specific fungus ¿ a species of Pseudomassaria. The assayists found that it produced an interesting activity.¿ Salituro is a senior research fellow in the department of natural product drug discovery at the Merck Research Laboratories in Rahway, N.J.

¿Colleagues in Spain pulled the fungus,¿ Salituro went on, ¿which we entered in our natural products screening pro gram. We work with various countries at various sites,¿ he told BioWorld Today, ¿to collect these types of natural product specimens. From Madrid, this one was brought back here to Rahway, where it was further isolated and cultured, then screened, together with over 50,000 other mixtures of samples. As the products moved forward in the screening process,¿ he went on, ¿they were further characterized.¿

Salituro is one of 17 Merck scientists whose article in today¿s Science, dated May 7, 1999, sums up the outcome to date of that product-winnowing process. Its title: ¿Discovery of a small-molecule insulin mimetic with antidiabetic activity in mice.¿ The paper¿s lead author is molecular endocrinologist Bei Zhang.

¿Our ultimate goal,¿ she told BioWorld Today, ¿is to develop an oral insulin-mimicking drug.¿ (See BioWorld Today, Dec. 15, 1997, p. 1, and Feb. 28, 1997, p. 1.)

If and when Merck reaches that goal, it will give millions of diabetic suffers a new lease on lifestyle. Those with early-onset, insulin-dependent, Type 1 diabetes mellitus (IDDM) must self-inject insulin up to four times a day to supply the hormone their bodies are unable to make. The non-insulin-dependent (NIDDM), Type 2 version usually begins in people who are over 40 and overweight. They take drugs to lower their blood glucose levels, with strictly regulated diet, weight and exercise programs.

Diabetics of both types have been holding their breath for years, as reports of oral insulin pills come and go.

¿The assays that we developed to look at insulin mimickers,¿ Zhang recalled, ¿were ready about five years ago. That¿s when our group actually started screening for insulin-like compounds from the 50,000 natural and synthetic-product mixtures that we had.¿

When the co-authors put the Congolese fungal extract into a culture of transgenic Chinese hamster ovary cells programmed to bind to a specific site on insulin¿s cell-surface receptor, the product turned those target receptors on to full activity.

Merck¿s candidate compound, by then bearing the reference number L-783,281, also stimulated glucose uptake in the fat cells of rats, by 263 percent over base level.

By then it was time to test L-783,281 in living, warm-blooded animals that replicate the classic symptoms of human diabetes. Two such mouse models, db/db and ob/ob, have long been available as test benches for studying the ailment in vivo and trying out new drug candidates.

Oral Product Corrected Mouse Insulin Performance

A single oral dose of L-783,281 lowered blood glucose briefly in both models, in a dose-dependent fashion. Long-term daily medication with the oral drug over seven days ¿also resulted in significant correction of hyperglycemia in db/db mice, independent of food intake,¿ the Science paper reported.

Even longer-term treatment, up to 15 days, encountered no adverse side effects. Food intake, body and organ weight, and liver function remained normal. But Zhang made the point that, ¿We did not do any further toxicity testing at this point, because it¿s so early in our research.¿

Quinone, the chemical class to which the original fungal extract¿s refined formulation belongs, is an oddball small molecule. It¿s neither a protein nor a peptide. With poetic justice, one of quinone¿s numerous industrial uses is as a fungicide. The L-783,281 molecule¿s peculiar size and shape qualify it to resist the corrosive effects of stomach acid; ergo, it¿s an orally available candidate drug.

How it acts as an insulin copycat goes back to that hormone¿s own mode of action in the body.

As Biology 101 teaches, insulin is secreted from beta cells in the pancreas¿ islets of Langerhans. A sudden dollop of glucose ingested during a meal alerts insulin to its job of either converting the sugar into bodily energy or storing it away in fat cells for future reference.

But just as a car needs a starter to do its job, insulin needs its receptors. These molecules poke through the outer membrane of nearly every cell in the body. ¿Those insulin receptors,¿ Zhang pointed out, ¿do not bind to glucose directly. Their activation triggers a series of biochemical reactions within the cell, leading it to take up glucose.¿

All that tells the healthy, nondiabetic story. But in Type 1, an autoimmune disease, anti-receptor antibodies destroy 90 percent of the beta cells as fast as they can make insulin. Type 2 diabetes patients produce plenty of the hormone; however, there¿s something wrong with the cells. They are not responsive to the insulin.

How Fungal-Derived Compound Does Its Job

The Merck co-authors figured out that their candidate oral drug binds to the cell¿s insulin receptor and then activates the first enzyme in the domino-like cascade of events that carries the glucose news into the cell, and leads to increased insulin activity.

The ignition in that starter is the chemical process of phosphorylation, which adds phosphate groups to key enzymes and switches them into action. The Merck researchers determined their L-783,281 molecule stimulated the phosphorylating action of the insulin receptor by up to 100 times more than the other natural products they tested.

¿The result of our studies,¿ Zhang observed, ¿demonstrated that it is possible to selectively activate the insulin receptor with a small molecule. We continue to do research in this area. However,¿ she concluded, ¿at this stage it is too early to predict when or if this particular compound will be in clinical trials.¿ n