Only a fool would neglect to investigate further the abundant flora of the world for the presence of new drugs that will benefit not only the world, but also the fool. --Norman Farnsworth
In March 1992 Professor Djaja Djendoel Soejarto traveled halfway around the world to look at a stump. In the Malaysian state of Sarawak, on the island of Borneo, he found it. The tree was gone, and none like it were growing nearby.
It hadn't been a rare or endangered species. Known as Calophyllum lanigerum to botanists and as "bintangor" to locals, it doesn't grow big enough to interest loggers. But this particular tree--variety austrocoriaceum, now very dead--was special. Apparently it had been the only one of its kind to produce a hitherto unknown chemical compound called calanolide A. And in preliminary laboratory tests, calanolide A stopped the AIDS virus in its tracks.
Soejarto (Soo-ee-yar-toe) is professor of pharmacognosy (literally "knowledge of drugs"--more on this later) in the College of Pharmacy at the University of Illinois at Chicago. According to the college's interim dean, Geoffrey Cordell, UIC is "number one in the United States, maybe in the world" among academic institutions devoted to collecting and testing plants for medicinal use.
Soejarto grew up in Indonesia, on the island of Java. A Dutch botanist at his agricultural college first got him interested in plant taxonomy, and a graduate professor at Harvard in the 1960s turned him toward economic botany and ethnobotany--studying not just the difference between one kind of plant and another, but their value to people as well.
When he first came to UIC in 1979, Soejarto cared for one of UIC's claims to medicinal-plant fame, making sure that the correct plant names were entered in its Natural Products Alert database, NAPRALERT (which now contains information on some 38,000 plant species and is available on-line worldwide). In 1985 he applied for and got a five-year, $1.4 million grant from the National Cancer Institute to look for plants with anticancer properties in the tropical rain forests of Southeast Asia. (The grant has been renewed for a second five years. The New York Botanical Garden is surveying Latin America, and the Missouri Botanical Garden, Africa.) One year into the project, NCI decided to screen the plant samples for anti-HIV activity as well.
Soejarto arranged to collect in partnership with local colleagues in Indonesia, Malaysia, the Philippines, Papua New Guinea, Thailand, and Taiwan. This was partly an attempt to divorce Western science from its imperialistic heritage ("I had no intention of going there as an independent researcher and shipping stuff back here"). It was also a realistic recognition that third world governments regulate their natural resources more carefully than in the past: local institutions can often get the necessary paperwork done better than someone in Chicago. If a marketable drug results from the search, the country it came from will share in the profits.
That's a big if. From 1986 to 1991, Soejarto's multinational team sent more than 10,000 samples--each about a kilogram (2.2 pounds)--back to the U.S. For each sample, they also sent a voucher specimen (one documenting the plant from which it was taken) to the top herbariums worldwide, including the one at the Field Museum of Natural History, where Soejarto wears the additional hat of research associate. Of the first 3,000 Southeast Asian samples NCI tested in the lab, only 106 slowed or stopped the growth of the human immunodeficiency virus. And the most effective of those 106 was an extract from the light-brown twigs and round green fruit of Calophyllum lanigerum austrocoriaceum collected in October 1987 in a swamp in Sarawak.
Clearly, the odds of coming up with a new drug in this way are not great. "When I give a talk [about] this," says Soejarto's senior colleague Norman Farnsworth, "the last slide I show is a Las Vegas craps table. . . . All drug development, no matter how well it has been theoretically conceived, is a big gamble. Probably more drugs have been developed through serendipity than through any planned attack. But serendipity is educated observation of abnormal events that are picked up and followed through." Soejarto, of course, knew this when he started the NCI project, but he was about to experience it firsthand.
Why roll 'em in nature's pharmaceutical casino when organic chemists can whip up any compound you like in the lab? Because they can't really, according to pharmacy dean Geoffrey Cordell. Twigs and fungi and microbes contain more different chemicals than human synthesizers can dream up. "There is no such thing as cranking out new synthetic drugs at random. Typically you do a chemical reaction with a particular compound in mind. But if you pick a few leaves from a tree, right there you have 300 or 400 compounds to test."
The NCI strategy is to check out as many different kinds of plants as possible, before population growth and deforestation kill them off. In theory, Soejarto and colleagues could improve the odds by spending time with natives, becoming accepted in their cultures, and collecting and testing the plants they use as medicines. (In a 1992 paper, Soejarto, Farnsworth, and UIC colleague Charlotte Gyllenhaal reported that 18 percent of traditionally used medicinal plants showed some anti-HIV activity, while 10 percent of other plants did so.) But there isn't time.
Once the extract of Calophyllum lanigerum showed up as a promising HIV-stopper, NCI chemists took a closer look. Using the laborious process of fractionation, they found the active ingredient, a new compound they called calanolide A, C22H26O5. (It acts on the same enzyme in the human immunodeficiency virus as does AZT, but at a different location.) And in the laboratory it proved "100% effective in preventing the replication of HIV-1 and the killing of human immune cells by the virus," Soejarto wrote. That's still a long way from being a useful drug. But an NCI official, quoted in the Field Museum newsletter, called calanolide A "not another ho-hum screening lead, [but] a very intriguing development."
In the spring of 1991, NCI asked Soejarto for another ten kilograms (about 22 pounds) of leaves and twigs for further testing. Members of his research team returned to Sarawak to find that the original tree had been cut down, probably by a local villager. They couldn't find any other C. lanigerum trees in the swamp, but they were able to locate some in the uplands, and packed off a load of leaves and twigs to NCI. Everything seemed to be going fine.
Soejarto's work is in part very traditional and in part so new that a lot of people haven't caught up to it yet. And it's safe to say he wouldn't be doing it from Chicago--and perhaps not at all--if it weren't for Norman Farnsworth.
In person, Farnsworth could almost pass for a Chicago alderman, with his shock of white hair, his suspenders, the cigar clenched in his teeth, and his blunt talk. He is in fact head of the university's Program for Collaborative Research in Pharmaceutical Sciences, and in his 24 years at UIC he's hired most of his colleagues and some of his now-superiors. He is not afflicted with modesty or reticence: "I'm the university character, the eccentric, only because I speak my mind. I was one of the first 100 Americans to go to China in 1974. I've talked with the king and queen of Thailand and Robert Mugabe in Zimbabwe. And I come back here and see these little administrators running around."
Farnsworth has put UIC in the forefront of medicinal plant research. And he's lived long enough to see that work hailed as foresighted planning and not a quirky immersion in an academic backwater.
When Farnsworth emerged from graduate school in 1959, pharmacognosy looked as dead as Soejarto's stump in Sarawak. From the dawn of human civilization until the 1940s and 50s, every drug came from nature, and most came from plants. Pharmacists (by whatever name) had to be able to tell one plant product from another. Pharmacognosy was that study, one of those memory courses would-be druggists had to slog through. "Typically you'd get a couple hundred samples," recalls Farnsworth, "and by the end of the year you'd have to be able to identify them by sight and also in powdered form under the microscope." Plants had given us aspirin, quinine, codeine, morphine, and many less familiar drugs.
Plant-derived compounds still make up about a fourth of all prescriptions. But once organic chemists began to be able to synthesize chemicals in the laboratory, plants seemed passe. In 1962 pharmacists quit dispensing herbal medicines (a decision Farnsworth and colleagues deplore to this day, since it leaves the selling of herbal preparations in the hands of people who can't tell carcinogenic sassafras root bark from the beneficial tranquilizer valerian, and who wouldn't be allowed to say anything to their customers if they could). Every time the expanding pharmacy curriculum had to add a new course, pharmacognosy took a beating. In the early 1960s, Gordon Svoboda of Lilly Research Laboratories derived two safe and effective drugs from the Madagascan periwinkle--vincristine, used leukemia, and vinblastine, against Hodgkin's disease. But even that coup failed to spur new interest in plants. (Says Farnsworth, "Annual sales were only $200 million a year. Companies don't get excited about that--forget about the alleviation of human suffering.") In 1980 the National Cancer Institute phased out an unsuccessful 30-year search for anticancer drugs from plants. Surely the pharmaceutical future was in the lab, not in the woods.
Farnsworth didn't abandon pharmacognosy; he updated it. He began UIC's NAPRALERT database in 1975. Over time he improved the school's ability to do both chemical and biological tests on potential drugs. And he produced a constant stream of articles insisting that medicinal plants are still worth studying. One was titled, "How Can the Well Be Dry When It Is Filled With Water?" In another, a lecture delivered at the Smithsonian, he pleaded his case with statistics: "The 119 plant- derived drugs in use throughout the world today are obtained from less than 90 species of plants. How many more can be reasonably predicted to occur in the more than 250,000 species of plants on Earth?"
Farnsworth acknowledges past fiascos that have scared drug companies away from plants, stories that could be called Nightmares of Interdisciplinary Research: the nonbotanist who carefully identified each newly collected plant by writing in ballpoint pen on the leaves, which dried and crumbled during shipment, rendering his samples worthless; the nonpharmacist who listed the traditional medicinal use of a plant as "contraceptive," without mentioning which sex it was for.
There's no sure cure for these snafus, but eventually several changes converged to help Farnsworth and his renovated discipline out of the wilderness. One, of course, was the rising public interest in all things natural. (This can be an uneasy alliance for scientists. Says Cordell, "You sometimes see crazy things, like foods labeled, "All natural, no chemicals."') Another change was more detailed knowledge of how drugs work. Once scientists know exactly what a chemical must do to a particular cell in order to relieve pain (for instance), they can test new drug candidates at the cellular level--much more quickly and precisely than before, and bypassing much (not all) animal testing, with its ethical and political quandaries and scientific uncertainty. And they can automate the testing process itself. "Ten years ago," says Cordell, "pharmaceutical companies would brag about testing fifteen to twenty thousand compounds per year. Now they do a million a year and more."
Where will they find enough different compounds to test? In nature, of course. But Cordell draws a simple graph with two curves, one rising and one falling. The rising curve is our increasing capacity to test compounds; the falling curve is the number of species available in nature for us to test. "Biodiversity is going down, testing technology is going up," he says. "The question is, at what point are we on this graph? Every second you and I sit here, we lose another football field of rain forest"--a significant loss because it is so biologically diverse. In the same amount of space, a tropical rain forest can hold up to 200 different species of good-size trees while a temperate forest can support only about a dozen (35 at most).
Improved test technology makes the new NCI drug-search project more hopeful than the old one. In the first--which batted 0 for 35,000--plant compounds were tested against one particular strain of mouse leukemia, using real mice. But "cancer" is many diseases, not one. Now NCI has 60 different kinds of cancer cell cultures to test possible drugs against.
In practice, synthetic and natural drug development now often work together. Having seen the original natural compounds, chemists can often make them more effective or less toxic. And if they can find an inexpensive way to synthesize a natural product, we don't have to choose between decimating other species and our own health.
In the fall of 1991, Soejarto got bad news from the NCI laboratory: those new Calophyllum samples didn't contain any of the HIV-killing calanolide A the first one had. "It came as a shock to me," he says. He realized that not being able to produce an identical sample could put his entire project in jeopardy. It's one thing to have the results of an experiment surprise you. It's another--much worse--to be unable to produce the same results from the same process. But surely somewhere he could find another specimen of Calophyllum lanigerum that produced calanolide A!
Given the crisis and the fact that his original local collaborator in Sarawak had moved on, he decided to go there himself. In March 1992, after a collecting stint on the Philippine island of Palawan, and with help from the Sarawak Forest Department, Soejarto went to see the stump.
Staring at it, and checking out other trees nearby, he began to wonder whether stubborn pursuit of Calophyllum lanigerum was really the best strategy. "On the spot I decided that some other species of the genus Calophyllum might have the same compound."
Sarawak alone has more than 60 known species of Calophyllum. Returning to Sarawak in July and October, Soejarto collected samples from 15. Getting them is physically complicated, because the leaves and twigs and fruit are often so high up you need binoculars to see them--and someone has to shinny 60 feet or so up the tree trunk after them. The search is also biologically complicated because of the rain forest's very diversity: trees of the same species tend not to grow near each other. "Here, you see a maple and the whole forest is maples," says Soejarto. "In the tropical rain forest, species might be separated by a kilometer."
Worse yet, the task is taxonomically complicated--species of Calophyllum are hard to define, and very hard to tell apart. One standard reference on the genus is 700 pages long; some species can be told apart only by microscopic differences in the shapes of the hairs on the terminal buds. Nor do taxonomists always agree on where one species ends and another begins. Nonbiologists often have trouble realizing that species aren't like Hondas and Toyotas and Fords--the dividing lines are not always clear-cut. But if you ever resembled a cousin more than your own brother or sister, then you have an inkling of the taxonomists' dilemma.
As he was collecting, Soejarto became interested in the saplike "latex" the Calophyllum trees produce. When he would slash the bark on a tree for sampling, exposing the reddish or brownish wood inside, tiny dots of liquid formed on the sliced surface and united into blobs. Overnight they might overflow and run down the trunk. Some Calophyllum species had a faster flowing, more watery latex. Some had white latex in the twigs and yellow latex in the trunk. Different trees of the same species growing in different habitats even had latexes of different colors.
But latex, of whatever kind, was a lot handier than leaves. Collecting it would be easier on the trees and on the field-workers, Soejarto thought, especially if commercial quantities ever had to be taken. "If we could get an active ingredient from the latex, then harvesting it would be more sustainable." So, even though it hadn't been part of his plan with NCI, Soejarto started scraping latex from the new species as he found them and popping it into the bar-coded plastic bags used for chemical testing. Six months later, he was pleased to find that trees he had scraped latex from in July 1992 had healed.
That fall, both of Soejarto's new strategies paid off. NCI reported that latex from the species Calophyllum teysmannii, variety inophylloide, contained another active compound, previously known but never tested. Called costatolide, it's as potent against HIV-1 as calanolide A at 50 percent higher dosages, and more than makes up for that by being about a thousand times more concentrated in the tree. Whereas calanolide A makes up only one-tenth of one percent of the leaves and fruit of lanigerum when it appears there at all, costatolide comprises an astonishingly bountiful 20 percent of the latex of C. teysmannii.
Ironically, after all this work, calanolide A came back into the picture. Some C. lanigerum trees were found to yield calanolide A (in smaller amounts than the original tree), and a group of chemists managed to synthesize it efficiently in the laboratory. According to Gordon Cragg, chief of natural products research at the National Cancer Institute, the agency hasn't decided which compound to pursue in the next stage of drug development--"preclinical" testing on animals. (This stage is necessary in order to see if the compound works in a living organism as well as it did against isolated cells: Is it excreted or metabolized too fast to do any good? Does it get bound up with proteins and become unavailable? Does it have too many toxic side effects? Only after these questions are answered in animals can the would-be drug be tried on people.)
According to Cragg, NCI has so far tried 30,000 different plant extracts against HIV-1. So far just four compounds--calanolide A, costatolide, and two others--have made it to preclinical testing. Cragg says, "Our experience with cancer suggests that the chance of a candidate drug getting into clinical use is one in forty or fifty thousand," meaning that if one of the four makes it through, he'll be pleased. So will a lot of other people.
Back in Sarawak, the government in June 1993 declared C. lanigerum and C. teysmannii protected species--both to maintain the forest and to deter people from hawking bundles of plant parts to gullible (or desperate) American tourists. And Soejarto has gotten two small additional grants from NCI. With one, UIC graduate student Marian R. Kadushin is studying the variability of Calophyllum trees from one population, season, and tree to the next. "If [the drug] is commercially feasible," explains Soejarto, "we want to multiply only the best varieties."
The second grant has helped the Sarawak Forest Department start a small plantation of the promising species of Calophyllum from a cluster of seedlings found by Soejarto and his team. Soejarto says the transplanted seedlings "are growing well and putting out new leaves after eight months in their new site. Now it's up to the laboratory people to determine whether costatolide has any toxicity. In terms of the field aspect, I can guarantee we will have no problem."
Scientists don't agree on exactly how much danger tropical rain forests are in, or even on how many extinctions are likely to result from a given amount of forest clearing. Many species (especially of plants and insects) have yet to be discovered and systematically described--and in genera like Calophyllum, botanists can't even agree on how many species there are.
Farnsworth and his colleagues usually cite an estimate made by Peter Raven of the Missouri Botanical Garden at a 1986 symposium (preserved by editor Edward O. Wilson as the anthology Biodiversity): "This episode [of extinctions] could amount to the loss of perhaps 10% of the world's species by the end of the century and to more than a 25% loss within the next couple of decades."
They do not mention the more skeptical opinion of Ariel Lugo of the U.S. Forest Service in Puerto Rico, who writes in the same volume that alarming estimates like Raven's leave out mitigating factors. For instance, "The very wet life zones support the highest number of plant species and are subjected to the lowest rate of deforestation. . . . Those who calculate species extinction rates must not assume that all tropical forests are subjected to equal rates of deforestation, respond uniformly to reductions in area, contain the same density of species, or turn into sterile pavement once converted." Taking these factors into consideration, Lugo suggests, might reduce the true extinction rate to 4 percent.
But whatever the true extinction rate is, the cost could be great. Undiscovered medicines are by definition difficult to put a price tag on. But in 1985, Soejarto published an article in the journal Economic Botany. They estimated the average value of each drug-producing species of plant, in terms of what consumers paid for prescription drugs (1980 dollars) alone: roughly $203 million per species per year.
"Estimates are all we have," says Soejarto when asked about extinction rates. But he is sure that plants are not being collected and tested fast enough. From the Calophyllum story it's easy to see why. It may not be enough to preserve a species. It may not be enough to define every tiny variation as a separate species and save them all. When individual plants vary in their chemical makeup as much as some Calophyllum trees seem to, who knows what unheard-of miracle drug may disappear forever with the next cry of "timber"?
Art accompanying story in printed newspaper (not available in this archive): photos/Cynthia Howe.