May 29, 2003
Contact: Robert Sanders
(510) 643-6998
rls@pa.urel.berkeley.edu

Transplanting genes from yeast into bacteria, with an added pinch from wormwood,
yields cheap way to make antimalarial "miracle" drug

Berkeley - By combining genes from three separate organisms into a single bacterial factory, University of California, Berkeley, chemical engineers have developed a simpler, less expensive way to make an antimalaria "miracle" drug that is urgently needed in Third World countries.

The drug, artemisinin, is one of the most promising next-generation antimalarials because of its effectiveness against strains of the malaria parasite now resistant to front-line drugs. It is now too expensive for broad use in countries such as Africa where it is most needed.

"By inserting these genes into bacteria, we've given them the ability to make artemisinin quickly, efficiently and cheaply, and in an environmentally friendly way," said Jay D. Keasling, professor of chemical engineering at UC Berkeley. His research is being published online June 1 in Nature Biotechnology and is scheduled to run in the journal's printed edition in July.

Keasling's technique for transplanting yeast and plant genes to construct an entirely new metabolic pathway inside bacteria can be used generally to produce a broad family of so-called isoprenoids -
chemical precursors to many plant-derived drugs and chemicals of interest to industry, including the anticancer drug taxol and various food additives. Isoprenoids, found widely in microbes, plants and marine organisms, currently are very expensive for the chemical industry to synthesize from scratch and nearly as expensive to extract from plant material.

"This process could be of interest to everybody - drug companies making cancer agents, the government producing antibiotics against bioterror agents, or industries making flavors and fragrances," Keasling said. "A company could tweak the bacteria a bit, adding any number of plant genes involved in making the chemical of interest, to get pretty much any isoprenoid. It would be easy to do now."

Keasling's achievement is a big advance over the day-to-day practice in today's biotechnology industry. There, protein drugs are produced primarily through fermentation by recombinant yeast that seldom have
more than one gene inserted in them, perhaps with an additional piece of DNA fused to that gene to make the yeast spit out the protein.

Scientists have found it much harder to transplant entire gene systems to build new chemical assembly lines. Keasling, however, assembled 10 genes, including control elements, from three different
organisms - bacteria, yeast and wormwood- and got them to work together successfully.

The goal of Keasling's group was to create bacteria capable of producing chemicals that can be used to make many kinds of isoprenoids, a class of some 30,000 known compounds of immense
interest to the chemical and pharmaceutical industries. Isoprenoids are expensive to synthesize, however, and natural isoprenoids like taxol are costly to isolate from plant material. Often, too, these
plants are rare and endangered, so that harvesting causes environmental damage.

Keasling's approach leapfrogs the bulk of the laborious synthesis necessary today, leaving only a few additional chemical alterations to obtain the desired drug or chemical. The development took more
than three years and involved numerous people, primarily Keasling's UC Berkeley coauthors: post-doctoral fellows Vincent J. J. Martin and Jack D. Newman and graduate students Douglas J. Pitera and
Sydnor T. Withers.

Other laboratories have tried to engineer the common intestinal bacteria, E. coli, to make isoprenoid precursors that could be used to produce drugs or industrial chemicals, but the methods involved
hijacking the cell's own production factory. E. coli produce chemicals that can be used to make isoprenoid precursors, but diverting these chemicals to make more of them entails overcoming control mechanisms within the bacterial cell that are not fully understood.

Keaslings innovation was to leave E. coli's isoprenoid pathway alone, but transplant a similar pathway from yeast. This pathway takes in a common chemical in E. coli, mevalonate, and sends it down
a cascade of reactions resulting in isoprenoid precursors, primarily isopentenyl pyrophospahate (IPP) and dimethylallyl pyrophospate (DMAPP). In initial experiments, these isoprenoids accumulated in
the cells and threw then out of whack - the cells either stopped growing or mutated to avoid the toxins - so he stuck in a wormwood gene for an enzyme that converts them to amorphadiene, a chemical
precursor of artemisinin that the cells can deal with.

Artemisinin has been known to the Chinese for 2,000 years as an herbal medicine, qinghaosu. Though highly effective at killing the malaria parasite, currently it is expensive to manufacture because
of the costs of chemical extraction from wormwood (Artemisia annua, a relative of the herb used to make absinthe) or total laboratory synthesis. The expense stands in the way of its use in Africa, where
resistance is rapidly spreading to the first-line antimalarial drugs- chloroquine and sulfadoxine-pyrimethamine.

Since their first success, Keasling and his laboratory colleagues have improved yield from the bacteria 10,000 fold, nearly to the level at which industrial production of the antimalarial drug would be cost effective. Another order of magnitude is doable, he said.

Keasling noted, however, that it is feasible to insert another chemical step into the bacteria to produce a compound, artemisinic acid, that is even closer to artemisinin. And one possibility is to let the bacteria grow and evolve in a petri dish and see if they can produce derivatives of artemisinin that have similar or improved effects on the malaria parasite.

"With the ability to produce taxol or amorphadiene in E. coli, we can easily encourage the bacteria to evolve a molecule not found in nature that could be more effective in human disease," he said.

IPP and DMAPP are precursors to all isoprenoids, which means that the bacterial strains Keasling's group produced "can serve as platform hosts for the production of any terpenoid compound for which the biosynthetic genes are available," they write in their paper.

The family of isoprenoids includes chemicals called terpenoids, which give plants their aroma and which also include taxol from the Pacific yew tree, and carotenoids, such as the compounds that give
plants their color. Aside from their importance in flavorings, colorings and perfumes, isoprenoids from organisms as diverse as coral and fungi are being identified as potential drugs.

"The ability to produce amorphadiene in a simple organism like E. coli opens up a whole realm of possible molecular backbones that can later be functionalized to make drugs," Keasling said.

Keasling's lab concentrates on metabolic engineering of microbes to do complex chemical syntheses to replace current methods that are expensive, polluting and wasteful of resources.

"Enzymes are very specific catalysts that can accomplish in far fewer steps what takes us many complex steps in the laboratory," Keasling said. "We are trying to put enzymes inside cells to create
a biosynthetic cascade using the cell's metabolites as starting material, to provide essentially complete molecules in an aqueous environment using no toxic reagents. We're taking an organism,
co-opting its metabolism, and using it for our benefit now."

"This is just a start," he emphasized. "I really want to push the limits of the organism."

Keasling's work was supported by the National Science Foundation, UC BioSTAR and Maxygen, Inc. of Redwood City, Calif.

###

NOTE: Jay Keasling can be reached at (510) 642-4862 or
keasling@socrates.berkeley.edu.