The Atlantic Home

Last Monday, a group of New Yorkers assembled for a dinner at my apartment to hear Professor Jay Keasling share his vision for the future. Dr. Keasling is a professor of chemical engineering and bioengineering, and the founder of the synthetic biology department at the University of California at Berkeley. He is also the CEO of the Joint BioEnergy Institute in nearby Emeryville, and the director of the physical biosciences division of the Lawrence Berkeley National Laboratory, where he is temporarily filling in for Dr. Steven Chu, now the Secretary of Energy. At my apartment, he unveiled a new and surprising vision, both for the pharmaceutical industry and renewable energy.

Dr. Keasling's group has an innovative approach: rather than solving one problem at a time, they seek to create a set of genetic tools that can be used as interchangeable genetic switches to regulate genes. This approach is very much like that of the electronics industry: using standardized components to build diverse products, A small group of like-minded scientists, working at other universities including MIT and Stanford, have since dubbed this new field "synthetic biology." I prefer the less threatening and more descriptive name "constructive biology."

When I first met Dr. Keasling, he was undertaking his first practical application of these methods. I was a trustee of the not-for-profit organization One World Health, and our group was administering a $42 million grant Dr. Keasling had received from The Gates Foundation to produce the life-saving malaria drug artemisinin.

The story of artemisinin is itself remarkable. The healing properties of the artemesia--or sweet wormwood--plant were first recognized by the Chinese 2000 years ago who prescribed a tea made from its leaves as a cure for recurrent fevers and other maladies. In the 1960s, Chinese scientists isolated and characterized the active ingredient. Artemisinin and other closely related compounds soon became the best available treatment for malaria. The organisms that cause the disease have developed resistance to all other effective therapies, including quinine and chloroquine.
 
The idea Dr. Keasling proposed was straightforward: to transfer the process the artemesia plant uses to make the drug to an organism that is simple to grow in large, reproducible batches. Ultimately, the single-cell organism yeast was chosen, as it has been used for many years to produce ethanol for drinking and for fuel.  Yeast is also a favorite of scientists. The sequence of its entire genome is known, and genes can be added and subtracted at will.

In cells, complex molecules such as artemisinin are assembled one step at a time from simpler components. The problem, therefore, was to find a set of enzymes that would act in tandem to convert chemicals found in yeast to artemisinin. Most of the genes for these enzymes were quickly identified in the artemesia genome itself. Each gene was isolated and inserted into the yeast genome. In all, twelve genes were needed, each expressed in just the right amount.

The initial stages of the work were done in Dr. Keasling's university laboratories. The concluding steps were done by his biotechnology company, Amyris. The final product was a strain of yeast that produced large amounts of a pure compound that required only one final chemical modification to yield pure artemisinin. The entire project was completed in four years.

The artemisinin-producing yeast was then transferred to the French pharmaceutical giant Sanofi-Aventis. Under the terms of the agreement, Sanofi-Aventis will provide the drug to poor countries at its manufacturing cost. The company may sell the drug to the international traveler's market and in developed countries at a higher price, provided the profit is used to subsidize the cost of the drug for poorer countries. It is expected that over time the cost of artemisinin will be reduced tenfold. Moreover, constant supply of high quality artemisinin will be available. Those who receive the drug must pledge to use it only in approved combinations to reduce the likelihood of resistance.

Success emboldened Dr. Keasling and his Berkeley colleagues to apply the technology to other problems. Their newest initiative is to create microorganisms that efficiently convert cellulose to diesel, jet fuel, and gasoline. In the early planning stages of this project, I worked with Dr. Keasling to assemble a board of trustees for a new Institute of Synthetic Biology, a group I now chair. Dr. Keasling and his colleagues at Berkeley and other universities and research centers received a U.S. Department of Energy grant for $134 million spread over five years, as well as a British Petroleum grant for $500 million spread over 10 years.

Biofuels hold the potential to dramatically reduce carbon-dioxide-driven global warming. Unlike fossil fuels such as coal and oil--whose burning releases carbon that has been buried deep in the earth for thousands of years--biofuels contain carbon derived from living plants, which convert atmospheric carbon dioxide to cellulose and sugar. When these fuels burn, they simply return to the atmosphere what was already there.

Despite this promise, the biofuels currently on the market have several disadvantages. Ethanol is highly corrosive, and in order to be used in traditional engines, pipelines, and tankers, it must be diluted with gasoline by 90 percent. Ethanol also mixes too readily with water, and distilling its alcohol is a process that itself requires energy. Gasoline and diesel do not mix with water but float as a distinct layer. Moreover, diesel and gasoline are far more energy-dense, which means that cars running on gasoline get 50 percent more mileage per gallon than cars running on ethanol.

The Berkeley group, now centered at the Joint Bio Energy Institute, has devised a comprehensive strategy for developing more practical and efficient fuels. The first task, improving the yield of usable cellulose from plants, is driven by the desire avoid using food crops for fuel. Plants that thrive on marginal land, require little to no fertilizer, and can survive arid and saline conditions are favored. The Berkeley group now includes plant geneticists and agronomists who have joined the world wide search for suitable plants. Two current favorites are Miscanthus giaganticus, an Asian grass that grows rapidly 10 to 12 feet high, and switchgrass, a native of the U.S. plains states.

The group's second task is to improve the efficiency of conversion of cellulose to sugars for fermentation. The preparation of cellulose into a form suitable for conversion to sugars for fermentation is a complex process that requires large amounts of energy and expensive solvents. The woody material of plants is comprised of three primary components: cellulose, hemi-cellulose and lignin tightly bound to one another.

The Berkeley group has several possible solutions. One is to breed plants that have more cellulose and less hemi-cellulose and lignin. This approach may ultimately be limited, as both components provide needed stiffness. A second approach is to create plants that produce altered forms of lignin and hemi-cellulose--structures can be readily separated from the cellulose itself. The first step of this process is to discover all of the enzymes pants use to make lignin and hemi-cellulose. Thanks to powerful new methods of gene analysis, this work is proceeding rapidly. The second phase of this work is to produce plants with forms of altered lignin and hemi-cellulose that separate from the cellulose dissolve readily during processing.

The Berkeley scientists, and many others, are also actively searching for enzymes that efficiently convert cellulose to sugars. Such enzymes exist in nature and are found in plants and animals that feed on decaying wood. Promising candidates have been isolated from the bacteria that live in the gut of termites, and from fungi that feed on dead and dying trees. Once isolated, ever-more efficient enzymes can be created by randomly altering the gene of each enzyme and selecting the best performing variants. This work is also well advanced.

Notably, the process Dr. Keasling and others are using to create new fuels is similar to the one they used to create artemisinin. In fact, the several of the initial enzymatic steps are identical for both processes. Prototype strains of yeast capable of producing biofuels have already been made. Some of these new compounds have proven to be highly effective fuel additives, increasing the efficiency of diesel fuels.

Constructive biology has come of age. It represents a next logical step in biotechnology. The early progress of biotechnology was made possible by splicing a single gene of one organism--the gene for human insulin, for example--into the genome of a much simpler organism such as a bacterium or yeast. Dr. Keasling has demonstrated the practicality and utility of recreating entire complex biochemical pathways in organisms that are easy to grow in large batches. These advances open the possibility that a wide variety of useful chemicals, heretofore available only from rare natural sources, can be made in virtually limitless amounts. Constructive biology not only opens the door to entirely new pharmacopeias but may provide the means to build a fossil fuel free carbon economy.