Stanford News


CONTACT: Janet Basu, News Service (415) 723-7582;
Patricia Craig, Carnegie Institution News Office
(202) 745-0780; e-mail

Genes of a "useless weed" reveal clues to plant biology, with a possible bonus: a plastic potato?

The first time Chris Somerville mentioned the plastic potato, it was strictly as a joke.

A reporter asked what practical use could come from mapping out the genome of a plant. "Better crops," Somerville replied, and he chose an off-the-wall example. "Maybe we'll grow a plastic potato, to go with the plastic tomatoes we already get in the supermarket."

That was in 1989. Three years later, Somerville's lab borrowed genes from a bacterium and taught a modest weed called Arabidopsis thaliana to grow beads of plastic in its seeds. Suddenly it was not such a wild idea to imagine a field of lush green plants growing industrial grade, biodegradable plastic. Monsanto Corp. has assigned a research and development team to try to make this technology work in a viable crop, perhaps a high-yield plastic soybean.

Meanwhile, since 1994, Somerville has been a Stanford professor of biological sciences and director of the Carnegie Institution of Washington's Department of Plant Biology, an independent research organization located on the university campus. Working with graduate students from Stanford and postdoctoral fellows from around the world, he tackles basic biological questions that sometimes have an angle to them as wild ­ possibly as wildly practical ­ as a plastic plant.

Take the oil potato, for example. This month, postdoctoral fellow Joe Ogas of Somerville's lab published research showing that Arabidopsis roots can be genetically instructed to do something that roots don't normally do: produce vegetable oil instead of starch. It may be a long step from this to an oil-producing root crop, perhaps a potato or a monounsaturated sugar beet. But from there, persuading the root crop to produce plastic should be easy. Someday, the punchline to Somerville's joke could be a real plastic potato.

Others in his lab are working on ways to instruct Arabidopsis to make more nutritious vegetable oils; to grow renewable raw materials for coatings and lubricants; to grow a natural precursor of nylon. There's an environmental angle to many of their pursuits: One group is trying to teach the plant to make extra cellulose in its cell walls. That could be a first step toward tree plantations that would need half as much land as usual to supply the demand for pulp in paper mills.

All of these projects are just spin-offs of Somerville's real passion, the work that earned him election last year to the exclusive National Academy of Sciences. He is on a quest to learn as much as possible about plants by learning about their genes. For nearly 20 years, he has championed Arabidopsis thaliana, a small, unprepossessing cousin of the mustard plant, as the laboratory mouse of the plant kingdom, the organism that scientists will study to understand the genes of all flowering plants.

Growing a genome database

That effort is now a big success. Thousands of scientists study the little plant, and hundreds of labs worldwide are working on the Arabidopsis thaliana Multinational Genome Project, mapping out the sequence of 100 million base pairs, the rungs on the ladder that make up the 20,000 to 25,000 genes in Arabidopsis DNA. Stanford's team of genome sequencers, led by biochemistry Professor Ron Davis, are leaders in this effort.

Somerville said that thanks to these scientists, Arabidopsis may be the second organism (after yeast) for which scientists learn the function of all the genes. "When we have the whole sequence, we'll be able to knock out genes one at a time and find out what they do," he said.

The value of seeking those genes is already becoming obvious. Flowering plants began to evolve only about 150 million years ago. All are closely related, and the order of genes in their DNA is similar. "That means anything we can do in Arabidopsis, we can do in almost any plant," Somerville said. This is why his students can work on cellulose in the lab, confident that what they learn may have a real application in trees. Already some of the knowledge has gone beyond plants: A gene that controls the chemical composition that makes cell walls flexible in plants led to the discovery of a similar gene at work in human cells.

James Watson became a booster for the Arabidopsis genome project because of that human link. Watson, the Nobel laureate and co-discoverer of the spiral structure of DNA, is the godfather of the Human Genome Project. In 1989, he joined Somerville and Elliott Meyerowitz of Caltech in persuading the National Science Foundation to fund the plant genome project, after it had been turned down by other agencies. Watson argued that there must be good genetic maps of many organisms, from yeast to plants to mice, if scientists are ever to learn what human genes do. When a similar gene, called a homolog, is found in humans and in plants, scientists can test the plant in ways they could never use to experiment upon a person.

Until the genome project officially began, Somerville, Meyerowitz and a handful of other scientists seemed to be the only ones who saw the potential of their little mustard plant. Many botanists preferred to continue their work on corn, peas and other organisms that have been studied for more than a century. The U.S. Department of Agriculture found no way they could justify Arabidopsis genome studies to their constituency among farmers.

"They called it a useless weed," Somerville said.

Useless weed makes good

"So you ask, why Arabidopsis?" Joe Ogas said he hears this question often ­ why not study something directly useful like peas? In the Carnegie Institution greenhouse, Ogas opened a lab appliance that looks like a tall refrigerator but is really just the opposite ­ a warm, lighted incubator to promote growth. He picked up a palm-sized agar plate, a transparent square holding 36 tiny seedlings, each with four bright green leaves and a wispy network of nearly transparent roots. On a nearby lab bench, he showed how the seedlings look when full-grown: Each compact cluster of foliage is small enough to fit in the two-inch-square plastic pots that nurseries use to sell starter plants. The plant's life cycle, from the day a pinpoint-sized seed is planted in agar to the day when new seeds can be harvested and planted, takes less than three months.

Those are the factors that make this an ideal lab plant: its size, its quick life-cycle, and the fact that it has a relatively small, easy to study genome. Coincidentally, one common name for this lab mouse of plant genetics is "mouse-ear cress."

Ogas was screening thousands of plants when he happened to notice the mutation that led to his discovery of roots that grow oil. Then he grew thousands of genetic copies of the mutant to figure out what caused the phenomenon. "Imagine doing that with tomatoes or peas," he said. "You would need lots of land."

A 1986 Stanford alum in chemistry who did his graduate work at the University of California-San Francisco in yeast genetics, Ogas was drawn to Somerville's lab in part by what he describes as "Chris's strong vision, his belief that basic research can make fundamental advances in agriculture." He also was drawn by the chance to answer some fundamental questions about how cells choose and maintain an identity. Plants are a good model to study this because many plants regenerate easily, forming roots, leaves, flowers and seeds from a cutting of a leaf or a twig. In effect, a plant cell can change its identity from one type to another, even from an adult stage.

Screening the roots of seedlings under a microscope, Ogas found an example of mistaken identity. The central taproots of some plants stopped branching out with smaller roots. Instead, the taproots became thick, opaque and bright green. They looked like miniature kosher pickles, and the mutation that caused them was promptly dubbed "pickle."

Colleagues at Berkeley independently found the same mutation, and noticed that the oddly shaped roots packed in extra starch. Ogas found that the pickle root tissues also produced lots of oil ­ unheard-of for this type of root. And he found out why.

The seedlings with "pickle" roots are defective in their response to a hormone, gibberellin, that normally instructs the plants to change from one developmental stage to another. In this case, the plant began to change from an embryo inside a seed into a seedling. But the taproot reverted to its previous identity, as if it were still a cluster of cells inside a seed. The confused root cells began to act like embryonic seed cells and started producing oil.

Ogas and his Berkeley colleagues published their discovery in the July 4 issue of Science. Now Ogas is looking for the gene that normally responds to gibberellin. Eventually, he says, the trail of this search could lead to the genes that instruct seeds to produce oil. Thanks to the serendipitous discovery of "pickle," scientists know that plants can survive if those oil-producing instructions are expressed in their roots. Eventually, Ogas or some other scientist will try to use the instructions to grow an oil-producing potato.

Even if such a crop never becomes practical, however, the discovery should allow Ogas to figure out the cascade of genetic instructions that tell a plant to change from one developmental state to another. And that should tell something about how cells determine their identity, in plants and other organisms, including humans.

Basic science, real-world results

Answers to such basic questions remain the major goal of Somerville's research and his work to establish the genetics of Arabidopsis. His first work with the plant was a study conducted in the late '70s with his wife, Shauna. They isolated genes and used them to map out a basic biochemical pathway that had eluded plant biologists for years. In the process, they answered a long-standing riddle: why plants exhale carbon dioxide when they could use the carbon as a nutrient.

Somerville did not start as a plant biologist; he did his graduate work at the University of Alberta in bacterial genetics. He said that two things drew him to work on plants. One was the pioneering work of several earlier scientists who had begun mapping out the genetics of Arabidopsis; from those studies, he was convinced that the plant was a good model organism for molecular biology.

His other inspiration was Shauna Somerville, now a Carnegie plant biologist studying how plants resist disease. Her work convinced him that research on plants is a productive way for a laboratory scientist to contribute to human problems like hunger and environmental degradation. "I think of this as molecular environmental science," he said.

So the oil-producing potato is an attractive idea, though Somerville and Ogas admit that there is a long way from a laboratory discovery to a useful crop. Potatoes, sugar beets and other root crops yield far more plant matter than above-ground crops like soybeans. It is easy to dream of a high-yield crop of oil potatoes enriching a village. It is easy to imagine the potato converting that oil to plastic ­ an improvement on methods for making plastic from petroleum but also an improvement on the plastic seed crop, an under