Computable Plant Increases Scientific Understanding
Team members Pawel Krupinski and Todd Johnson
meet with Mjolsness to discuss results.
Irvine, Calif., Nov. 15, 2007 -- It isn’t particularly fashionable or trendy, but it’s a model nevertheless. A computer model, that is.
Arabidopsis, a small flowering plant in the mustard family, is the diva of a computer modeling experiment that will help scientists better understand how and why plants grow the way they do. This insight can lead to advances in global sustainability, ecology, agriculture and energy.
Principal investigator Eric Mjolsness, a UC Irvine Calit2-affiliated computer scientist, and co-PI Elliott Meyerowitz, a Caltech plant biologist, are collaborating on the “Computable Plant,” a $5 million NSF Frontiers in Biological Research grant that combines computing, microscopy and molecular biology. They are developing a mathematically-based computer model of Arabidopsis’ growth cycle, specifically its arrangement of leaves – known as phyllotaxis – and its shoot meristem, the sets of stem cells at the tip of each shoot, which ultimately provide the cells that make the stem, leaves and flowers.
Biology Meets Computing
Meyerowitz’ team captures live images of the plant’s leaf growth with a confocal laser scanning microscope.
Scientists "mesh" the images of the plant meristem to define
simpler geometric units that are easier to manipulate in computing.
Mjolsness and his UCI colleagues convert the biological processes to a mathematical format by designing a series of computer models to reflect the plant’s cell growth, cell communication and molecular regulation. The models are visually brought to life on a computer, allowing the scientists to gain insight into the plant’s growth process.
The team also designs algorithms to quantify those images. “Image analysis algorithms allow us to see where the growth is happening, where it is slow and where it is fast,” Mjolsness says, adding that compared to human analysis, a mathematical approach saves hundreds of hours.
Finally, Mjolsness and his collaborators compare the computer models with time-lapse-microscope movies of actual plant development to see if their theories check out.
“Modeling allows biologists to make hypotheses about complicated systems containing many cells and interacting molecules,” says Mjolsness. “You can run mathematical models forward and see what would happen if your hypotheses were true. Then you can compare that to what you’re actually observing.”
In particular, the collaborators are developing a model based on the movement of the growth hormone auxin, which influences how the plant’s shoots, roots and leaves are patterned.
Because placement of the plant’s organ may be influenced by localized concentrations of the hormone, it’s important to understand how the auxin moves between cells.
New research suggests that auxin can manipulate its own movement by regulating which cell membranes contain the necessary proteins to direct it. If scientists can confirm this “autoregulated transport,” it would offer a new look at the workings of cell communication in biological development.
Mjolsness says his group wants to understand in greater detail the plant forms and structures that are visible to the eye. Why do many plants have a spiral pattern of flowers, leaves and branches? Why do the flowers grow in specific locations on the plant?
“We can explain the morphology in terms of molecules, genes and proteins that we know. But if you change one of these components, everything about the morphology could change,” he says. “We are trying to understand why.”
The researchers have noticed that plant tissue makes room for new floral buds by moving the older primordia, which compete for auxin, out of the way. This helps to explain why new leaves at the shoot tip occur as far as possible from old ones.
Understanding spatial patterning in plants can set scientists on the path to solving universal problems. “If you look at the terrestrial biosphere, almost all of it is plant life,” Mjolsness says. “All the questions that arise from this biosphere will benefit from an understanding of the way plants give rise to their adult form.”
The associate professor of information and computer science was originally a physicist. He pursued research in computer science to help him solve complex physics problems computationally.
Now he’s applying his expertise to the plant world. “The ability to reengineer the architecture of plants by controlling their basic spatial patterning mechanisms could be important in redesigning plants for energy, food or growth in altered environments,” he says.
After that, who knows? Says Mjolsness: “We’re finding that image analysis can be used for fish and flies, bacteria and more, so there’s no reason to stop with plants.”