A supercomputer to design better plants?
Is it possible to create more productive crops than nature does without growing hybrids or genetically modified plants? According to researchers at University of Illinois at Urbana-Champaign (UIUC), the answer is yes . They've simulated photosynthesis, the process by which plants convert light to energy, with the help of supercomputers at the National Center for Supercomputing Applications (NCSA). Their models suggest that soybean productivity could be increased by 40 to 60 percent. They're also working with wheat and rice and expect that biotechnology companies will use this research to design more productive plants without altering their genes.
In order to better calibrate their computer model, which simulates the photosynthetic behavior of actual leaves, you can see above a gas exchange system used by the researchers to measure the rate of carbon dioxide and electron transport in real leaves. (Credit: Don Hamerman/UIUC). Here is a link to a a larger version of this photo.
This research project was led by Steve Long, a professor of plant biology and crop sciences at the University of Illinois. Long also is the deputy director of the Energy Biosciences Institute and an affiliate of the Institute for Genomic Biology at UIUC. Long was helped by Xin-Guang Zhu, a research scientist in his lab who co-developed the computer model, and by Eric de Sturler, a former associate professor of computer science at UIUC who recently moved to the Department of Mathematics at Virginia Tech.
Let's start with what Long says about photosynthesis. "Photosynthesis, the process by which plants convert light to energy to power their growth and make food we eat, is the least explored path to improving plant productivity, Long said recently. 'Photosynthesis has really been untouched,' added Long, a UI plant biologist and crop sciences professor. 'There's been almost no improvement.'"
But why? "In part, that's because of the difficulty of doing anything to improve a process happening on a molecular level and driven by genes and proteins, absent recent advances in the science and technology needed to understand and work with things at that scale. Now, we know the genes involved in producing the more than 100 proteins that drive photosynthesis, opening the door for molecular biologists to manipulate the molecules advantageously, Long said."
This is why the researchers decided to use computers to design a better leaf (UIUC, November 9, 2007). "The researchers first had to build a reliable model of photosynthesis, one that would accurately mimic the photosynthetic response to changes in the environment. [...] After determining the relative abundance of each of the proteins involved in photosynthesis, the researchers created a series of linked differential equations, each mimicking a single photosynthetic step. The team tested and adjusted the model until it successfully predicted the outcome of experiments conducted on real leaves, including their dynamic response to environmental variation."
Now, let's go back to the first UIUC article to learn more about the computer model. "The computer model, and the computing power at the NCSA, allows the researchers to test lots of combinations simultaneously and quickly, particularly when compared with the time it takes to grow a new plant. The result is, in effect, a better plant design, where photosynthesis is concerned anyway. 'The supercomputer can work through those permutations in hours,' Long said. 'It can pick out which are the best targets for the experimental biologists. We've given them a guide. Maybe these five proteins are the ones where you get a big return.' The researchers also had the computer simulate a variety of environmental conditions, such as the higher carbon dioxide levels expected by scientists in the future.
This research work has been published in Plant Physiology under the name "Optimizing the Distribution of Resources between Enzymes of Carbon Metabolism Can Dramatically Increase Photosynthetic Rate: A Numerical Simulation Using an Evolutionary Algorithm" (Volume 145, Number 2, Pages 513-526, October 2007). Here are two links to the abstract and to the full paper (PDF format, 14 pages, Open Access article). But read it at your own risk. If you don't have a Ph.D. in Live Sciences, I doubt you'll follow. Personally, I couldn't.
Sources: Greg Kline, Urbana/Champaign News-Gazette, November 25, 2007; and various websites
You'll find related stories by following the links below.