Aug. 6, 2009

Maize findings could lead to vigorous new varieties and insights into human genetics

Two new large-scale studies by researchers at Cornell and the U.S. Department of Agriculture (USDA), published in this week's Science (Aug. 7) journal, report major discoveries in maize genetics that could revolutionize maize breeding and may help researchers better predict complex traits in humans.

In a study using more than 1 million maize plants to identify the genes involved in flowering time, the researchers found that the trait is influenced by the combined effects of more than 40 genes. Flowering time influences whether a plant can adapt to new environments and is the main hindrance to exchanging crops internationally.

The study is relevant to humans, because such traits as height, for example, are likely to also involve many genes.

"We looked as hard as we could for big genes and big effects, but they don't exist," said Ed Buckler, the lead author of the maize flowering time paper, a USDA-Agricultural Research Station (ARS) research geneticist in Cornell's Institute for Genomic Diversity and an adjunct professor in plant breeding and genetics.

"Instead, there are lots of smaller genes around the genome that contribute to flowering time," Buckler added. "What we see in maize is probably what is going on in humans as well."

Buckler and Stephen Kresovich, Cornell's vice provost for life sciences and a professor of plant breeding and plant biology, were lead co-authors of both papers, along with USDA-ARS research geneticists James Holland at North Carolina State University and Michael McMullen at the University of Missouri. Researchers from many other institutions also contributed to both studies.

While some genes are known to interact with other genes with unexpected results, the researchers noted that the effects of the genes they studied all added together to influence maize-flowering time, which can vary by months in the 5,000 maize varieties that were crossed in the study.

Using a computer model based on the additive effects of these genes, the researchers predicted with 93 percent accuracy when a maize plant would flower. The research offers promise for designing new maize varieties with computers and then predicting their flowering time in the field, Buckler said.

In human genetics, researchers can only explain 4 percent or 5 percent of variation, but the researchers believe the new maize model may help geneticists better predict the effects of genetic variations in humans.

In the second large-scale study by the same group, also in this week's Science, the researchers uncovered for the first time an important pattern in gene recombination (the ability to shuffle genetic variation), where large parts of the genome fail to recombine near the center of a hybrid corn's chromosome. Known as the centromere, this area looks like a knot between two strands of yarn tied together near the middle.

Over the last 100 years, breeders have created vigorous and diverse hybrids that have increased U.S. yields eightfold. But after crossing 25 diverse lines of maize and producing 136,000 recombinations since 2001, the researchers uncovered a consistent pattern that genes near the centromeres are often in arrangements that produce less vigorous plants. The researchers believe this pattern may contribute to the survival and vigor of hybrids, but that it has also prevented breeders from arriving at optimal genetic combinations.

"Now, combining gene variants near the centromere will be a future direction for plant breeders," said Buckler. "Breeding schemes designed to exploit this knowledge could accelerate plant breeding worldwide."

For this study, the researchers created 5,000 different varieties of maize, each capturing a wide range of trait variation. The seeds are available to the public.

The studies were funded by the National Science Foundation and the USDA-ARS.