CoPIs: John Doebley (University of Wisconsin - Madison), Sherry Flint-Garcia and Michael D. McMullen (University of Missouri - Columbia/USDA-ARS), James Holland (North Carolina State University/USDA-ARS), Stephen Kresovich and Qi Sun (Cornell University)
Senior Personnel: Jeffrey C. Glaubitz and Theresa Fulton (Cornell University)
Genetic architecture is the constellation of gene effects and interactions that underlie variation in a quantitative trait. Essentially, genetic architecture is the map between phenotype and genotype. Understanding variation in genetic architecture is key to understanding evolution, manipulating species for a sustainable agriculture, and preserving variation as species adapt. This project will improve our understanding of the genetic architecture of complex traits in maize and its wild relative, teosinte. Maize has a combination of life history, economic and societal value, and genetic tools that make it uniquely suited to studying genetic architecture. The project will identify genes that control domestication traits and three key agronomic traits: flowering time, plant height, and kernel quality. Genetic linkage, association, and fine mapping analyses will be performed on the largest and most diverse set of mapping families publicly available for any species. A large series of isogenic lines will be used to characterize allelic series and epistatic interactions. The genetic architecture of each of the four trait groups will be compared and contrasted, and the influence of recombination and past domestication bottlenecks on the genomic distribution of functional diversity will be examined. Finally, the ability of genetic architecture-based models to predict phenotype will be evaluated in a broad range of germplasm, including elite US hybrids. This project will take a step toward the ultimate goal of predicting phenotype from genotype.
Broader Impacts Maize has the highest production of any crop in the world, and plays a central role in all of US agriculture and food production. Maize also has the greatest molecular and phenotypic diversity among crop species. This genetic diversity enabled domestication and is key for future maize improvement. Understanding maize genetic architecture will accelerate the breeding of future crops. In addition, this project will generate valuable germplasm resources and develop genomic tools to access and utilize maize and teosinte diversity. These resources will be used by many other research groups to dissect numerous other traits and facilitate marker assisted breeding, allele mining, and genetic analysis. Project resources will be made available to the public through a project website (www.panzea.org), integration with community websites (Gramene, www.gramene.org; MaizeGDB, www.maizegdb.org), and stock centers (Maize Genetics Cooperation Stock Center, maizecoop.cropsci.uiuc.edu; CIMMYT, www.cimmyt.org; North Central Regional Plant Introduction Station). Maize is also an excellent system for teaching about evolution, genetics, and agriculture. Outreach activities will target four audiences: (1) the general public and students through a traveling museum exhibit on maize domestication, diversity and improvement, (2) high school teachers through an enrichment course with North Carolina Agriculture & Technical State University, (3) collaborative science through an African Scientist Fellowship at Cornell's Institute of Genomic Diversity (www.igd.cornell.edu), and (4) undergraduate students through mentoring and research opportunities.
This project has helped to position maize (corn), the most economically important of U.S. crops, at the forefront of modern approaches to crop breeding and improvement, and at the cutting edge for studying evolution, artificial selection, and the relationship between genotype (DNA sequence) and phenotype (measurable traits). Future crop breeding relies on mastery and synergy of groundbreaking developments in four key areas: (1) high throughput genotyping, (2) high throughput phenotyping, (3) computational tools to handle "big data", and (4) sophisticated and efficient statistical algorithms to analyze those data. This project has made significant contributions to all of these areas. We developed a streamlined laboratory protocol, dubbed "Genotyping by Sequencing" (GBS) that harnesses the incredible power of modern DNA sequencing technologies to quickly and cost-effectively genotype large numbers of individuals at high density. In collaboration with many other maize researchers, we applied GBS to more than 50,000 maize plants, encompassing nearly all of the important maize lines from around the world. This unification of diverse maize experimental and breeding populations through a common genotyping method has enabled powerful new analyses to be conducted at unprecedented scale. In order to store, manipulate, and make sense out of such large-scale genotypic and phenotypic data, we have provided optimized and efficient computational tools and statistical algorithms in TASSEL and GAPIT which are open-source, freely available software programs commonly used by geneticists around the world. A major scientific accomplishment of this project was the development of an integrated approach, and an accompanying set of community resources, used by numerous maize researchers around the world to pinpoint the key genes underlying economically and agriculturally important traits (e.g., yield, height, flowering time, nutrient levels, and disease resistance). We have employed this integrated approach to study the "genetic architecture" of maize and its wild ancestor, teosinte. Genetic architecture refers to (1) the number of genes affecting traits and the number of variants at each of those genes, (2) the degree to which those genes interact with each other, and (3) the nature of the DNA sequence variation underlying differences in traits. The DNA sequence variation may be either within genes or intergenic, and may consist of either single "letter" (nucleotide) changes or presence/absence (cutting and pasting) of larger chunks of DNA. We found that most of the economically and agriculturally important traits in maize are controlled by large numbers of genes, each of small effect. However, for domestication traits that contrast markedly between maize and teosinte, single genes with strong effects are more common. We uncovered more than 1000 segments of the maize genome that carry evidence of their own involvement in the domestication of maize from teosinte, and in the subsequent breeding and improvement of maize into the highly productive modern varieties grown today. This tells us which genes have historically been the most important for agricultural productivity. Encouragingly, for many important maize traits, we are able to accurately predict phenotype from genotype without accounting for epistasis (the myriad of interactions between genes). This makes it easier to use these predictive models to accelerate breeding. Our results also indicate that intergenic DNA sequence variation, through its ability to regulate gene expression, has disproportionate influence on trait variation: a DNA variant outside of a gene stands a somewhat greater chance of being statistically associated with trait variation then a variant inside of a gene. Similarly, presence/absence variation, common in the genome of maize and many other crop species, influences trait variation more than would be expected given its abundance relative to single nucleotide changes. With abundant diversity present in or near most maize genes, future acceleration of the pace of breeding will depend on our ability to identify the handful of "broken" (deleterious) genes present in each elite variety, and to remove them from breeding populations. Our work has shown that these deleterious genes tend to be concentrated in chromosomal regions where recombination (gene shuffling) is suppressed. Exploitation of the knowledge gained from this project is leading to more efficient breeding, not only for maize, but other crops species as well. In addition, our work has influenced how genetic studies are conducted in many other organisms, from fruit flies to pine trees. This project published 80 articles in peer reviewed scientific journals, which have been cited a total of 4,771 times. To improve public awareness of how the power of genetics can be used to (1) better understand the origin of the crop species grown today, and (2) accelerate the future improvement of these crops in the face of rapid global change, we constructed a mobile museum exhibit entitled Maize: Mysteries of an Ancient Grain. This travelling exhibit has been highly successful, as it has been viewed by more than 250,000 people to date.
