Statement of Dr. Brian A. Larkins, Harry W. and Elsie M. Porterfield Professor of Plant Sciences, University of Arizona

Before the Senate Committee on Agriculture, Nutrition and Forestry

October 6, 1999

Thank you Mr. Chairman for the invitation to appear today before the Committee. My name is Brian Larkins, and I am a professor in the Department of Plant Sciences at the University of Arizona. I am a member of the National Academy of Sciences, and for the past year I have served as President of the American Society of Plant Physiologists, a professional society of nearly 6000 scientists who do fundamental and applied research on how plants grow and develop.

I appreciate being invited here today to describe my research applying genomics and molecular genetic techniques to improve the nutritional quality and milling properties of corn. As you may be aware, cereal grains, such as corn, provide 50% of the dietary protein for humans and can comprise 70% of the protein intake for people in developing countries. In the US, the major cereal grain produced is corn, and most of this is fed to livestock. The protein in corn has nutritional limitations for humans, but these also apply to several types of livestock, especially swine and poultry.

Typically, the protein in corn seeds contains around 2% lysine, while we require 5% lysine in our diet to avoid protein deficiency disorders. Globally, nearly 195 million children younger than five years are undernourished for protein, and in 1992 an estimated 12 million American children were estimated to have diets that were significantly lower in protein than what is recommended by the National Academy of Sciences. Poor nutrition leads to a number of health problems in children, including stunted growth, weakened resistance to infection and impaired intellectual development.

For many years, there has been research based on conventional breeding methods to create cereals with higher contents of essential amino acids, but these efforts have been largely unsuccessful. Currently, we deal with this problem by supplementing grain with essential amino acids produced by bacterial fermentation. Although this approach works well for feeding animals, it is expensive. The approximate annual cost of amino acid supplements for livestock feed is around $300-$400M. Furthermore, amino acids are lost from foods processed from corn meal, such as tortillas, and for this and other reasons it is valuable to have the essential amino acids incorporated into proteins.

In the early 1960s, researchers at Purdue University discovered a mutation in corn called opaque 2 that significantly increases the lysine content. opaque2 corn was shown to not only provide a superior source of protein for swine and poultry, but to also overcome kwashiorkor, the common protein deficiency disease in children. There was great optimism following the discovery of the so-called "high lysine" corn, because many people believed it would lead to the development of more nutritious cereals. However, the soft, starchy nature of opaque2 seeds prevented the development of this mutant into agronomically useful cultivars, and within a few years of its discovery, breeding work with opaque2 was largely abandoned. But some plant breeders who worked with opaque2 were able to identify genes that modified the starchy seeds, creating kernels with a normal texture. While these modifiers are difficult to work with, several corn breeders were successful in creating a new type of high lysine corn called Quality Protein Maize, or QPM. QPM is comparable to normal varieties in protein content and seed quality, and it approaches the nutritional quality of typical opaque2 mutants. Currently, there are multinational efforts to create QPM genotypes that can be grown in various regions of the world. Nevertheless, further enhancement of the lysine content is required in order to meet the level recommended for young children by FAO.

My research has focussed primarily on two questions related to Quality Protein Maize. What is the origin of the increased lysine content in opaque2 mutants, and what is the mechanism by which opaque2 modifiers convert the soft, starchy kernel to a normal phenotype? By understanding the answers to these questions, we hope to increase the protein quality of the kernel, while producing a seed that performs well agronomically and also has excellent properties for shipping and handling. Most hybrid corn grown in the US is a soft, dent type, and it tends to crack and chip when shipped, reducing the value of the grain. A harder kernel also has greater value for dry millers, as it is the vitreous part of the grain that produces corn grits and corn flakes.

We used a genomics technology to identify the nature of the lysine-rich proteins synthesized in the storage cells of the seed. First, we isolated gene sequences corresponding to the most lysine-rich proteins in the developing seed, and then we determined which of these are increased in opaque2 mutants. As a result, we identified a lysine-rich protein known as elongation factor-1 (EF-1). This protein is increased two to three-fold in the opaque2 mutant. Since EF-1 contains a high percentage of lysine (10%), we reasoned that this protein could provide an index of the lysine content of the seed. Indeed, the concentration of EF-1 has an exceedingly high correlation (r= 0.9) with the grain's lysine content. Because EF-1 can be detected with an antibody, we were able to create a simple immunological assay to estimate its content. In our most recent studies, we have shown there is significant variation in EF-1 levels among corn hybrids, and that a high grain lysine content can be obtained by using EF-1 as a marker for high lysine in a conventional breeding program.

Plant breeders throughout the world have worked for more than 30 years to improve the protein quality of maize and other cereals. However, the genetic complexity of this trait, as well as the cost associated with measuring it, precluded significant progress. Using molecular genetic and genomics approaches, we were able to unravel the complex problem of the inheritance of lysine-rich proteins in corn. Furthermore, it appears our findings are applicable to other types of cereal grains, including sorghum and wheat, and thus it may be possible to generally improve the protein quality cereals through this strategy.

This work serves as an example of how plant genomics techniques can provide insight about the nature of complex genetic traits. The knowledge gained from understanding the molecular basis of such traits can be applied to crop improvement through conventional breeding programs. Alternatively, it might also prove effective to over-produce a lysine-rich protein by genetic engineering, and several laboratories are exploring this possibility. Recently, researchers at Du Pont used genetic engineering techniques to increase the lysine content of corn seeds by over-producing the free amino acid. Thus, there are multiple technical approaches to improve the nutritional value of this grain, and I believe it will be only a short time before we will be able to produce corn that has the protein quality of milk. Think of the value of this to the US, and, especially, for alleviating malnutrition and human suffering for millions of people in developing countries!

We are also using genomics techniques to investigate the patterns of gene expression that are associated with the formation of hard and soft kernels. We initiated these experiments several years ago in collaboration with researchers at Pioneer Hi-Bred International in Johnston, Iowa. I might add that we did this because Pioneer was proactive about implementing genomics techniques in their research, and we were anxious to learn to apply these approaches. Furthermore, Pioneer was interested in the studies we were doing. There are at least 12 different mutations that lead to the formation of a soft, starchy kernel. We know the molecular basis of only two of these mutations, opaque2 and floury2, and they affect two different types of genes involved in the synthesis of seed storage proteins. It is our hypothesis that by understanding the global pattern of gene expression in these mutants, we will be able to decipher a set of genes for which changes in expression underlie the phenotypic basis of kernel texture. Likewise, by monitoring the effect of opaque2 modifiers on the pattern of gene expression in these mutants, we can determine compensatory changes in gene expression that restore the normal kernel texture. These experiments are ongoing, and I cannot report their outcome today. However, we are optimistic that we will have some insights within a few months. Perhaps we will know the secret of producing better corn flakes with higher protein quality in the coming year!

In closing, I want to emphasize that the resolution of gene expression that is possible through genomics techniques, where you can simultaneously monitor the expression of thousands of genes in a given cell or tissue, provides insight into the basis of agronomic traits to a degree never before possible. It is not unlike what van Leeuwenhoek's microscope did for seeing life in a drop of water. And as was true for the microscope, improvements in genomic techniques are rapidly increasing the resolution of this technology. Eventually, these procedures will allow us to understand what were heretofore undecipherable, complex agriculturally important traits such as hybrid vigor, stress tolerance, nutrient composition and nutritional quality. This information will allow the plant breeder, either through conventional breeding or genetic engineering techniques, to create more efficient, more productive and more nutritious crops that can be grown on fewer acres and with less impact on the environment. This will clearly be advantageous to the farmer, the consumer, the country, and the world in general.