Strength in Diversity

MU’s plant geneticists are harnessing the power of maize’s genetic diversity. By Melody Kroll and Charles E. Reineke. Photos by Rob Hill.

Just a few months after Columbus moored his caravels on the island of Hispaniola, the folks back home began hearing about an amazing new food discovery. Pedro Martyr de Angleria, an Italian teacher whose work in Spain gave him access to news of the voyages, recorded perhaps the oldest extant account of the buzz.

“The islanders make bread with a kind of millet, similar to that which exists plenteously among the Milanese and Andalusians,” Martyr de Angleria wrote in a letter to the Vatican dated 1493. “The millet is a little more than a palm in length, ending in a point. The grains are about the form and size of peas. When ground they are whiter than snow. This kind of grain is called maiz.”

Little could he have known, but by the time Pedro Martyr put ink to paper, “maiz” was already among the most important foodstuffs in the world—albeit a world unknown to Europeans.

The beginnings of maize, or corn, are shrouded in myth and fable. But genetic investigations by plant scientists have determined that the indigenous people of central Mexico probably began cultivating it around 6,000 BC at a site near the Nevado de Toluca volcano. These stalks, relatives of a domesticated grass called teosinte, were carefully sown and reaped for close to 7,500 years before Columbus was born.

Early farmers soon learned what we take for granted today: that maize is a miraculously diverse plant, its ability to change and adapt nearly unlimited. Mesoamericans may not have understood the science behind maize diversity, but they knew a gift from the gods when they saw one. With time, patience, and careful cultivation, they coaxed maize into growing productively in a variety of soil and climate conditions.

Reliable sources of sustenance can give rise to great civilizations, and so it happened in Mexico and Central America. Little wonder that the Maya and their successors believed the gods created humanity, quite literally, from a tasty paste of water and ground corn.

King of Comestibles

Maize has never surrendered its place as the king of New World comestibles, nor has interest abated in tinkering with its myriad forms. Today, in fact, few would dispute the claim that maize is one of the most-studied plants on the planet. Much of that study is geared, like the attentions of the ancients, toward understanding how traits from the bewildering diversity of maize varieties might be employed to create healthier, more productive plants.

A team of maize geneticists with MU’s Plant Genetics Unit in the U.S. Department of Agriculture’s Agricultural Research Service, ARS, are among those investigators. The team’s most recent work, published in a series of articles in the journal Science, is the culmination of a multiyear, multi-institutional investigation aimed at determining how various maize genes work in tandem to influence flowering and, by extension, other complex traits.

“We picked flowering time, among other traits, because it is of critical importance to this history of maize,” says Michael McMullen, who leads the MU-based ARS team. An important aspect of that history, biologically speaking, is bound up in how much light it needs to flower and set seeds. Since maize began in the tropics, McMullen explains, early varieties were “short-day” plants, meaning they flowered only as nights grew longer. As people dispersed maize farther and farther north, it had to evolve away from this short-day restriction. The maize we see today is mostly “day-neutral,” meaning it doesn’t need variations in light to flower.

“We’re interested in flowering time because we’d like to know what genes were selected to get from a short-day plant to a day-neutral plant. This was critical in making maize into the crop it is today, as grown in Argentina and Iowa,” continues McMullen, who is also an MU adjunct professor of plant sciences.

Before the advent of contemporary science, people cultivated maize by selecting the best plants among those growing in open-pollinated fields. As farming moved north, people brought with them varieties with traits that grew best in the new environments. Among those traits were maize varieties that flowered earlier in the year and under longer-day conditions. Over the centuries, this deliberate selection was extraordinarily effective in helping maize cultivation to span the globe.

In the early 20th century, pioneering geneticists such as E.M. East and George Shull observed that crossing two “inbred” lines of genetically related field corn resulted in “heterosis,” or hybrid vigor. The hybrids that resulted from these crosses out-yielded conventional cultivars and led to an explosion in corn production that continues to this day. In terms of flowering, this selective form of breeding also allowed farmers to tailor maize to local conditions, so that “the corn you see in Minnesota has a different maturity than the corn you see growing in Missouri,” says McMullen.

This greater specificity is both good and bad. While it has made farming more predictable, it has also locked maize out of its diverse heritage. “Breeders avoid bringing in exotic germplasm to broaden their base,” McMullen says, “because they have to ‘rebuild’ the flowering time and other complex traits.”

The biggest problem for geneticists and breeders has been getting a handle on these all-important complex, or quantitative, traits—like flowering time—that involve a continuous spectrum of gene interactions and environmental influences. The reason, McMullen says, is that technological limitations restricted scientists’ ability to conduct anything other than single experiments on single genes.

The advent of more powerful computers changed that. Genomics and advanced statistical capabilities, McMullen says, meant “people could stop doing all this one-at-a-time stuff and go back to thinking about complex traits, such as plant disease,” he says.

“[It was] a paradigm change on how one approaches problems, from a single experiment, where the focus was on just one gene at a time, to going back to thinking about how genes interact with biological processes.”

Next came the ability to identify qualitative trait loci, or QTLs in the parlance of geneticists. This is where the genomic rubber hits the phenotypic road: by identifying the stretches of DNA that most closely correspond to an observable characteristic, researchers can zero in on the genetic basis for specific traits. These might include, for instance, a particularly plump ear of corn or a leaf that resists blight.

“Most traits of agricultural value, for example milk production in dairy cattle, yield in any crop, fiber quality in sheep or cotton, are quantitative traits,” McMullen says. “You just can’t go into a herd of cattle and look at them and say, ‘That’s my best, that cow is going to give me more milk than anybody else; that’s the one I’m going to breed for the next generation,’ because you have this continuum.”

“About 20 years ago, we got the tools, in the sense of molecular markers and statistical approaches, to do basic QTL analysis,” McMullen adds. “This was done by taking two, in the case of plants, inbred lines and crossing them together. That made a small population. You could then plant that population, measure a trait of interest, genotype the individual, and then run an analysis. Then you started all over again.” This method yielded, at most, a few QTLs for a complex trait, and only those traits associated with genes that had large effects.

Big Science

McMullen and his team, which includes ARS colleague and fellow MU adjunct assistant professor Sherry Flint-Garcia, joined up with fellow ARS colleagues Ed Buckler of the Robert W. Holley Center for Agriculture and Health in Ithaca, N.Y., and Jim Holland at the ARS Plant Science Research Unit in Raleigh, N.C., to develop a more efficient way forward. The technique they’ve come up with is a specially created maize population, called the maize “nested association mapping,” or NAM for short.

NAM is an attempt to combine the strengths of two powerful genetic tools: linkage analysis, showing the relationships of trait-producing alleles along the chromosome, and association mapping, a way to link genetic markers and quantitative traits. The goal, for McMullen’s maize project, is nothing short of uncovering the ‘genetic architecture’ of any complex trait in maize.

The National Science Foundation-funded project involves more than just intellectual heavy lifting. All together, the researchers planted and visually assessed close to one million maize plants. No published study of plants has ever been larger. “It is big,” McMullen says with a laugh. He keeps an aerial photo of his local NAM field on his computer desktop.

McMullen and his colleagues created the NAM population for the study by crossing 25 diverse corn lines, called founder lines, with a common parent, B73, one of the most widely used commercial lines and the first maize variety to have its entire genome sequenced.

The 25 founder lines, McMullen notes, were selected by a computer. “We didn’t include a line because it had good disease resistance or because it was our favorite. By using a computer, our goal was to capture the widest possible diversity to maximize variation for two things: genes and traits.”

The maize B73 reference genome

The maize B73 reference genome, in all its colorful complexity, as published in the November 20, 2009 edition of Science. The concentric circles show aspects of the genome, including its chromosome structure, linkages across the genome (as depicted on 6363 genetically and physically mapped markers), insertions of the so-called mutant genes (used by scientists to develop more desirable traits), and other aspects of B73 genetic structure. The image and caption information were provided courtesy of the American Association for the Advancement of Science.

After crossing the founders to B73, the offspring were then self-pollinated under controlled conditions, otherwise known as “inbreeding,” for five generations. This created 200 recombinant inbred lines for each founder, or 5,000 lines in total.

Each of the 5,000 lines was then genotyped, the process by which scientists determine the genetic composition of an individual, to create a high-resolution genetic map. This allowed the scientists, for the first time, to identify all the loci, or stretches of DNA, that influence a quantitative trait in maize, including genes with small effects. They could also say, with high precision, where these loci are located in the genome.

Finally, the 5,000 NAM lines were planted in four locations: Missouri, Illinois, North Carolina and New York. “This was done to control for environmental effects,” McMullen says, referring to variations in soil or weather conditions. The researchers then collected flowering-time data from each of the 5,000 lines at each site over two growing seasons.

A ‘Major Advance’

The findings, published in two papers appearing in the August 7, 2009, issue of Science, showed that flowering time is influenced by the combined effects of more than 40 genes and that, contrary to what the scientists thought, most genetic effects are small, affecting flowering time by a day or less.

Patrick Schnable, director of the Center for Plant Genomics and a professor of agronomy at Iowa State University, was the lead author on the recent publication of the B73 maize genome, a landmark achievement in plant biology. “I think we all expected there would be a lot of genes that control for the [flowering] trait,” Schnable says. “It was important to do this experiment, however, to get it documented. I think one of the nice findings they came up with is that there are alleles out there in the NAM population—the same gene—that lengthen the flowering time or shorten the flowering time. So the alleles in maize for these important traits, like flowering time, can either make it better or worse. There’s a lot of diversity out there, and NAM highlights that.”

Harnessing this genetic diversity and using it to improve these important traits is, of course, key to what the NAM technique offers for the future. It’s that future that’s generating excitement among scientists.

Mel Oliver is a supervisory research geneticist for the USDA-ARS Plant Genetics Unit at MU and adjunct professor of plant sciences. He calls the maize NAM population “a big leap forward” for genetics.

“Put it this way: in genetics, certainly in plant genetics and to some extent in the animal world, the ’holy grail’ has always been that you want to take a phenotype and go straight from that phenotype to the gene,” Oliver says. “And the faster you can do it, the better. By doing what they did in maize, it allows them to make the QTLs much, much narrower and they can quickly go to a single gene. That’s the theory. And it looks like it’s working. It’s really exciting. It’s the major advance in plant genetics in my lifetime. It is that huge.”

Schnable too, says the group’s work could be a game changer. Sequencing the B73 genome is less like uncovering a maize instruction manual—an analogy he says he has often used in the past—than like discovering a wiring diagram, he says. That’s where NAM comes in.

“If I showed you a wiring diagram for my iPhone, it wouldn’t help you very much in terms of understanding how to use the iPhone. It would be a step, but it wouldn’t be as useful as a user manual. So the NAM population really is a key step between the DNA sequence and actually identifying what that sequence does: how it is that specific genes that were discovered during the genome sequencing project, how is it that those genes control growth and development, and which genes control the expression of traits that we care about.”

With NAM as a user manual, growers can easily rebuild their lines after introducing a new trait. For flowering time, McMullen explains, “once you know the 40 genes and you know, for each of the lines, what allele you have, then it’s just a matter of taking those allelic effects—be they zero, plus a half-a-day, or minus half-a-day—and adding them up.”

Schnable goes on to call the NAM project “both a creative step and clearly a very extraordinary and valuable resource.” And not only for maize geneticists. “The information we learn from maize will benefit people who work on other plants, and probably animals,” he says.

Maybe even the human animal, McMullen adds. “We may be coming to the same model as human geneticists studying disease. Humans display a continuous spectrum in disease susceptibility, such as developing diabetes or cardiovascular disease, and the alleles that cause large effects and severe susceptibility are quite rare.”

At the moment, however, McMullen and his team are content to explore NAM’s potential for more accurately predicting phenotypes in maize, a development that would reopen to growers maize’s vast diversity.

“That’s been the promise for a few decades,” he says. “It may take a few more decades. But I think there’s promise.”

Back to Top

University of Missouri

Published by the Office of Research

© 2017 The Curators of the University of Missouri