BEIJING — The windows of China's National Maize Improvement Center may look out onto the busy streets of this capital city, but for the scientists within, the focus is always on the countryside, and finding ways to use genomic research tools to improve the productivity of what has become in recent years the most important crop in China.
Last week, BioArray News visited the center at China Agricultural University, where Jinsheng Lai, a professor at the university, laid out the issues facing maize researchers in the country. Rising living standards in China have led to greater demand for animal-derived products, from meat to milk, Lai told BioArray News. That in turn has led to an increased demand for maize to feed the animals, prompting an increase in maize production that led the plant to surpass rice as the nation's most widely cultivated crop five years ago.
"Seventy percent of maize goes to animal feed," Lai said. "Compared to that, direct human consumption is very limited, perhaps similar to US consumption," he said. Yet about 10 percent of the Chinese maize crop is lost to disease each year — "maybe even more," Lai added. And, to keep up with demand, the country has been forced to import maize from the US and Argentina — including a record of 5.5 million tons projected for this year, according to the US Department of Agriculture.
In addition to losses caused by fungal and bacterial diseases, another issue affecting Chinese maize yield is climate, as about 60 percent of the arable land in China is subject to drought, according to Mingliang Xu, another professor at the center. All of these issues — increased demand for maize driven by the livestock industry, consistent losses caused by disease and drought, and reliance on imports — have led the Chinese government to invest more heavily in maize research in recent years, allowing researchers like Lai and Xu to use the best genomic research tools available to identify markers that can be used to breed drought- and disease-resistant corn.
"Right now maize is the most important crop, so the government provides a lot of support for maize research," Xu told BioArray News. "This is very good news for us."
Microarrays are Xu's technology of choice. For the past few years, he has been using two Illumina products — a custom 1,536-SNP GoldenGate assay panel and its catalog MaizeSNP50 BeadChip — to identify quantitative trait loci linked to disease resistance and to facilitate marker-assisted selection of lines for breeding.
According to Xu, the diseases affecting China's annual maize crop differ from region to region. For instance, head smut, a fungal disease that causes tumors to form on the body of the corn, is the greatest cause of crop loss in northern China. Meanwhile, in other areas of maize cultivation in China, the sugarcane and maize dwarf mosaic viruses, stalk rot, gray leaf spot, and southern leaf blight similarly impact yield.
"My main research is to identify [quantitative trait loci] that show resistance to all of these different major diseases," said Xu. "Then you can use the QTL to introduce them via marker-assisted selection into new lines."
One benefit of using array technology, Xu said, is that it has allowed him to reduce the amount of time it takes to identify and validate a marker that can be used in selection. "SNP chips have saved us a lot of time in QTL analysis," he said. "Before, when we were using microsatellites, it could take a year to look at a marker in a segregating population, so you would spend a lot of time on the analysis," he said.
So far, Xu and fellow researchers have identified and validated a number of disease-resistant QTLs, and Xu is a coauthor on several recent publications describing the QTL mapping of resistance to gray leaf spot in maize and fine-mapping a QTL for resistance to stalk rot in the crop.
But while Xu and colleagues at the National Maize Improvement Center are identifying QTLs to breed better crops, he said that most maize seed companies in China are only using array technology at the research level, and that both the cost of the technology and the structure of the seed industry are hindering the broader adoption of microarray-based selection in maize breeding.
"The problem is that the price is too high," said Xu. He said that it currently costs about $200 to assay one sample on the MaizeSNP50 and about $70 to analyze a sample using the custom GoldenGate panel. "You can imagine for a company that wants to run thousands of samples, to use this technology to check their germplasm, it would be a lot of money," Xu noted. "We sometimes tell the breeders that they can use our [GoldenGate] array to check their germplasm, but they always complain that the price is too high."
He estimated that for the Chinese maize industry to seriously consider adopting an array-based approach, the cost of an assay would have to be between a fifth and a tenth of current prices. "I think that ¥100 ($16) for one sample would be acceptable for Chinese companies," he said. "If the price was lower, they could accept the technology."
Another issue encumbering the use of arrays by Chinese maize breeders is the nature of the industry. Whereas in other countries maize cultivation is dominated by a number of large seed providers, such as Monsanto or Syngenta, in China there are many smaller breeders, which don't have the resources to use higher cost genomic tools as part of day-to-day selection programs.
"Companies are not as big as in the US, the industry is not at the same scale, so they cannot spend a lot of money on this," said Xu.
Even though most maize breeders are using array technology for research, Xu said that some partners have integrated disease-resistant QTLs his lab has identified, such as for head smut, into susceptible inbred lines.
One of the main uses of the technology, however, is to identify the origins of breeders' germplasms.
"In China, maize varieties often have a very complicated pedigree, and often the breeders don't know where the germplasm comes from, so they need to use this technology to check their germplasm to know its origin," said Xu. Such information "would be very helpful" in selecting the parental lines to make new cross hybrids and in molecular marker-assisted selection.
Genotyping by Sequencing
Arrays may be useful for QTL analysis and marker-assisted selection, but both Xu and Lai agree that the MaizeSNP50 does not offer sufficient resolution to perform association studies across the many different inbred lines that exist in China.
"Maize is very diverse across different lines and recombination is very frequent," Lai said. According to Lai, there could be "thousands" of different varieties of maize in China alone, not because of different breeding initiatives, but because de novo mutations occur rapidly in the form of SNPs, copy number variants, and insertion/deletions.
As an example of how diverse maize is, Lai noted that the difference between B73 and MO17, two of the most commonly cultivated parental lines of maize, "is probably much larger than between humans and mice."
In order to better understand the genetic makeup of such diverse lines, Lai and colleagues have turned to whole-genome sequencing. "By doing that, we get a jump start," he said. "We don't have to wait for a custom chip to become available, and at the same time the sequencing will give us more SNPs for each genome," he said. While he agreed that a 5-million-marker maize chip could provide high enough resolution to conduct similar studies, he said it "would take some time" to produce such an array. Instead, Lai and colleagues mainly rely on the Illumina HiSeq 2000 for resequencing.
"We all want to understand the genetic makeup of maize, and are trying to explore the inbred lines as much as possible," said Lai. "Right now, we are fortunate to have these tools and to be able to address these issues at a larger scale at a lower cost," he said. "We can compare the inbred lines at very high resolutions."
Lai's work has been featured in numerous publications in recent years. He coauthored several studies that appeared in Nature Genetics in June. One paper discussed the sequencing of 278 temperate maize inbred lines from different stages of breeding history, including four lines with known pedigree information; another provided a comprehensive assessment of the evolution of modern maize based on the genome-wide resequencing of 75 wild, landrace, and improved maize lines; and a third detailed the identification of 55 million SNPs in 103 lines across pre-domesticated and domesticated maize varieties.
Lai coauthored another sequencing-driven study that appeared in the same journal two years ago. In that paper, the authors discussed genetic variation among a group of six elite maize inbred lines, including the most productive commercial hybrid in China. The effort uncovered more than a million SNPs, 30,000 indel polymorphisms, and 101 low-sequence-diversity chromosomal intervals in the maize genome.
"Only within the last century did we start to breed maize with scientific guidelines, and that process has dramatically increased the yield," Lai said of these recent papers. "Eighty years ago the yield would have been less than one fifth of our current yield," he said, "but maize is still maize."
According to Lai, the main issue he and colleagues have been trying to address in these and other studies is how the population has changed in that time.
"You have older material that gives less production and newer material that gives higher production," he said. "So population genetics can show what has been changed, and that can be used to guide future maize breeding."
Lai said that the research underway at the National Maize Improvement Center and other institutions in China is "already making a huge impact" on the breeding of maize, noting that new genomic tools, such as arrays and sequencing, "make it so much easier to identify disease-resistant loci," and that pedigrees constructed via resequencing should be "directly applicable" to maize-breeding programs in China.
"I think we will have a much higher return on breeding gains," said Lai. "Traditionally, the breeding gains can increase by maybe 1 percent per year, which has been the case for the last 50 years," he said, adding that with molecular tools these gains could double.
Lower-multiplex genotyping is part of the equation, but Lai said it is unclear what technology will be most widely adopted. "Each platform has advantages and limitations," Lai said, mentioning both microarrays and Douglas Scientific's Array Tape screening tool. The Array Tape consists of a continuous polypropylene strip serially embossed with reaction wells in customized volumes and formats. Earlier this year, Douglas Scientific agreed to optimize the platform for use with Life Technologies' TaqMan SNP Genotyping Assays. The companies said at the time that the deal would be particularly beneficial to the plant agricultural biotechnology and animal breeding industries.
"Most of the time I think we will use a combination of different tools to fulfill our goals," Lai said of the tools available.
While China's use of genomic research tools to improve its maize crop appears to be paying off, Lai offered the same caveats that Xu did about technology adoption, noting that China's maize is largely produced by small companies, making it harder to implement the more costly approaches into day-to-day breeding operations. "We want them to merge and be more competitive," Lai said of the industry. "Then maybe the technology can happen."
Another issue is talent. According to Lai, there are not enough scientists in China with the skills to focus on basic maize research, which has left the country struggling to catch up to other markets where corn is an important crop. "In China, we started late," Lai said. "It has only been in the last decade that maize has been given more attention, and so we are pretty much following the US and Europe and we do not have enough scientists in China to focus on basic research."
To combat that problem, Lai said that he and others are actively recruiting Chinese nationals to return to their homeland, and are looking to entice foreign researchers to relocate as well, to take part in CAU's Crop Biology Center for maize, rice, and wheat research. "We hope that whatever the nationality we can get the talent to come," said Lai. "I would like to see three to five foreign researchers join my maize program," he said. "That would be great, that would make my job a success."