This story was originally published Feb. 17.
A team of Japanese researchers has developed a method that combines crossbreeding with next-generation sequencing to speed up the identification of genes that cause agronomically important traits in mutant crop lines.
The researchers described the method, called MutMap, in a study published in Nature Biotechnology last month. They also outlined their use of the method to rapidly identify the locations of genes associated with desirable traits in mutants of an elite rice cultivar.
MutMap combines a strategy of cross-breeding a mutant plant with a desirable trait back to its wild-type parent with whole-genome sequencing of a sample of bulked DNA from several of the resulting crossbreeds. The method's simplicity and efficiency relative to previous methods could help accelerate the improvement of rice and other crops through genomic marker-assisted selective breeding, the group reported.
Ryohei Terauchi, the team's leader, told In Sequence that marker-assisted selection has become more and more widespread in crop breeding, but it remains challenging to locate markers for agronomically important traits like yield, flowering time, plant height, and biomass.
“In order to evaluate phenotype, you have to grow all the plants in the paddy field,” he said. "So, you have to have a huge field and huge manpower. But if you can mark some desired gene by DNA markers, you can just select this preferred individual in the juvenile stage [when it is still a] seedling by extracting a small amount of DNA from a leaf and genotyping it.”
Terauchi said previous methods for marker-assisted selection have used linkage analysis, which requires crosses between genetically distant lines. But such methods struggle to find markers for many agronomically important traits like plant height and flowering time because these changes are controlled by quantitative trait loci.
“Because the effect of each gene locus is small, it is difficult to identify each gene by linkage analysis of progeny derived from a distant cross, since other genes mask the effect of the focused gene,” he said.
MutMap allows the isolation of a causative gene region much more quickly and efficiently using next-generation sequencing, the researchers reported.
Terauchi explained that with MutMap, researchers must first identify a particular plant line showing some desirable change. "After mutagenesis, we raise and propagate the seeds to evaluate performance, comparing [the mutant plants'] traits to the wild-type parental line," he said.
"If you find a mutant having desired traits, you single it out and from that point, the MutMap procedure starts."
In their study, Terauchi and his colleagues first tested the method in a proof-of-concept using two pale-leafed mutants. They then targeted rice with a semi-dwarf phenotype, desirable for its ability to insure a high yield by withstanding strong winds. The group applied MutMap to four semi-dwarf mutants to identify the genomic regions responsible for the phenotype, the authors wrote.
"What we do," Terauchi said, "is cross this semi-dwarf mutant to the parental line." This creates a first-generation hybrid, which the group then self-breeds to create a second generation.
Because this second generation is derived from a cross between mutant and wild-type, "if the trait is recessive, in the [second filial, or F2] generation you will see the segregation of phenotype between wild-type and semi-dwarf in a three-to-one ratio, a simple Mendelian segregation," Terauchi said.
In their experiments with pale-leafed and dwarfed rice mutants, the researchers collected 20 of these parentally-crossed individuals with the desired phenotype, bulking this DNA in an equal ratio and then sequencing it using an Illumina Genome Analyzer.
Comparing sequencing reads to a reference sequence for the wild-type parent, the team measured the gene changes between mutant and parent as SNPs and indels.
In their description of MutMap, the researchers explained that "among the F2 progeny, the majority of SNPs will segregate in a 1:1 mutant/wild type ratio. Because the responsible SNP is homozygous in crossed progeny showing a mutant phenotype, with substantial coverage (greater than 10x) SNPs unlinked to the mutant phenotype should show 50 percent mutant and 50 percent wild-type sequence reads, while the causal SNPs and closely linked SNPs should show 100 percent mutant and no wild-type reads."
For MutMap, the group defined a "SNP index" as the "ratio between the number of reads of a mutant SNP and the total number of reads corresponding to the SNP." This index should equal one near the causal gene, and 0.5 at unlinked loci, the researchers wrote. So by scanning the genome to find regions with a SNP index of one, they could identify the location of genes responsible for the mutant phenotype.
In their semi-dwarf experiment, after sequencing the 20 dwarfed parental crossbreeds, the researchers scored SNP indices and found that "in all cases, only a single genomic region contained a cluster of SNPs with SNP index of 1… [and found] at most four SNPs that could have caused nonsynonymous changes of protein-coding genes," concluding that these regions must correspond to the locations of the causal mutations for the desired dwarf phenotype.
Other methods have been developed to use sequencing to identify agronomic trait genes for marker-assisted selection, Terauchi and his colleagues reported in the study, mentioning two recent approaches: SHOREmap, described by a European team in 2009 in Nature Methods and a "next generation mapping" technique published by Canadian researchers in Plant Journal last July.
Terauchi and his team wrote that these methods, which involve crossing a mutant with a wild-type from another distantly related plant, require the consideration of many more SNPs, which limits the ability to identify the causal SNP itself. In addition, applications of the two have required many more F2 progeny — 500 for SHOREmap and more than 50 for NGM — than MutMap, which required 20 in the current study. "Growing so many F2 progeny in the field can be impractical," they wrote.
According to Terauchi, the speed of MutMap is its main advantage. With all the crossbreeding steps included, he said the process of isolating a cultivar with a particular beneficial trait could be done quickly enough to respond to rapid changes in the environment or agricultural needs.
"We can deliver the good variety of cultivar to farmers in a short time, maybe two years or so,” he said. “So we can fine-tune already adapted cultivars to [things like] abrupt change in climate, or some other urgent need and we can make a very rapid response to the demands of local agriculture."
Along these lines, Terauchi’s team is now using MutMap to breed a strain of rice that can withstand high salinity, a concern for agricultural areas affected by the 2011 tsunami in Japan.
"A large area of paddy fields were flooded by salt water and it is very difficult to plant the normal rice in these paddy fields, so we have started our screen of salt-tolerant mutants and we have already identified a couple of these candidates and put them in the pipeline of MutMap to identify the causative genes," Terauchi said.
The method should also find use in other countries where there may be "a much larger demand" for new crop varieties in response to environmental changes or natural disasters.
With MutMap, "this kind of quick adaptation and responsiveness is now possible, so if any kind of natural disaster happens we should be able to develop new varieties to help people to feed themselves," he said
One limitation of the method, Terauchi said, is that it may not be efficient or affordable for plants with larger genomes, like maize, sorghum, soybean, barley, and wheat, because it requires relatively deep sequencing.
Since the rice genome is only around 380 megabases, the group could get more than 10x coverage in a single lane using the Illumina GA, Terauchi said. "We can use one lane to treat each sample, so this is very cheap actually. We can identify one gene by investing only $3,000 to $4,000, apart from investment in the facility."
He estimated that the whole genome of wheat or maize would currently cost at least ten times that amount.
As the price of sequencing continue to fall, however, "maybe five or ten years from now, we could do 50 or 100 times more sequencing with the same cost, and then it would be possible to analyze crops with larger genome size."
The researchers have not sought any patents on the method, and Terauchi said it can be used by anyone with a sequencing infrastructure in place.
Have topics you'd like to see covered in In Sequence? Contact the editor at mashford [at] genomeweb [.] com.