NEW YORK (GenomeWeb) – Scientists at the French National Institute for Agricultural Research (INRA) have developed a number of new methods that could be used in genomic selection and association mapping in wheat and potentially other crops.
After genotyping several hundred lines of a multiparental wheat population using an Illumina microarray for wheat, they created a method to estimate the contribution of each parent in the evolved population, as well as a new approach to identify quantitative trait loci under selection.
The methods, as well as the findings of the INRA study, are discussed in a recent Genetics paper.
According to the authors, INRA has focused its wheat research on a multiparent advanced generation inter-cross, or MAGIC, population, rather than a traditional approach of crossing two varietals and searching for associations in the resulting generation.
In comparison, INRA's MAGIC population was generated from 60 founder wheat lines cycled through 12 generations, which provided not only more diverse progeny, but a higher yield of relevant associations, they wrote in the paper. To ensure that the wheat lines would cross with other lines, rather than self pollinating, a male sterility gene was introduced and maintained.
Stéphanie Thépot, an INRA researcher and corresponding author on the paper, told GenomeWeb that the outcrossed nature of the wheat population studied is one "originality" of the study, noting that the uniqueness of the population led directly to the development of the new analysis methods.
"To our knowledge, this is the first long-term use of nuclear male sterility to modify a plant's reproductive biology, turning wheat from selfing to outcrossing during twelve generations," said Thépot. "The specific characteristics of the population and of its history led us to develop various methods to describe its evolution, and the genetic basis of this evolution."
INRA's MAGIC wheat population is from a long-term dynamic management program started in the 1980s, Thépot noted. The aim of the program has been to conserve the adaptive potential of wheat crops through the repeated cultivation of numerous populations in different locations. As part of the project, three populations were developed: two self-pollinating populations based on eight parents and one outcrossing population based on 60 parents. This final, outcrossing population is the one studied in the paper.
According to the authors, available parents from the MAGIC population and a subset of 380 single-seed-descent lines from the same population were first phenotyped for earliness, an important trait in wheat breeding that relates to the amount of time it takes for the wheat to flower and head. The same samples were then genotyped using a 9,000-SNP Illumina iSelect BeadChip. An international research team developed the array last year.
"At the beginning of my PhD, the Illumina 9K array was the largest available chip on bread wheat," said Thépot about the choice of platform. She acknowledged that other products and methods for genotyping wheat have become available recently, including a 817,000-marker chip produced by Affymetrix. INRA's MAGIC population has also been genotyped using a separate, 420,000-marker Affymetrix chip developed specifically for France's Breedwheat project, which involved 26 public and private research groups from across the country, Thépot said.
In terms of genotyping by sequencing options, Thépot commented that "it is important to notice that NGS technology is in its early stages on bread wheat."
Following genotyping on the Illumina iSelect array and analysis, Thépot and fellow researchers wrote in the paper that 12 generations of strict outcrossing "rapidly and drastically reduced linkage disequilibrium to very low levels even, at short map distances, and also greatly reduced the population structure exhibited among the parents."
To better understand the effects of outcrossing, they devised two new methods of analysis to identify genomic areas under selection and to estimate selection coefficients. According to Thépot, these methods assume an initial population composed of an unknown mixture of multiple known parent lines with subsequent evolutionary change in composition over time.
"The first method allows us to reconstitute the initial unknown population with a Bayesian algorithm based on allelic frequency," Thépot said. "The second detects areas under selection, areas with a [greater] evolution over time than expected."
And the new methods could find applicability in future array-driven agricultural biotechnology studies. According to Thépot, these approaches can be used "with any populations sampling over time, or with any population sampling once and with known parent lines."
The new analytical approaches allowed the INRA research team to identify 26 genomic areas under selection. Using association tests between flowering time and polymorphisms, six of the 26 genomic areas appeared to carry flowering time QTLs, one of which corresponded to Ppd-D1, a gene known to be involved in photoperiod sensitivity. Frequency shifts at four out of the six identified genomic areas were consistent with earlier flowering of the evolved population relative to the initial population, the authors reported.
Going forward, Thépot said the INRA team would like to harness some of the new, higher-density microarrays that have become available.
"With this new genotyping density, we hope to find new areas associated with the adaptive trait of earliness and to work with haplotypes instead of markers," said Thépot. "Since the number of panmictic generations guarantees very low linkage disequilibrium and the absence of structure, these evolutionary mapping populations with more individuals and a higher density of markers should be an invaluable platform for trait discovery and validation in the future."