An Improved Genome-Wide Association Procedure Explores Gene–Allele Constitutions and Evolutionary Drives of Growth Period Traits in the Global Soybean Germplasm Population

Can Wang, Xiaoshuai Hao, Xueqin Liu, Yanzhu Su, Yongpeng Pan, Chunmei Zong, Wubin Wang, Guangnan Xing, Jianbo He, Junyi GaiInternational Journal of Molecular Sciences 2023 · DOI

TL;DR

This study introduces an improved genome-wide association method (GASM-RTM-GWAS) to analyze gene-allele systems controlling soybean growth periods. It reveals how soybeans adapted to different environments and maturity groups through mutation and selection, and highlights the potential for breeding by predicting optimal crosses.

What Problem Does This Paper Address?

Soybean growth periods, crucial for yield and adaptation, are influenced by day-length and temperature. While previous studies identified genes affecting these traits, they often relied on single-locus models or inferred gene functions from quantitative trait loci (QTLs). This left a gap in understanding the complete gene-allele landscape and how these systems evolve across diverse geographic regions and maturity groups. The authors aim to directly identify gene-allele systems and explore their evolutionary dynamics, providing a more comprehensive view for breeding and adaptation studies.

Methods

The core innovation is the GASM-RTM-GWAS (Genome-Allele Sequence Marker - Restricted Two-Stage Multi-Locus Genome-Wide Association Study). This method uses gene-allele sequences as markers, offering a more direct link to causal genes compared to traditional SNP-based markers. It employs a two-stage approach: first, a single-locus analysis to pre-select candidate genes, followed by a multi-locus analysis to identify the complete gene-allele system. This approach is designed to be more powerful in detecting complex traits and avoiding false positives and negatives. The study also leverages protein-protein interaction (PPI) network analysis to understand gene functional relationships.

Datasets

The study utilized a dataset of 354 soybean varieties from 27 countries, representing a global soybean germplasm population (WSGP). These varieties were grown in Nanjing, China, across four seasons. Environmental data (day-length and temperature) were obtained from the Nanjing Meteorological Bureau.

Key Insights and Ideas

  • Direct Gene-Allele Identification: GASM-RTM-GWAS allows for the direct identification of gene-allele systems, bypassing the inference step from QTLs to genes, leading to a more accurate understanding of genetic architecture.
  • Evolutionary Drivers: Geographic adaptation in soybeans is driven by allele emergence (mutation), while expansion across maturity groups is characterized by allele exclusion (selection) and inheritance (migration).
  • DSF vs. DFM Dynamics: Growth period traits like days to flowering (DSF) are more strongly influenced by accumulated day-length (ADL) and active temperature (AAT) than days to maturity (DFM). This suggests different genetic control mechanisms for these two phases.
  • Recombination Potential: The study predicts significant potential for transgressive segregation through recombination, indicating that crossing diverse soybean varieties can yield offspring with superior traits.
  • Trait-Specific Gene Networks: While genes involved in similar biological functions are found across traits, the specific gene sets and their interactions are largely trait-specific, highlighting the complexity of growth period regulation.

Key Results

  • Extensive Gene Identification: The study identified a substantial number of genes and alleles for the six traits studied (DSF, DFM, ADLDSF, AATDSF, ADLDFM, AATDFM), ranging from 124 to 141 genes and 362 to 406 alleles per trait.
  • Environmental Influence: ADL and AAT explained a significant portion of the phenotypic variance for growth period traits, with AAT having a greater impact on DSF and ADL on DFM.
  • Geographic Adaptation Patterns: Analysis of gene-allele matrices revealed distinct evolutionary patterns for geographic adaptation (allele emergence) and maturity group expansion (allele exclusion).
  • Breeding Potential: The prediction of recombination potential indicated a wide range of possible transgressive segregations, suggesting ample opportunities for developing improved soybean cultivars.
  • Functional Categorization: Genes were categorized into four main functional groups, with a significant proportion involved in flowering, light/circadian rhythm, temperature response, and primary metabolism.

Limitations and Future Work

  • Gene Annotation: While the study identified candidate genes, their precise functions often require further experimental validation.
  • Model Limitations: The study acknowledges that the GASM-RTM-GWAS, while powerful, is a statistical model and further refinement might be needed for specific applications.
  • Future Directions: The authors suggest that the identified gene-allele systems can be used for gene network exploration, population evolutionary genetic studies, and optimal cross prediction. They also highlight the potential for further research into the evolutionary drives and the role of allele recombination in soybean improvement.

Summary updated

Abstract

In soybeans (<i>Glycine max</i> (L.) Merr.), their growth periods, DSF (days of sowing-to-flowering), and DFM (days of flowering-to-maturity) are determined by their required accumulative day-length (ADL) and active temperature (AAT). A sample of 354 soybean varieties from five world eco-regions was tested in four seasons in Nanjing, China. The ADL and AAT of DSF and DFM were calculated from daily day-lengths and temperatures provided by the Nanjing Meteorological Bureau. The improved restricted two-stage multi-locus genome-wide association study using gene-allele sequences as markers (coded GASM-RTM-GWAS) was performed. (i) For DSF and its related ADL<sub>DSF</sub> and AAT<sub>DSF</sub>, 130-141 genes with 384-406 alleles were explored, and for DFM and its related ADL<sub>DFM</sub> and AAT<sub>DFM</sub>, 124-135 genes with 362-384 alleles were explored, in a total of six gene-allele systems. DSF shared more ADL and AAT contributions than DFM. (ii) Comparisons between the eco-region gene-allele submatrices indicated that the genetic adaptation from the origin to the geographic sub-regions was characterized by allele emergence (mutation), while genetic expansion from primary maturity group (MG)-sets to early/late MG-sets featured allele exclusion (selection) without allele emergence in addition to inheritance (migration). (iii) Optimal crosses with transgressive segregations in both directions were predicted and recommended for breeding purposes, indicating that allele recombination in soybean is an important evolutionary drive. (iv) Genes of the six traits were mostly trait-specific involved in four categories of 10 groups of biological functions. GASM-RTM-GWAS showed potential in detecting directly causal genes with their alleles, identifying differential trait evolutionary drives, predicting recombination breeding potentials, and revealing population gene networks.
BibTeX
@article{wang2023improved,
  author = {Can Wang and Xiaoshuai Hao and Xueqin Liu and Yanzhu Su and Yongpeng Pan and Chunmei Zong and Wubin Wang and Guangnan Xing and Jianbo He and Junyi Gai},
  title = {{An Improved Genome-Wide Association Procedure Explores Gene–Allele Constitutions and Evolutionary Drives of Growth Period Traits in the Global Soybean Germplasm Population}},
  journal = {International Journal of Molecular Sciences},
  year = {2023},
  doi = {10.3390/ijms24119570},
}
Figure