Genome-Wide Association Study of Lint Percentage in Gossypium hirsutum L. Races

Yuanyuan Wang, Xinlei Guo, Xiaoyan Cai, Yanchao Xu, Runrun Sun, Muhammad Jawad Umer, Kunbo Wang, Tengfei Qin, Yuqing Hou, Yuhong Wang, Pan Zhang, Zihan Wang, Fang Liu, Qinglian Wang, Zhongli ZhouInternational Journal of Molecular Sciences 2023 · DOI

TL;DR

This study uses genome-wide association mapping (GWAS) on a diverse collection of upland cotton ( Gossypium hirsutum ) races to uncover the genetic underpinnings of lint percentage, a crucial trait for yield. By identifying significant SNPs and pinpointing candidate genes like Gh_D12G0934 and Gh_A08G0526, the research provides valuable genetic resources for accelerating high-yield cotton breeding programs.

What Problem Does This Paper Address?

Cotton is a vital global crop, and maximizing its yield is paramount. Lint percentage (LP), the proportion of fiber weight to seed-cotton weight, is a key component of overall fiber yield and a significant economic indicator. While LP is known to be a complex trait influenced by genetics and environment, its precise genetic basis, particularly in the diverse wild and semi-wild "races" of G. hirsutum, remains poorly understood. Traditional breeding methods struggle with the negative correlations between LP and other desirable traits, and the genetic diversity in modern cultivated cotton has been reduced by extensive selection. This research aims to bridge this knowledge gap by systematically dissecting the genetic architecture of LP in a broad population of G. hirsutum races, which possess greater genetic variation than cultivated varieties.

Methods

The researchers employed a multi-faceted approach:

  • Phenotyping: Lint percentage was evaluated across five different environments over two years for 189 G. hirsutum accessions (188 races and one cultivar, TM-1).
  • Genotyping: A high-density CottonSNP80K chip was used to genotype the accessions, yielding over 51,000 high-quality Single Nucleotide Polymorphisms (SNPs).
  • Genome-Wide Association Study (GWAS): Two popular GWAS software packages, GAPIT 3.6.0 and Tassel 5.0, were utilized with multiple models (e.g., GLM, MLM, FarmCPU, BLINK) to identify SNPs significantly associated with lint percentage. A stringent significance threshold was applied.
  • Candidate Gene Discovery:
    • SNPs consistently identified across environments and/or models were prioritized.
  • The 550 Kb up- and downstream regions of these stable SNPs were analyzed for candidate genes using the G. hirsutum TM-1 genome annotation.
  • RNA sequencing data from ovule and fiber tissues at different developmental stages were used to assess gene expression patterns.
  • Quantitative Real-Time PCR (qRT-PCR) was performed to validate gene expression patterns.
  • Protein-protein interaction analysis (using Arabidopsis homologs via STRING and Cytoscape) and promoter <i>cis</i>-element analysis (using PlantCARE) were conducted.
  • Potential microRNA (miRNA) targeting was predicted using psRNATarget.

Datasets

  • Natural Population: A collection of 189 G. hirsutum accessions, comprising 188 G. hirsutum races and one cultivar (TM-1), was used for association mapping.
  • Genomic Data: High-density SNP data from the CottonSNP80K chip.
  • Reference Genome: The G. hirsutum TM-1 genome (NAU v1.1) was used for gene annotation and physical mapping.
  • Expression Data: RNA-seq data from ovule and fiber tissues of TM-1 at various developmental stages were obtained from NCBI (accession no. PRJNA248163).
  • Previous Studies: Data from 54 previous publications on lint percentage QTL and markers were used for validation and comparison.

Key Insights and Ideas

  • Rich Genetic Variation in Races: The study highlights that G. hirsutum races, often considered semi-wild, harbor significantly more genetic variation for lint percentage compared to cultivated varieties. This underscores their importance as a resource for breeding.
  • Complex Genetic Architecture: The identification of numerous SNPs distributed across all 24 chromosomes suggests that lint percentage is a polygenic trait, influenced by many genes with small to moderate effects.
  • Stability of Associated SNPs: The identification of 45 SNPs consistently associated with lint percentage across multiple environments and/or models indicates their robustness and potential for marker-assisted selection (MAS).
  • Intergenic Regions as Key Regulators: A significant proportion of the associated SNPs were found in intergenic regions, suggesting that regulatory elements controlling gene expression might play a crucial role in determining lint percentage.
  • Convergent Evolution of Candidate Genes: The identification of candidate genes (Gh_D12G0934 and Gh_A08G0526) involved in fundamental processes like cell differentiation, division, and cell wall biogenesis, and their conservation across species (Arabidopsis homologs), strengthens their candidacy.

Key Results

  • 274 Significant SNPs: A total of 274 SNPs were found to be significantly associated with lint percentage across the studied population and environments.
  • 45 Stable SNPs: Forty-five of these SNPs were detected in at least two environments or by at least two GWAS models, indicating their stability.
  • 584 Overlapping Markers: The regions surrounding these 45 stable SNPs (within 5 Mb) contained 584 markers previously associated with lint percentage in other studies, validating the findings.
  • 335 Candidate Genes: The 550 Kb regions surrounding 11 of the most stable SNPs harbored 335 candidate genes.
  • Two Key Candidate Genes: Through integrated analysis of gene expression, protein interactions, and promoter elements, Gh_D12G0934 (linked to fiber initiation) and Gh_A08G0526 (linked to fiber elongation) were identified as crucial candidate genes.
  • A and D Sub-genomes Contribution: The study found that the A and D sub-genomes contributed relatively equally to the genetic variation of lint percentage.

Limitations and Future Work

  • Functional Validation: While candidate genes were identified and their expression patterns analyzed, direct functional validation through transgenic approaches (e.g., gene overexpression or RNA silencing) is still needed to definitively confirm their roles in lint percentage.
  • Arabidopsis Homologs: The prediction of protein interactions relied on Arabidopsis homologs, which is a common practice but may not perfectly reflect the specific interactions in cotton.
  • Environmental Influence: Although multiple environments were used, the complex interplay between genotype and environment for lint percentage could be further explored.

The authors suggest that these identified SNPs and candidate genes provide a solid foundation for future research into the genetic architecture of lint percentage and can be directly applied to accelerate high-yield breeding programs for G. hirsutum through marker-assisted selection.

Summary updated

Abstract

Lint percentage is one of the most essential yield components and an important economic index for cotton planting. Improving lint percentage is an effective way to achieve high-yield in cotton breeding worldwide, especially upland cotton (<i>Gossypium hirsutum</i> L.). However, the genetic basis controlling lint percentage has not yet been systematically understood. Here, we performed a genome-wide association mapping for lint percentage using a natural population consisting of 189 <i>G. hirsutum</i> accessions (188 accessions of <i>G. hirsutum</i> races and one cultivar TM-1). The results showed that 274 single-nucleotide polymorphisms (SNPs) significantly associated with lint percentage were detected, and they were distributed on 24 chromosomes. Forty-five SNPs were detected at least by two models or at least in two environments, and their 5 Mb up- and downstream regions included 584 makers related to lint percentage identified in previous studies. In total, 11 out of 45 SNPs were detected at least in two environments, and their 550 Kb up- and downstream region contained 335 genes. Through RNA sequencing, gene annotation, qRT-PCR, protein-protein interaction analysis, the <i>cis</i>-elements of the promotor region, and related miRNA prediction, <i>Gh_D12G0934</i> and <i>Gh_A08G0526</i> were selected as key candidate genes for fiber initiation and elongation, respectively. These excavated SNPs and candidate genes could supplement marker and gene information for deciphering the genetic basis of lint percentage and facilitate high-yield breeding programs of <i>G. hirsutum</i> ultimately.
BibTeX
@article{wang2023genome,
  author = {Yuanyuan Wang and Xinlei Guo and Xiaoyan Cai and Yanchao Xu and Runrun Sun and Muhammad Jawad Umer and Kunbo Wang and Tengfei Qin and Yuqing Hou and Yuhong Wang and Pan Zhang and Zihan Wang and Fang Liu and Qinglian Wang and Zhongli Zhou},
  title = {{Genome-Wide Association Study of Lint Percentage in Gossypium hirsutum L. Races}},
  journal = {International Journal of Molecular Sciences},
  year = {2023},
  doi = {10.3390/ijms241210404},
}
Figure