A Genome-Wide Association Study Coupled With a Transcriptomic Analysis Reveals the Genetic Loci and Candidate Genes Governing the Flowering Time in Alfalfa (Medicago sativa L.)

Abstract

The transition to flowering at the right time is very important for adapting to local conditions and maximizing alfalfa yield. However, the understanding of the genetic basis of the alfalfa flowering time remains limited. There are few reliable genes or markers for selection, which hinders progress in genetic research and molecular breeding of this trait in alfalfa. We sequenced 220 alfalfa cultivars and conducted a genome-wide association study (GWAS) involving 875,023 single-nucleotide polymorphisms (SNPs). The phenotypic analysis showed that the breeding status and geographical origin strongly influenced the alfalfa flowering time. Our GWAS revealed 63 loci significantly related to the flowering time. Ninety-five candidate genes were detected at these SNP loci within 40 kb (20 kb up- and downstream). Thirty-six percent of the candidate genes are involved in development and pollen tube growth, indicating that these genes are key genetic mechanisms of alfalfa growth and development. The transcriptomic analysis showed that 1,924, 2,405, and 3,779 differentially expressed genes (DEGs) were upregulated across the three growth stages, while 1,651, 2,613, and 4,730 DEGs were downregulated across the stages. Combining the results of our GWAS and transcriptome analysis, in total, 38 candidate genes (7 differentially expressed during the bud stage, 13 differentially expressed during the initial flowering stage, and 18 differentially expressed during the full flowering stage) were identified. Two SNPs located in the upstream region of the Msa0888690 gene (which is involved in isop renoids) were significantly related to flowering. The two significant SNPs within the upstream region of Msa0888690 existed as four different haplotypes in this panel. The genes identified in this study represent a series of candidate targets for further research investigating the alfalfa flowering time and could be used for alfalfa molecular breeding.

Keywords: GWAS; SNP; alfalfa; flowering time; haplotypes.

FIGURE 1

Boxplots showing the variation in the flowering time evaluated in each association panel. (A) Boxplot of the flowering time of plants in the subgroups (wild, landrace, and cultivar groups) according to the breeding status. (B) Boxplot of the flowering time of plants in the subgroups (America, China, Europe, and Turkey) according to geographical origin. The different asterisks above the boxplots in (A,B) indicate significant differences at the P < 0.05 level according to Duncan’s multiple comparison test.

FIGURE 2

Manhattan plot of the flowering time in different years. (A) Manhattan plot of the average flowering time over three years. (B) Manhattan plot of the flowering time in 2019. (C) Manhattan plot of the flowering time in 2020. (D) Manhattan plot of the flowering time in 2021. The GWAS was performed by BLINK C software, and the threshold for significantly associated loci was an LOD score ≥5 (blue line). The red arrow indicates the SNP sites co-located with two years of phenotypic data.

FIGURE 3

(A–H) Raincloud plots of the highest distribution of the flowering time of the plants with relevant SNP genotypes. The top plots represent the kernel density estimation, the middle plots represent box diagrams, and the bottom plots are dithering scatter diagrams. The different asterisks on the right side of the diagram indicate significant differences at the P < 0.05 level according to Duncan’s multiple comparison test. The abscissa represents flowering time and the ordinate represents SNP genotype.

FIGURE 4

Functional annotations of candidate genes. The number on the left represents the percentage of candidate genes. The text on the right represents the functional classification of the genes.

FIGURE 5

Transcriptomic analysis. (A) Number of upregulated genes (orange) and downregulated genes (blue) during alfalfa flowering at three different periods. (B) Upregulated and downregulated DEGs in the three periods. (C) Heatmap clustering of all DEGs according to their expression level. The different colors indicate different levels of expression. The blue box above represents the early flowering varieties, and red represents the late-flowering varieties.

FIGURE 6

qRT–PCR analysis of three candidate genes in three stages associated with the flowering time in alfalfa.

FIGURE 7

Msa0888690 involvement in the flowering time in alfalfa. (A) Association analysis of SNPs located within the Msa0888690 gene region. (B) Haplotype analysis of significant SNPs associated with the flowering time. (C) Comparison of the flowering time among the four haplotypes. A boxplot of the four haplotypes is shown on the left, and the symbols represent the correlations. (D) Allele frequency of the four haplotypes in the Turkish, American, European and Chinese subgroups.