Resequencing and phenotyping of the first highly inbred eggplant multiparent population reveal SmLBD13 as a key gene associated with root morphology

Abstract

The MEGGIC (Magic EGGplant InCanum) population here presented is the first highly inbred eggplant (Solanum melongena) multiparent advanced generation intercross (MAGIC) population developed so far, derived from seven cultivated accessions and one wild Solanum incanum from arid regions. The final 325 S5 lines were high-throughput genotyped using low-coverage whole-genome sequencing (lcWGS) at 3X, yielding 293 783 high-quality SNPs after stringent filtering. Principal component analysis (PCA) and neighbor-joining clustering revealed extensive genetic diversity driven by the unique genetic profile of the wild founder, and lack of genetic structure, suggesting a well-mixed population with a high degree of recombination. The eight founders and a core subset of 212 lines were phenotyped for above- and belowground traits, revealing wide phenotypic diversity. Root morphology traits displayed moderate heritability values, and strong correlation were found between root and aerial traits, suggesting that a well-developed root system supports greater aboveground growth. Genome-wide association studies (GWAS) identified a genomic region on chromosome 6 associated with root biomass (RB), total root length (RL), and root surface area (SA). Within this region, SmLBD13, an LOB-domain protein involved in lateral root development, was identified as a candidate gene. The S. incanum haplotype in this region was linked to reduced lateral root branching density, a trait that may enhance deeper soil exploration and resource uptake. These findings provide key insights into root genetics in eggplant, demonstrating MEGGIC potential for high-resolution trait mapping. Furthermore, they highlight the role of exotic wild germplasm in breeding more resilient cultivars and rootstocks with improved root architecture and enhanced nutrient uptake efficiency.

Figure 1

MEGGIC population genotyping results. (A) SNP filtering pipeline from the 3X lcWGS dataset of 31 673 278 biallelic SNPs to the final subset of 293 783 high-confident SNPs used for the subsequent analysis. The workflow illustrates the impact of each filtering step and the downstream selection of marker subsets. Dashed lines indicate the steps where SNP sets derived from the founders’ 20X resequencing data were used as the gold standard and as the reference panel. (B) Distribution of the final SNP subset along the 12 eggplant chromosomes. Color code indicates the SNP density per Mb.

Figure 2

MEGGIC population structure analysis. (A) PCA including the genotyping data of the founders and the MEGGIC population, highlighting the set of lines selected for this study over the entire population. The eight founders A–H are indicated with triangles using a color code. (B) Zoom in on the top left-hand corner of the PCA. (C) Dendrogram indicating founders’ location with colored branches and triangles with the color code. (D) Chromosome-wide and genome-wide allele contribution of parental lines. Numbers in the legend indicate the overall genome-wide contribution of each parental line.

Figure 3

Boxplots illustrating the diversity among MEGGIC founders (A–H) for different aerial (aerial biomass, AB; plant height, HE; leaf number, LN; and leaf area, LA) and root (root biomass, RB; total root length, RL; surface area, SA; maximum depth, MD; and maximum width, MW) traits.

Figure 4

Statistical analysis of the MEGGIC seedlings phenotyping. (A) Histograms and density plots of the phenotype values among MEGGIC core subset lines for different aerial (aerial biomass, AB; plant height, HE; leaf number, LN; and leaf area, LA) and root (root biomass, RB; total root length, RL; surface area, SA; maximum depth, MD; and maximum width, MW) traits. (B) PCA score plot and (C) loading plot on the first principal components based on all the studied traits for the selected MEGGIC lines. (D) Pairwise phenotypic (lower-left matrix) and genetic (upper-right matrix) correlations among the studied traits. Pearson’s correlation coefficient (r) is shown using a Bonferroni correction at the significance level of 0.05.

Figure 5

GWAS results for (A) prickles (PR), (B) anthocyanin pigmentation (AN), and (C) different root-related traits, including root biomass (RB), total root length (RL), and surface area (SA). The horizontal lines represent the Bonferroni threshold at P = .05 (LOD = 5.77). The vertical dashed line indicates the associated genomic position. Highlighted dots covered the neighboring positions to the top significantly associated SNP. Genes close to the top significantly associated SNP position are shown, with the proposed candidate gene indicated.