Natural variation in Fatty Acid 9 is a determinant of fatty acid and protein content

Abstract

Soybean is one of the most economically important crops worldwide and an important source of unsaturated fatty acids and protein for the human diet. Consumer demand for healthy fats and oils is increasing, and the global demand for vegetable oil is expected to double by 2050. Identification of key genes that regulate seed fatty acid content can facilitate molecular breeding of high-quality soybean varieties with enhanced fatty acid profiles. Here, we analysed the genetic architecture underlying variations in soybean seed fatty acid content using 547 accessions, including mainly landraces and cultivars from northeastern China. Through fatty acid profiling, genome re-sequencing, population genomics analyses, and GWAS, we identified a SEIPIN homologue at the FA9 locus as an important contributor to seed fatty acid content. Transgenic and multiomics analyses confirmed that FA9 was a key regulator of seed fatty acid content with pleiotropic effects on seed protein and seed size. We identified two major FA9 haplotypes in 1295 resequenced soybean accessions and assessed their phenotypic effects in a field planting of 424 accessions. Soybean accessions carrying FA9^H2 had significantly higher total fatty acid contents and lower protein contents than those carrying FA9^H1 . FA9^H2 was absent in wild soybeans but present in 13% of landraces and 26% of cultivars, suggesting that it may have been selected during soybean post-domestication improvement. FA9 therefore represents a useful genetic resource for molecular breeding of high-quality soybean varieties with specific seed storage profiles.

Keywords: Domestication; Natural variation; Seed fatty acid content; Seed protein content; Soybean.

Figure 1

Population structure analysis of 547 soybean accessions used in this study. (a) Geographic distribution of 547 soybean accessions, including 365 cultivars (yellow) and 182 landraces (green). A, Heilongjiang. B, Jilin. C, Liaoning. D, Inner Mongolia. E, northwest China. Thirty landraces from North America and one cultivar from Japan were also included. (b) A neighbour‐joining phylogenetic tree constructed using the high‐quality SNPs from all accessions. Landraces are shown in yellow and cultivars in cyan. (c) Bar plots showing population structure analyses with K values of 2 and 3. Different ancestral components are shown in different colors. (d) The PCA plot showing the first two components for all accessions. The red, blue, and cyan dots represent Group 1, Group 2, and Group 3, respectively. Triangles and circles represent landraces and cultivars, respectively. (e) Linkage disequilibrium (LD) decay plots for Group 1 (red), Group 2 (black), and Group 3 (blue).

Figure 2

GWAS results for oil content in 547 accessions. (a) Manhattan plot for oil content using best linear unbiased predictions (BLUPs). The red solid line represents the significant P‐value threshold (P = 5.38 × 10⁻⁸), and the red dotted line represents the suggestive P‐value threshold (P = 1.08 × 10⁻⁶). (b, c) Local Manhattan plot (b), gene location map (c, top), and linkage disequilibrium plot (c, bottom) for SNPs surrounding the peak SNP on chromosome 9. The red dashed lines indicate the candidate region for the peak SNP, and the red dot indicates the peak SNP (chr9: 47 115 317). The red box in the gene location map indicates the candidate gene FA9. The red solid line represents the significant P‐value threshold (P = 5.38 × 10⁻⁸), and the grey dotted line represents the suggestive P‐value threshold (P = 1.08 × 10⁻⁶). (d) Gene expression heatmap of candidate genes in the linkage disequilibrium block region. (e) Expression patterns of Glyma.09G250400 in different tissues. (f) Haplotype detection in the promoter (2 kb) and genic regions of FA9: the three main haplotypes were named FA9 ^H1 (light green), FA9 ^H2 (light blue), and FA9 ^H3 (grey). ‘P’ indicates proline, and ‘S’ indicates serine. (g) Comparison of total fatty acid contents of the three haplotypes using BLUPs (Student's t‐test). (h) Statistical analysis of total fatty acid content in WT (Col‐0), FA9 ^H1 ‐OE, and FA9 ^H2 ‐OE Arabidopsis plants. Data are presented as mean ± SD (n = 3, Student's t‐test).

Figure 3

Functional validation of the effect of FA9 knockout on seed fatty acid and protein contents. (a–c) Statistical analysis of (a) oil content, (b) protein content, and (c) fatty acid components of DN50 and fa9 KO transgenic plants. Data are presented as mean ± SD (n = 3; Student's t‐test). (d) Lipid droplets and protein bodies (PBs) in seeds of DN50 and fa9 KO transgenic plants at four seed developmental stages: early maturity (EM), middle maturity (MM), late maturity (LM), and DS. Scale bars = 5 μm. Red arrows indicate lipid droplets, and blue arrows indicate PBs. (e, f) Comparison of (e) seed length and (f) seed width of DN50 and fa9 KO transgenic plants. (g) Statistical analysis of seed length in DN50 and KO lines. (h) Statistical analysis of seed width in DN50 and KO lines. Data are presented as mean ± SD (for e and f, three plants with 15 independent seeds for each; g and h, three plants with 15 independent seeds for each). (Student's t‐test).

Figure 4

Aspects of the regulatory network by which FA9 influences seed fatty acid contents, as revealed by the integration of transcriptomic, lipidomic, and proteomic data. Alterations in lipid metabolism associated with knockout of FA9 are shown; heatmaps show fold changes in gene expression (green to red), lipid metabolite abundance (blue to red), and protein abundance (turquoise to orange).

Figure 5

Geographic differentiation of FA9 haplotypes and their effects on soybean seed fatty acid contents. (a) Haplotype analysis of FA9. (b) Geographic distribution of FA9 ^H1 and FA9 ^H2 alleles in China. (c) Distribution of FA9 ^H1 and FA9 ^H2 alleles in soybean wild accessions, landraces, and cultivars. (d) Tajima's D and π statistics of the region surrounding FA9. (e) Fatty acid profiles of soybean varieties carrying FA9 ^H2 and FA9 ^H1 . (f) Proposed model of how the distribution of two FA9 haplotypes influences the seed storage profiles of soybeans from different regions.