Discovery of Hub Genes Involved in Seed Development and Lipid Biosynthesis in Sea Buckthorn (Hippophae rhamnoides L.) Using UID Transcriptome Sequencing

Abstract

Sea buckthorn is a vital woody oil species valued for its role in soil conservation and its bioactive seed oil, which is rich in unsaturated fatty acids and other compounds. However, low seed oil content and small seed size are the main bottlenecks restricting the development and utilization of sea buckthorn. In this study, we tested the seed oil content and seed size of 12 sea buckthorn cultivars and identified the key genes and transcription factors involved in seed development and lipid biosynthesis via the integration of UID RNA-seq (Unique Identifiers, UID), WGCNA (weighted gene co-expression network analysis) and qRT-PCR (quantitative real-time PCR) analysis. The results revealed five cultivars (CY02, CY11, CY201309, CY18, CY21) with significantly higher oil contents and five cultivars (CY10, CY201309, CY18, CY21, CY27) with significantly heavier seeds. A total of 10,873 genes were significantly differentially expressed between the S1 and S2 seed developmental stages of the 12 cultivars. WGCNA was used to identify five modules related to seed oil content and seed weight/size, and 417 candidate genes were screened from these modules. Among them, multiple hub genes and transcription factors were identified; for instance, ATP synthase, ATP synthase subunit D and Acyl carrier protein 1 were related to seed development; plastid-lipid-associated protein, acyltransferase-like protein, and glycerol-3-phosphate 2-O-acyltransferase 6 were involved in lipid biosynthesis; and transcription factors DOF1.2, BHLH137 and ERF4 were associated with seed enlargement and development. These findings provide crucial insights into the genetic regulation of seed traits in sea buckthorn, offering targets for future breeding efforts aimed at improving oil yield and quality.

Keywords: Hippophae rhamnoides L.; UID RNA-seq; WGCNA; lipid biosynthesis; seed development; transcription factors.

Figure 1

The oil content, thousand-seed weight (TSW) and size of developing sea buckthorn seeds. (A) Oil contents in seeds from twelve cultivars, ‘CY04’, ‘CY03’, ‘CY02’, ‘CY32’, ‘CY10’, ‘CY11’, ‘CY201309’, ‘CY13’, ‘CY18’, ‘CY21’, ‘CY27’, and ‘SQH’, at two development stages. (B) TSW in seeds from twelve cultivars at two development stages. (C) The transverse diameters from twelve cultivars at two development stages. (D) The longitudinal diameters from twelve cultivars at two development stages. (E) The developmental progress of fruits from 12 cultivars at S1 and S2. Error bars indicate standard deviations of three biological replicates.

Figure 2

Gene ontology analysis of DEGs.

Figure 3

KEGG enrichment analysis of DEGs. Top 20 KEGG up (A) and down (B) pathways of S1 vs. S2. The abscissa is the P value, and the ordinate is the KEGG path name. The greater the rich factor (p < 0.05), the more significant the enrichment level of DEGs in this pathway. The color of the circle represents the degree of enrichment of the rich factor. The size of the circle represents the number of DEGs contained in that KEGG pathway.

Figure 4

Heatmap analysis of significant DEGs associated with seed development from two developmental stages in twelve cultivars. The expression value (in RPKM) for the DEGs during S1 and S2 development in both cultivars was log2 transformed, and the total RPKM value was greater than 20.

Figure 5

Heatmap analysis of significant DEGs associated with lipid biosynthesis from two developmental stages in twelve cultivars. The expression value (in RPKM) for the DEGs during S1 and S2 development in both cultivars was log2 transformed, and the total RPKM value was greater than 20.

Figure 6

Heatmap analysis of significant TFs associated with lipid biosynthesis from two developmental stages in twelve cultivars. The expression value (in RPKM) for the DEGs during S1 and S2 development in both cultivars was log2 transformed, and the total RPKM value was greater than 20.

Figure 7

Weighted gene co-expression network analysis and visualization; clustering tree of gene systems based on TOM and the division of modules.

Figure 8

Module–trait correlation heatmap. Each row corresponded to a module, and each column corresponded to a trait. OC: oil content; TSW: thousand-seed weight; TD: transverse diameters; LD: longitudinal diameters.

Figure 9

Gene co-expression network analysis in five modules of yellow4 (A), darkmagenta (B), navajowhite2 (C), coral2 (D) and thistle1 (E).