Transcriptomic profiling of taproot growth and sucrose accumulation in sugar beet (Beta vulgaris L.) at different developmental stages

Abstract

In sugar beet (Beta vulgaris L.), taproot weight and sucrose content are the important determinants of yield and quality. However, high yield and low sucrose content are two tightly bound agronomic traits. The advances in next-generation sequencing technology and the publication of sugar beet genome have provided a method for the study of molecular mechanism underlying the regulation of these two agronomic traits. In this work, we performed comparative transcriptomic analyses in the high taproot yield cultivar SD13829 and the high sucrose content cultivar BS02 at five developmental stages. More than 50,000,000 pair-end clean reads for each library were generated. When taproot turned into the rapid growth stage at the growth stage of 82 days after emergence (DAE), eighteen enriched gene ontology (GO) terms, including cell wall, cytoskeleton, and enzyme linked receptor protein signaling pathway, occurred in both cultivars. Differentially expressed genes (DEGs) of paired comparison in both cultivars were enriched in the cell wall GO term. For pathway enrichment analyses of DEGs that were respectively generated at 82 DAE compared to 59 DAE (the earlier developmental stage before taproot turning into the rapid growth stage), plant hormone signal transduction pathway was enriched. At 82 DAE, the rapid enlarging stage of taproot, several transcription factor family members were up-regulated in both cultivars. An antagonistic expression of brassinosteroid- and auxin-related genes was also detected. In SD13829, the growth strategy was relatively focused on cell enlargement promoted by brassinosteroid signaling, whereas in BS02, it was relatively focused on secondarily cambial cell division regulated by cytokinin, auxin and brassinosteroid signaling. Taken together, our data demonstrate that the weight and sucrose content of taproot rely on its growth strategy, which is controlled by brassinosteroid, auxin, cytokinin, and gibberellin.

Fig 1. Taproot growth and sucrose accumulation analyses.

The E-type cultivar SD13829 and the Z-type cultivar BS02 were grown in field. (A) The sampling area. (B) The growth curve of taproot. (C) The growth rate of taproot. (D) Phenotypes of taproot at different growth stages. (E) Sucrose contents of taproot.

Fig 2. Venn diagrams of co-expressed and specifically expressed genes at five growth stages of taproot.

(A) The Venn diagram of E-type. (B) The Venn diagram of Z-type.

Fig 3. The numbers of differentially expressed genes (DEGs).

DEGs obtained in each paired comparison were shown. (A) DEGs obtained in each stepwise comparison. (B) DEGs obtained in each genotype comparison.

Fig 4. Venn diagrams to show the concurrently and specifically expressed DEGs in different comparison.

(A) Significantly up-regulated genes identified at the growth stages of 59 DAE, 82 DAE, 113 DAE and 134 DAE. (B) Significantly down-regulated genes identified at the growth stages of 59 DAE, 82 DAE, 113 DAE and 134 DAE. The numbers of concurrently and specifically expressed DEGs at each development stages were respectively marked in the overlapped and non-overlapped regions.

Fig 5. GO enrichment analyses.

(A) The enriched terms of DEGs gained from the comparisons of E-37-vs-E-59, E-59-vs-E-82, E-82-vs-E-113 and E-113vs-E-134. (B) The enriched terms of DEGs gained from the comparisons of Z-37-vs-Z-59, Z-59-vs-Z-82, Z-82-vs-Z-113 and Z-113vs-Z-134. (C) The enriched terms of DEGs gained from the comparisons of Z-37-vs-E-37, Z-59-vs-E-59, Z-82-vs-E-82, Z-113-vs-E-113 and Z-134-vs-E-134. No enriched term in the comparison of Z-59-vs-E-59 was identified. The names of GO terms were replaced with GO accession numbers and shown in S8 Table.

Fig 6. Expression pattern of genes involved in the brassinosteroid signaling.

(A) Orthologues of genes encoding the key enzymes for the biosynthesis and metabolism of brassinosteroid. (B) Orthologues of genes encoding the components in the signal transduction pathways of brassinosteroid. Z score transformation [26] was calculated by subtracting the mean FPKM of all the “main Beta vulgaris orthologs” in all of the ten samples, and dividing that result by the standard deviation based on all the “main Beta vulgaris orthologs” in all of the ten samples for each gene. The “main Beta vulgaris orthologs” included the DEGs or the genes with FPKM ≥1 in any library. Gene IDs were marked on the left. +, up-regulated; -, down-regulated; T, up- and down-regulated compared the previous growth stage; C, E-type compared with Z-type.

Fig 7. Expression pattern of genes involved in auxin signaling.

(A) Orthologues of genes encoding the key enzyme for the biosynthesis of auxin. (B) Orthologues of genes encoding the components in the signal transduction pathways of auxin. Gene IDs were marked on the left. +, up-regulated; -, down-regulated; T, up- and down-regulated compared to the previous growth stage; C, E-type compared with Z-type.

Fig 8. Expression pattern of genes involved in cytokinin signaling.

(A) Orthologues of genes encoding the key enzymes for the biosynthesis and metabolism of cytokinin. (B) Orthologues of genes encoding the components in the signal transduction pathways of cytokinin. Gene IDs were marked on the left. +, up-regulated; -, down-regulated; T, up- and down-regulated compared to the previous growth stage; C, E-type compared with Z-type.

Fig 9. Expression pattern of genes involved in the biosynthesis and metabolism of gibberellin and sucrose.

(A) Orthologues of genes encoding the key enzymes for the biosynthesis and metabolism of gibberellin. (B) Genes involved in sucrose biosynthesis and metabolism. Gene IDs were marked on the left. +, up-regulated; -, down-regulated; T, up- and down-regulated compared to the previous growth stage; C, E-type compared with Z-type.

Fig 10. Comparisons of the gene expression as determinated by RNA-Seq and qRT-PCR analyses, respectively.

(A) Expression pattern validation of transcriptomic data in SD13829 or BS02 by qRT-PCR. (B) Expression pattern validation of selected genes in another E-type cultivar ST13092. (C) Expression pattern validation of selected genes in another Z-type cultivar ND0905. (D) Comparisons of the selected gene expression in the comparison of Z-82-vs-E-82 determined by RNA-Seq and ND0905-82-vs-ST13092-82 determined by qRT-PCR. Gene ID, abbreviation and primer of target gene were shown in S9 Table.