Transcriptional profiling of sweetpotato (Ipomoea batatas) roots indicates down-regulation of lignin biosynthesis and up-regulation of starch biosynthesis at an early stage of storage root formation

Abstract

Background: The number of fibrous roots that develop into storage roots determines sweetpotato yield. The aim of the present study was to identify the molecular mechanisms involved in the initiation of storage root formation, by performing a detailed transcriptomic analysis of initiating storage roots using next-generation sequencing platforms. A two-step approach was undertaken: (1) generating a database for the sweetpotato root transcriptome using 454-Roche sequencing of a cDNA library created from pooled samples of two root types: fibrous and initiating storage roots; (2) comparing the expression profiles of initiating storage roots and fibrous roots, using the Illumina Genome Analyzer to sequence cDNA libraries of the two root types and map the data onto the root transcriptome database.

Results: Use of the 454-Roche platform generated a total of 524,607 reads, 85.6% of which were clustered into 55,296 contigs that matched 40,278 known genes. The reads, generated by the Illumina Genome Analyzer, were found to map to 31,284 contigs out of the 55,296 contigs serving as the database. A total of 8,353 contigs were found to exhibit differential expression between the two root types (at least 2.5-fold change). The Illumina-based differential expression results were validated for nine putative genes using quantitative real-time PCR. The differential expression profiles indicated down-regulation of classical root functions, such as transport, as well as down-regulation of lignin biosynthesis in initiating storage roots, and up-regulation of carbohydrate metabolism and starch biosynthesis. In addition, data indicated delicate control of regulators of meristematic tissue identity and maintenance, associated with the initiation of storage root formation.

Conclusions: This study adds a valuable resource of sweetpotato root transcript sequences to available data, facilitating the identification of genes of interest. This resource enabled us to identify genes that are involved in the earliest stage of storage root formation, highlighting the reduction in carbon flow toward phenylpropanoid biosynthesis and its delivery into carbohydrate metabolism and starch biosynthesis, as major events involved in storage root initiation. The novel transcripts related to storage root initiation identified in this study provide a starting point for further investigation into the molecular mechanisms underlying this process.

Figure 1

Microscopic view of adventitious root cross sections 26 days after transplanting. A. Fibrous root (FR) cross section. B. Initiated storage root (ISR) cross section. Tissue sections were taken at 3 cm from the proximal root part and stained with toluidine blue. LC – lignified cells; AC – anomalous cambium; RVC – regular vascular cambium. Scale bars = 0.1 mm.

Figure 2

Sequence length distributions of contig sequences in the sweetpotato root transcriptome.

Figure 3

E-value distribution of contig sequences’ BLAST results in the sweetpotato root transcriptome.

Figure 4

Similarity distributions of the contigs to their BLAST results.

Figure 5

Gene ontology classification of assembled contigs.

Figure 6

Kyoto Encyclopedia of Genes and Genomes classification of assembled contigs.

Figure 7

Starch levels in developing adventitious roots around the timing of storage root initiation. Root samples of Georgia Jet were pooled from five to seven plants at 1, 2, 3 and 4 weeks after transplanting, spanning the period of storage root (SR) initiation. Data represent an average of three biological replicates ± SE. “Fibrous roots” represent fibrous root samples derived from plants at 8 weeks after transplanting (at least 4 weeks past the period of SR initiation).

Figure 8

Validation by quantitative on-line RT-PCR analyses of the differential expression between sweetpotato initiating storage and fibrous roots revealed by the use of Illumina-based sequencing at the cDNA level. Expression levels were measured in initiating storage root (ISR) and fibrous root (FR) samples of Georgia Jet sweetpotato variety, using at least four biological replicates. Each replicate consisted of cDNA representing pooled root tissue from 30 plants. Quantitative RT-PCR was performed and values were normalized relative to the expression levels of 18S rRNA in the same cDNA sample. Expression data are the means (± SE) of at least four replicates and are presented as relative expression values of the respective gene in the ISR sample relative to its expression in the FR sample (ISR/FR). The Y axis has a logarithmic scale. ADP glucose pyrophosphorylase – AGPase; coumaroyl-CoA synthase – 4CL; caffeoyl-CoA O-methyltransferase – CCoAOMT; cinnamyl alcohol dehydrogenase – CAD.

Figure 9

GO-term enrichment in the initiating storage root sample (‘sample’) relative to the root transcriptome database (‘reference’). Initiating storage root sample histograms are indicated in black, while the ‘reference’ histograms are indicated in yellow.

Figure 10

GO-term enrichment in the fibrous root sample (‘sample’) relative to the root transcriptome database (‘reference’). Fibrous root sample histograms are indicated in black, while the ‘reference’ histograms are indicated in yellow. A. Biological process (BP). B. GO terms included in the ‘Secondary metabolic process’ BP category. C. Molecular function (MF) and cellular component (CC) categories.

Figure 11

Changes in the phenylpropanoid biosynthesis pathway map between fibrous roots (FRs) and initiating storage roots (ISRs). Enzymes exhibiting up-regulated expression in ISRs and FRs are marked in green and in red, respectively. Marked in light green are enzymes representing gene sequences that exhibit up-regulated expression in both ISRs and FRs (in most cases, a larger number of contigs exhibited higher expression in FRs compared to ISRs (Table 5)). Marked in white are enzymes representing gene sequences that were not detected in the Illumina-generated transcription profiles (exhibited less than 10 reads) or general enzyme categories representing an enzyme class (such as 4.1.1.-, 2.1.1- and 5.2.1-, representing lyases, methyltransferases and isomerases, respectively). Enzyme annotation was obtained from the sequence annotation and GO classification data.