Comprehensive analysis of RNA-seq data reveals the complexity of the transcriptome in Brassica rapa

Abstract

Background: The species Brassica rapa (2n=20, AA) is an important vegetable and oilseed crop, and serves as an excellent model for genomic and evolutionary research in Brassica species. With the availability of whole genome sequence of B. rapa, it is essential to further determine the activity of all functional elements of the B. rapa genome and explore the transcriptome on a genome-wide scale. Here, RNA-seq data was employed to provide a genome-wide transcriptional landscape and characterization of the annotated and novel transcripts and alternative splicing events across tissues.

Results: RNA-seq reads were generated using the Illumina platform from six different tissues (root, stem, leaf, flower, silique and callus) of the B. rapa accession Chiifu-401-42, the same line used for whole genome sequencing. First, these data detected the widespread transcription of the B. rapa genome, leading to the identification of numerous novel transcripts and definition of 5'/3' UTRs of known genes. Second, 78.8% of the total annotated genes were detected as expressed and 45.8% were constitutively expressed across all tissues. We further defined several groups of genes: housekeeping genes, tissue-specific expressed genes and co-expressed genes across tissues, which will serve as a valuable repository for future crop functional genomics research. Third, alternative splicing (AS) is estimated to occur in more than 29.4% of intron-containing B. rapa genes, and 65% of them were commonly detected in more than two tissues. Interestingly, genes with high rate of AS were over-represented in GO categories relating to transcriptional regulation and signal transduction, suggesting potential importance of AS for playing regulatory role in these genes. Further, we observed that intron retention (IR) is predominant in the AS events and seems to preferentially occurred in genes with short introns.

Conclusions: The high-resolution RNA-seq analysis provides a global transcriptional landscape as a complement to the B. rapa genome sequence, which will advance our understanding of the dynamics and complexity of the B. rapa transcriptome. The atlas of gene expression in different tissues will be useful for accelerating research on functional genomics and genome evolution in Brassica species.

Figure 1

UTRs analysis of B. rapa genes. (A) Scatter-plot and histogram showing the length distributions of identified 3' and 5' UTRs based on RNA-seq data. (B) Transcripts with significantly larger (diamonds) or smaller (square) UTRs for selected GO categories. Vertical dashed lines represent median UTRs length.

Figure 2

Expression pattern of novel transcripts. (A) A comparison of expression levels between novel and general transcripts, as well as between novel transcripts with EST support and those without EST support. (B) Tissue-specific expression profiles of novel transcripts.

Figure 3

Dynamics of B. rapa gene expression in tissues. (A) The expression levels, defined as numbers of reads per kilobase per million mapped reads within each 100 Kb window, are shown, along B. rapa ten chromosomes (1–10) in each tissue. a: Callus, b: Root_1, c: Root_2, d: Stem, e: Leaf_1, f: Leaf_2, g: Flower, h: Silique; G: Gene density within each 100 Kb non-overlapping window represented as log2-tranformed total length of genes. (B) The number of highly (FPKM > 50), medium (5 < FPKM ≤50), and lowly (FPKM ≤ 5) expressed genes in each tissue. The black line shows the cumulative expressed gene number as the tissue number increased. (C) The dendrogram of tissue transcriptomes based on clustering of log2-transformed FPKM values of constitutively expressed genes. (D) Nine modules (ID: 1–9) from WGCNA analysis. The mean gene FPKM value within each module is marked by figures in blocks, and also represented as the color (red) depth of the block. The mean FPKM values of all expressed genes in each tissue are shown in the first row of blocks for comparison. The detailed information of genes in each module is provided in Additional file 2: Table S10.

Figure 4

Alternative splicing (AS) events in B. rapa. (A) Size distribution of AS-retained introns (purple) compared with introns in IR genes (yellow) and introns in general (green). (B) The tissue breadth for novel and annotated exon-exon splice junctions. The percentage of splice junctions shared by different number of tissues was plotted. (C) Some AS-enriched GO categories associated with regulation and signal transduction. (D) TFs subfamilies with high AS frequency.