Transcriptome Analysis of Diurnal Gene Expression in Chinese Cabbage

Abstract

Plants have developed timing mechanisms that enable them to maintain synchrony with daily environmental events. These timing mechanisms, i.e., circadian clocks, include transcriptional/translational feedback loops that drive 24 h transcriptional rhythms, which underlie oscillations in protein abundance, thus mediating circadian rhythms of behavior, physiology, and metabolism. Circadian clock genes have been investigated in the diploid model plant Arabidopsis thaliana. Crop plants with polyploid genomes-such as Brassica species-have multiple copies of some clock-related genes. Over the last decade, numerous studies have been aimed at identifying and understanding the function of paralogous genes with conserved sequences, or those that diverged during evolution. Brassicarapa's triplicate genomes retain sequence-level collinearity with Arabidopsis. In this study, we used RNA sequencing (RNAseq) to profile the diurnal transcriptome of Brassicarapa seedlings. We identified candidate paralogs of circadian clock-related genes and assessed their expression levels. These genes and their related traits that modulate the diurnal rhythm of gene expression contribute to the adaptation of crop cultivars. Our findings will contribute to the mechanistic study of circadian clock regulation inherent in polyploidy genome crops, which differ from those of model plants, and thus will be useful for future breeding studies using clock genes.

Keywords: Arabidopsis; Brassica rapa; circadian-related gene; polyploid genome; transcriptome.

Figure 1

Chinese cabbage cultivation. Chinese cabbage seedlings were grown in a growth chamber in individual pots under a 16/8-h light/dark cycle for 2 weeks, with a cool-white fluorescent lamp. Plants were entrained for 1 week under a 12/12 h light/dark cycle. Aerial parts of three-week-old plants were harvested at ZT 0, 4, 8, 12, 16, 20, and 24 h, with two biological replicates. The plants were watered every three days, and the temperature was maintained at 23 °C until 3–5 true leaves emerged. White and black boxes indicate day (light on) and night (light off), respectively. ZT: zeitgeber time; ZT0: sunrise.

Figure 2

Statistical analysis of differentially expressed genes across samples. MA plots showing pairwise comparisons of transcript levels across samples. ZT0, 4, 8, 12, 16, 20, and 24 indicate time points after the light-on time. Y-axis: log2 fold change (logFC) between the two samples; X-axis: log2 average count normalized to the size factor. Red dots: transcripts with a logFC significantly higher than 2 or lower than −2; black dots: transcripts with a logFC of −2 to 2 (A). Clustered heatmap showing the Pearson correlation matrix for pairwise sample comparisons. The color key was adjusted based on the log2-centered values for optimal visual detection of differences, and the dendrograms show the distance between samples (B). The number of significantly expressed genes (FDR p <0.05) at each time point versus ZT0 (C).

Figure 3

Clustering of 11,699 clock-regulated genes according to their diurnal cycle expression patterns in Chinese cabbage seedlings using JTK-CYCLE [51]. Thick boxes indicate groups that peaked during night time. Using JTK-CYCLE, 11,699 circadian-regulated genes were identified in leaves of Chinese cabbage seedlings growing under a 12/12 h light/dark cycle. These genes were clustered based on their diurnal expression. The k-means clustering was performed on fragments per kb of exon per million (FPKM) average values of the two repeats at 4 h intervals over 24 h using the MeV v. 4.8.1 software [52]. The number of genes in each cluster is shown.

Figure 4

Classification of transcription factors (TFs) in the B. rapa transcriptome data set. (A) Total number of differentially expressed genes (DEGs) encoding TFs. (B) Forty-four transcription factors are circadian-regulated, where the white bar graphs represent genes upregulated in the daytime (ZR0–ZT12), and the black bar graph represents genes upregulated in the night time (ZR12–ZT24).

Figure 5

Expression of candidate clock-related genes over time in B. rapa. Heat map shows expression patterns of candidate clock-related genes. Candidate circadian-regulated genes have significantly different expression in B. rapa over time. The orthologs of Arabidopsis circadian-related genes identified based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were screened in B. rapa (right). Dendrogram: hierarchical clustering of unigenes based on their expression pattern similarities. Gene ID numbers of the B. rapa sequence (Phytozome V9.0) (left). ZT0–24, time points after the light-on time (bottom).

Figure 6

Circadian clock-related genes conserved in B. rapa. On the Arabidopsis clock network, all orthologous genes of B. rapa overlapped. The Arabidopsis clock network was adapted and modified from Lou et al. [41] and Hsu and Harmer [57]. Regulatory relationships in other networks are extrapolated from the Arabidopsis network rather than being derived empirically. Arrowheads and t-shaped arrows indicate positive and negative regulatory relationships, respectively. Yellow characters represent genes with free-running periods under continuous light. Colored tiles in bars indicate expression levels over 1 day. Expression levels decrease as the color changes from dark orange to white.

Figure 7

Expression pattern of glucosinolate-related genes in B. rapa. Heatmap shows significant differentially expressed candidate genes involved in the glucosinolate pathway in B. rapa [61]. Dendrogram: hierarchical clustering of unigenes based on their expression pattern similarities. Gene sequence IDs of the B. rapa (Phytozome V9.0) (left). ZT0–24: time points after the light-on time.