Anthocyanin is inhibited by shading in stalk of XH1
Based on previous observations, we found that shading stress inhibited dramatically anthocyanin biosynthesis following 6 days, in stalk of XH1 (Brassica rapa L), but didn’t affect the biomass accumulation (Fig. 1A). Therefore, we sampled the stalk of XH1 from 3 days shading treatments for transcriptomes and metabolite analysis to decipher the molecular mechanism that responsible for anthocyanin accumulation in Brassica rapa L. The samples within same treatment perform well repeatability either for transcriptome and metabolism (Fig. 1B-C).
Scaffolds assemble and quality control of transcriptomes
Based on transcriptome analysis, we found 46,836,588 raw reads and 44,703,004 clean reads, with 47% GC content, 98% Q20 and 96% Q30 in average across different samples (Table S2). The reads mapped and unique reads account 87.9% and 85.6% for total reads, respectively (Table S3). The reads error rates maintain 2% (Fig. S1A), and base distribution on different position along reads is ~150 bp (Fig. S1B). Frequencies of nucleotide A and T were higher than C and G (Fig. S1B). After reads assemble, clean reads account for 95%, while only 3% events characterized with low quality (Figs. S1C). For reads alignments, we found that the coverage depts upon different chromosomes are 20× (Fig. S2A), and 80% percentile covers at downstream part of gene body (Fig. S2B). There are 98% reads mapped to the exon of genes and only 2% reads mapped to intergenic region (Fig. S2C). After gene annotation process, we found that more than 4000 genes are related to general function prediction only, posttranslational modification, protein turnover, signal transduction mechanisms (Fig. S3A). In addition, more than 70% genes are annotated to brassica rapa #28437, while only 22% are mapped to brassica napus #8961 and only 0.6% are mapped to Arabidopsis thaliana according to NR database (Fig. S3B). The values of FPKM across global genes show normal distribution (Fig. S4B).
Global gene expression induced by shading treatments
Principal component analysis shows that the samples within either control or shading treatments were clustered closely, and the explanation rates for top PC1 and PC2 are 29.7% and 27.4%, respectively (Fig. 2A). In addition, we found that there are 1906 significantly differently expressed genes (DEGs) in total (two fold up or down), and among these genes, 886 genes are upregulated and 1020 genes are down-regulated caused by shading treatments on the stalk of XH1 (Fig. 2B, Table S4).
Top 50 terms of GO analysis shows that circadian rhythm, rhythmic process, cellular response to light stimulus, detection of light stimulus, pigment biosynthetic process, response to gibberellin, starch catabolic process, anthocyanin-containing compound biosynthetic process are significantly enriched in the list of DEGs (Fig. 3A). From KEGG enrichments analysis, we found that zeatin biosynthesis, circadian rhythm, biosynthesis of secondary metabolites, anthocyanin biosynthesis and plant hormone signal transduction were significantly enriched in the list of DEGs (Fig. 3B). These extensive biological pathways were altered suggest that transcription factors could act key roles. Consistently, we found that there are 236 transcription factors are differentially expressed due to shading effects, and majority of them include AP2/ERF, bHLH, bZIP, C2C2-Dof, MYB, NAC, WRKY, TCP and DBB gene family (Fig. 4A-B. Table S5). We found that the top two ranking in upregulated transcription factors with account for 25% and 18% belongs to bHLH and C2C2-Dof gene family, respectively (Fig. 4A), while the top two ranking in downregulated transcription factors with account for 24% and 20% belongs to MYB and bHLH gene family, respectively (Fig. 4B). Interestingly, we found that all TCP gene family was up-regulated due to shading treatments (Fig. 4B). Therefore, we further detected the known genes related to anthocyanin biosynthesis pathway (Fig. 4C). Results show that there are 16 DEGs related to anthocyanin biosynthetic pathway including 7 transcription factors of interest that used for further analysis (Fig. 4C, Table S6). These transcription factors include MYB4 (Bra008570) gene and TCP15 (Bra007875) gene that were downregulated and upregulated, respectively (Fig. 4C, Table S6).
To confirm the expression of genes of interest inhibited by shading treatment, we conducted qRT-PCR experiments on the six downregulated genes related to anthocyanidin biosynthesis as shown by transcriptomes analysis (Table S6). Results show that the expression levels of genes such as BrLDOX (Bra019350), BrDFR (Bra027457), Br5MAT (Bra036208) and BrUGT75C1 (Bra038445) were inhibited by 2~4 times, in contrast, the expression level of BrTCP15 (Bra007875) was stimulated by two times (Fig. 5A-F). These results further confirm that transcript abundances from transcriptome data regarding the downregulated genes related to anthocyanidin biosynthesis.
Extremely differentially abundant metabolites due to shading effects
Considering anthocyanin metabolic pathways were detected by both GO and KEGG analysis, we further performed non-targeted metabolism analysis. Results first show that samples within same treatment were tightly closed and the explanation rates for top PC1 and PC2 are 45.71% and 18.83%, respectively (Fig. 6A), and volcano plot shows that there are 290 differentially abundant metabolites (DAMs), and 118 and 172 were down and up-regulated, respectively (Fig. 6B-C, Table S7). According to KEGG classification analysis, we found that the pathways of biosynthesis of secondary metabolites, phenylalanine metabolism, lysine degradation, glutathione metabolism, fructose and mannose metabolism, flavone and flavanol biosynthesis, citrate cycle, anthocyanin biosynthesis were significantly enriched in the list of DAMs (Fig. 6D).
For DAMs related to anthocyanin biosynthesis or anthocyanin related pathway according to GO analysis, we found 19 DAMs from anthocyanins, flavanone, terpene, amino acids and derivatives class (Fig. 7, Table S8). In particular, peonidin 3-O-glucoside chloride, and cyanindin O-syringic acid were upregulated, while Cyanidin 3-O-rutinoside, Quercetin 7-O-rutinoside and rutin were downregulated due to shading effects (Fig. 7, Table S8).