Background & Summary

Rose (Rosa sp.) is the most popular flower crop in the world. With its long history of cultivation, the rose has been endowed with cultural connotations in both the eastern and western world. In 2016, the worldwide turnover of cut roses was 4.96 billion euros, accounting for 29.7% of total cut flowers, which was the largest in flower crops (AIPH, 2016).

The vase life refers to the duration of a cut flower retaining its appearance in a vase. It is therefore the most important trait to the ornamental crops that are used as a cut flower, including rose. During the early vase life period, flowers open and appeal flavor, later they start the senescence, dehydration and abscission of petals. In addition, cut flowers are also subject to postharvest diseases, such as gray mold disease caused by the necrotrophic fungus Botrytis cinerea. All these physiological changes occur at a specific stage and in a highly synchronized manner, involving the balance of phytohormones in the flower and up- or down-regulation of numerous genes in various hormones pathways.

Therefore, the effects of plant hormones that extend vase life are major component of floriculture research and have been studied at the physiological and biochemical levels for several decades. Many of chemicals involved in plant hormones and their inhibitors have been added into the preservatives to improve the post-harvesting fresh keeping and vase life of cut rose. In rose, a number of studies have reported that flower opening, petal senescence, dehydration and abscission can be affected by abscisic acid (ABA), cytokinins (CKs), ethylene (ET) and gibberellins (GAs)1,2,3,4,5,6,7,8. In addition, recently, role of brassinosteroids (BRs), jasmonic acid (JA) and ET in rose petal defense against B. cinerea infection have been reported9,10.

However, the global gene expression pattern of rose flowers behind hormonal treatment has not been well-studied yet. To date, the only involved transcriptome data is screening of ethylene responsive genes from rose flowers (SRA045958), which is currently presented in the NCBI Sequence Read Archive (SRA)11.

The information on the molecular mechanism of phytohormones regulating flower traits remains scarce due to the lack of transcriptome. Therefore, transcriptome data from rose petals after different hormone treatments will be useful for studying the expression patterns of hormone-related genes and excavating key genes that regulate flower traits. Using RNA-seq, we recently investigated the transcriptomic dynamics of rose flower under the treatment of eleven natural or synthetic hormones, including auxin, BRs, CKs, GAs, ABA, JA, salicylic acid (SA), ethylene (ET) as well as ethylene inhibitors. We obtained approximately 240 Gb data and dissected the transcriptional network with the aim of exploring the transcriptional variation of rose responses towards those plant hormones. Our data will be useful to all those working with the analysis of rose gene expression.

Methods

Plant materials

Rosa hybrida ‘Samantha’, a classic hybrid tea rose cultivar and frequently used for cut flowers, has a red colour and mild fragrance. The ‘Samantha’ plant was grown under a plastic cover in Changping District (40°139 N, 116°129E), Beijing, China. In Spring of 2017, cut flower samples were harvested at developmental stage 2 of flower opening1.

Exogenous phytohormone treatment

Flowering rose stems were cut into lengths of 30 cm and placed in aqueous solutions of 1-naphthaleneacetic acid (NAA), 2,4-dichlorophenoxyacetic acid (2,4-D), 2,4-epibrassinolide (BR), 6-benzylaminopurine (6-BA), ABA, gibberellic acid 3 (GA3), JA, SA, as well as ethylene inhibitor AgNO3, respectively, for 24 h under the controlled conditions of 22 °C with 30% to 40% relative humidity and 16 h/8 h day/night periods. ‘Samantha’ treated with deionized water were used as the control. For ethylene (ET) and 1-methylcyclopropene (1-MCP) treatment, rose flowers were exposed to ethylene, 1-MCP, or regular air as the control, for 24 h and 1 M NaOH was used to absorb CO2 released by respiration. The detailed information and final concentration of chemical agents were listed in Table 1. For each treatment, three replicates were harvested and 4 flowers were randomly collected for each replicate.

Table 1 The information of exogenous phytohormone in vase treatments.
Full size table

RNA extraction, library construction, and Illumina sequencing

Total RNA was extracted from the outer layer of rose petals using the hot borate reagent following the previous description3. The quality of the RNA was verified by agarose gel electrophoresis, NanoDrop 2000 spectrophotometers (Thermo Fisher Scientific) and Agilent Technologies 2100 Bio-analyzer, all samples QC results were shown in Table 2. The libraries were sequenced on the Illumina HiSeq™ 2500 system (Illumina Inc., San Diego, CA), according to the manufacturer’s instructions. Illumina sequencing was conducted at Novogene, Beijing, China.

Table 2 RNA QC Results Summary.
Full size table

Data Records

Our 39 raw data of RNA-seq were deposited into the NCBI database at Sequence Read Archive (SRA) with the accession number PRJNA52266412.

Average gene expression (fragments per kb per million reads, FPKM) information of each experiment was deposited in Gene Expression Omnibus (GEO) in NCBI, number GSE14069613.

Technical Validation

Our 39 raw data were achieved on Illumina HiSeq. The raw data was cleaned by removing the adaptor sequences, reads containing N > 10% (N represents the base cannot be determined), as well as low quality (Qscore <  = 5) base which is over 50% of the total base (Table 3)14. FastQC was used to test the quality of 78 paired-end clean data15. We have shown the quality control results of BR-3 clean data as an illustration (Fig. 1). The FPKM (fragments per kb per million reads) of 39 samples representing different treatments were subjected to principal component analysis (PCA), and the clear separation between the treatment and mock was detected (Fig. 2).

Table 3 Data Quality Summary.
Full size table
Fig. 1
figure 1

Quality control result of BR-3 clean data. (A) Quality score of per position in read. (B) Quality score of mean sequence. (C) GC content distribution. (D) Sequence length distribution.

Full size image
Fig. 2
figure 2

The principal components analysis (PCA) of RNA-seq data following hormone treatments. (A–I). represented the PCA of 2,4-D, NAA, ABA, BR, 6-BA, GA3, JA, SA and AgNO3 compared with mock treatment (H2O). (J) and (K) represented the PCA of ET and 1-MCP compared with air (ET-CK).

Full size image

Clean reads were mapped to reference genome Rosa chinensis ‘Old blash’ (RchiOBHm-V2, GCF_002994745.1)16 by default parameters of Tophat2(version 2.1.1). Then use Cufflinks (version 2.2.0) to generate transcriptome with the Tophat2 (version 2.1.1) resulting alignment files, the assemblies were merged with the Cuffmerge, which is included in the Cufflinks package. These merged results provide a uniform basis for calculating gene and transcript expression. Then the merged assembles were provided to Cuffdiff, which calculated expression levels and tested the statistical significance of observed changes.

Average gene expression information of each experiment was deposited in Gene Expression Omnibus (GEO) in NCBI, number GSE140696. The differentially expressed genes (DEGs) were analyzed by DESeq (Anders et al., 2013) and defined as genes with |log2 fold change (FC)| ≥ 0.5, and an adjusted P-value < 0.05. The number of DEGs for 11 treatments compared with the control, was shown in Fig. 3.

Fig. 3
figure 3

The number of different expression genes (DEGs) in rose petals under exogenous phytohormone treatments. The DEGs were determined with |log2 fold change (FC)| ≥ 0.5, and an adjusted P-value < 0.05.

Full size image

In addition, we have analyzed the DEGs under auxin-related plant growth regulators (2,4-D and NAA) and ethylene-related chemicals (ET, 1-MCP and AgNO3). The results showed that 132 overlapped DEGs were identified under 2,4-D and NAA treatments (Fig. 4A). Among the 132 DEGs, 84.09% showed a similar expression pattern under the two different auxin-related plant growth regulators (Fig. 4C). In ET-, 1-MCP- and AgNO3 treatments, there were 245 overlapped DEGs were screened out (Fig. 4B). Although both 1-MCP and AgNO3 are the inhibitors of ET, they work in different ways. We have identified that 81.3% and 26.1% of the 245 DEGs played an opposite expression pattern in 1-MCP and AgNO3 treatment compared with it in ET treatment, respectively (Fig. 4D).

Fig. 4
figure 4

The DEGs among auxin related hormone (2,4-D and NAA), ethylene and its inhibitors (AgNO3 and 1-MCP). (A). Venn diagram depicted the number and overlap DEGs from 2,4-D- and NAA- treatments. (B). Venn diagram depicted the number and overlap DEGs from ET-, AgNO3- and 1-MCP- treatments. (C). The heatmap of overlap DEGs by hierarchical cluster analysis among 2,4-D- and NAA- treatments. (D). The heatmap of overlap DEGs by hierarchical cluster analysis among ET-, AgNO3- and 1-MCP- treatments.

Full size image