Skip to content


  • Research
  • Open Access

De novo transcriptome sequencing and comparative analysis to discover genes related to floral development in Cymbidium faberi Rolfe

  • 1,
  • 1Email author,
  • 1,
  • 1,
  • 1 and
  • 1

  • Received: 18 November 2015
  • Accepted: 17 August 2016
  • Published:


Cymbidium faberi is a traditional orchid flower in China that is highly appreciated for its fragrant aroma from its zygomorphic flowers. One bottleneck of the commercial production of C. faberi is the long vegetative growth phase of the orchid and the difficulty of the regulation of its flowering time. Moreover, its flower size, shape and color are often targeting traits for orchid breeders. Understanding the molecular mechanisms of floral development in C. faberi will ultimately benefit the genetic improvement of this orchid plant. The goal of this study is to identify potential genes and regulatory networks related to the floral development in C. faberi by using transcriptome sequencing, de novo assembly and computational analyses. The vegetative and flower buds of C. faberi were sampled for such comparisons. The RNA-seq yielded about 189,300 contigs that were assembled into 172,959 unigenes. Furthermore, a total of 13,484 differentially expressed unigenes (DEGs) were identified between the vegetative and flower buds. There were 7683 down-regulated and 5801 up-regulated DEGs in the flower buds compared to those in the vegetative buds, among which 3430 and 6556 DEGs were specifically enriched in the flower or vegetative buds, respectively. A total of 173 DEGs orthologous to known genes associated with the floral organ development, floral symmetry and flowering time were identified, including 12 TCP transcription factors, 34 MADS-box genes and 28 flowering time related genes. Furthermore, expression levels of ten genes potentially involved in floral development and flowering time were verified by quantitative real-time PCR. The identified DEGs will facilitate the functional genetic studies for further understanding the flower development of C. faberi.


  • Cymbidium faberi Rolfe
  • RNA-seq
  • Flower development
  • Flowering
  • Genes


Orchidaceae is one of the largest families in the angiosperms with more than 25,000 species, which displays a great biodiversity resulting from adaptation to diverse habitats (Pridgeon et al. 2005). Genomic information on orchids were mainly focused on Phalaenopsis (Su et al. 2011; Cai et al. 2014), Dendrobium (Yan et al. 2015; Zhang et al. 2016), Cymbidium ensifolium (Li et al. 2013), Cymbidium sinense (Zhang et al. 2013). Cymbidium faberi Rolfe., common named as “Hui Lan”, is one of the oldest and most popular orchids species cultivated in China, which is highly appreciated mainly because of its beautiful flower posture and fragrant aroma (Wolff, 1999). However, large scale commercial production of C. faberi was often hindered due to the long vegetative growth phase (usu. 5–7 years) and difficulties in flowering time control.

Plant flowering involves a transition process from vegetative growth to reproductive development with a series of conserved underlying metabolic or external phenotypic changes taking place in the shoot apical meristem. In Arabidopsis thaliana, there are four major pathways controlling the timing of flowering, including photoperiod, vernalization, gibberellin (GA) and autonomous pathways (Mouradov et al. 2002). Major genetic elements involved in this pathways have been defined as the key switches to control flowering, such as CONSTANS (CO) in the photoperiodic pathway and FLOWERING LOCUS C (FLC) in the autonomous and vernalization pathways (Mouradov et al. 2002). The FLOWERING LOCUS T (FT) gene activates the expression of a number of flower developmental genes (Mouradov et al. 2002). In the orchid plant Phalaenopsis aphrodite, PaFT1 can suppress the delayed flowering caused by SHORT VEGATATIVE PHASE (SVP) and FRIGIDA (FRI) (Jang et al. 2015). The Cymbidium FT orthologous gene was also cloned and ectopic expression of CgFT resulted in early flowering in transgenic Arabidopsis (Huang et al. 2012). Over expression of DnAGL19, a SOC1-1/TM3-like ortholog in Arabidopsis resulted in a slightly accelerated flowering time under normal growth conditions (Liu et al. 2016).

During the transition from vegetative growth to reproductive growth, MADS-box family genes play important roles in regulating floral organ specification, development and evolutionary in higher plants (Weigel and Meyerowitz 1994; Purugganan et al. 1995; Münster et al. 1997). Although conserved flowering pathways and multiple key genes were revealed in model plants, there is limited information of C. faberi, the very unique and important plant species featured with a 5–7 year-long vegetative growth phase. Unlike most tropical orchids (e.g. Phalaenopsis and Cattleya), C. faberi flowers are not brightly showy but of strong aromas to attract pollinating insects. It is of strong interest for orchid breeders to improve the appearance of its flower size, color and shape with uncompromised aromas. The flowers of C. faberi are zygomorphic consisting of four whorls: The first whorl is comprised of the petal-like sepals, the second whorl is of two petals and one highly specialized labellum in the middle, the third whole is of stamens, and fourth whorl is of pistils in the form a highly specialized united gynostemium (Rudall and Bateman 2002). The well-known ABC model clarify that floral development is determined by five kinds of floral organ identity genes in diverse plant groups (Coen and Meyerowitz 1991; Weigel and Meyerowitz 1994). Sepal formation is specified by the expression of A-class function genes. Expression of AB and BC determine the development of petals and stamen formation, respectively. The development of carpel is determined by C-class genes function alone, and D-class genes specify ovules. While class E function redundantly specify petals, stamens, and carpels as well as floral determinacy (Pelaz et al. 2000, 2001; Anusak et al. 2003). The floral morphology of orchid species is unique with gynostemium, labellum and resupination caused by 180° torsion of the pedicel (Rudall and Bateman 2002). ‘The Perianth (P) code’ clarifies the mechanisms of sepal/petal/lip determination in Oncidium and Phalaenopsis orchids. The competition between different APETALA3/AGAMOUS-LIKE6 (AP3/AGL6) homologues determines the formation of the complex perianth patterns in orchids. The formation of sepal/petal were specified by the higher-order heterotetrameric SP (sepal/petal) complex (OAP3-1/OAGL6-1/OAGL6-1/OPI), whereas the lip formation required the L (lip) complex (OAP3-2/OAGL6-2/OAGL6-2/OPI) exclusively (Hsu et al. 2015). Other MADS-box function genes participating in the sepal and petal development were also isolated from Dendrobium madame, D. crumenatum, Oncidium Gower Ramsey (Yu et al. 2000, 2002; Hsu and Yang 2002; Hsu et al. 2003; Xu et al. 2006). Genes involved in flower identity and floral organ specification are still unknown in C. faberi yet.

Besides petal shape and size, floral symmetry significantly affects the ornamental value of flowers. So far, three transcription factors (TFs) that determine the floral symmetry are identified, including CYCLOIDEA (CYC) from the TCP family (TEOSINTE BRANCHED1/CYCLOIDEA/PROLIFERATING CELL NUCLEAR ANTIGEN FACTOR), DIVARICATA (DIV) and RADIALIS (RAD) from the MYB family (Luo et al. 1996; Doebley et al. 1997; Almeida et al. 1997; Galego and Almeida 2002; Corley et al. 2005; Costa et al. 2005). Of these genes, CYC and its orthologues establish the floral monosymmetry through specifying dorsal flower identity (Luo et al. 1996, 1999; Feng et al. 2006; Wang et al. 2008). Studies on floral symmetry genes mainly focused on species with highly derived morphologies in eudicots, whereas only a few studies are available for monocots (Luo et al. 1996, 1999; Feng et al. 2006; Wang et al. 2008; Yuan et al. 2009; Bartlett and Specht 2011; Preston and Hileman 2012; Hoshino et al. 2014). Orchidaceae is characterized by highly specialized zygomorphic flowers. Studies on orchid flower symmetry is very limited (Paolo et al. 2015). Overall, the detailed molecular mechanism or the genetic elements in the regulation of floral organ specification and development in C. faberi remains elusive.

Transcriptome analysis is an useful tool to discriminate differences in transcript abundance among different cultivars, organs and different treatment conditions in model and non-model plants (Cheung et al. 2006; Trick et al. 2009; Li et al. 2013; Hyun et al. 2014; Zhang et al. 2014). In order to excavate genes that might regulate the floral development in C. faberi, we used high-throughput Illumina sequencing and bioinformatics analysis to compare the de novo transcriptomes of vegetative and flower buds from C. faberi. The vegetative transcriptome can be used to find genes related to the vegetative growth. The floral transcriptome was sufficiently comprehensive for gene discovery and analysis of major metabolic pathways associated with flower traits. Genes expressed differently in the vegetative buds and flower buds may play important roles in the vegetative growth or floral development. Through transcriptome analysis, large numbers of genes related to floral organ initiation, flower symmetry patterning and flowering were identified in C. faberi. These results provide fundamental information for further studies on the molecular mechanism of flower development in C. faberi.


Plant materials

Mature plants of C. faberi with light green flowers, originally introduced from Henan province of China, were grown in the greenhouse at Department of Horticulture, Nanjing Agricultural University (Nanjing, China) under natural light and a controlled temperature of 22–28 °C. The vegetative buds were collected from lateral buds, and the flower buds (0.5–1.0 cm in length) were collected from the peduncle (Fig. 1). Organs from three individual plants were pooled as one sample. The fresh samples were frozen immediately in liquid nitrogen and stored at −80 °C.
Fig. 1
Fig. 1

The mature plants and flowers of C. faberi. a Flowering plants of C. faberi; b Inflorescence with flower buds (f); c Vegetative buds (v); d blooming flowerers; e single flower; f different parts of a single flower, s sepal; p petal; l labellum; c column

RNA extraction, cDNA library construction and de novo assembly

Total RNA of flower and vegetative buds were extracted using EASYspin plant RNA rapid extraction kit according to the manufacturer’s protocol (Yuan Ping Hao Biotechnology Co. Ltd, Beijing, China). Then, the concentration was detected by a NanoDrop 2000™ micro-volume spectrophotometer (Thermo Scientific, Waltham, MA, USA) and the quality was tested by gel electrophoresis. cDNA library construction for the flower and vegetative buds and Illumina deep sequencing were performed on the HiSeq™ 2000 platform according to the manufacturer’s instructions at Hangzhou Woosen Biotechnology Co. Ltd. (Hangzhou, Zhejiang, China). The Illumina reads were assembled to obtain the contigs and unigenes using Trinity software and Cap3 after removing the short raw reads and quality inspection by fastQC (Grabherr et al. 2011).

Functional annotation of unigenes

The unigenes were annotated with the National Center for Biotechnology Information non-redundant databases (NR), Gene Ontology (GO), Clusters of Orthologous Groups (COG) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases using BLASTX searches (E-value ≤ 1.0e−5). The Blast2GO software package and WEGO software were employed to compare and determine unigenes’ GO annotations and obtain GO functional classifications for all annotated unigenes (Götz et al. 2008; Ye et al. 2006).

Identification of differentially expressed genes (DEGs)

Expression levels of all unigenes were calculated and then compared between the two tissue samples using the fragments per kilobase of transcript per million reads of library method (FPKM). The false discovery rate (FDR) was adopted to determine the threshold of P-values in multiple tests. DEGs were also annotated with GO assignments, COG assignments and KEGG pathways. The criteria FDR < 0.05 was used to identify DEGs and acted as a threshold of significant difference of gene expression in GO terms, COG classification and KEGG annotation.

Quantitative real-time PCR (qRT-PCR) analysis

Total RNA of flower buds and vegetative buds were extracted as above and the first strand cDNA was synthesized using PrimeScript RT reagent Kit With gDNA Eraser (Takara Bio Inc.). All primers used in this study were designed by the Primer 5 software according to the RNA-Seq data. CfGAPDH (JX560732) was selected as an internal reference (Additional file 1: Table S1).

The qRT-PCR analysis was performed on ABI 7500 Real-Time PCR Detection System (Applied Biosystems) using the SYBR® Premix ExTagTM reagent kit (Takara Bio Inc.) according to the manufacturer’s protocol. The PCR reactions were 40 cycles (95 °C for 15 s; 55 °C for 15 s; 72 °C for 20 s) according to the instruction manual. A melting curve was generated to test the specificity of products after the qRT-PCR. The relative expression levels of the selected unigenes were normalized to CfGAPDH gene and calculated using the 2−ΔΔCt method. Data were derived from three independent replicates.


Transcriptome sequencing and de novo assembly

Equal amount of RNAs from the vegetative and flower buds were used to constructed cDNA libraries separately, and then sequenced. A total of 35,511,583 and 32,514,423 raw reads were obtained in the flower and vegetative buds, respectively. After removing the short raw reads and quality inspection by fastQC, the RNA-seq produced 35,510,239 and 32,513,288 clean reads for the flower and vegetative buds, respectively (Table 1). All of these reads were employed for further de novo assembly. And 189,300 contigs with an average length of 755 bp were generated using the Trinity software package. These contigs were assembled into 172,959 unigenes (200 to >10,000 bp) with an average length of 698 bp and an N50 (N50 represents weighted median length of all contigs) of 1340 bp (Table 2). Among them, 59,505 (34.4 %) unigenes were more than 500 bp. The size distribution of the assembled unigenes is shown in Additional file 2: Fig. S1. And this transcriptome shotgun assembly project has been deposited at DDBJ/EMBL/GenBank under the accession of GDHD00000000.
Table 1

Statistical summary of C. faberi transcriptome sequencing data

Statistics of data production

Flower buds

Vegetative buds

Raw reads



Raw bases



Clean reads



Clean bases



Clean data rate (%)



Table 2

Summary of assembly quality from transcriptome in C. faberi




Total number



Average length(bp)



Max length(bp)



Min length(bp)



GC percentage (%)



Length of N50 (bp)



Functional annotation of C. faberi transcriptome

To validate and annotate the assembled unigenes, the 172,959 unigenes generated were subjected to BLASTX searches (E-value ≤ 1.0e−5) against public protein databases such as NCBI NR, GO, COG and KEGG. As a result, a total of 65,577 (37.91 %) unigenes were predicted to be coding sequence and 66,000 (38.16 %), 50,161 (29.00 %), 27,443 (15.87 %), 19,715 (11.40 %) unigenes had homologous sequences respectively in NR, GO, COG and KEGG databases (Table 3).
Table 3

Summary statistics of functional annotation for C. faberi vegetative and flower buds unigenes in public databases

Public protein database

No. of unigene hits

Percentage (%)

Predict protein number






NR (E-value < 10−5)









According to the NR annotation, 66,000 (38.16 %) unigenes had homologous sequence in the database, of which 14,365 (21.77 %), 4309 (6.53 %), 4173 (6.32 %) unigenes were annotated with homologous genes from Vitis vinifera, Oryza sativa Japonica Group and Prunus persica, respectively (Fig. 2a). The similarity distribution indicated that 54.87 % of the unigenes showed a similarity higher than 60, and 40.04 % unigenes had similarity between 60 and 80 % (Fig. 2b). For the E value distribution, 53.79 % of the top hits had high homology with the E-value <1.0e−50 (Fig. 2c).
Fig. 2
Fig. 2

Characteristics of sequence homology of C. faberi BLASTED against NCBI non-redundant (Nr) database. a E-value distribution of BLAST hits for matched unigene sequences, using an E-value cutoff of 1.0E−5. b Similarity distribution of top BLAST hits for each unigene. c Species distribution of the top BLAST hits

As an international standardized gene functional classification system, GO describes properties of genes and their products in many organisms and provides a comprehensive description of gene properties across species and databases (Hao et al. 2011). Based on the BlastX results, a total of 50,161 (29.00 %) unigenes were assigned into 49 GO term annotations, including three categories, i.e., molecular function, biological process, and cellular component (Fig. 3). Within the molecular function category (70,280 GO terms), “binding” (32,021, 45.6 %) and “catalytic activity” (26,742, 38.0 %) were the most highly represented terms. For biological process (108,214 GO terms), “cellular process” (28,941, 26.7 %) and “metabolic process” (29,813, 27.6 %) were the highly represented terms. Among the cellular component category (87,253 GO terms), “cell” (28,076, 32.2 %), “cell part” (13,723, 15.7 %) and “organelle” (28,076, 32.2 %) were the three main categories.
Fig. 3
Fig. 3

GO classification of all annotated unigenes. GO classification of all annotated 50,161 unigenes. All terms belonged to the three main GO categories: biological process, cellular component and molecular function. The x-axis indicated the subcategories, the right y-axis indicated the number of genes in each category, the left y-axis indicated the percentage of a specific category of genes in the corresponding GO category. Red column indicated all annotated unigenes

A total of 27,443 unigenes were classified into 24 COG functional categories (Fig. 4). And the largest category was “general function prediction only” (6177, 22.5 %), followed by “signal transduction mechanisms” (2220, 8.1 %) and “replication, recombination and repair” (2196, 8.0 %). Unigenes annotated as the “signal transduction mechanisms” in our study may allow for the identification of novel genes involved in signal transduction pathways. Approximately 16.9 % of the unigenes were associated with biochemical synthesis and metabolism, such as “carbohydrate transport and metabolism” (1242, 4.5 %), “amino acid transport and metabolism” (1184, 4.3 %) and “secondary metabolites biosynthesis, transport and catabolism” (544, 2.0 %). A total of 1225 unigenes (4.46 %) was “function unknown”.
Fig. 4
Fig. 4

COG annotations of putative proteins. All putative proteins were aligned to the COG database and were functionally classified into at least 24 molecular families. The capital letters in x-axis indicates the COG categories as listed on the right of the histogram and the y-axis indicates the number of unigenes in the corresponding COG category

Furthermore, we classified 19,715 (11.40 %) unigenes into 274 KEGG pathways by searching against the KEGG database (Additional file 3: Table S2). Among these pathways, “metabolic pathways”(3684, 18.7 %) represented the largest group, followed by “biosynthesis of secondary metabolites”(1616, 8.2 %). The other highly represented pathways included “microbial metabolism in diverse environment (803, 4.1 %)”, “ribosome (509, 2.6 %)” and “protein processing in endoplasmic reticulum (507, 2.6 %)” (Table 4). A total of 301 unigenes involved in plant hormone signal transduction were found, including 9 pathways controlling the signal transduction of auxin, cytokinin, gibberellin acid, abscisic acid, ethylene, brassinosteroid, jasmonic acid and salicylic acid (Table 5). Besides, through the KEGG pathway annotation, a circadian rhythm pathway including 58 unigenes was also found, which can be used for further studies on the flowering-related genes of the photoperiod pathway.
Table 4

Significantly enriched pathways of all unigenes and DEGs in C. faberi


No. of all genes with pathway annotation (19,715)

No. of DEGs with pathway annotation (66)

Pathway ID

Up-regulated genes

Down-regulat-ed genes

Metabolic pathways





Biosynthesis of secondary metabolites





Microbial metabolism in diverse environments










Protein processing in endoplasmic reticulum










RNA transport





Starch and sucrose metabolism





Cell cycle










Pyrimidine metabolism





RNA degradation





Table 5

The pathways involved in plant hormone biosynthesis



Pathway ID

Unigene number

Cysteine and methionine metabolism




Phenylalanine metabolism

Salicylic acid



Tryptophan metabolism




alpha-Linolenic acid metabolism

Jasmonic acid



Diterpenoid biosynthesis




Carotenoid biosynthesis

Abscisic acid



Zeatin biosynthesis




Brassinosteroid biosynthesis




Indole alkaloid biosynthesis

Indole-acetic acid



Based on the annotation of unigenes, 118 MADS-box genes were also discovered (Additional file 4: Table S3), these including the sepal development related genes such as APETALA1 (AP1), CAULOFLOWER (CAL); petal development related genes DEFICIENS (DEF)and PISTILLATA (PI); the C/D/E class function genes AGAMOUS (AG), SEEDSTICK (STK) and FLORAL BINDING PROTEIN-LIKE (FBP-like) were also identified. TCP gene family play important role in the development of plants and can be divided into two classes: PCF class and TCP-C class (which involes the CYC/TB1 clade and CIN clade) (Howarth and Donoghue 2006; Navaud et al. 2007). The CYC/TB1 clade includes genes mainly involved in the development of axillary meristems and floral bilateral symmetry (Luo et al. 1996; Doebley et al. 1997). To determine the TCP gene family in C. faberi, we analyzed the transcriptome database generated by this study and found 35 unigenes annotated as TCP TFs, including 14 CIN-like genes, 20 PCF-like genes and 1 CYC/TB1-like genes. As shown in Additional file 5: Table S4, they showed homology with 13 Arabidopsis TCP genes.

A total of 240 genes associated with flowering time were obtained (Additional file 6: Table S5). These include floral meristem identity genes LEAFY (LFY) and APETALA1 (AP1); autonomous pathways genes FCA, FPA, FLOWERING LOCUS (FLD), FY, FVE, FLOWERING LATE KH MOTIF (FLK); vernalization pathways genes FRI and VERNALIZATION INSENSITIVE (VIN); photoperiod pathway genes such as FT, Phytochrome A (PHYA), Phytochrome B (PHYB), PIF3, ELF3, LHY, SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (SOC1), CIRCADIAN CLOCK ASSOCIATED1 (CCA1) and CO; Gibberellin (GA) pathway genes such as GIBBERELLIC ACID INSENSITIVE (GAI). All these unigenes provide important resources for future study of floral organ development, floral bilateral symmetry and flowering time.

DEG analysis of vegetative and flower buds

The FPKM methods were used to analyze the gene expression in the two libraries: 123,128 and 125,862 unigenes were obtained in the flower buds and vegetative buds, of which 39,549 and 42,283 genes expressed specifically in the flower buds and vegetative buds, respectively (Fig. 5). To analyze different gene expression in the two libraries, 13,484 DEGs were identified using FPKM methods. Among them, 7683 were down-regulated and 5801 were up-regulated in the flower buds when compared to those in the vegetative buds, including 3430 and 6556 genes specifically enriched in flower and vegetative buds, respectively. To validate and annotate the assembled DEGs, the 13,484 DEGs were subjected to BLASTX comparisons (E-value ≤10−5) against GO, COG and KEGG database to identify putative functions of these unigenes. As a result, 5348 (39.7 %), 2829 (21.0 %) and 3927 (29.1 %) DEGs had homologous sequences in GO, COG and KEGG databases, respectively. In addition, 5348 DEGs were classified into 44 GO term annotations, including cellular component (1399), biological process (3921) and molecular function (7006), with multiple terms assigned to the same transcript (Fig. 6). A total of 2829 DEGs were classified into 23 functional COG categories (Fig. 7). “general function prediction only” (506) was the most, followed by “posttranslational modification, protein turnover, chaperones” (248) and “transcription” (222). To further understand the function in biological processes, 3927 DEGs were classified into 245 KEGG pathways. These results showed that most of the DEGs were involved in the “metabolic pathways”(300), “biosynthesis of secondary metabolites”(252), “microbial metabolism in diverse environment (142)” and “ribosome”(89) (Table 4). After transcriptome analysis, TFs involved in the floral differentiation were revealed and their expression levels in the flower and vegetative buds were calculated and compared using the FPKM method.
Fig. 5
Fig. 5

The numbers of specific genes and shared genes between the flower and vegetative buds of C. faberi

Fig. 6
Fig. 6

GO classification of DEGs

Fig. 7
Fig. 7

COG annotations of DEGs. 2829 DEGs were aligned to the COG database and were functionally classified into at least 23 molecular families. The capital letters in x-axis indicates the COG categories as listed on the right of the histogram and the y-axis indicates the number of unigenes in the corresponding COG category

A total of 173 DEGs related to the floral development were discovered in our transcriptome data (Additional file 7: Table S6). These TFs were attributed to different gene families, including MADS (34 DEGs), TCP (12 DEGs), MYB (34 DEGs), ARF (9 DEGs), NAC (21 DEGs), b-ZIP (17 DEGs), WRKY (18 DEGs) (Table 6). Twenty-eight DEGs related to the flowering time were identified (Additional file 8: Table S7) and some shown in Table 7. Most of these TFs showed higher expression in the flower buds indicating that these genes might involve in the flower development.
Table 6

DEGs related to floral differentiation in the flower and vegetative buds

Genes related to floral differentiation

Total No. of DEGs

No. of DEGs in flower buds

No. of up-regulated DEGs in flower buds

No. of DEGs in vegetative buds

No. of up-regulated DEGs in vegetative buds











































Table 7

Some DEGs related to flowering time in the flower and vegetative buds

Genes related to flowering

Total No. of DEGs

No. of DEGs in flower buds

No. of up-regulated DEGs in flower buds

No. of DEGs in vegetative buds

No. of up-regulated DEGs in vegetative buds











































qRT-PCR validation of selected genes

To verify the reliability of RNA-Seq data and explicit the expression patterns of genes related to floral development, ten genes mainly related to the floral organ development, floral symmetry and flowering time associated genes were selected to perform qRT-PCR analysis. Our results indicated that DEF, PI, AG, AP1, BRANCHED2 (BRC2) and CCA1 had higher expression and TCP4, CO, SOC1 and LFY displayed lower expression in the flower buds (Fig. 8a-j ). PI and DEF, the two B class function genes controlling the petal development, showed significantly higher expression in the flower buds than the vegetative buds. While CCA1 participating in the circadian rhythm showed significantly higher expression in the flower buds than the vegetative buds. Overall, all of the 10 genes performed the same expression patterns in our qRT-PCR results as in the RNA-Seq data, verifying the reliability of the Illumina sequencing data.
Fig. 8
Fig. 8

qRT-PCR analysis of ten selected genes in different organs of C. faberi. aj indicated the relative expression level of DEF, PI, AG, AP, TCP4, BRC2, CO, SOC1, CCA1, LFY gene, respectively. In these pictures, F flower buds; V vegetative buds; Asterisk indicated significantly higher expression


Capacity of the transcriptome database

In recent years, transcriptome analysis based on deep RNA sequencing has been applied to many plants to identify differences in gene expression levels among different cultivars, organs and different treatment conditions (Chen et al. 2014; Yates et al. 2014). In this study, we obtained 189,300 contigs and identified 172,959 unigenes by de novo assembly through RNA-Seq technology in the vegetative and flower buds of C. faberi. According to the Nr protein database, 66,000 (38.16 %) unigenes were successfully annotated. Furthermore, 173 DEGs involved in floral organ development, floral zygomorphy and flowering time were found. These results supported that plant conservative genes, C. faberi-specific genes, and C. faberi tissue-specific genes all were identified in our transcriptomic analysis. Our research will provide valuable information for future study to inquire the flower development mechanisms of C. faberi.

Genes related to the floral organ development

MADS-box genes are known for their roles in the flower organ development in Arabidopsis and Antirrhinum (Coen and Meyerowitz, 1991; Weigel and Meyerowitz, 1994). While some monocot species like the orchid family possess distinct floral structures, thus floral patterning in Arabidopsis may not be comparable to such flowering plants. In recent years, a few MADS-box genes have been identified and characterized in D. thyrsiflorum (Reichb. f.) (Martin et al. 2006), Oncidium Gower Ramsey (Hsu and Yang 2002), Phalaenopsis (Tsai et al. 2004) and C. faberi (Xiang et al., 2011). According to the “Orchid code” theory, the identity of the lateral petals and floral lip were determined by four different AP3/DEF-like genes, whereas the PI/GLO-like genes retained the function unchanged like class A, C, D and E genes (Mondragón-Palomino and Theissen 2008, 2009, 2011; Aceto and Gaudio 2011). In this study, 34 DEGs of MADS-like genes were found. Meanwhile, 29 DEGs were up-regulated in the flower buds (26 DEGs specifically expressed in the flower buds and most of them were class B and class C genes) and only 5 DEGs were up-regulated in the vegetative buds. Meanwhile, the expression level of PI and DEF showed significantly higher in the flower buds than the vegetative buds, proving that PI and DEF played a pivotal role in the flower development.

Genes related to the floral zygomorphy

The ornamental value of C. faberi is determined by many factors, especially the novel flower color, shape and fragrance. Floral zygomorphy have evolved multiple times from radially symmetrical (actinomorphic; polysymmetric) ancestors in different angiosperm lineages (Endress 1999; Stebbins 1974). The mechanism of TCP model for bilateral flower symmetry has been well established in model species, such as snapdragon (A. majus), Lotus japonicus and rice (Luo et al. 1996; Feng et al. 2006; Yuan et al. 2009). In A. majus, CYC and DICH control the dorsoventral asymmetry (Luo et al. 1999). In L. japonicus, floral dorsoventral asymmetry is regulated by three LjCYC genes (LjCYC1, LjCYC2, LjCYC3), especially the LjCYC2. In monocot, RETARDED PALEA1 (REP1) controls palea development and floral zygomorphy in rice, whose flower is different from that of A. majus or L. japonicus. In our C. faberi transcriptome data, we identified 35 TCP genes, of which 10 CIN-like genes and one PCF-like gene showed significantly higher expression levels in the vegetative buds compared to the flower buds. qRT-PCR analysis revealed that the expression level of BRC2 was higher in the flower buds than the vegetative buds (Fig. 8e), suggesting that it might play important role in the regulation of floral zygomorphy of C. faberi.

Genes related to the regulation of flowering

Previous studies showed that the flowering of Arabidopsis was regulated by four pathways, including autonomous pathway, vernalization pathway, gibberellic acid (GA)-dependent pathway and the photoperiod pathway. And CCA1 influenced the circadian rhythm, overexpression of CCA1 resulted in long hypocotyls and late flowering (Wang and Tobing 1998). CO involved in the photoperiod pathway in Arabidopsis and acted as a floral promoter, which was regulated by the circadian clock (Yano et al. 2000). The temporal and spatial regulation of CO turned out to be important to the photoperiod-dependent induction of flowering (An et al. 2004). Early in a day, the expression of CO was low, then increased rapidly 10 h after dawn and peaked at around 15 h (Suárez-López et al. 2001). LFY is a major floral meristem regulator, its overexpression caused early flowering in transgenic Arabidopsis (Benlloch et al. 2007). A single PHYA gene, LHY gene and three CCA1 genes involved in the photoperiod pathways showed more highly expressed in the vegetative buds than in the flower buds. Two VRN genes playing important roles in the vernalization pathway were detected in our study, and were found to be more highly expressed in the vegetative buds than in the flower buds. Besides, DEGs related to the autonomous pathway were also detected, including two FCA genes, a FLD gene and a FY gene. qRT-PCR revealed that expression of LFY was higher in the vegetative buds and expression of CO, SOC1, CCA1 was higher in the flower buds, which was consistent with the RNA-Seq data (Fig. 8g–j). Expression of CCA1 was significantly higher in the flower buds than the vegetative buds. This observation suggesting that the mechanism of flowering regulation in C. faberi might be similar to Arabidopsis.

In conclusion, the current RNA-seq and DEG analysis revealed 393 genes associated to the floral development, including 35 TCP transcription factors (TFs), 118 MADS-box genes and 240 flowering time related genes. A total of 173 DEGs were identified, including 12 TCP genes, 34 MADS-box genes and 28 flowering time related genes. The transcriptome database in the present study will be a valuable supplement to the genomic sequence dataset of C. faberi, and will serve as an important public information platform for further studies on the flower development mechanism in C. faberi.


Authors’ contributions

The experimental plan was conceived and designed by GW. YS, YL and SG participated in sample collection, RNA preparation and experiments performing. YS, JL and YY contributed to data analysis and manuscript preparation. GW participated in drafting and revising the manuscript. All authors read and approved the final manuscript.


This work was supported by the National Natural Science Foundation of China (Grant No. 31372101).

Competing interests

The authors declare that they have no competing interests.

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Authors’ Affiliations

Department of Horticulture, Nanjing Agricultural University, Nanjing, 210095, China


  1. Aceto S, Gaudio L (2011) The mads and the beauty: genes involved in the development of orchid flowers. Curr Genomics 12:342–356View ArticleGoogle Scholar
  2. Almeida J, Rocheta M, Galego L (1997) Genetic control of flower shape in Antirrhinum majus. Development 124:1387–1392Google Scholar
  3. An H, Roussot C, Suárez-López P, Corbesier L, Vincent C, Piñeiro M, Hepworth S, Mouradov A, Justin S, Turnbull C, Coupland G (2004) CONSTANS acts in the phloem to regulate a systemic signal that induces photoperiodic flowering of Arabidopsis. Development 131:3615–3626View ArticleGoogle Scholar
  4. Anusak P, Ditta DS, Beth S, Liljergren SJ, Elvira B, Wisman E, Yanofsky MF (2003) Assessing the redundancy of MADS-box genes during carpel and ovule development. Nature 424:85–88View ArticleGoogle Scholar
  5. Bartlett ME, Specht CD (2011) Changes in expression pattern of the teosinte branched1-like genes in the Zingiberales provide a mechanism for evolutionary shifts in symmetry across the order. Am J Bot 98:227–243View ArticleGoogle Scholar
  6. Benlloch R, Berbel A, Serrano-Mislata A, Madueño F (2007) Floral initiation and inflorescence architecture: a comparative view. Ann Bot 100:659–676View ArticleGoogle Scholar
  7. Cai J, Liu X, Vanneste K, Proost S, Tsai WC, Liu KW, Chen LJ, He Y, Xu Q, Bian C, Zheng Z, Sun F, Liu W, Hsiao YY, Pan ZJ, Hsu CC, Yang YP, Hsu YC, Chuang YC, Dievart A, Dufayard JF, Xu X, Wang JY, Wang J, Xiao XJ, Zhao XM, Du R, Zhang GQ, Wang M, Su YY, Xie GC, Liu GH, Li LQ, Huang LQ, Luo YB, Chen HH, Van de Peer Y, Liu ZJ (2014) The genome sequence of the orchid Phalaenopsis equestris. Nat Genet 47:65–72View ArticleGoogle Scholar
  8. Chen X, Zhang J, Liu QZ, Guo W, Zhao TT, Ma QH, Wang GX (2014) Transcriptome sequencing and identification of cold tolerance genes in Hardy Corylus species (C. heterophylla Fisch) floral buds. PLoS One 9(9):e108604View ArticleGoogle Scholar
  9. Cheung F, Haas BJ, Goldberg SM, May GD, Xiao Y, Town CD (2006) Sequencing Medicago truncatula expressed sequenced tags using 454 Life Sciences technology. BMC Genom 7:272View ArticleGoogle Scholar
  10. Coen ES, Meyerowitz EM (1991) The war of the whorls: genetic interactions controlling flower development. Nature 353:31–37View ArticleGoogle Scholar
  11. Corley SB, Carpenter R, Copsey L, Coen E (2005) Floral asymmetry involves an interplay between TCP and MYB transcription factors in Antirrhinum. Proc Natl Acad Sci 102:5068–5073View ArticleGoogle Scholar
  12. Costa MMR, Fox S, Ai Hanna, Baxter C, Coen E (2005) Evolution of regulatory interactions controlling floral asymmetry. Development 132:5093–5101View ArticleGoogle Scholar
  13. Doebley J, Stec A, Hubbard L (1997) The evolution of apical dominance in maize. Nature 386:485–488View ArticleGoogle Scholar
  14. Endress PK (1999) Symmetry in flowers: diversity and evolution. Int J Plant Sci 160:3–23View ArticleGoogle Scholar
  15. Feng X, Zhao Z, Tian Z, Xu S, Luo Y, Cai Z, Wang Y, Yang J, Wang Z, Weng L, Chen J, Zheng L, Guo X, Luo J, Sato S, Tabata S, Ma W, Cao X, Hu X, Sun C, Luo D (2006) Control of petal shape and floral zygomorphy in Lotus japonicus. Proc Natl Acad Sci 103:4970–4975View ArticleGoogle Scholar
  16. Galego L, Almeida J (2002) Role of DIVARICATA in the control of dorsoventral asymmetry in Antirrhinum flowers. Genes Dev 16:880–891View ArticleGoogle Scholar
  17. Götz S, García-Gómez JM, Terol J, Williams TD, Nagaraj SH, Nueda MJ, Robles M, Talón M, Dopazo J, Conesa A (2008) High-throughput functional annotation and data mining with the Blast2GO suite. Nucleic Acids Res 36:3420–3435View ArticleGoogle Scholar
  18. Grabherr MG, Haas BJ, Yassour M, Levin JZ, Thompson DA, Amit I, Adiconis X, Fan L, Raychowdhury R, Zeng QD, Chen ZH, Mauceli E, Hacohen N, Gnirke A, Rhind N, Palma FD, Birren BW, Nusbaum C, Lindblad-TohK Friedman N, Regev A (2011) Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nat Biotechnol 29:644–652View ArticleGoogle Scholar
  19. Hao QN, Zhou XA, Sha AH, Wang C, Zhou R, Chen SL (2011) Identification of genes associated with nitrogenuse efficiency by genome-wide transcriptional analysis of two soybean genotypes. BMC Genom 12:1–15View ArticleGoogle Scholar
  20. Hoshino Y, Igarashi T, Ohshima M, Shinoda K, Murata N, Kanno A, Nakano M (2014) Characterization of CYCLOIDEA-like genes in controlling floral zygomorphy in the monocotyledon Alstroemeria. Sci Hortic 169:6–13View ArticleGoogle Scholar
  21. Howarth DG, Donoghue MJ (2006) Phylogenetic analysis of the ‘ECE’ (CYC/TB1) clade reveals duplications predating the core eudicots. Proc Natl Acad Sci USA 103:9101–9106View ArticleGoogle Scholar
  22. Hsu HF, Yang CH (2002) An orchid (Oncidium ‘Gower Ramsey’) AP3-like MADS genes regulate floral formation and initiation. Plant Cell Physiol 43:1198–1209View ArticleGoogle Scholar
  23. Hsu HF, Huang CH, Chou LT, Yang CH (2003) Ectopic expression of an orchid (Oncidium Gower Ramsey) AGL6-like gene promotes flowering by activating flowering time genes in Arabidopsis thaliana. Plant Cell Physiol 44(8):783–794View ArticleGoogle Scholar
  24. Hsu HF, Hsu WH, Lee YI, Mao WT, Yang JY, Li JY, Yang CH (2015) Model for perianth formation in orchids. Nat Plants 46:1–8Google Scholar
  25. Huang WT, Fang ZM, Zeng SJ, Zhang JX, Wu KL, Chen ZC, Silva TS, Duan J (2012) Molecular cloning and functional analysis of three FLOWERING LOCUS T (FT) homologous genes from Chinese Cymbidium. Int J Mol Sci 13:11385–11398View ArticleGoogle Scholar
  26. Hyun TK, Lee S, Kumar D, Rim Y, Kumar R, Lee SY, Lee CH, Kim JY (2014) RNA-seq analysis of Rubus idaeuscv. Nova: transcriptome sequencing and de novo assembly for subsequent functional genomics approaches. Plant Cell Rep 33:1617–1628View ArticleGoogle Scholar
  27. Jang S, Choi SC, Li HY, An G, Schmelzer E (2015) Functional characterization of Phalaenopsis aphrodite flowering genes PaFT1 and PaFD. PLoS One 10(8):e0134987View ArticleGoogle Scholar
  28. Li XB, Luo J, Yan TL, Xiang L, Jin F, Qin DH, Sun CB, Xie M (2013) Deep sequencing-based analysis of the Cymbidium ensifolium floral transcriptome. PLoS One 8(12):e85480View ArticleGoogle Scholar
  29. Liu XR, Pan T, Liang WQ, Gao L, Wang XJ, Li HQ, Liang S (2016) Over expression of an orchid (Dendrobium nobile) SOC1/TM3-like ortholog, DnAGL19, in Arabidopsis regulates HOS1-FT expression. Front Plant Sci 7:99Google Scholar
  30. Luo D, Carpenter R, Vincent C, Copsey L, Coen E (1996) Origin of floral asymmetry in Antirrhinum. Nature 383:794–799View ArticleGoogle Scholar
  31. Luo D, Carpenter R, Copsey L, Vincent C, Clark J, Coen E (1999) Control of organ asymmetry in flowers of Antirrhinum. Cell 99:367–376View ArticleGoogle Scholar
  32. Martin S, Johansen LB, Pederson KB, Signe F, Bo BJ (2006) Cloning and transcription analysis of an AGAMOUS- and SEEDSTICK ortholog in the orchid Dendrobium thyrsiflorum (Reichb. f.). Gene 366:266–274View ArticleGoogle Scholar
  33. Mondragón-Palomino M, Theissen G (2008) MADS about the evolution of orchid flowers. Trends Plant Sci 13:51–59View ArticleGoogle Scholar
  34. Mondragón-Palomino M, Theissen G (2009) Why are orchid flowers so diverse? Reduction of evolutionary constraints by paralogues of class B floral homeotic genes. Ann Bot 104:583–594View ArticleGoogle Scholar
  35. Mondragón-Palomino M, Theissen G (2011) Conserved differential expression of paralogous DEFICIENS- and GLOBOSA-like MADS-box genes in the flowers of Orchidaceae: refining the ‘orchid code’. Plant J 66:1008–1019View ArticleGoogle Scholar
  36. Mouradov A, Cremer F, Coupland G (2002) Control of flowering time: interacting pathways as a basis for diversity. Plant Cell 14(Suppl):111–130Google Scholar
  37. Münster T, Pahnke J, Di Rosa A, Kim JT, Martin W, Saedler H, Theissen G (1997) Floral homeotic genes were recruited from homologous MADS-box genes preexisting in the common ancestor of ferns and seed plants. Proc Natl Acad Sci 94:2415–2420View ArticleGoogle Scholar
  38. Navaud O, Dabos P, Carnus E, Tremousaygue D, Hervé C (2007) TCP transcription factors predate the emergence of land plants. J Mol Evol 65:23–33View ArticleGoogle Scholar
  39. Paolo SD, Gaudio L, Aceto S (2015) Analysis of the TCP genes expressed in the inflorescence of the orchid Orchis italica. Sci Rep 5:16265View ArticleGoogle Scholar
  40. Pelaz S, Ditta GS, Baumann E, Wisman E, Yanofsky MF (2000) B and C floral organ identity functions require SEPALLATA MADS-box genes. Nature 405:200–203View ArticleGoogle Scholar
  41. Pelaz S, Gustafson-Brown C, Kohalmi SE, Crosby WL, Yanofsky MF (2001) APETALA1 and SEPALLATA3 interact to promote flower development. Plant J 26:385–394View ArticleGoogle Scholar
  42. Preston JC, Hileman LC (2012) Parallel evolution of TCP and B-class genes in Commelinaceae flower bilateral symmetry. Evodevo 3:6View ArticleGoogle Scholar
  43. Pridgeon AM, Cribb PJ, Chase MW, Rasmussen FN (2005) Genera Orchidacearum: Epidendroideae (part one). Oxford University Press, OxfordGoogle Scholar
  44. Purugganan MD, Rounsley SD, Schmidt RJ, Yanofsky MF (1995) Molecular evolution of flower development: diversification of the plant MADS-box regulatory gene family. Genetics 140:345–356Google Scholar
  45. Rudall PJ, Bateman RM (2002) Roles of synorganization, zygomorphy and heterotopy in floral evolution: the gynostemium and labellum of orchids and other lilioid monocots. Biol Rev 77:403–441View ArticleGoogle Scholar
  46. Stebbins GL (1974) Flowering plants: evolution above the species level. Harvard University Press, CambridgeView ArticleGoogle Scholar
  47. Su CL, Chao YT, Yao-Chien AC (2011) De Novo assembly of expressed transcripts and global analysis of the Phalaenopsis aphrodite transcriptome. Plant Cell Physiol 52(9):1501–1514View ArticleGoogle Scholar
  48. Suárez-López P, Wheatley K, Robson F, Onouchi H, Valverde F, Coupland G (2001) CONSTANS mediates between the circadian clock and the control of flowering in Arabidopsis. Nature 410:1116–1120View ArticleGoogle Scholar
  49. Trick M, Long Y, Meng JL, Bancroft I (2009) Single nucleotide polymorphism (SNP) discovery in the polyploidy Brassica napus using Solexa transcriptome sequencing. Plant Biotechnol J 7:334–346View ArticleGoogle Scholar
  50. Tsai WC, Kuoh CS, Chuang MH, Chen WH, Chen HH (2004) Four DEF-like MADS box genes displayed distinct floral morphogenetic roles in Phalaenopsis orchid. Plant Cell Physiol 45:831–844View ArticleGoogle Scholar
  51. Wang ZY, Tobing EM (1998) Constitutive expression of the CIRCADIAN CLOCK ASSOCIATED l (CCA1) gene disrupts circadian rhythms and suppresses its own expression. Cell 93:1207–1217View ArticleGoogle Scholar
  52. Wang Z, Luo Y, Li X, Wang L, Xu S, Yang J, Weng L, Sato S, Tabata S, Ambrose M, Rameau C, Feng X, Hu X, Luo D (2008) Genetic control of floral zygomorphy in pea (Pisum sativum L.). PNAS 105(30):10414–10419View ArticleGoogle Scholar
  53. Weigel D, Meyerowitz EM (1994) The ABCs of floral homeotic genes. Cell 78:203–209View ArticleGoogle Scholar
  54. Wolff E (1999) Chinese Cymbidium species. Orchids: the Magazine of the American Orchid Society 68(7):682–693Google Scholar
  55. Xiang L, Qin DH, Li XB, Li BJ, Guo FQ, Wu C, Sun CB (2011) Cloning and expression analysis of B Class MADS-box genes from Cymbidium faberi. Acta Hortic Sin 38:143–147Google Scholar
  56. Xu YF, Teo LL, Zhou J, Prakash PK, Yu H (2006) Floral organ identity genes in the orchid Dendrobium crumenatum. Plant J 46:54–68View ArticleGoogle Scholar
  57. Yan L, Wang X, Liu H, Tian Y, Lian J, Yang R, Hao S, Wang X, Yang S, Li Q, Qi S, Kui L, Okpekum M, Ma X, Zhang J, Ding Z, Zhang G, Wang W, Dong Y, Sheng S (2015) The genome of Dendrobium officinale illuminates the biology of the important traditional Chinese orchid herb. Mol Plant 8(6):922–934View ArticleGoogle Scholar
  58. Yano M, Katayose Y, Ashikari M, Yamanouchi U, Monna L, Fuse T, Baba T, Yamamoto K, Umehara Y, Nagamura Y, Sasaki T (2000) Hd1, a major photoperiod sensitivity quantitative trait locus in rice, is closely related to the Arabidopsis flowering time gene CONSTANS. Plant Cell 12:2473–2484View ArticleGoogle Scholar
  59. Yates SA, Swain MT, Hegarty MJ, Chernukin I, Lowe M, Allison GG, Ruttink T, Abberton MT, Jenkins G, Skøt Leif (2014) De novo assembly of red clover transcriptome based on RNA-Seq data provides insight into drought response, gene discovery and marker identification. BMC Genom 15:453View ArticleGoogle Scholar
  60. Ye J, Fang L, Zheng H, Zhang Y, Chen J, Zhang Z, Wang J, Li S, Li R, Bolund L, Wang J (2006) WEGO: a web tool for plotting GO annotations. Nucleic Acids Res 34:293–297View ArticleGoogle Scholar
  61. Yu H, Yang SH, Goh CJ (2000) DOH1, a class 1 knox gene, is required for maintenance of the basic plant architecture and floral transition in orchid. Plant Cell 12(11):2143–2159View ArticleGoogle Scholar
  62. Yu H, Yang SH, Goh CJ (2002) Spatial and temporal expression of the orchid floral homeotic gene DOMADS1 is mediated by its upstream regulatory regions. Plant Mol Biol 49(2):225–237View ArticleGoogle Scholar
  63. Yuan Z, Gao S, Xue DW, Luo D, Li LT, Ding SY, Yao X, Wilson ZA, Qian Q, Zhang DB (2009) RETARDED PALEA1 controls palea development and floral zygomorphy in rice. Plant Physiol 149:235–244View ArticleGoogle Scholar
  64. Zhang JX, Wu KL, Zeng SJ, Silva JT, Silva JA, Xia HQ, Tian CE, Duan J (2013) Transcriptome analysis of Cymbidium sinense and its application to the identification of genes associated with floral development. BMC Genom 14:1–17View ArticleGoogle Scholar
  65. Zhang HN, Wei YZ, Shen JY, Lai B, Huang XM, Ding F, Su ZX, Chen HB (2014) Transcriptomic analysis of floral initiation in litchi (Litchi chinensis Sonn.) based on de novo RNA sequencing. Plant Cell Rep 33:1723–1735View ArticleGoogle Scholar
  66. Zhang GQ, Xu Q, Bian C, Tsai WC, Yeh CM, Liu KW, Yoshida K, Zhang LS, Chang SB, Chen F, Shi Y, Su YY, Zhang YQ, Chen LJ, Yin Y, Lin M, Huang H, Deng H, Wang ZW, Zhu SL, Zhao X, Deng C, Niu SC, Huang J, Wang M, Liu GH, Yang HJ, Xiao XJ, Hsiao YY, Wu WL, Chen YY, Mitsuda N, Ohme-Takagi M, Luo YB, Van de Peer Y, Liu ZJ (2016) The Dendrobium catenatum Lindl. genome sequence provides insights into polysaccharide synthase, floral development and adaptive evolution. Sci Rep 6:19029View ArticleGoogle Scholar


© The Author(s) 2016