Introduction

The genus Calanthe, a member of the Orchidaceae, is widely distributed in tropical and subtropical regions, including Africa, the Pacific Islands, Southeast Asia, China, Japan, and Korea (Gale and Drinkell 2007). Although this genus comprises about 171 species worldwide, only four species (C. discolor Lindl, C. sieboldii Ohwi, C. reflexa Maximowicz, and C. aristulifera Hayata) and a natural hybrid between C. discolor and C. sieboldii (C. bicolor Lindl) are reported to occur naturally in Korea (Lee and Kwack 1983; Kim and Kim 1989; Hyun et al. 1999a, b). Depending on taxonomic and nomenclatural aspects, C. sieboldii is often classified into C. discolor f. sieboldii (Dence) Ohwi (Iwatsuki 1995; Kim et al. 2008; Korea Plant Name Index 2007). Among them, the most popular and captivating species are C. discolor and C. sieboldii. While these two species do not exhibit significant morphological differences during vegetative growth, the flowers of C. sieboldii are slightly larger and exude a stronger fragrance than those of C. discolor. Furthermore, the flowers of C. sieboldii are uniformly yellow, whereas those of C. discolor bear bicolored flowers with brownish outer tepals (often called sepals) and lateral inner tepals (petals), and a white median inner tepal (lip).

Flower color is one of the most important traits in ornamental plants. It is well established that flavonoids, carotenoids, and betalains are largely responsible for the broad range of flower colors, from yellow to red, violet, and blue. Carotenoids mainly contribute to the production of yellow and orange pigments, such as those present in sunflowers and tomatoes (Bartley and Scolnik 1989). Betalains are yellow to red nitrogenous compounds distributed only in Caryophyllales (Stafford 1994). Anthocyanins, a class of flavonoids, are the major floral pigments in higher plants and accumulate in vacuoles (Goto and Kondo 1991). The biosynthetic pathways underlying the production of these pigments are generally conserved among plant species, and the genes that encode the enzymes involved in pigment biosynthesis have mostly been isolated (Holton and Cornish 1995; reviewed in Tanaka and Ohmiya 2008). Flower coloration is also thought to be specified by temporal and spatial expression patterns of the regulatory genes (Ludwig and Wessler 1990; Holton and Cornish 1995), suggesting that detailed studies of the transcriptional control of biosynthetic genes may lead to the identification of novel regulatory components.

Flowers of the Calanthe vary greatly in terms of color, shape, fragrance, and longevity. Therefore, conventional breeding programs have been conducted to improve the horticultural traits and genetic diversity within a species of Calanthe or a range of closely related Calanthe species. Recently, a molecular linkage map and quantitative trait loci (QTL) analysis identified four QTLs that are closely associated with flower and lip color expression in C. discolor, C. sieboldii, and variants (Cho et al. 2009). However, the molecular mechanisms underlying key biological processes and the identification of the critical components contributing to phenotypic variation remain largely unknown. Most molecular studies of Calanthe have been limited to taxonomic classification (Hyun et al. 1999a, b; Gale and Drinkell 2007; Silveira et al. 2008). Indeed, all DNA sequences deposited for this genus in Genbank thus far belong to molecular markers with applications in phylogenic studies. To identify the genes that play key roles in floral development and encode proteins that participate in the underlying metabolic pathways, we analyzed the differentially expressed genes (DEGs) that contribute to the distinct characteristics of flowers from the two Calanthe species, C. discolor and C. sieboldii. Based on differential display RT-PCR (DDRT-PCR) and gene-specific RT-PCR, a subset of key regulatory genes, presumably involved in a signaling pathway, secondary metabolism, and hormone pathways, were identified as being differentially expressed in the floral organs of the two species. Our data presented here may be a useful resource for scientists focusing on orchid floral development and biotechnology.

Materials and Methods

Plant Materials

Two Calanthe species, C. discolor Lindl and C. sieboldii Ohwi, were collected from Jeju Island in Korea and cultivated for over 5 years in the greenhouse facility of Chonbuk National University. Young flower buds were harvested just before flowering in late April from five individual plants of each species, and the buds from each species were combined and kept at −70°C until used. Total RNA was extracted from the two sets of combined samples.

Total RNA Extraction, DDRT-PCR, and Semiquantitative RT-PCR

Total RNA was extracted using Tri reagent (MRC, USA) and then treated with RQ1-DNase I (Promega, Germany) to remove DNA contamination, according to the manufacturers’ instructions. Samples were then analyzed by DDRT-PCR technology using the Gene Fishing Kit (Seegene, Inc., Korea), according to the manufacturer’s instruction. Briefly, 3 μg of total RNA was reverse transcribed using the dT-ACP1 primer and M-MLV reverse transcriptase. Then, 1/30 of the first-stranded cDNA samples was used as a template for PCR in combination with 120 arbitrary ACPs and the dT-ACP2 primer set. PCR conditions were as follows: 1 cycle of 94°C for 5 min, 50°C for 3 min, and 72°C for 1 min; 40 cycles of 94°C for 40 s, 65°C for 40 s, and 72°C for 40 s; and a final extension period of 72°C for 5 min. The resulting PCR products were separated and visualized on a 1.2% agarose gel. After ethidium bromide staining, the differentially expressed products were excised from the gel and cloned into a pBluescript-T vector that is modified for cloning of PCR products (kind gift from Dr. Hwang Inhwan, POSTECH, Korea) for DNA sequencing.

Semiquantitative RT-PCR was performed to confirm the differential expression of the identified cDNA clones using the gene-specific primers listed in Table 1. Total RNA (3 μg) was used for cDNA synthesis, and 1/10 of the resulting volume of cDNA was used as template for the PCR. PCR conditions were as follows: 25 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 30 s for the actin cDNA clone; and 30 cycles of 94°C for 30 s, 55 ∼ 62°C for 30 s depending on clones, and 72°C for 30 s for other cDNA clones. A partial cDNA fragment of the actin gene was obtained from both species by RT-PCR using degenerate primers, 5′-GARAARATGACNCARATHATG-3′ and 5′-TCNACRTCRCAYTTCATDAT-3′. After the PCR products were sequenced, primers specific for the actin gene were designed (Table 1). The constitutively expressed actin mRNA was used as a control. PCR products were separated on a 1.2% agarose gel, and the gel was photographed using a digital imaging system.

Table 1 Sequences of the gene-specific primers used for RT-PCR analysis
Full size table

DNA Sequencing and Sequence Analysis

The nucleotide sequences of PCR products cloned into pBluescript-T vector were determined using a directional vector-specific T7 primer. Sequence data were processed with Vector NTI 7 software (InforMax, Inc.) to remove the primer sequence, the polyA tail, and ambiguous sequences. By assembling the sequence data, overlapping cDNA clones were excluded from further analysis. The BLASTn and BLASTx algorithms in the NCBI database (http://www.ncbi.nlm.nih.gov/database, accessed on Jan 13, 2009) were used to detect similarities between all edited sequences and previously deposited sequences. Similarities to known sequences were considered significant when the E values were less than 10–4 for sequences of more than 50 nucleotides in length. Finally, the unique sequences reported in this paper were deposited in NCBI’s GenBank database under the accession numbers provided in Tables 2 and 3.

Table 2 List of clones upregulated in C. discolor flower buds and the results of a BLAST homology search
Full size table
Table 3 List of clones upregulated in C. sieboldii flower buds and the results of a BLAST homology search
Full size table

Results and Discussion

Analysis of Differentially Expressed Genes in Flowers of Two Calanthe Species

Modulation of floral development and of metabolic pathways related to shape, color, fragrance, and lifespan are tightly regulated by endogenous and environmental factors, which undoubtedly involve complex combinatorial gene expressions and specific function of genes (reviewed in Tanaka and Ohmiya 2008; Mondragón-Palomino and Theissen 2009). Although floral development and the acquisition of floral traits have been extensively characterized in model species, such as rice and Arabidopsis, they have yet to be deciphered in orchids, which produce complex flowers that differ markedly from those of all other plant families.

To investigate the regulatory mechanisms underlying orchid flower development and the metabolic pathways that dictate floral traits, we compared the expression pattern of genes in the floral organs of two closely related Calanthe species, C. discolor and C. sieboldii. While two species show high similarities in vegetative morphology, the flowers differ in terms of color, size, shape, and fragrance. Since gene expression can be influenced by subtle changes or differences among individuals and/or by environmental conditions, RNA was extracted from pools of flower buds derived from five individuals of each species. For an accurate and reproducible analysis, we employed annealing controlled primer (ACP)-based RT-PCR technology (Kim et al. 2004) to reduce false-positive signals resulting from the nonspecific annealing of short arbitrary primers. Using 120 ACPs and anchored oligo-dT (dT-ACP2) primer combinations, a total of 110 differential PCR fragments were observed on the agarose gels. Representative gel images are shown in Fig. 1. Of these, PCR bands consisting of more than 150 nucleotides of cDNA were excised from gels and cloned into the pBluescript-T vector. Clones harboring less than 150 nucleotides were excluded from further analysis since we found that the short cDNA fragment from 3′-end of polyA tail generally showed no significant similarity in a BLAST search analysis. In some cases, a second round of PCR amplification was performed to yield a sufficient amount of DNA for the cloning. The nucleotide sequences of the cDNA fragments cloned into the pBluescript-T plasmid were determined by single-pass sequencing using the vector-specific T7 primer. After removing poor-quality sequence data and overlapped sequences, 26 DEGs from C. discolor and 40 from C. sieboldii were obtained (Tables 2 and 3, respectively). A BLAST homology search revealed that 48 clones (72%) exhibited significant similarity to known genes from other organisms. However, 18 clones (28%) had no significant similarity to any sequences in the public database. This might be due to the absence of sequence information or due to the unique genes within Calanthe species. Taken together, our analysis suggests that a wide range of genes, potentially involved in diverse functions, is differentially expressed in floral organs of these two Calanthe species.

Fig. 1
figure 1

Representative agarose gel images of DD RT-PCR analysis. Total RNA from flower buds of C. discolor (D) and C. sieboldii (S) were used for RT-PCR using ACPs (SeeGene, Inc.), as described in “Materials and Methods.” The actin gene was amplified as a control. Cloned DEGs are labeled as follows: a Cs003; b Cs004; c Cs006; d Cs007; e Cd006; f Cd007; g Cd008; h Cs015; i Cs019; j Cs020; k Cd024; l Cd025; m Cs012; n Cs021; o Cs034; and p Cs039

Full size image

Confirmation of DEGs by Gene-specific RT-PCR

Semiquantitative RT-PCR using gene-specific primers was performed to confirm the differential expression patterns of 14 representative DEG clones. Since most clones are partial fragments corresponding to the 3′ region of cDNA, which contains the polyA tail and a high AT content, cycle parameters and annealing temperatures of the RT-PCR were optimized independently for each clone. All primers could be used to amplify the target mRNA from both species. RT-PCR results demonstrated that the differential expression patterns were reproducible for the tested clones.

A significant proportion of DEGs from this study share similarity with genes that encode enzymes involved in primary and secondary metabolism, hormonal pathways, and signal transduction pathways. The phenylpropanoid pathway contributes to the synthesis of thousands of compounds, including flavonoids, such as flavonols, anthocyanidins, and tannins (Hamberger and Hahlbrock 2004), and is known to be involved in the control of floral fragrances (Knudsen et al. 1993). The 4-coumarate:CoA-ligase protein (4CL, EC 6.2.1.12) catalyzes the conversion of hydroxycinnamic acids to hydroxycinnamoyl-CoA thioesters, precursors of a variety of phenylpropanoid biosynthetic derivatives. The Cs004 clone shares high-sequence similarity with a gene that encodes 4-coumarate:CoA-ligase, and this gene is preferentially expressed in the floral buds of C. sieboldii and only slightly expressed in those of C. discolor (Figs. 1 and 2). Since C. sieboldii produces uniformly yellow color flower with a stronger fragrance than those of C. discolor, the implication of Cs004 regarding to these phenotypes remains to be interested. Furthermore, we identified the Cd025 clone, which encodes a protein with high sequence similarity to flavonoid 3′-hydroxylase (F3′H), a member of the cytochrome P450 family that controls flower color. Unlike the Cs004 clone, its expression was higher in floral buds of C. discolor than in those of C. sieboldii (Figs. 1 and 2), suggesting its possible roles in the biosynthetic pathways contributing to floral phenotypes in C. discolor.

Fig. 2
figure 2

Semiquantitative RT-PCR analysis of representative clones obtained by DD RT-PCR analysis. RT-PCR was performed using the gene-specific primers listed in Table 1. Clone names are represented by the protein names that showed the highest similarity in a BLAST homology search. Actin was used as the internal control for RT-PCR

Full size image

S-adenosylmethionine synthase (SAMS, EC 2.5.1.6) is a key enzyme in the biosynthesis of polyamines and ethylene (Yang and Hoffman 1984) that catalyzes the biosynthesis of SAM from Met and ATP. It has been reported that SAMS genes play critical roles during ovary development and floral senescence and also in the plant’s response to treatment with phytohormones (Gómez-Gómez and Carrasco 1998). Interestingly, the Cd024 and Cs012 clones, which encode putative SAMS, were found to be differentially expressed, with greater expression in the floral buds of C. sieboldii and C. discolor, respectively (Figs. 1 and 2). Further sequence analysis using the PCR products demonstrated that a high level of species conservation of Cs012 and Cd024 exists between two species, as both the nucleotide and deduced amino acid sequences share 99.5% and 100% identity, respectively (Fig. 3a). However, the nucleotide and deduced amino acid sequences of Cd012 and Cd024 clones share about 73% and 85% identity with each other, respectively (Fig. 3b), suggesting that these two clones belong to isoforms of putative SAMS gene family. Therefore, it is interesting to evaluate that the unique expression patterns of these genes may impact on metabolic regulation during floral development in each of the Calanthe species.

Fig. 3
figure 3

Alignment of the nucleotide (a) and deduced amino acid (b) sequences corresponding to the coding regions of the Cs012 and Cd024 clones. Differences in sequence between the two species are represented in shadow boxes. dCs012 and sCs012 indicate the Cs012 clone isolated from each C. discolor and C. sieboldii, respectively. dCd024 and sCd024 indicate the Cd012 clone isolated from each C. discolor and C. sieboldii, respectively. Asterisks indicate the S-adenosylmethionine synthetase C-terminal domain (pfam02773) with E value of 8e-44 identified by NCBI conserved domain search

Full size image

The clones Cs015, Cs019, Cs034, Cd002, and Cd008 showed strong similarity to genes encoding indole-3-acetic acid 1 (IAA1), ARF, SAR GTPases, and two calmodulin isoforms, respectively. The functions of these proteins in controlling floral development are largely unknown at present. However, auxin plays an essential role in the initiation of floral primordia and organ outgrowth, the identity of floral organs, and the development of the female reproductive organs of angiosperms (Vernoux et al. 2000; Nemhauser et al. 2000; Okada et al. 1991; Pfluger and Zambryski 2004). The IAA1 is believed to have a role in auxin signal transduction. GTPase and calmodulin are known to act as upstream regulators in diverse signal transduction pathways (Memon 2004; Yang and Poovaiah 2003). The differential expression pattern of these genes (Figs. 1 and 2) is of great interest, since it can help in assigning their possible functions in floral development of two Calanthe species.

Although the subset of genes identified in this study makes an initial contribution to the field of functional genomics in Calanthe orchids, the identification of differentially expressed genes may help in the understanding of metabolic pathways possibly involved in flowering and flower architecture, biosynthesis of pigments and fragrances, and the lifespan of flowers. The described genes in this study are currently being studied in search for candidates that might regulate floral development allowing us to step forward in the genetic manipulation of Calanthe species. Therefore, further research into the biological functions of these genes may greatly enhance our understanding of floral development and contribute to the pool of available plant biotechnology resources.