Introduction

The developmental transition from the vegetative to reproductive phase is a major developmental stage in the plant life cycle and critical to reproductive success. In higher plants, four major pathways: photoperiod, vernalization, hormone and autonomous regulation act to promote flowering (Levy and Dean 1998; Simpson et al. 1999; Shen et al. 2012). These multiple pathways form a regulatory network that integrates the developmental state of plants with responses to environmental signals in order to control flowering time (Kardailsky et al. 1999; He and Amasino 2005; Baurle and Dean 2006). These signals are transmitted via integrator genes, such as FLOWERING LOCUS T (FT) and SUPPRESSOR OF OVER-EXPRESSION OF CONSTANS1 (SOC1) to the floral meristem identity genes APETALA1 (AP1) and LEAFY (LFY) at the apical meristem. Of these two genes, FT, a well known floral integrator gene plays an important role in controlling flowering time (Bradley et al. 1997; Boss et al. 2004).

FT encodes for a small protein similar to human phosphoethanolamine binding protein (PEBP) and Raf kinase inhibitor protein (RKIP) which belongs to PEBP superfamily. The plant PEBP genes group into six FT-like sub-families in Arabidopsis including FT, TWIN SISTER OF FT (TSF), BROTHER OF FT (BFT), ARABIDOPSIS THALIANA RELATIVE OF CENTRORADIALIS (ATC), MOTHER OF FT (MFT) and TERMINAL FLOWER1 (TFL1) (Bradley et al. 1997; Mimida et al. 2001; Yoo et al. 2004; Yamaguchi et al. 2005; Franziska et al. 2008; Hedman et al. 2009; Mohamed et al. 2010). Thus, the six genes comprise a small gene family (herein referred to as the FT/TFL1 gene family) in the Arabidopsis genome. It is likely that the members of the family are important regulators for flowering time because constitutive expression of some of these members modulated flowering time (Mimida et al. 2001). FT is a floral promoter and acts in the photoperiod-dependent pathway (Kardailsky et al. 1999). FT is regulated by CONSTANS (CO). Expression of CO is upregulated, which in turn induces expression of FT (An et al. 2004). Ectopic expression of either CO or FT causes early flowering. Recent report suggest that FT protein is the mobile flower-inducing signal that moves from an induced leaf through phloem to the shoot apex (Jager and Wigge 2007; Lin et al. 2007; Tamaki et al. 2007). FT protein is produced within the phloem of leaves and acted in the shoot apex where it interacts with the bZIP transcription factor FLOWERING LOCUS D (FD) to form a FT/FD heterodimer, that promotes the transition to flowering by activating suppression of over-expression of SOC1 and AP1 (Abe et al. 2005; Wigge et al. 2005). In Arabidopsis, FT is not only an output of the photoperiodic pathway, but also of the autonomous and vernalization pathways. The effect of vernalization is downregulation of FLOWERING LOCUS C (FLC), thus allowing FT to be expressed in the leaves and increasing the ability of the shoot apex to respond to FT (Michaels et al. 2005; Searle et al. 2006; Zeevarrt 2008).

TFL1 has an antagonistic regulatory effect on FT which in turns affects flowering. Mutant plants where TFL1 has ceased to function resulted in early flowering independently of day length led to the germination of a terminal flower, similar to FT-over-expressing plants (Bradley et al. 1997). MFT sequence is almost equally similar to both FT and TFL1, making it difficult to predict MFT function. Enaction of MFT was unclear, there were no obvious phenotypic changes in the MFT-1 mutant, but over-expression of MFT-1 slightly accelerated flowering time in Arabidopsis, suggesting it might act redundantly to control flowering time (Yoo et al. 2004).

Dendrobium nobile Lindl is an economically important orchid in flower markets globally with a high ornamental and medicinal value. Little research has been reported on the functions of flowering time genes of orchids (Hou and Yang 2009). In this study, we presented the isolation of the DnFT and DnMFT genes and their expression in D. nobile Lindl and, their role in transgenic Arabidopsis plants were also studied. The results showed that ectopic expression of DnFT led to early flowering in Arabidopsis through inducing the expressions of AP1 and LFY. Over-expression of DnMFT in transgenic Arabidopsis plants gave rise to a slightly late-flowering phenotype.

Materials and methods

Plant materials and growth conditions

The orchid D. nobile Lindl, was grown in greenhouses at the Orchid Research Center at the South China Normal University. Plants with finished floral bud differentiation were selected. The selected orchid plants were transferred to a growth chamber under a low-temperature environment (12 ± 2 °C during the day, 9 ± 1 °C at night) for 40 days. The buds and leaves were collected on days: 0, 5, 10, 20, 30 and 40 under cold treatment for RNA isolation. Meanwhile, the buds and leaves without cold treatment were selected as a control.

The Arabidopsis (ecotype Columbia) seedlings were grown in a greenhouse under long days (LD) conditions (16 h light/8 h dark at 22 °C) for future transformation.

Cloning of DnFT and DnMFT genes in D. nobile Lindl

DNA fragments of DnFT and DnMFT genes were obtained from a previously constructed cDNA library by sequencing (unpublished). The 3′ and 5′ end of these genes were then isolated with a 3′ and 5′RACE cDNA Amplification Kit (TaKaRa, Japan). The gene-specific primers for 3′ and 5′RACE of DnFT were 3′RACE outer: 5′-TCCAGATGCTCCAAGTCCAA-3′, 3′RACE inner 5′-CAACGACCAGCGCGACATTT-3′; 5′RACE outer: 5′-AAATGTCGCGCTGGTCGTTG-3′, 5′RACE inner: 5′-TTGGACTTGGAGCATCTGGA-3′, respectively. Gene-specific primers for 3′ and 5′RACE of DnMFT were 3′RACE outer: 5′-ATCAATGGCGGTCAACCCAC-3′, 3′RACE inner: 5′-GTGAGCCCAGAATGAGGGAA-3′; 5′RACE outer: 5′-TTCCCTCATTCTGGGCTCAC-3′, 5′RACE inner: 5′-GTGGGTTGACCGCCATTGAT-3′, respectively. Full-length cDNA for DnFT was obtained by PCR amplification using the forward primer: 5′-AAATCTAGAATGAGTAGAGAGAGAGACCC-3′ and reverse primer: 5′-AAAGGATCCCAGTCTTGCATTCTTCTTCCG-3′. The cDNA for DnMFT was obtained by PCR amplification using the forward primer: 5′-AAATCTAGAATGGCCTCAACCTTTGTGGA-3′ and the reverse primer: 5′-AAAGGATCCACTAGCGGCGGCGGTTGTT-3′. The forward and reverse primers for DnFT and DnMFT contained a BamHI(GGATCC) and a XbaI(TCTAGA) recognition site respectively to facilitate the cDNAs subsequent cloning.

Sequence analysis

Sequences of PEBP proteins were downloaded from GenBank for comparison. The alignment was performed using programs from DNAssist Package version 2 and CLUSTAL W. (http://www-igbmc.u-strasbg.fr/BioInfo/ClustalX). Sequence relatedness was analyzed using CLUSTAL X following the neighbor-joining method (Goloveshkina et al. 2012).

Quantitative real-time RT-PCR analysis

Quantitative real-time RT-PCR was carried out using SYBR Premix Ex Taq Kit (TaKaRa, Shiga, Japan). The amplification conditions were 95 °C for 30 s, followed by 40 cycles of amplification (95 °C for 5 s, 60 °C for 20 s) with plate reading after each cycle. Primer pairs used for quantitative real-time RT-PCR detection were as follows: DnFT: forward primer (5′-GCCAAGCCTAGGCATACATCGC-3′) and reverse primer (5′-CAGTCTTGCATTCTTCTTCCGCC-3′); DnMFT: forward primer (5′-ACCGCTGGGCATCCACCG-3′) and reverse primer (5′-GGTTGTTGGGCTCTCGCTGG-3′). UBQ (Ubiquitin) was used for normalization with forward primer (5′-CGCCGTCAACCTCATTCCAT-3′) and reverse primer (5′-GTGAGGTAGCGACCGTGGC-3′). All the experiments were performed in triplicates. Semi-quantitative RT-PCR was carried out to detect AP1 and LFY RNA in the transformed plant. AP1: forward primer (5′-ATGGGAAGGGGTAGGGTTCAATTG-3′) and reverse primer (5′-ATGCTGTTTTGCTCCTGTATGG-3′); LFY: forward primer: 5′-GCTAAAGACCGTGGCGAA-3′ and reverse primer: 5′-GCATCCACCACGTCCAGA-3′.

Plant transformation and transgenic plant analysis

XbaI and BamHI fragments containing the full length cDNA for DnFT and DnMFT were cloned into the binary vector pCAMBIA1390 under the control of the CaMV35S promoter. Arabidopsis columbia plants were transformed by a floral dip method as described earlier (Clough and Bent 1998). Putative transformants that survived in a medium containing hygromycin (30 mg L−1) were further selected by RT-PCR analyses. The hygromycin-resistant seeds obtained from the T1, T2 or T3 generations (primary transformants were defined as T1) were sown on agar plates followed by stratification at 4 °C for 3 days and then grown in a growth chamber controlled at 22 °C under LD.

Southern blot analysis

For Southern blot analysis, 10 μg of genomic DNA per sample was digested with HindIII restriction enzyme at 37 °C overnight, separated on a 0.8 %(w/v) agarose gel, and transferred to Hybond N+ membranes. A digoxigenin (DIG) labeled DnFT and DnMFT cDNA fragment was synthesised by PCR and used as the probe. Prehybridization, washing, and chemiluminescent detection of the blots was performed according to the manufacturer’s instructions (Roche Diagnostics GmbH, Mannheim, Germany). The primers used for amplification of the DnFT fragment were forward 5′-AAATCTAGAATGAGTAGAGAGAGAGACCC-3′, reverse 5′-AAAGGATCCCAGTCTTGCATTCTTCTTCCG-3′; and DnMFT fragment were forward 5′-AAATCTAGAATGGCCTCAACCTTTGTGGA-3′, reverse 5′-AAAGGATCCACTAGCGGCGGCGGTTGTT-3′.

Results

Protein sequence analysis of DnFT and DnMFT

Two initial fragments from the cDNA library that exhibited sequence similarities to other PEBP/RKIP genes (E ≤ e−19) were selected, and 5′ and 3′RACE were used to isolate their full-length cDNAs. DnFT cDNA encoded for a-176-amino acid protein that showed 94, 91, 83, 81 and 74 % identity to Cymbidium goeringii CgFT, Oncidium Gower Ramsey OnFT, Populus nigra PnFT, rice Hd3a and Arabidopsis FT, respectively (Fig. 1). The conserved key amino acid residue Tyr in FT homologs was indentified at position 84 of the DnFT protein (Fig. 1), DnFT also contained two highly similar sequence to Arabidopsis FT in the 14-AA stretch referred as “segment B” and in the LYN triad in “segment C” (Ahn et al. 2006) (Fig. 1). The cDNA sequence of DnMFT encoded for a-173-amino acid protein that showed 67, 64, 61 % identity to Citrus unshiu CuMFT, Hordeum vulgare HvMFT and Arabidopsis thaliana AtMFT, respectively. DnFT and DnMFT showed 50.9 % sequence similarity.

Fig. 1
figure 1

Alignment of amino acid sequences of DnFT and DnMFT with other closely related PEBP protein. Shaded in purple are amino acid positions identical in all sequences. Shaded in green are amino acid positions identical in two or three sequences. Blue squares indicate the 14-amino-acid-stretch (segment B) and the LYN triad in segment C, and a highly conserved amino acid Tyr among FT orthologs are indicated. AtMFT (Arabidopsis thaliana), AtFT (Arabidopsis thaliana),CgFT (Cymbidium goeringii), DnFT (Dendrobium nobile), DnMFT (Dendrobium nobile), HaFT1 (Helianthus annuus), Hd3a (Oryza Sativa), HvMFT (Hordeum vulgare), OnFT (Oncidium Gower Ramsey), PnFT (Populus nigra). (Color figure online)

Full size image

To investigate the phylogenetic relationships of PEBP proteins, a phylogenetic tree was constructed with several other FT/MFT orthologs (Fig. 2). The results revealed that DnFT had a close relation to OnFT in Oncidium Gower Ramsey, while DnMFT was closely related to HvMFT in Hordeum vulgare.

Fig. 2
figure 2

Phylogenetic relationships among selected PMBP proteins. The phylogenetic tree was constructed based on the full sequences of FT/MFT proteins by the maximum likelihood method using the CLUSTAL X software. In the scale, the bar 0.1 is equal to 10 % sequence divergence

Full size image

Expression pattern of DnFT and DnMFT

To determine whether the transcriptional regulation of DnFT and DnMFT was involved in floral transition, quantitative RT-PCR was performed at different developmental stages of the buds and leaves. At the same time, the expression of the two genes was investigated under a low temperature (10 °C for 40 days) treatment that mimics the vernalization which induced flowering in D. nobile Lindl. The two genes transcripts were detectable both in buds and leaves under field conditions (Fig. 3). The expression of DnFT decreased in buds during flower bud development (Fig. 3b), while its transcript level increased in leaves (Fig. 3a). Under low temperature treatment, the DnFT transcript in buds accumulated to a lesser degree than under normal conditions (Fig. 3b), while higher DnFT expression was detected in the leaves compared to that found under normal conditions (Fig. 3a). The expression of DnMFT increased in buds and decreased in leaves during flower bud development compared to normal conditions and its transcription did not change significantly under cold treatment (Fig. 3c, d).

Fig. 3
figure 3

Detection of expression of DnFT and DnMFT by real-time PCR. a Quantification of the expression of DnFT during leaf development. b Quantification of the expression of DnFT during bud development. c Quantification of the expression of DnMFT during leaf development. d Quantification of the expression of DnMFT during bud development

Full size image

Constitutive expression of DnFT led to early flowering in Arabidopsis

To test whether DnFT was involved in the regulation of flowering time, DnFT cDNA driven by a CaMV35S promoter was transformed into wild-type Arabidopsis columbia plants. After antibiotic selection, transgenic seedlings were used to perform a RT-PCR experiment to check if DnFT was overexpressed in those plants (data not shown). Thirty independent 35S::DnFT transgenic Arabidopsis T1 plants were obtained. Three lines of 35S::DnFT in the T3 generation were grown under LD conditions and their phenotypes were examined. In addition, we performed Southern-blot with the DNA extracted from the 3 lines. As we used HindIII for DNA restriction and there was only one HindIII site in the T-DNA region, the number of hybridization band(s) represented transgene copies. Thus we found that all 3 lines of 35S::DnFT transgenic plants had single copy T-DNA integration (Fig. 4a).

Fig. 4
figure 4

Southern-blot analysis of the transgene integration and phenotypes of transgenic Arabidopsis ectopically expressed DnFT and DnMFT under long-day conditions. a Southern-blot with DnFT probe for DnFT transgenic plants, b Southern-blot with DnMFT probe for DnMFT transgenic plants, c 25-day-old wild plant (right), 35S::DnFT-3, 35S::DnFT-10 and 35S::DnFT-24 d 29-day-old wild plant (right), 35S::DnMFT-2, 35S::DnMFT-3 and 35S::DnMFT-8

Full size image

Transgenic plants flowered (derived from 35S::DnFT-3) 14.44 ± 0.73 days after sowing by producing 4.67 ± 0.5 rosette leaves; transgenic plants (35S::DnFT-10) flowered at 17 ± 1.07 days after sowing by producing 6.5 ± 1.07 rosette leaves; transgenic plants (35S::DnFT-24) flowered at 13.11 ± 0.78 days after sowing by producing 5.33 ± 0.71 rosette leaves (Table 1) while the flowering times of wild-type plants were about 24 ± 1.38 days, producing 10.5 ± 1.26 rosette leaves (Table 1). In contrast to wild-type plants, the inflorescence was terminated by two or three flowers being produced at the end of the inflorescence in the 35S::DnFT plants (Fig. 4c). These results revealed that constitutive expression of DnFT accelerated phase transition in Arabidopsis. To explore whether the early flowering phenotype was correlated with DnFT expression in 35S::DnFT transgenic plants, RT-PCR analysis was performed. As shown in Fig. 5, higher DnFT expression was observed in the 35S::DnFT transgenic plants. Further analysis indicated that the promotion of flowering time in 35S::DnFT transgenic plants was also correlated with significant up-regulation of the flower meristem identity gene AP1 and LFY in transgenic plants (Fig. 5).

Table 1 Comparison of flowering time, leaf number of wild-type and transgenic Arabidopsis plants and ectopic expression of DnFT and DnMFT
Full size table
Fig. 5
figure 5

Ectopically express DnFT activates downstream genes AP1 and LFY in Arabidopsis. Three independent transgenic lines were used for the analysis

Full size image

Constitutive expression of DnMFT delayed flowering in Arabidopsis

To explore whether DnMFT was also able to regulate flowering transition, DnMFT driven by the CaMV35S promoter was transformed into wild-type Arabidopsis. Sixteen independent 35S::DnMFT transgenic T1 plants were obtained. Among them, three lines of 35S::DnMFT in the T3 generation were selected and grown under LD conditions. Southern-blot analysis of the 3 transgene lines revealed that the number of T-DNA copies in each line were between 2 and 3 (Fig. 4b). In these transgenic lines, the flowering time was delayed by 1–2 days compared to the wild-type. 35S::DnMFT-2 plants flowered at 11.25 ± 1.04 rosette leaves, 35S::DnMFT-3 plants flowered at 11.38 ± 0.52 rosette leaves and 35S::DnMFT-8 plants flowered at 10.75 ± 0.89 rosette leaves. In contrast, wild-type plants flowered at 10.63 ± 0.74 rosette leaves (Fig. 4d; Table 1). Over-expression of DnMFT in transgenic Arabidopsis plants gave rise to a slightly late-flowering phenotype.

Discussion

FT plays an important roles in floral induction, and its functions are conserved across different plant species. There are varying numbers of FT homologs present in different plant species however. The number of FT homologs varies from two in Arabidopsis (FT and TSF) to thirteen in rice (Izawa et al. 2002) and fifteen in maize (Danilevskaya et al. 2008). Most of FT functions remain undetermined.

In this study, a FT and a MFT homolog were identified in the orchid D. nobile Lindl and their expression and function in regulating the transition from vegetative to reproductive stage was investigated. These results showed that the expression of DnFT in leaves increased during plant growth and over-expression of DnFT in Arabidopsis produced an early flowering phenotype. These results indicated that DnFT may act as a promoter of flowering in D. nobile Lindl.

D. nobile Lindl is a biennial plant, its flowering requires a winter vernalization stage. In Arabidopsis, vernalization promotes flowering by suppressing FLC activity (Chanvivattana et al. 2004). FT is not only the output of the photoperiodic pathway, but also of the autonomous and vernalization pathways. In cold-requiring accessions of Arabidopsis, the MADS-box transcriptional regulator FLC represses expression of FT in the leaf and FD and SOC1 in the shoot apex (Searle et al. 2006). FT was specifically induced by CO in phloem cells (Takada and Goto 2003; An et al. 2004). In rice, a very low accumulation of Hd3a mRNA was detected in the leaf blade tissue and sheath (Tamaki et al. 2007; Wu et al. 2008). The DnFT was expressed at a low level in young leaves and increased during leaf development. After 40 days cold treatment, the expression level was higher than in the control (Fig. 3a). The effect of vernalization might be to down-regulate the FLC in D. nobile Lindl, thus allowing DnFT to be expressed in the leaves and increasing the ability of the shoot apex to respond to DnFT. The DnFT transcription in the axillary buds was also examined and it was observed that DnFT transcription was higher in the early stage of axillary bud development but then gradually decreased after the 40 day point (Fig. 3b). A similar expression pattern for OnFT was also reported in Oncidium Gower Ramsey (Hou and Yang 2009). This was different from that in some plant species where FT expression gradually increased during flower maturation (Kobayashi et al. 1999). Despite the conserved functions of FT homologs, their expression may therefore be controlled by different systems in different species. These differences were thought to be important in the variation of flowering time and in the distinctive morphological characteristics among plant species (Andersen et al. 2004). The results from this study indicated that the DnFT produced in a leaf might act as florigen in D. nobile Lindl, while the high expression of DnFT in the axillary buds during the early vegetative stage was not sufficient to induce transition in D. nobile Lindl.

In contrast to DnFT, DnMFT had a different tissue-type expression pattern. The expression of DnMFT increased in buds and decreased in leaves during flower bud development. Their different expression patterns suggested functional diversification. The higher expression of DnMFT in developing buds suggested that it might maintain inflorescence meristem identity. The expression of DnMFT was not regulated by temperature.

The function of DnFT in flower transition was also revealed by functional analysis. The early flowering phenotype and the loss of inflorescence indeterminacy observed in 35S::DnFT transgenic Arabidopsis (Fig. 4c) suggested that DnFT was potentially a FT ortholog,that regulated flower transition in D. nobile Lindl as in other plant species (Kardailsky et al. 1999; Kobayashi et al. 1999; Lifschitz et al. 2006; Hayama et al. 2007; Lgasaki et al. 2008; Komiya et al. 2008; Takahashi et al. 2009). The expression of AP1 and LFY, a marker for floral initiation and the downstream gene of FT, was up-regulated in 35S::DnFT transgenic Arabidopsis plants (Fig. 5). This indicated that constitutive expression of DnFT acted similarly to Arabidopsis FT in regulating flower transition by activating AP1 and LFY.

The DnMFT sequence had almost same homologous to FT and TFL1. Although the sequences of FT/TFL1 gene family members were homologous, their effects on flowering time were different. FT over-expression caused early flowering while ectopic expression of TFL1 showed the opposite phenotype (Kardailsky et al. 1999). In this study, the role of DnMFT in Arabidopsis was also examined. Constitutive expression of DnMFT delayed phase transition in Arabidopsis and the transgenic plants flowered 2–3 days later than wild-type plants. Compared with the phenotypes in Arabidopsis that ectopically express TFL1 orthologs, the 35S::DnMFT transgenic plants did not have more rosette leaves (Mimida et al. 2001; Hou and Yang 2009). Furthermore, over-expression of MFT and PopMFT in Arabidopsis caused slightly early flowering (Yoo et al. 2004; Franziska et al. 2008). These results indicated that the role of DnMFT in regulation of flowering time was slightly different from its homologs in other plant species.

Conclusion

In summary, this paper described two PEPB genes, DnFT and DnMFT of D. nobile Lindl, and elucidated the function of each gene by transcription analysis and over-expression of the genes in Arabidopsis. These results suggested that DnFT is potentially a FT ortholog that regulated flower transition in D. nobile Lindl. DnMFT can influence flowering time but functions slightly differently from the homologs of other plants. Further studies will be undertaken in orchids to characterize more PEPB genes and to utilize these genes to control flowering time in orchids.