Introduction

During floral transition the shoot apical meristem of angiosperms acquires inflorescence identity to generate flowers rather than vegetative tissues. This phase transition from vegetative to floral development in plants (Arabidopsis) is mediated by a multitude of genetic pathways in response to various endogenous and environmental signals [1–6]. These pathways include the photoperiod pathway that integrates information about light quantity and circadian rhythm and affects flowering according to these. The vernalization pathway influences flowering time in response to low temperature exposure. A very crucial pathway involved in flowering time regulation is the autonomous pathway. Instead of responding to environmental fluctuations, this pathway integrates only signals from the developmental state of the plant. Furthermore, in short-day conditions, the gibberellin pathway is very important for flowering time regulation. Besides, various other pathways that are triggered in responses to different wavelengths of light and temperature changes above a critical threshold are also implicated in flowering time determination. All these pathways affect flowering through a numbers of genes belonging to various families. In Arabidopsis around 173 genes are reported to be involved in floral transition through these pathways. However, the input from numerous flowering time pathways converges to regulate the expression of a small number of so-called “floral integrators”. Major genes with floral integrator activity in Arabidopsis include AGAMOUS-LIKE 24 (AGL24), SOC1, LEAFY (LFY), AND FLOWERING LOCUS T (FT) [7–9]. These floral integrators function co-operatively to activate the expression of floral meristem identity genes LFY and AP1, which in turn direct floral development [1].

Among the 173 genes involved in flowering, 19 are members of the MADS-box family including as well several of the above-mentioned key regulators. Hence, the MADS-box gene family constitutes a major class of regulators mediating floral transition. MADS-box genes encode a large family of transcription factors in plants (109 members in Arabidopsis) that share a highly conserved DNA binding MADS-domain of 56–58 aas [10]. These genes encode transcriptional regulators otherwise involved in diverse aspects of plant development ranging from root to flower and fruit to seed [10–15]. A high proportion of MADS-box genes are involved in the determination of flowering time, floral meristem, and floral organ identity [16, 17]. SOC1 is a floral promoter, while SHORT VEGETATIVE PHASE (SVP) acts as a repressor of flowering. Moreover, these two genes along with AGL24 are repressed by the meristem identity gene AP1, which consequently terminates shoot development thereby contributing to specification of floral meristem identity [9, 18, 19].

In Arabidopsis, AGL24 is one of the MADS-box genes found to promote flowering [20, 21]. AGL24 expression is detectable in the vegetative shoot apex and is up-regulated in the inflorescence apex during floral transition. Even though AGL24 and SVP act in the same pathway leading to floral transition and are closely related, they have contrasting functions; SVP is a floral repressor and AGL24 is a floral promoter. Interestingly, AGL24 and SVP are orthologs of MPF2-like (MADS-box gene of Physalis floridana 2-like) MADS-box genes of Solanaceous plants [22, 23]. These genes belong to the STMADS11 superclade of the MADS-box gene family. Probably due to allotetraploidization, MPF2-like genes are duplicated into MPF2-like-A and MPF2-like-B in Withania and Tubocapsicum [24]. In Withania somnifera these genes are designated as WSA206 and WSB206, whereas in Tubocapsicum anomalum as TAB 201. An MPF2-like-A gene (TAA201) does not exist in this species probably due to its loss from the genome of Tubocapsicum. MPF2-like-A, being expressed in floral tissues controls calyx inflation in Withania. By contrast, MPF2-like-B is not expressed in flowers and its function remains obscure.

The genus Withania consists of 11 species that are worldwide in distribution. The most common species, W. somnifera, is medicinally very important and is known as “Indian Ginseng” due to its tonic properties. The genus Tubocapsicum comprises two species, which are mainly endemic to humid regions of Eastern Asia. Phylogenetically, Withania and Tubocapsicum are sister taxa and are placed in the sub-tribe Withaninae along with seven other genera [25]. As already mentioned the major function of MPF2-like genes is inflated calyx syndrome (ICS) formation [23, 24]. Withania and Tubocapsicum exhibit very different phenotypes with respect to the degree of calyx inflation. Withania feature a huge calyx, while Tubocapsicum on the other hand is a natural evolutionary “loss mutant” of ICS. Being members of the STMADS11 superclade and orthologous to AGL24, MPF2-like genes might control flowering time in these two genera, as well. Therefore MPF2-like genes from Withania and Tubocapsicum were selected for phase transition experiments.

MPF2-like proteins, like other MIKC MADS proteins, are characterized by a conserved structural organization. The MADS-domain (M), an intervening (I) region, a keratin-like domain (K), and a C-terminal domain are present from the N-terminus to the C-terminus in a MADS-box gene [25]. Obvious differences exist in M-, I-, K- and C-domains of MPF2-like-A and -B proteins. But the most dramatic changes are the presence of a 3 aa deletion and an 8 aa extension in the C-terminal region of MPF2-like-A proteins [24]. This extension increases the length of MPF2-like-A protein. The C-terminal region of MADS-domain proteins in general and of MPF2-like proteins in particular is involved in transcriptional activation and interaction with other proteins [26].

Yeast 2-hybrid screening of an Arabidopsis oligo-dT cDNA library using WSA206 as bait revealed that this protein interacts with MAF1, MAF3, MAF4, MAF5, and SOC1. These proteins are implicated in regulating flowering time. Similarly, TAB 201 also interacts with MAF1, MAF5, and SOC1. On the other hand WSB206 showed no interactions with any of the above mentioned flowering time proteins (Table 1). Among MPF2-like interactors MAF1 [27], which is known as AGL27, has also been described by a few groups as FLOWERING LOCUS M [28]. MAF1 is a dosage-dependent inhibitor of flowering in Arabidopsis that acts independently of the photoperiod. MAF1 transcript is detectable at the seedling stage and later on becomes confined to the root and shoot apex. After transition from vegetative to flowering phase, MAF1 expression is visible in the floral primordia. MAF1 mutants flower earlier and ectopic expression of this gene induces delayed flowering. It can promote flowering only in late flowering ecotypes, like Pitztal or Stockholm, when overexpressed [27, 28].

Table 1 MADS-domain interactors of MPF2-like proteins from Arabidopsis oligo-dT cDNA yeast library screening
Full size table

Early flowering due to overexpression of CONSTANS (CO) seems to be partially suppressed by mutation in SOC1 [29]. SOC1 is a common component of the flowering time pathway, regulated by environmental conditions including long days via CO, vernalization via FLOWERING LOCUS C (FLC) and autonomous factors like the age of the plant [30, 31]. Three members of the MADS-box family of transcription factors, SVP, AGL24 [32, 33], and SOC1 (also known as AGL20) exhibit a close relationship in terms of flowering time control. Moreover, AGL24 and SOC1 also interact with each other [1].

The interaction of MPF2-like proteins with flowering time proteins (SOC1 and MAF1) hints at their potential role in floral transition. From an applied perspective, induction of early flowering in plants through gene transformation technology can be a useful strategy to obtain early-maturing crops. But before incorporating MPF2-like genes in crop plants for achieving early maturity can be successfully applied, it is imperative to study in detail the expression pattern of these genes and their functions with regard to flowering time. As MAF1 and SOC1 were the most abundantly obtained prey proteins for MPF2-like baits in Arabidopsis yeast 2-hybrid library screening, these two genes were selected for detailed expression analysis using real-time RT-PCR. Transformation of Solanaceous plants is demanding, hence MPF2-like transgenic Arabidopsis plants were used for flowering time assay and real-time RT-PCR expression analysis of MAF1 and SOC1 genes.

Phylogenetically, MPF2-like genes are the orthologs of AGL24 and one of the sub-clades of STMADS11 genes, along with SVP-like genes. Based on functional analysis we propose that WSA206 (MPF2-like-A), like AGL24, accelerates flowering. Concomitantly, the role of WSA206 as a floral promoter is supported by higher expression of SOC1 and lower expression of MAF1 in WSA206 overexpressing plants compared to wild type plants. The 8 aa extension characteristic for MPF2-like-A seems to have a complex influence on the proteins’ ability to affect flowering time, as evident from flowering time assay and alterations in expression profiles of SOC1 and MAF1 in transgenic Arabidopsis plants transformed with MPF2-like truncated and elongated gene constructs.

Materials and Methods

Plant Materials and Growth Conditions

Plant materials selected for flowering time assay and expression analysis comprised two types of Arabidopsis thaliana plants (ecotype Columbia). These included wild type and transgenic Arabidopsis plants transformed with MPF2-like genes (WSA206, WSB206, TAB 201) under the control of CaMV 35S promoter or a corresponding empty T-DNA, respectively, as control. In order to decipher the role of the 8 aa extension in flowering and interaction with other proteins, truncated WSA206ΔC246 constructs (WSA206 having 8 aa (SASKGRLY) deleted from the C-terminus) and elongated constructs, WSB206∇C257 and TAB 201∇C257 (WSB206 and TAB 201 having the respective 8 aa (SASKGRLY) added to the C-terminus) were generated. MPF2-like genes native to the Solanaceous plants Withania and Tubocapsicum were previously overexpressed in Arabidopsis driven by the 35S CaMV promoter in pBAR-A vector [24]. The T3 transgenic seeds of homozygous lines were sown in pots in the greenhouses of the Max Planck Institute for Plant Breeding Research, Cologne, Germany and the National Institute for Genomics and Advanced Biotechnology (NIGAB), NARC, Islamabad, Pakistan. These plants were grown at 22 °C under long-day (16 h light/8 h dark) conditions.

Phylogenetic Reconstruction

STMADS11 super-clade MADS-box homologs from A. thaliana, Solanum tuberosum, Petunia hybrida, Ipomoea batatas, P. floridana, W. somnifera, T. anomalum, Witheringia coccoloboides and Solanum lycopersicon were selected for phylogenetic reconstruction. The coding sequences of the MPF2-like protein interactors SOC1 and MAF1 from Arabidopsis were also included. These sequences were retrieved from TAIR (http://www.arabidopsis.org) and NCBI (http://www.ncbi.nlm.nih.gov/) databases. A Neighbor-joining (NJ) algorithm from the PAUP4.0b10 [34] program with default settings and uncorrected P distance was employed [25].

Genetic Diversity Analysis of MPF2-Like-A and MPF2-Like-B Genes

To survey the molecular diversity differences between MPF2-like duplicates, we compared the nucleotide diversity among groups of MPF2-like-A and MPF2-like-B genes. cDNA sequence nucleotide diversity was compared between MPF2-like-A and MPF2-like-B genes from W. somnifera (WSA206 and WSB206) and T. anomalum (TAB 201) using the DnaSP v5.0 software [35]. The dataset used as input for DnaSP contained sequence fragments from start till stop codon. These sequences were aligned [25] using the program MegAlign from the Lasergene DNAstar package.

Multiple Alignments of the C-Terminal Region of MPF2-Like Proteins

As the C-terminal region of MPF2-like proteins is highly variable, this region was used for revealing the sequence divergence between MPF2-like-A and -B proteins. For this purpose the coding sequences of MPF2-like-A and -B from Withania aristata, W. coagulans, W. riebeckii, W. somnifera, W. frutescens, and T. anomalum were acquired from the NCBI database. MPF2, STMADS16, AGL24 and SVP from P. floridana, S. tuberosum and A. thaliana, respectively, served as reference genes. An alignment of the partial C-terminal region of the selected sequences was performed with ClustalW from the MacVector™ 7.2.3 software (Accelrys Inc.; gcg/Wisconsin Package, University of Wisconsin).

Flowering Time Assay of MPF2-Like Transgenic Arabidopsis Plants

For flowering time analysis, three independent homozygous T3 lines of transgenic Arabidopsis plants along with empty vector and wild type control were grown under long-day conditions. There was no change in the germination time of all the seeds. The number of rosette and cauline leaves of at least 30 plants were marked and recorded at 10, 14 ,and 18 days after stratification and final results were obtained at the bolting stage (20–26 days after stratification). The data were statistically analyzed for student’s test using Minitab version 13. Finally, photographs were taken with a digital camera.

Primer Design

For confirmation of transgenic Arabidopsis plants and expression analysis through semi-quantitative RT-PCR, primer sequences were designed from divergent regions of MPF2-like genes using the “Primers” tool of the MacVector™ 7.2.3 software (Accelrys Inc.; gcg/Wisconsin Package, University of Wisconsin). For quantitative real-time RT-PCR analysis of SOC1 and MAF1 genes, primer sequences were designed in the K- and the C-domains. The primer sequences of internal control genes (18S rRNA, PP2A, and EF1) were also designed with MacVector™ 7.2.3 software. Primer sequences used for gene expression analyses are listed in Table 2.

Table 2 List of primer pair sequences used for gene expression analysis
Full size table

Nucleic Acids Extraction and First Strand cDNA Synthesis

DNA extraction from transgenic Arabidopsis was carried out using the DNAeasy Plant Mini kit (Qiagen) or according to the method of Zhang and Stewart [36]. RNA contamination in these DNA samples was removed by treatment with DNase-free RNase (Qiagen). For RNA extraction the upper part of the growing seedlings was harvested. Total RNA was isolated from transgenic and wild type Arabidopsis seedlings at 10, 14, and 18 days after germination using the PureLink RNA mini Kit (Invitrogen). These RNAs were mixed together. For quality assessment, DNA and RNA were run on 1 % and 2 % agarose gels, respectively. Nano spectrophotometer was used for the quantification of DNA and RNA.

The first strand cDNA synthesis was carried out using M-MuLV Reverse Transcriptase (Fermentas) with 1 μg of total RNA in a 20 μl reaction volume.

Semi-Quantitative RT-PCR

Multiplex RT-PCR was carried out in 20 μl reaction volumes using Taq polymerase (Fermentas) for MPF2-like primer pairs for 14 cycles and continued to another 16 cycles after the addition of 18S rRNA primers. Normalization was done using constitutively expressed 18S rRNA genes. Amplified PCR products were run on 2 % agarose gels and photographed. The intensity of amplification (MPF2-like expression) was determined after scanning the gel. All the fragments were sequenced to confirm their identity as MPF2-like genes. The experiment was repeated three times for each of the technical as well as biological replicates.

Quantitative Real-Time RT-PCR Analysis

Real-time RT-PCR was carried out using the Bio-Rad iq5 real-time PCR detection system. The PP2A and EF1 genes were used as internal controls. Reactions were carried out in 25-μl reaction volumes including 125 nmol/l of each gene-specific primer and 1× iQ SYBR Green Supermix solution (Bio-Rad). The program conditions were:95 °C for 10 min to activate the polymerase, followed by 45 cycles of 95 °C for 30 s, 60 °C for 30 s, 72 °C for 1 min, and a final melting curve analysis from 60 °C to 95 °C. Amplification efficiency was determined and only primer pairs that showed 86–112 % amplification efficiency were used for expression studies. The experiments were repeated three times for each of the biological as well as technical replicates. Relative expression of the flowering time genes was normalized with respect to PP2A and EF1 expression. Expression data analysis was carried out according to the Pfaffl method (Bio-Rad [37]). Data were statistically analyzed with Minitab version 13.

Results

The MPF2-Like Along with the SVP-Like Clade Belongs to the STMADS11 Superclade

Best known for their effects on calyx inflation, MPF2-like genes are also implicated in fertility and floral transition [23, 38]. These genes were duplicated in Withania and Tubocapsicum into MPF2-like-A (WSA206) and MPF2-like-B (WSB206 and TAB 201). In order to infer phylogenetic relationship of these genes with STMADS11 superclade members, a NJ tree based on STMADS11 homologs from A. thaliana, S. tuberosum, P. hybrida, I. batatas, P. floridana, W. somnifera, T. anomalum, W. coccoloboides, and S. lycopersicon were constructed. Figure 1 shows that the STMADS11 superclade is separated into three distinct clades designated as STMADS11, SVP-like, and MPF2-like. The Arabidopsis AGL24 gene is associated with the MPF2-like clade. SOC1 and MAF1 occupy distinct positions and are not clustered within the STMADS11 superclade. Interestingly, WSA206 is sister to MPF2 instead of WSB206, though these are assumed to be paralogs. This phylogenetic differentiation reflects differences in the coding sequences of MPF2-like genes after genome duplication of these genes through allotetraploidization [24].

Fig. 1
figure 1

Phylogenetic reconstruction of STMADS11 superclade MADS-box genes. A NJ tree showing the phylogenetic relationships of STMADS11 superclade genes is depicted. Sequences of STMADS11 homologs from A. thaliana, S. tuberosum, P. hybrida, I. batatas, P. floridana, W. somnifera, T. anomalum, W. coccoloboides, and S. lycopersicon were retrieved from databases. Numbers on the branches show the bootstrap support probability of 10,000 times calculations. Gene names in the solid blue box indicate the duplicated MPF2-like genes (WSA206 and WSB206) of W. somnifera (Color figure online)

Full size image

MPF2-Like C-Terminus is Highly Polymorphic

To survey the molecular diversity differences between duplicates in M-, I-, K-, and C- domains, we compared the nucleotide diversity among groups of MPF2-like-A and MPF2-like-B genes spanning a region in which the two groups of genes could be well aligned (750 bp). An equal number of genes from both groups were selected for polymorphism detection. Both MPF2-like-A and MPF2-like-B genes did not show similar values for the average number of nucleotide differences per site in pairwise comparisons (Pi). It could also be observed from the total number of substitutions (S) and average number of nucleotide differences (K) that both groups exhibited different patterns. In Fig. 2a, a gliding window analysis based on the comparison of nucleotide diversity (Pi) shows that the two groups of genes in different domains, i.e., M-domain, I-domain, K-box, and C-terminal domain, have regions that are specifically highly polymorphic. Both MPF2-like-A and -B genes exhibit islands of relatively high nucleotide sequence divergence in the I-region and C-terminal part. Within the last two exons of the C-terminal region, a very low level of sequence divergences is observed in comparison with the first two exons where maximum polymorphism is detectable. However, a striking alteration has occurred in the fourth exon of the C-terminal region. At the C-terminal end following a divergent sequence of 4 aas, a dramatic change in form of an extension of 8 aas (24 nucleotides) is observed (indicated by a solid red box at the end of the alignment in the MPF2-like-A group of genes in Fig. 2b). This 8 aa extension is specific to MPF2-like-A proteins and is the characteristic feature of this group of proteins. Thus, nucleotide diversity analysis has revealed the C-terminal region to be highly polymorphic [24].

Fig. 2
figure 2

Nucleotide diversity analysis of MPF2-like duplicates. a MPF2-like cDNA sequence nucleotide diversity analysis shows sequence divergence between MPF2-like-A and -B genes of Withania (WSA206 and WSB206) and Tubocapsicum (TAB 201 only; TAA201 is lost from the genome). Nucleotide diversity is shown on the Y-axis, while the X-axis marks the position along the cDNA sequences. The four important domains of MADS-box genes are also indicated below the diagram. b ClustalW alignments of the C-terminal region of MPF2-like proteins. The C-terminal region of MPF2-like-A and -B genes from W. aristata, W. coagulans, W. riebeckii, W. somnifera, W. frutescens and T. anomalum were aligned. MPF2, STMADS16, AGL24 and SVP from Physalis floridana, S. tuberosum and A. thaliana, respectively, served as reference genes. Shaded boxes depict amino acids conservation. The solid black vertical line underneath the alignment indicates the start of the fourth exon of the C-terminal domain. The blue rectangle marks the 3 amino acid deletions and the red rectangle indicates the extension of 8 amino acids in MPF2-like-A (WSA206) proteins (Color figure online)

Full size image
Fig. 3
figure 3

Flowering time assay of transgenic Arabidopsis plants overexpressing MPF2-like genes. Photographs showing the phenotypes of MPF2-like transgenic Arabidopsis thaliana plants and lines transformed with an empty constructs in comparison to WT (Col). a Three MPF2-like genes (WSA206, WSB206, and TAB 201) were ectopically expressed in Arabidopsis using the pBAR-A vector. The transgenic lines used were 1, 3 and 4 for WSA206, WSB206, and TAB 201, respectively. B) In order to study the role of the 8 amino acids on flowering time, MPF2-like truncated (WSA206∆C246) and elongated genes (WSB206∇C257 and TAB 201∇C257) were also expressed under control of the 35S CaMV promoter. The transgenic lines used were 1, 2 and 2 for WSA206∆C246, WSB206∇C257 and TAB 201∇C257, respectively. The T3 transgenic plants were grown under long-day condition

Full size image

The role MPF2-like proteins and the 8 aa extension in flowering time control was investigated by transforming A. thaliana with constructs carrying 35S:WSA206, 35S:WSB206, 35S:TAB 201 and their derivative constructs with the sequence encoding for the 8 aa extension either removed (35S:WSA206ΔC246) or added (35S:WSB206∇C257 and 35S:TAB 201∇C257).

MPF2-like-A (WSA206) Promotes Flowering

AGL24, the subsumed ortholog of MPF2-like proteins, promotes flowering in Arabidopsis, while MPF1 (a clade member of STMADS11 superclade) delays it [39]. Flowering time control exerted by MPF2-like genes was analyzed by counting the number of rosette leaves at the bolting stage (20–26 days after stratification) in the transgenic Arabidopsis plants grown under long-day conditions. Table 3 and Fig. 3a show the flowering time assay of three independent homozygous T3 transgenic lines transformed with MPF2-like genes (WSA206, WSB206, and TAB 201) along with wild type plants. Clearly, 35S:WSA206 transgenic plants show early flowering, which is revealed not only by early bolting but also by a reduction in the number of rosette leaves. In the extreme case bolting time is 19.72 ± 0.91 days for 35S:WSA206 # 11 with only 6.75 ± 0.39 rosette leaves in comparison with Wild-type Col # 1 plants for which 24.1 ± 1.4 days are required for bolting and which produce more than 10.68 ± 0.2 rosette leaves. Thus early flowering of 35S:WSA206 transgenic plants in comparison to WT control plants is statistically highly significant. The 35S:WSB206 and 35S:TAB 201 plants, in contrast, displayed a bolting time and rosette leave number similar to wild type control plants. These results illustrate that WSA206 overexpression moderately accelerates the flowering time in transgenic Arabidopsis.

Table 3 Flowering time assay of 35S:MPF2-like transgenic Arabidopsis plants grown under long-day conditions
Full size table

The 8 Amino Acid Extension of MPF2-Like-A Influences Flowering Time Enigmatically

The most prominent structural feature of MPF2-like-A proteins is an 8 aa extension at the C-terminus. To evaluate that the revealed 8 aa extension is implicated in flowering, the transgenic plants carrying derivative versions of 35S:MPF2-like constructs, in which either the 8 aa encoding region was deleted or inserted for WSA206 and WSB206/TAB 201, respectively, were analyzed for flowering time. When bolting time and number of rosette leaves for the plants containing these derivative constructs were recorded, remarkably, early flowering (19 day for bolting with 7 rosette leaves) was observed in 35S:TAB 201∇C257 plants (Table 3 and Fig. 3b). Loss of the 8 aa extension from WSA206, however, did not substantially suppress the flowering promotion exerted by 35S:WSA206, and resulted only in a slight increase in flowering time (21 days for bolting with 8 rosette leaves) as compared to the native protein version. On the other hand, 35S:WSB206∇C257 expressing plants showed no deviation from 35S:WSB206 lines with respect to days for bolting and number of rosette leaves, i.e., flowering time. Hence, the function of 8 aa in promoting floral transition remains obscure.

MADS-box genes were involved in autoregulation as well as regulated by other MADS-box genes [40, 41]. As WSA206 and TAB 201 were shown to interact with SOC1 (floral promoter) and MAF1 (floral repressor) proteins in yeast 2-hybrid library screening (Table 1), therefore, the expression patterns of these genes were determined in transgenic Arabidopsis in MPF2-like background.

SOC1 and MAF1 are Differentially Expressed in MPF2-like Transgenic Arabidopsis Plants

In order to confirm the transgenic nature of the generated Arabidopsis lines and to study their expression patterns, semi-quantitative RT-PCR was performed. Total RNA from 10-day-old seedlings of three independent homozygous T3 transgenic Arabidopsis lines containing 35S:WSA206, 35S:WSB206, 35S:TAB 201, 35S:WSA206ΔC246, 35S:WSB206∇C257, and 35S:TAB 201∇C257 and of the wild type control was extracted and converted to cDNA. Semi-quantitative RT-PCR was carried out using 18SrRNA as the internal control. Figure 4 depicts that gene expression at the transcription level is easily detectable in all the transgenic lines in case of 35S:WSA206, 35S:WSB206, and 35S:TAB 201. Similarly, in the truncated and elongated versions of MPF2-like transgenic plants (35S:WSA206ΔC246, 35S:WSB206∇C257, and 35S:TAB 201∇C257), transcript signals are also present. The intensity of expression is similar in all the transgenic lines irrespective of the type of construct. These data suggest that native MPF2-like (WSA206, WSB206 and TAB 201), truncated (WSA206ΔC246), as well as elongated genes (WSB206∇C257 and TAB 201∇C257) are transcribed normally.

Fig. 4
figure 4

RT-PCR expression analyses of MPF2-like genes in transgenic A. thaliana. Seedlings of transgenic Arabidopsis plants along with wild type grown under long-day conditions after 10, 14, and 18 days were used for RNA extraction. The upper part of the seedlings was harvested for nucleic acid extraction. The numbers from 1 to 3 represent three different lines of MPF2-like (35S:WSA206, 35S:WSB206, 35S:TAB 201) plants, while 4–6 marks the three different lines of MPF2-like derivative expressing plants (35S:WSA206∆C246, 35S:WSB206∇C257 and 35S:TAB 201∇C257; see also Table 3 for detailed description of the lines used). Three plant samples were used for RT-PCR for each gene construct and the experiment was repeated thrice technically. The upper band shows the expression of the MPF2-like genes whereas the lower band serves as a control and represents 18S rRNA

Full size image

The expression patterns of flowering time regulators SOC1 and MAF1 [27, 29] were probed by real-time RT-PCR analysis. Figure 5 shows statistically significant differences in the expression of SOC1 and MAF1 in 35S:WSA206 plants in comparison with wild type plants. The increase in SOC1 expression and decrease in MAF1 levels might allow SOC1 to promote flowering. This is indeed in line with the early flowering phenotype of 35S:WSA206 plants (Fig. 3a). In case of 35S:WSB206 and 35S:TAB 201 plants, although the transcript signals for SOC1 and MAF1 are deviating from wild type, the ratio of SOC1 and MAF1 expression (1.2 ± 0.18 and 1.5 ± 0.22 normalized fold expression for 35S:WSB206 against 0.9 ± 0.17 and 0.8 ± 0.2 of 35S:TAB 201) does not seem to be changed sufficiently for the affected flowering time. This means, the more the increase in SOC1 expression in comparison with MAF1, the more the decrease in flowering time. In general, the plants transformed with derivative constructs the expression of SOC1 and MAF1 was decreased except for 35S:TAB 201∇C257 plants, where SOC1 expression was tremendously (almost fourfold increase in comparison with MPF1) elevated. These expression patterns substantiate the role of 5S:TAB 201∇C257 plants in executing early flowering (Fig. 3b). A decrease in expression of SOC1 and MAF1 in plants expressing the truncated WSA206 (35S:WSA206ΔC246) coincides with a slight delay in flowering time in comparison with 35S:WSA206 plants (Fig. 3b). These results support the role of the 8 aa extension in affecting SOCI and MAF1 expression thereby inducing early flowering. But this coincidence might be context dependent and partner specific as well.

Fig. 5
figure 5

Quantitative real-time RT-PCR analysis for SOC1 and MAF1. Real-time RT-PCR analysis of the Arabidopsis genes SOC1 and MAF1 in 35S:MPF2-like transgenic Arabidopsis lines compared to wild type plants. Both wild types as well as transgenic T3 homozygous lines were grown under long-day conditions. The lines used are identical to the ones depicted in Fig. 3a and b for the flowering time assay. RNA from 10-, 14-, and 18-day-old plants was extracted and mixed. The bars show the normalized fold expression of SOC1 (empty) and MAF1 (light gray). The values given are average expression levels based on three independent experiments for each of the biological as well as technical replicates. Error bars represent standard deviation of the mean

Full size image

Discussion

Genetic manipulation of developmental control genes such as MADS-box genes is a useful strategy to bring about a radical change in plant genetic programs for a particular trait that may improve crop productivity. Early flowering can be desirable, as it may lead to early maturation of crops. However, a complete expression and functional analysis of MADS-box genes are of the utmost importance before their application in biotechnology can be considered.

Apart from floral organ identity and meristem specification, MADS-box genes are implicated in flowering time determination [1]. A large number of MADS-box genes are involved in the transition to flowering in Arabidopsis [30, 42–45]. The svp mutant shows early flowering without displaying other obvious phenotypic deviations [33]. Loss or reduction of AGL24 activity by double-stranded RNA (dsRNA)-mediated interference results in dosage-dependent late flowering, while over-expression of MPF2, STMADS16 and AGL24 under the CaMV 35S promoter leads to early flowering [32, 46, 47]. Suppression of flowering time genes such as SVP, AGL24, and SOC1 by AP1 or LFY results in the transition from vegetative to floral phase [1, 9]. The flowering time control genes AGL24 and SVP belonging to the STMADS11 superclade of the MADS-box gene family are orthologous to MPF2-like genes in Arabidopsis. Therefore, in addition to their role in ICS, fertility and leaf development, a further role of Withania MPF2-like-A (WSA206) and Tubocapsicum MPF2-like-B (WSB201) proteins in controlling flowering time is conceivable. Such a role is supported not only by their interactions with SOC1 and MAF1, MAF3, MAF4, and MAF5 proteins (Table 1), but also by conservation of many putative binding sites for INDETERMINATE1 (ID1), a transcription factor in all the MPF2-like promoters and in the first large intron of the coding sequence (data not shown). These observations imply a role of MPF2-like proteins in flowering time control. In order to validate this hypothesis in planta, previously overexpression of WSA206, WSB206, and TAB 201 are carried out in the heterologous host Arabidopsis. In this study, a flowering time assay in transgenic Arabidopsis revealed that the Withania MPF2-like-A (WSA206) protein is recruited in accelerating flowering time, unlike WSB206 protein, for which also no affinity for proteins involved in flowering time control could be detected using the yeast 2-hybrid system. Similar to WSA206, MPF2, a close ortholog of WSA206, also promotes flowering [24, 38]. The role of WSA206 as a floral promoter is also supported at the transcriptome level by stronger expression of SOC1 and weaker expression of MAF1 genes as revealed by real-time RT-PCR analysis.

In general, the C-terminal region of MADS-domain proteins is involved in transcriptional activation or the formation of multimeric transcription factor complexes [26, 47, 48]. In case of MPF2-like-A and –B duplicates, the C-terminus is highly divergent also due to the presence of a 3 aa deletion and an 8 aa extension in MPF2-like-A [24]. A number of yeast two hybrid experiments using site-directed mutagenesis and domain/motif swapping of MPF2-like proteins revealed no effect of the 3 aa deletion on MPF2-like protein function in calyx inflation, floral reversion, fertility, inflorescence architecture, pedicel length and floral transition (unpublished results). By contrast, the loss and addition of 8 aa in MPF2-like proteins did alter the calyx inflation phenotypes (Fig. 6). The incorporation of the 8 aa in TAB 201 induced secondary sepal growth in transgenic plants. The size of the secondary calyx was not increased up to the extent of WSA206 sepals but sepals became persistent and did not drop off earlier. Surprisingly, the 35S:WSA206ΔC246 transformed plants lost the capacity to generate and retain the secondary sepals and looked like 35S:WSB206ΔC25 plants or wild type that did not show any increase in sepal growth. Hence, the 8 aa extension influences secondary sepal growth and persistence. However, in this study we could not find an unambiguous correlation between the presence of the 8 aa extension and early flowering in MPF2-like transgenic plants. The extension may modulate flowering time but is neither sufficient nor strictly necessary for early flowering. The 8 aa extension might exert its influence on the formation of MADS complexes. These complexes in turn might alter the ratio of SOC1 to MAF1 and loss of MAF1 and gain of SOC1 expression consequently might decrease flowering time.

Fig. 6
figure 6

Morphological analysis of transgenic A. thaliana after transformation with MPF2-like genes having the C-terminus 8 amino acids deleted and added. The picture shows the comparison of flowers and fruits of wild type, 35S:WSA206ΔC246, 35S:WSB206∇C257, and 35S:TAB 201∇C257 transgenic plants. Note the growth and retention of sepals by the flowers and fruits of 35S:TAB 201∇C257plants. The sepals of 35S:WSA206ΔC246 are persistent to some extent till the flowering stage but preclude at the maturity. Scale bar 2 mm

Full size image

Early flowering induction by addition of the 8 aa to 35S:TAB 201∇C257 is understandable but ambiguity of the role of the 8 aa in inducing early flowering becomes apparent as 35S:WSA206∆C246 plants are still able to promote the flowering. These results imply that there are other regions in WSA205 in addition to the ultimate C-terminus that are mediating interaction with other proteins in the flowering regulatory network. This is not surprising, as for instance, the K domain has been shown to be involved in interaction with other MADS-domain proteins [49]. Furthermore it acts as a determinant for interacting spectrum of MPF2-like proteins (unpublished results).

It is important to note that in transgenic Arabidopsis semi-quantitative RT-PCR revealed similar levels of transcriptional signals for all the MPF2-like genes and their derivative constructs. This strongly supports, that indeed divergent functional properties of the proteins, and not differences in expression levels, cause the observed phenotypic differences. In addition, it could be that there are differences in the levels of the final products, i.e., the proteins. It is possible that some proteins are not stable enough to form persistent complexes and affect the expression of other genes. Therefore, MPF2-like protein stability might play a role in flowering time determination via direct and/or indirect interaction with SOC1 and MAF1, as well. In the future, to determine the effects of protein stability, it will be interesting to analyze protein levels by using antibodies against the respective proteins or by means of fluorescent fusion proteins.

In conclusion, we posit that MPF2-like genes, specifically WSA206, promote flowering in transgenic Arabidopsis, which is supported by higher expression of its interactor SOC1, a floral promoter. This study also adds to the knowledge about the functionally active role played by C-terminal motifs in influencing flowering time by changing expression of SOC1 and MAF1 genes.