Introduction

Increasing research on flowering mechanisms in model plants has revealed that the developmental transition of plants from the vegetative to the reproductive phase is controlled by many genes, with several most important players (Blázquez 2005; Tan and Swain 2006; Mutasa-Göttgens and Hedden 2009; Xu et al. 2010; Hemming and Trevaskis 2011; Wang et al. 2011; Chang et al. 2012). Among them, FLORICAULA (FLO) from Antirrhinum (Coen et al. 1990) and its Arabidopsis orthologue, LEAFY (LFY) (Weigel et al. 1992), play a central role in the determination of flower meristem identity (Moyroud et al. 2010). LFY encodes a plant-specific transcription regulator that is expressed at low levels during the vegetative phase but increased upon the transition to flowering. Mutation of LFY results in the conversion of lateral flowers into indeterminate secondary shoots. The overexpression of 35S::LFY in Arabidopsis accelerates reproductive transition, causing early flowering and converting the lateral secondary shoots into flowers. This phenomenon occurs not only in Arabidopsis but also in species such as aspen (Weigel and Nilsson 1995), rice (Kyozuka et al. 1998), tobacco (Ahearn et al. 2001) and citrus (Peña et al. 2001). In addition, heterologous expression of LFY homologs from other species in Arabidopsis lfy mutant can partially or fully complement the mutation (Mouradov et al. 1998; Maizel et al. 2005). LFY expression is necessary to activate the expression of some of the floral organ identity genes (Ng and Yanofsky 2000; Moyroud et al. 2010), suggesting that LFY integrates environmental and endogenous signals to control flower timing, and acts as a link between the upstream flowering time genes and downstream floral homeotic genes in the flower development pathway. These reports implied that controlling the expression LFY, or its homologs, could enable control the induction of flowering in plants.

The regulation of flowering in fruit trees is remarkably different from that in herbaceous species (i.e., long juvenile phases, season-dependent bud dormancy, and alternation between vegetative and reproductive phases, which are disadvantages for breeding and stable annual crop production) (Wilkie et al. 2008). Thus, understanding the genetic mechanisms underlying flower transition is essential for establishing breeding programmes and promoting genetic research. Recent studies have reported that some FLO/LFY orthologues and partial or total FLO/LFY-like sequences were cloned form woody species or other basal angiosperms and gymnosperms (Mellerowicz et al. 1998; Rottmann et al. 2000; Wada et al. 2002; Yu et al. 2005; Shitsukawa et al. 2006; Allnutt et al. 2007). However, the role of the gene species in controlling the characteristic features of tree reproductive development varies; the functional information and universality still need to be elucidated.

To help understand the genetic and molecular mechanisms underlying the flowering process in peach, the model plant in woody perennial fruit tree research (Baid et al. 1996; Ginannino et al. 2003), we report the cloning and characterisation of PpLFL, a LFY/FLO homolog in peach. The sequence feature shared great similarity and identity with LFY homologs from other dicotyledons, especially woody perennials. Our results show that PpLFL was expressed in both vegetative and floral meristem, and could fully complement the Arabidopsis lfy mutation, even to the point of accelerating flowering and giving rise to normal flower organs. The results suggested that PpLFL is involved in flower initiation in peach, and also a functional homolog of the Arabidopsis LFY gene.

Materials and Methods

Plant Material

Peach (Prunus persica (L.) Batsch. cv. Bayuecui), grown in the orchard (Beijing, China), was used in this study. Developing axillary buds borne on the shoots were collected from 30 May to 10 October at 10 day intervals in 2005. The cut axillary buds, including meristems, were immediately frozen in liquid nitrogen and stored at −80 °C. For protein immunohistolocalisation, samples collected at various development stages were fixed and embedded in paraffin according to the method of Sung et al. (1999).

Arabidopsis (Arabidopsis thaliana) ecotype Columbia was used as a control in this study. A transgenic line (catalogue no. salk_057202) carrying a T-DNA insertion in Arabidopsis LFY gene was identified in the Salk Institute Genomic Analysis Laboratory line collection. Homozygous mutants were selected by ensuring that all of their progeny were resistant to kanamycin (50 mg L−1). The presence of the T-DNA insertion was confirmed by PCR using gene-specific primers (5′ primer, 5′-ACAGGGTTATCTGAGGAACCG-3′; 3′primer, 5′-TCTGTACAAGCAATGGCACAG-3′) and a T-DNA insertion primer (5′-ATTTTGCCGATTTCGGAAC-3′). The Arabidopsis seeds were surface-sterilized with 5 % (v/v) NaClO solution for 7 min and washed with sterile water, plated on Murashige and Skoog medium and vernalized at 4 °C in dark. The seedlings with three to four leaves were transferred to soil and grown in a controlled growth room with 20 ± °C to 22 ± °C, 90 to 120 μmol m−2 s−1, 16/8-h photoperiod, and 68 % to 78% humidity.

RNA Extraction and cDNA Cloning

Total RNA was extracted following the method of Chang et al. (1993), and the cDNA was constructed by oligo (dT15) primer with M-MLV reverse transcriptase (Promega, Madison, USA). Partial cDNA was obtained by a PCR-based strategy using the LFY specific degenerate primers LF (5′-CAYCCNTTYATH GTRACDGAGCL-3′) and LR (5′-TGNCGGAGCTTGGTNGGNACRTAC-3′), designed from the conserved region HPFIVTEP and WYVPTKLRQ of FLO/LFY. Full-length cDNA was obtained by the 5′ and 3′ RACE methods using the Generacer kit, following the manufacturer’s instructions (Invitrogen, Carlsbad, USA) with the gene-specific primers (GSP1: 5′-ATCAACAAGCCCAAGATGCGA-3′; GSP2: 5′-AGAGGCGAAAATGTGGGGGC-3′; GSPA1: 5′-CGCATCTTGGGCTTGTTGATGTA-3′; GSPA2: 5′-TTGTCGTAGAGATGGAAGAGGTAAT-3′). All fragments obtained were cloned at the EcoR V site of pMD18-T vector (Takara, Japan) and sequenced in both directions.

Southern Blotting

Genomic DNA was extracted using the cetyltrimethyl ammonium bromide (CTAB) method (Porebski et al. 1997). Genomic DNA was digested with the restriction enzymes Eco RI and Bgl II and blotted onto a nylon membrane (Hybond N+), and the digoxin (DIG)-labelled full-length PpLFL cDNA fragment was used as probe. Hybridisation was performed in DIG HYB solution (Mylab Corporation, China) overnight at 65 °C. The membrane was rinsed twice in 2 × SSC, 0.1 % SDS, at room temperature for 5 min, and washed twice in 0.1× SSC, 0.1 % SDS, at 50 °C for 15 min. Chemiluminescence was detected according to the manufacturer’s instructions (Mylab Corporation, China), and the resulting images were photographed.

Antibody Preparation

To express a truncated form of the PpLFL protein, the 235 bp Not I and Eco RI fragment was cloned to the expression vector pGEX-4T-1 (Amersham/Pharmacia, USA) and fused to the sequence of the GST gene to form the recombined protein PpLFL-GST. This construct was introduced into Escherichia coli strain BL21, and the production of the recombined protein was induced with 1 mmol L−1 Isopropyl β-d-1-thiogalactopyranoside. The expressed PpLFL protein was purified on a GST-affinity column according to the manufacturer’s instructions (Amersham/Pharmacia, USA). Isolated protein was further purified by electrophoresis on 12 % SDS-PAGE. After electrophoresis, the gel was soaked in 250 mmol L−1 cold KCl and incubated at 4 °C until protein bands were visualised (white colour). The band in size of 35.8 kDa corresponding to the truncated PpLFL protein was excised from the gel, and further purified by dialysing, and then concentrated. Rabbits were immunised initially with 250 μg mL−1 of protein together with adjuvant in a total volume of 2 ml. They received a total of four immunisations of 300 μg mL−1 of protein without adjuvant in a total volume of 2 ml at 1 week intervals. The antibodies for the PpLFL were purified with saturated ammonium sulphate precipitation and finally for using for protein immunolocalisation.

Protein Immunolocalisation

To determine the spatial expression pattern of PpLFL, the immunoreactive proteins at the shoot apical meristem (SAM) were detected. The embedded samples were sectioned into 8-μm slices, and dewaxed, rehydrated, washed for 10 min in PBS buffer. The slides were incubated in a blocking solution for 1 h at room temperature, then briefly washed with PBST buffer three times at 10-min intervals and incubated with primary antibodies for 2 h at room temperature. After incubation with primary antibodies, cross-sections were incubated with goat anti-rabbit IgG AP (Sigma, USA) for 1 h at room temperature. Detection of the amplified signals was carried out using the substrate 5-bromo-4-chloro-3-indolyl-phosphate/nitro blue tetrazolium. The stained sections were washed with distilled water for 5 min and then dehydrated and mounted with a cover glass for photographing.

Expression of PpLFL in Arabidopsis

The PpLFL open reading frame (ORF) was excised from pMD18-T Simple vector (Takara, Japan) by using the restriction enzymes Nco I and Bgl II and cloned into the same sites of the pCAMBIA 1301 vector (CAMBIA, Australia) under the control of CaMV 35S promoter. The positive recombined vector was transformed into Agrobacterium tumefaciens GV3101 strain, which was then used to transform Arabidopsis lfy mutant plants using the floral dip method (Clough and Bent 1998). Transgenic seedlings were first selected on Murashige and Skoog agar plates with 80 mg L−1 hygromycin as a selective agent. Antibiotic-resistant plants were then tested by PCR to confirm the presence of the transgene with the gene-specific primers PpLFLN (5′-CCATGGAACTCTGGGAAATATGGA-3′) and PpLFLB (5′-AGATCTGTGTAGGGTAGGTGATCACG-3′).

Results

Isolation and Sequence Analysis of the PpLFL cDNA, a LFY homolog in Peach

A 420-bp nucleotide sequence of the PpLFL gene was cloned using degenerate PCR primers, and then the entire ORF was obtained using both 3′ and 5′ RACE with lengths of 278 bp and 742 bp, respectively. The single PCR product of the entire ORF of PpLFL was then cloned using cDNA as templated and sequenced, as well as another single PCR product of PpLFL, which was isolated using genomic DNA as a template. The lengths of the entire ORF of PpLFL and the genomic sequence of PpLFL spanned 1,248 bp and 2,489 bp, respectively, which suggested that there were intron(s) in the PpLFL gene. According to the alignment results for the two sequences, PpLFL had three exons and two introns that spanned 270 bp and 901 bp, and the introns were located near the 5′ and 3′ ends of the nucleotide sequence, respectively (Fig. 1). The cDNA sequence was submitted to GenBank and the accession No. was EF 175 869. The deduced amino acid sequence of PpLFL gene contains 415 amino acids residues.

Fig. 1
figure 1

The genomic structure of peach PpLFL gene. The open boxes indicated the untranslated regions, the filled boxes indicated the protein coding regions and the disconnect lines indicated the introns. The large black triangle indicated the position of Eco RI in PpLFL gene. The numbers indicated size in base pairs

Full size image

Alignment of the predicted amino acid sequence of PpLFL to that of LFY and its homologs from other plant species revealed similar patterns of conserved and variable domains (Fig. 2) and an overall similarity of 50.48 % to 84.69 % (Fig. 3). There were two highly conserved regions in the PpLFL protein: an alanine-rich domain near the N-terminus, and an acidic domain in the middle (Fig. 2). Both domains were considered common features of transcription factors and were thought to be important in the function of LFY as a transcriptional activator (Gerber et al. 1994; Cho et al. 1999).

Fig. 2
figure 2

Alignment of predicted amino acid sequence of the PpLFL, with other LFY/FLO homologs. Accession numbers are as follows: AFL1(AB051658), PoLFY-1(AB162029), AFL2(AB056159), PpLFL (EF175869), CaLFY (DQ989225), LFY (M91208), PTLF (AY211519), SdLFY (AY230817), VFL (AF450278), CsLFY (AY338976), FLO (M55525), NFL2 (U16174), EgLFY1 (AY640314), LtLFY (AF321273), RFL (AB005620), ZmLFY (DQ343237). Shading indicated amino acids identical to the consensus amino acid sequence and the dashes indicated gaps included to obtain the best alignment. Conserved regions of alanine-rich and acidic domain were boxed. The positions of the two conserved introns in PpLFL were indicated by two large unfilled arrowheads. The residues involved in dimerization and sequence recognization were indicated by two black triangles

Full size image
Fig. 3
figure 3

Phylogenic relationship among PpLFL and other FLO/LFY proteins. Bootstrap analysis (1,000 replicates) was performed to assess the support of each branch. Branch lengths were shown on branches

Full size image

Copy Number of PpLFL in the Peach Genome

The copy number of PpLFL in peach genome was determined by southern blotting using full-length PpLFL cDNA as probe. The results revealed two Eco RI fragments of 3 kb and 2.5 kb and a 10 kb Bgl II fragment that hybridized to the probe (Fig. 4), which suggested that PpLFL was present as a single copy within the peach genome. This was in agreement with most LFY homolog genes but differed from some that are present as two or more copies in the genome (Kelly et al. 1995; Carmona et al. 2002; Wada et al. 2002, Bomblies et al. 2003; Dornelas et al. 2004).

Fig. 4
figure 4

Southern blotting. Probe used was DIG-labelled full-length cDNA of PpLFL

Full size image

Spatial Localisation of PpLFL

The peach floral apices developed in vegetative phase until the middle of June, and the typical changes of the apical meristem from dome to flat marked the beginning of flower differentiation (Li 1993). Our previous study showed that the expression of PpLFL was restricted in a highly regulated manner. The expression of PpLFL occurred just before the phenotypic change of the floral apices (An and Li 2008). Here, we used immunohistolocalisation to investigate spatial localisation of PpLFL protein during flower induction period in peach. In our experiments, PpLFL localisation was generally limited to the vegetative meristem (Fig. 5a–c) and the promordia that would give rise to the first floral whorl (Fig. 5d–k). The PpLFL was also localisation in the primorda destined to form petals (Fig. 5l).

Fig. 5
figure 5

In situ immunolocalisation of the PpLFL protein in developing shoot apex. Sections were hybridized with purified anti-PpLFL antibodies and photographed using a blue filter under bright-field optics. a–c The meristem with the domed apex; d–f the apex is flattening, marking the onset of floral bud development; g–i the beginning of the morphological differentiation of floral meristem; j–l the differentiation of the floral organs, and the arrows indicated the first whorl and second whorl flower organs. The blue–purple colour indicated the localization of PpLFL protein. The arrows indicated the expression of PpLFL in ‘se’ and ‘pe’ primordia in the SAM. se sepal; pe petal

Full size image

Ectopic Expression of PpLFL in Arabidopsis lfy Mutant

The PpLFL gene driven by the CaMV 35S promoter was transformed into the Arabidopsis lfy mutant to test whether overexpression of PpLFL in Arabidopsis could affect flowering time and rescue the lfy mutation phenotype. The results showed that compared to the wild-type, the transgenic plants showed the clear phenotypes of earlier flowering by approximately 3 days (Fig. 6a–c), and showed solitary shoots rising from rosette axils (Fig. 6d). In the case of 35S:LFY-transformed Arabidopsis or constitutive expression of other LFY homologs in Arabidopsis, the flowers often show abnormal phenotypes and are often observed with five petals and five sepals (Wada et al. 2002); in our experiments, however, in 15 independent transgenic lines, all flowers of the transgenic plants showed normal phenotypes (Fig. 6e).

Fig. 6
figure 6

Phenotypes of 35S:PpLFL transgenic plants. Flowering phenotype of lfy mutant (a), wild-type (b), and 35S:PpLFL transformed line PpL-6 (c); d several solitary shoots (arrows indicated) arising from rosette axils directly; e normal looking flower

Full size image

Discussion

The process of flower induction in fruit trees is very complex and is related to the integration of individual characteristics of plants and environmental conditions. Studies on the molecular mechanisms of this process are very important in agricultural breeding programmes and stable annual crop production. A large amount of research has demonstrated that Arabidopsis LFY and its homologs are key regulators of the integration of inductive signals during flower transition (Blázquez 2005; Corbesier and Coupland 2005; Moyroud et al. 2010; Wang et al. 2011), but the organisation and composition of the amino acid sequence, expression patterns of these genes are varied. Among cultivated woody fruit trees, the peach is considered as the model plant for isolating and characterising key agronomy genes because of its small genome size and structure (Baid et al. 1996). In this study, a LFY homolog PpLFL was isolated from peach axillary buds. Analysis of the nucleotide sequences revealed the gene structure, and the results of alignments showed the conserved regions (Figs. 2 and 3). The length and structure of PpLFL are very similar to LFY gene (Weigel et al. 1992; i.e., a small intron is located close to the 5′ end and a large one is near the 3′ end; Fig. 1), but they differ from some LFY-like genes in the number and size of introns (Allnutt et al. 2007). The alanine-rich domain near the N-terminus is similar to that in most woody perennials (Rottmann et al. 2000; Wada et al. 2002; Carmona et al. 2002; Dornelas et al. 2004) but different from that in herbaceous plants, which are proline rich (Coen et al. 1990; Weigel et al. 1992; Kelly et al. 1995). The acidic domain in the middle of the protein is rich in glutamate and aspartate, which is also in line with most LFY homologs. In particular, LFY executes its functions by inducing the expression of the floral homeotic genes. LFY can bind to the pseudo-palindromic sequence elements (CCANTGT/G) in the promoters of its target genes (Hamès et al. 2008). Two residues in the C-terminus of LFY are involved in such sequence recognization (Hamès et al. 2008). From the deduced amino acid sequence, we found these two residues were conserved in PpLFL (Fig. 2). These similar features imply that PpLFL might have the same function as LFY in flowering.

Our previous results showed that PpLFL was up regulated at the beginning of flower induction (An and Li 2008). Such expression pattern of PpLFL was similar to that of Arabidopsis LFY gene and apple AFL2 gene (Blázquez et al. 1997; Wada et al. 2002). In this study, we used immunohistolocalisation to detect the spatial expression pattern of PpLFL. Spatial expression patterns of PpLFL also revealed some interesting similarities to other LFY homologs. Strong immunohistochemical signals were detected in both the vegetative and floral meristem (Fig. 5). This pattern of hybridisation was most similar to NFL from tobacco, LFY from Arabidopsis, and NEEDLY from Pinus radiata (Kelly et al. 1995; Blázquez et al. 1997; Mouradov et al. 1998). In Arabidopsis, floral morphogenesis begins when the vegetative apical meristem switches to an inflorescence that forms the flower primordium, and eventually both primary and secondary inflorescences terminate in flowers, and LFY expression occurs throughout the inflorescence meristem and increases its intensity during development. We suggested that PpLFL may response for the switch of vegetative apical meristem to become floral meristem.

In Arabidopsis, LFY plays an important role in integrating flower induction signals (Blázquez and Weigel 2000). To determine the function of PpLFL, the 35S:PpLFL construct was transformed to lfy mutants and the transgenic plants were obtained by selection. We found that ectopic expression of PpLFL can not only completely rescue the late flowering phenotype of lfy mutant but also give rise to normal flower organs (Fig. 6). These phenotypes were consistent with the reports for transgenic Arabidopsis plants containing LFY or its homologs from other species (Mouradov et al. 1998; Wada et al. 2002; Chujo et al. 2003), excluding the ScLFY from Silene coeli-rosa, which could not complement the Arabidopsis lfy mutant (Allnutt et al. 2007). Of most interest was that all of our PpLFL transgenic lines had normal flowers. This was different from the apple LFY homolog AFL2. Constitutive expression of AFL2 in Arabidopsis caused abnormal flowers with open carpels, a greater number of petals and sepals, and an apetala1-like phenotype in Arabidopsis (Wada et al. 2002). Thus, our results suggest that ectopic expression of PpLFL can only promote flowering, but has no unusual effect on the development of floral organs.

In summary, a peach gene was isolated from axillary buds with degenerate LFY PCR primers and was highly identity to LFY and its homologs. The expression of PpLFL was limited to the SAM, including vegetative meristems and those floral organs primodium. In addition, ectopic expression showed complete rescue of lfy mutation and give rise to normal flower organs. As LEAFY and PpLFL share some of their biochemical properties, and as the PpLFL protein is detected in the developing floral meristem, PpLFL could act as a floral meristem identity gene in peach too, but this remains to be tested in peach.