The study of a SPATULA-like bHLH transcription factor expressed during peach ( Prunus persica) fruit development

Plant Physiology and Biochemistry 49 (2011) 654e663 Contents lists available at ScienceDirect Plant Physiology and Biochemistry journal homepage: www.elsevier.com/locate/plaphy Research article The study of a SPATULA-like bHLH transcription factor expressed during peach (Prunus persica) fruit development Eleni Tani a, Aphrodite Tsaballa b, Catalina Stedel c, Chrissanthi Kalloniati c, Dimitra Papaefthimiou a, Alexios Polidoros a, Nikos Darzentas a, Ioannis Ganopoulos b, Emmanouil Flemetakis c, Panagiotis Katinakis c, Athanasios Tsaftaris a, b, * a b c Institute of Agrobiotechnology (IN.A.), CERTH, 6th km Charilaou-Thermis Road, Thermi GR-570 01, Greece Department of Genetics and Plant Breeding, School of Agriculture, AUTh, Thessaloniki GR-541 24, Greece Department of Agricultural Biotechnology, Agricultural University of Athens, Iera Odos 75, 118 55 Athens, Greece a r t i c l e i n f o a b s t r a c t Article history: Received 24 August 2010 Accepted 11 January 2011 Available online 27 January 2011 Extensive studies on the dry fruits of the model plant arabidopsis (Arabidopsis thaliana) have revealed various gene regulators of the development and dehiscence of the siliques. Peach pericarp is analogous to the valve tissues of the arabidopsis siliques. The stone (otherwise called pit) in drupes is formed through lignification of the fruit endocarp. The lignified endocarp in peach can be susceptible to split-pit formation under certain genetic as well as environmental factors. This phenomenon delays processing of the clingstone varieties of peach and causes economical losses for the peach fruit canning industry. The FRUITFULL (FUL) and SHATTERPROOF (SHP) genes are key MADS-box transcription protein coding factors that control fruit development and dehiscence in arabidopsis by promoting the expression of basic helixloop-helix (bHLH) transcription factors like SPATULA (SPT) and ALCATRAZ (ALC). Results from our previous studies on peach suggested that temporal regulation of PPERFUL and PPERSHP gene expression may be involved in the regulation of endocarp margin development. In the present study a PPERSPATULA-like (PPERSPT) gene was cloned and characterized. Comparative analysis of temporal regulation of PPERSPT gene expression during pit hardening in a resistant and a susceptible to split-pit variety, suggests that this gene adds one more component to the genes network that controls endocarp margins development in peach. Taking into consideration that no ALC-like genes have been identified in any dicot plant species outside the Brassicaceae family, where arabidopsis belongs, PPERSPT may have additional role(s) in peach that are fulfilled in arabidopsis by ALC. Ó 2011 Elsevier Masson SAS. All rights reserved. Keywords: bHLH transcription factor Prunus persica Endocarp development 1. Introduction Fleshy fruits have been extensively examined due to their high economic value. While tomato became a model plant for fleshy fruit development of annual plants [1e3], peach (Prunus persica) is emerging as a model tree species for comparative genomics and identification of genes for flower and fruit development due to its small genome size of 300 Mb (about twice that of arabidopsis) and the relatively short reproductive time for a fruit tree (2e3 years until flowering). Particularly, peach cultivation of clingstone varieties is of fundamental importance for Greek agriculture and its canning industry, since next to USA, Greece is the largest canned * Corresponding author. Institute of Agrobiotechnology CERTH, 6th km Charilaou-Thermis Road, Thermi GR-570 01, Greece. Tel.: þ30 2310 498271; fax: þ30 2310 498270. E-mail address: tsaft@certh.gr (A. Tsaftaris). 0981-9428/$ e see front matter Ó 2011 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.plaphy.2011.01.020 peach producer worldwide, and the first exporter among EU countries. The commercial value of clingstone peach varieties depends on their resistance to split-pit. Split-pit of peaches is a physiological disorder characterized by a splitting or shattering of the pit and has been associated with genetic as well as environmental factors [4e6]. The magnitude of the disorder varies greatly among varieties indicating a strong role of the genetic factors involved, and it is usually worse in early-maturing varieties. Affected fruits ripen earlier and may exhibit skin splitting near the stem end of the fruit, permitting the easier entry of insects and disease-producing organisms. Spitted pits are crushed during removal of the pit in the canning factories dispersing pit fragments and resulting in economic losses and peach quality deterioration. Anatomical and physiological comparisons indicate that peach pericarp is an organ analogous to arabidopsis valves (both originating from the ovary carpels). Furthermore split-pit formation, in peach endocarp, is taking place exactly at the separation layer 655 E. Tani et al. / Plant Physiology and Biochemistry 49 (2011) 654e663 (suture), in a manner resembling the formation of the dehiscence zone of the valve’s endocarp. Different arabidopsis mutants that produce fruits with developmental defects revealed the mode of post-fertilization development of the arabidopsis silique fruit [7,8]. Different genetic tools were used to elucidate the molecular pathways involved in the development of the arabidopsis silique dehiscence zone. The first transcription factors shown to participate in dehiscence zone specification in arabidopsis were the MADS-box transcription factor proteins coding genes SHATTERPROOF1 (SHP1) and SHATTERPROOF2 (SHP2) [9]. The FRUITFULL (FUL) gene product, another MADS-box transcription factor which was detected in valves, represses SHP gene expression [10]. Moreover, REPLUMLESS (RPL), a member of the bell subfamily of homeodomain transcription factors, suppresses SHP1, 2 expression in the replum [11]. Thus both FUL and RPL restrict SHP activity to a narrow strip of cells sandwiched between the valve and the replum. Three basic helix-loop-helix transcription factors (bHLH) were also found to be involved in arabidopsis fruit dehiscence process: ALCATRAZ (ALC) [12], its closest relative, SPATULA (SPT) and INDEHISCENT (IND) [9]. IND, ALC, and SHP form together a regulatory network that orchestrates the differentiation of the valve margin, allowing seed dispersal to take place. On top of them, FUL negatively regulates IND, ALC and SHP1,2 to ensure that valve margin differentiation occurs at the edge of the valve [9,10]. SPT is expressed continuously in the margins of developing carpels, presumably supporting their growth in arabidopsis [13,14]. SPT expression was also localized in the stomium of the anther, the style and the stigma. Moreover, in spt mutants the transmitting tract is absent [15]. In the maturing silique, SPT expression is restricted to the edges of the valves in a pattern that resembles that of the SHATTERPROOF MADS-box genes. However there is no data available for its functional role(s) there [16e18]. Conversely, there is no evidence that ALC functions earlier in gynoecium development when SPT is active. SPT and ALC are apparently the products of recent gene duplication in an ancestor of the Brassicaceae family [15], since no ALC genes were found in species outside the Brassicaceae family. Further studies may uncover overlapping functions of these two closely related genes in arabidopsis. Since a similar network of transcription factors could control the opening of the peach endocarp margins, we studied genes affecting peach endocarp development [5,6]. As a continuation of these efforts, herein we cloned and characterized a PPERSPT-like gene from peach. Sequence characterization and phylogenetic analysis clearly classify this gene to SPT-like proteins. Interestingly, its expression profile during peach flower and fruit development shows that it may have additional functions as an ALC-like gene bearing in mind that no ALC-like gene has been identified in peach genomic database, as in any other dicot species examined, outside the Brassicaceae family. 2. Materials and methods 2.1. Plant material Flowers and developing fruits were collected from Veria, in Northern Greece, the main area of peach cultivation, in spring 2004, every week after anthesis until the collection of ripened fruits. Samples were taken from peach varieties ‘Andross’ (split-pit sensitive) and ‘Katherine’ (split-pit resistant), frozen in liquid nitrogen and stored at 80  C until used. 2.2. Cloning of PPERSPT gene We performed an improved inverse-RACE method, called Rolling Circle Amplification RACE (RCA-RACE), which employs CircLigaseTM (Epicentre, Madison, WI) for cDNA circularization followed by rolling circle amplification of the circular cDNA with Phi29 DNA polymerase and random primers [6,19]. A circular cDNA pool was constructed using total RNA from flowers and developing fruits of the ‘Andross’ variety. In order to isolate ALC/SPT-like genes from peach we designed degenerate primers based on the conserved domains found in arabidopsis ALC and SPT and its mostly related bHLH transcription factors proteins [20]. Nested PCR produced three different clones through the use of degenerate primers (ALC/ SPT-K-F and ALC/SPT-K-R). BLAST analysis showed sequence similarity of two clones to arabidopsis PIL5 and PIF3 transcription factors (data not shown) that were not further studied. The third clone (a 604 bp fragment) was homologous to ALC and SPT genes. Based on the obtained sequence we designed primers in order to obtain the missing 50 and 30 ends as well as the internal site between the degenerate primers. The PCR fragments were purified by agarose gel electrophoresis and cloned into pCR 2.1-TOPO vector using the TOPO TA cloning Kit (Invitrogen). The sequences of all primers used are presented on Table 1. 2.3. Protein sequence comparisons and phylogenetic analysis The deduced amino acid (a.a.) sequence of PPERSPT was employed for BLAST analysis on the NCBI database (http://www. ncbi.nlm.nih.gov). Among the BLAST hits, proteins that exhibited high degree of similarity were chosen for subsequent analysis (accession numbers are given in brackets). More similar sequences to PPERSPT were: Ricinus communis conserved hypothetical protein (XP_002523613) (62% identity), Vitis vinifera predicted hypothetical protein (XP_002277966) (61% identity), R. communis conserved hypothetical protein (XP_002510190.1) (50% identity), Populus trichocarpa predicted protein (EEE73912) (48% identity), and arabidopsis SPT (NP_568010) (44% identity). More SPT-like and ALC proteins were identified on the Uniprot database (http://www. uniprot.org): Zea mays (B6SNH3, B6SIB8, B6TPF3 and B6TPF3), Orysa sativa subsp. japonica (Q5VRS4), Capsicum annuum (Q6RJZ4) and the three isoforms of arabidopsis ALC (Q9FHA2, Q9FHA2-2 and Q9FHA2-3). Another predicted Brassica napus ALC-like a.a. sequence produced by the translated assembly of ESTs EE432874.1, Table 1 Primer sequences that were used in the experiments. Regarding the primers used for in situ hybridization experiments, consensus T3 sequence is shown in bold letters, preceded by a 9 base pair leader sequence in italic and followed after the slash (/) by the gene-specific sequence. Primer name Sequence PPERACTIN-2-F PPERACTIN-2-R PPERALC/SPT-K-F PPERALC/SPT-K-R PPERALC/SPT- F2 PPERALC/SPT 5end-1 PPERALC/SPT 5end-2 PPERALC/SPT-10-R1 PPERALC/SPT-R2 PPERALC/SPT-R5 PPPERALC/SPT-R13 PPERALC/SPT-F4 PPERALC/SPT-F5 PPERALC/SPT-F-INSITU PPERALC/SPT-R-INSITU 50 -GTGGGGATGGGACAGAAAGATG-30 50 -GAGGTCAAGCCGGAGGATGG-30 50 -ATGCTYGAYGARGCHATYGANTA-30 50 -GCYTTCATYTTYTCRTTDAT-30 50 -ATGYTKGATGARGCTATTGA-30 50 -ATGGGGGATACTTATGATC-30 50 -AGAAACCAGAGAGTGATGG-30 50 -TGAGCCCGCATTGACGATT-30 50 -TCGGATGAGTTTGGTAGTTG-30 50 -GATGAGTTTGGTAGTTGAA-30 50 -GCACTTGGAGCTGAAGCT-30 50 -ACCAATGCACCGATGCAAAC-30 50 -TGCTTATGAACCAGGAATCT-30 50 -AGCCGTCCGTGCCACTCCAT-30 50 -CAAGCTTCATTAACCCTCACTAAAGGGAGA/ TGCTTCAAGACCCT-30 656 E. Tani et al. / Plant Physiology and Biochemistry 49 (2011) 654e663 EE555580.1, EV007981.1, EV029403.1, GR448354.1, EL625410.1 (UGID: 2995763) along with the arabidopsis phytochrome interacting factor 3 protein (PIF3) (NP_849626), were also included in the analysis. More related sequences assigned by [15] as SPT orthologues were also included from diverse species (Table 2). The deduced a.a. sequence of PPERSPT was subjected to BLAST search, against the assembled genome of peach (v1.0), downloaded from the Phytozome database (http://www.phytozome.org/peach). Among the BLAST hits, a w2500 bp fragment of the genomic scaffold_7 was identified as the corresponding genomic fragment to PPERSPT. The predicted a.a. sequence of this fragment was also included in the analysis under the name PRUPE (P. persica) scaffold7. All sequences were aligned using the multiple sequence alignment program MAFFT [21]. The resulting alignment was edited with Jalview [22] and used for the protein subfamily identification, employing the SCI-PHY algorithm [23]. After subfamily identification, the multi-RELIEF feature weighting method [24] was used to detect specificity determining a.a. residues among subfamilies. Phylogenetic analyses of the complete PPERSPT mRNA sequence were performed using 36 highly similar homologous plant a.a. sequences carefully selected by sequence similarity searches from the PlnTFDB (3.0) [25] and UniProt databases (The Uniprot Consortium) (Table 2). Three algae sequences were selected for outgroup, based on a former phylogenetic analysis run for the plant bHLH family [26]. The final 910 a.a. sequence alignment was obtained using MAFFT [21], with settings for 1.53 gap opening and 0.13 gap extension. Model Generator [27] was Table 2 Proteins included in the protein sequence comparisons assigned as SPT-orthologues by [15], and proteins used for the phylogenetics analysis and not used in the protein sequence comparisons. For each protein, the species, the accession and the database from which it originated are also specified. Species Accession Database origin Gossypium hirsutum Citrus clementina Fragaria x ananassa Medicago truncatula Glycine max Populus trichocarpa contig of DT572940 and DN780013 DY289192 AY679615 AC144431 AC170860 eugene3.00400325 Vitis vinifera Solanum lycopersicum Oryza sativa Oryza sativa Picea glauca Vitis vinifera Arabidopsis thaliana Arabidopsis lyrata Arabidopsis lyrata Populus trichocarpa Ricinus communis Carica papaya Vitis vinifera Zea mays Sorghum bicolor Oryza sativa Sorghum bicolor Zea mays Ricinus communis Arabidopsis thaliana Zea mays Orysa sativa Arabidopsis thaliana Arabidopsis lyrata Arabidopsis thaliana Micromonas sp. Chlamydomonas reinhardtii Chlorella sp. AM434294 TC174571 Os02g56140, 11972.m10638 Os06g06900, 11976.m05415 TC30545 CBI21410.3 CAB16798.1 XP_002866965.1 fgenesh2_kg.8_2705_AT5G67110.1 XP_002327162.1 EEF52377.1 evm.TU.supercontig 21.173 XP002284880 B4F8E4 C5XV52 Q948F6 fgenesh1_pg.C_chr_100000471 ACF80365.1 EEF41301.1 BAC56979.1 NP_001147809.1 Q7FA23 Q8W2F3.1 XP_002878258.1 At1g09530 eugene.0300010394 fgenesh2_pg.C_scaffold_40000011 GenBank GenBank Genbank Genbank Genbank Joing Genome Institute (JGI) Genbank TIGR TIGR TIGR TIGR GenBank GenBank GenBank PlnTFD GenBank GenBank PlnTFD GenBank Uniprot Uniprot Uniprot PlnTFD GenBank GenBank GenBank GenBank Uniprot Uniprot GenBank GenBank PlnTFD PlnTFD IGS.gm_8_00085 PlnTFD employed for a.a. substitution model selection using the Bayesian information criterion to search against 96 existing models. Both distance and character based estimation methods were used for the calculation of sequence evolutionary history on the final tree topology. Distance and parsimony phylogenetic tree building methods were employed using the phylogenetic package PHYLIP v3.6 [28] whereas 1000 bootstrap replicates was the statistical inference method for topology support. For weighted Neighbor joining (NJ) method analysis [29], evolutionary distances were computed using the JTT matrix-based method [30]. Rate variation among sites was modeled with a gamma distribution (shape parameter ¼ 1.39). Both pairwise and complete a.a. deletion options were invoked, eliminating correspondingly pairwise and all gaps as well as unique positions. In pairwise deletion all sites were included, while for the complete deletion option a total of 51 positions were considered for the analysis, mainly corresponding to the bHLH domain. For the maximum parsimony (MP) method [31] all gaps were treated as missing data, thus 456 a.a. positions were included in the analysis as parsimony informative with 0.773 consistency index, 0.738 retention index and 0.57 composite index. Maximum likelihood (ML) [32] estimation of phylogenetic inference was calculated using PHYLIP and 1000 bootstrap replicates in order to access branch support. The amino acid substitution model employed was JTT with 1.39 discrete gamma distribution of 8 rates. Bayesian inference (BI) of phylogeny was constructed using the software program MrBayes 3.1 [33] where 10 a.a. models were tested for the final selection of the Jones substitution model with 1.37 mean gamma shape parameter (alpha). Four Metropolis-coupled Markov chain Monte Carlo chains were run on one million generations, with sampling rate of 100 generations. After discarding the first 3000 unstable trees as burn-in, a majority-rule consensus tree was constructed. Final tree topologies were almost identical for all tree reconstruction methods used and final representation was based on the ML topology as the backbone. Statistical topology support was represented by corresponding bootstrap values calculated for each branch by the ML, BI, MP and NJ phylogenetic inference methods. 2.4. DNA isolation and southern analysis Five mg of genomic DNA were digested with EcoRI, EcoRV, DraI, HindIII or BamHI, (TaKaRa, Otsu, Japan) and transferred to a positively charged Nylon membrane. The digoxigenin labeled PPERSPT gene-specific probe was prepared by PCR using primers PPERSPTF4 and PPERSPT-R13 and the PCR DIG Probe Synthesis Kit (Roche, Mannheim, Germany). Hybridization was performed using the DIG Easy Hyb buffer at 42  C according to the manufacturer and stringent washes were performed at 68  C in 0.5  SSC containing 0.1% SDS (twice). Detection was performed using the DIG Luminescent Detection Kit and finally chemiluminescense was monitored using the GeneGenome Bio Imaging System (Syngene, Cambridge, U.K.). 2.5. Expression analysis RNA was isolated from leaves, petals sepals, carpels (including ovules) and stamens of fully developed flowers from both varieties using the RNeasy plant RNA isolation kit (Qiagen, Crawley, U.K.). RNA was also isolated from the pericarp of the varieties ‘Andross’ and ‘Katherine’ from 2 to 12 weeks after full anthesis. For the RT-PCR, 0.5 mg of total RNA were used for first strand cDNA synthesis. The cDNA was synthesized using 1 mg of 30 RACE adapter Primer 50 -GCCACGCGTCGACTAGTAC(T) 17-30 (Invitrogen), 1 mM dNTPs and 200 U M-MuLV reverse transcriptase (NEB, Beverly, USA) in 50 ml total volume. This cDNA served as a template for the PCR reaction, E. Tani et al. / Plant Physiology and Biochemistry 49 (2011) 654e663 using 0.2 pmol of gene-specific primers, 0.2 mM dNTPs and 1 U DyNAzyme II DNA polymerase (Finnzymes, Espoo, Finland). PCR primers PPERSPT-F5 and PPERSPT-R1 were used for RT-PCR experiments and their sequences are given in Table 1. PPERACTIN-2 was used as the RT-PCR control. The thermocycler program involved 30 cycles of: 30 s at 94  C; 45 s at 55  C (or 53  C for PPERACTIN); and 1 min at 72  C, which were preceded by 3 min at 94  C and followed by 10 min at 72  C. Control PCR reactions contained the RNA that was used as a template in the cDNA synthesis. The PCR products were separated on 1% agarose gels where the amplification products of the expected size could be observed. 2.6. Real-time PCR Quantitative expression analysis of PPERSPT during fruit development was performed with real-time RT-PCR using an Opticon 3 (MJ Research, Waltham, MA) real-time PCR system. Real-time PCR was mainly used to provide a relative quantitative estimation of the PPERSPT expression ratio (see below) during the crucial period of pit hardening (almost 5 weeks after anthesis until completion of pit hardening). The PCR reactions were performed in 10 ml total volume using the Sybr-Green Kit (Invitrogen). The template was 1/50 of the cDNA synthesized from 0.5 mg of RNA extracted from developing fruits. The cycling parameters for PPERSPT were: incubation at 95  C for 2 min, followed by 34 cycles of incubation at 95  C for 30 s, 55  C for 30 s, 72  C for 30 s, plate read at 80  C and a final extension step of 5 min at 72  C. To identify the real-time PCR products, a melting curve was performed from 65 to 95  C with observation steps every 0.2  C and a 10-s hold between observations. The reactions were performed in triplicate. SYBR Green I fluorescence dye, which binds to double-stranded DNA, was used to monitor the newly synthesized PCR products [34]. The relative amount of RNA in each sample 657 was calculated by the 2 DDCT method using actin as internal control and the relative expression of ‘Andross’, two weeks after anthesis (A2) as calibrator [35]. Comparison of ‘Andross’ versus ‘Katherine’ was made by dividing the 2 DDCT of the two varieties (2 DDCT(A)/ 2 DDCT (K)) for every developmental stage. 2.7. In situ hybridization In situ hybridization experiments were performed as has it been described previously [36]. Briefly, fully developed buds, fruits before the onset of pit hardening (5 weeks after anthesis) as well as fruits in the middle phase of pit hardening (8 weeks after anthesis) were fixed in 4% (w/v) paraformaldehyde supplemented with 0.25% (v/v) glutaraldehyde in 10 mM sodium phosphate buffer (pH 7.4) for 4 h in a vacuum aspirator. Fixed tissues were block-stained in 0.5% (w/v) safranin, dehydrated through ethanol series, embedded in paraffin and cut into 8 mm-thin sections. Antisense RNA probes labeled with digoxigenin-11-rUTP (Boehringer Mannheim, Mannheim, Germany) were originated from PCR-generated templates incorporating T3 polymerase sites [37]. The probe was designed close to the 30 end of the gene and its length was 175 bp. Sections were prepared for hybridization as described (Scheres et al., 1990) and hybridized overnight at 42  C in 50% (v/v) formamide, 300 mM NaCl, 10 mM TriseHCl pH 7.5, 1 mM EDTA, 0.02% (w/v) Ficoll, 0.02% (w/v) polyvinylpyrrolidone, 0.025% (w/v) bovine serum albumin (BSA), 10% (v/v) dextran sulfate and 60 mM DTT. After hybridization, the sections were treated with a solution containing 500 mM NaCl, 1 mM EDTA, 10 mM TriseHCl and 50 lg/ml RNase A. Finally, sections were washed several times in a 2 SSC solution. Hybridization signals were visualized with anti-digoxigenin antibodies conjugated with alkaline phosphatase. Images were processed using Photoshop 6 software (Adobe Systems Inc., San Jose, CA, USA). Fig. 1. Multiple alignment of the PPERSPT predicted a.a. sequence (1_PRUPE_ADG56590) and its genomic translation (1_PRUPE_scaffold7) (both highlighted in yellow) with proteins from other species. Only informative areas of the alignment are shown that include the amphipathic helix domain (positions 53e63 of the alignment), the acidic domain (174e196), only present in subfamily 1, the two NLS regions (216e219 and 229e232), the bHLH domain (220e278) and the beta strand (269e276). Specificity determining residues identified with the multi-RELIEF algorithm are highlighted in black background. The name of each sequence consists of the number of subfamily, followed by the species (according to Uniprot) and finally, the accession number or the Uniprot ID or the TIGR annotation. Two areas of the alignment (indicated by small wedges and corresponding vertical lines: 66e170 and 199e214) have been hidden for clarity. Identically colored a.a. share similar biochemical properties. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article) 658 E. Tani et al. / Plant Physiology and Biochemistry 49 (2011) 654e663 3. Results 3.1. Cloning of the PPERSPT gene The PPERSPT gene was cloned following the RCA-RACE method developed recently from our group [19]. Using degenerate primers, we obtained a 604 bp sequence and then, with specific primers based on the obtained sequence, we obtained clones towards the 50 and 30 ends as well as the short internal site between the degenerate primers. Analysis of the sequencing results using the SeqMan software package (DNA Star, Madison, WI) revealed that the RCARACE clones and the obtained 604 bp PPERSPT sequence belong to the same contig. A PPERSPT 1730 bp transcript revealed an 1160 bp ORF encoding for 386 a.a. The sequence was deposited to the GenBank database (GU181270). Using the Spidey mRNA to genomic alignment program at NCBI (http://www.ncbi.nlm.nih.gov/spidey/), the PPERSPT mRNA was aligned with the w2500 bp fragment of peach’s scaffold_7, that was identified by BLAST as the corresponding genomic fragment. The alignment revealed the genomic organization of PPERSPT consisting of seven exons of 398, 105, 66, 66, 315, 70 and 140 bp respectively and six introns of 297, 140, 96, 619, 80 and 103 bp respectively. 3.2. Sequence characterization and analysis The PPERSPT prediction of a.a. sequence aligned with SPT-like and ALC-like proteins from other plant species allowed the identification of nine subfamilies. Four of them had only one member and were excluded from subsequent analysis. These subfamilies included the following proteins: C. annuum SPT-like (Q6RJZ4), arabidopsis PIF3 (NP_849626), P. glaucea SPT-like (TC030545) and Z. mays SPT-like (B6TPF3). The final dataset was composed of five subfamilies. The sequence alignment was re-calculated (Fig. 1). These five subfamilies were labeled as: 1, containing dicotyledonous SPT sequences; 2, containing rice SPT sequences; 3, maize Fig. 2. Phylogenetic analysis of 24 SPT-like and ALC-like and 7 PIF a.a. sequences from different monocots and dicots including the PPERSPT sequence (ADG56590.1) underlined in the tree. Numbers next to branches indicate bootstrap values (100 ¼ 100%) corresponding to ML, BI, MP, NJ analyses respectively. Three algal bHLH family sequences were used for rooting the tree. Scale represents one (1)(1) amino acid substitutions. Species abbreviations by Uniprot. E. Tani et al. / Plant Physiology and Biochemistry 49 (2011) 654e663 SPT sequences; 4, containing ALC sequences and 5, containing Solanaceous SPT-S sequences, according to [15]. The alignment confirmed the presence of four domains: a putative amphipathic helix (positions 52e63 of the alignment) and an acidic alpha helical domain present only in few SPT putative orthologs (positions 174e196), the bHLH domain (positions 220e268) and a conserved C-terminal extension in the form of a beta strand (positions 269e277) [15,20,38] (Fig. 1). The PPERSPT protein possesses both the amphipathic and the acidic domains, characteristic of SPTs [15]. Moreover, PPERSPT shares 59 out of 62 residues in the bHLH protein domain with the already characterized arabidopsis SPT [39]. Homologous genes were considered to be SPT orthologues if they matched this 62 a.a. sequence with greater identity than the closest relative of SPT in arabidopsis, ALC, which shares 51 out of 62 identical residues [12]. Several potential specificity determining residues, i.e., capable of separating the five subfamilies, were identified and are highlighted in black background at alignment positions 192, 218, 219, 220 and 222, on Fig. 1. Specifically: (a) The lysine residing in position 192 (K192) in subfamily 4 (ALCs) is substituted by glutamic acid (E192) in subfamilies 1, 2 and 5. (b) Respectively, N218, I219 and Q222 in subfamily 4 are substituted by S218, R219 and E222 respectively in all the other subfamilies. (c) In addition D220 in subfamily 4 is substituted by A220 in subfamilies 1, 2 and by a S220 in subfamilies 3 and 5. 3.3. Phylogenetic analysis Phylogenetic analyses of the complete PPERSPT mRNA were carried out. The highly related bHLH sequences dataset resulted in the topologically robust tree (Fig. 2) supported by both ML and the BI pairwise deletion analyses. Similar topologies were produced by the MP and NJ algorithms. The PPERSPT sequence was deeply rooted within other related dicotyledonous SPT and SPT-like sequences on Group 1. The three algal outgroup sequences provided strong support to the overall tree topology, where Group 5 containing PIF sequences was distinctively formed. PIF3, a bHLH sequence similar to SPT was also located within this branch. A monophyletic origin was attributed to the SPT and ALC lineage by all phylogenetic inference estimations, with variable bootstrap supports (ML: 69%, BI: 78%, MP: 64%, NJ: 47%). An exclusively cereal grass SPT-like sequences branch (Group 4) was formed as basal of the SPT/ALC lineage group with high topological support (92, 99, 96, 50% respectively for ML, BI, MP, NJ). The next level, deeper in the phylogeny, was occupied by a strongly supported branch (Group 3) formed by five dicotyledonous sequences, including two trees (P. trichocarpa, Carica papaya), R. communis, and V. vinifera. Diverging from this branch, the three ALC sequences formed a strongly supported group (Group 2), distinctly differentiated by all SPT sequences. All other dicotyledonous SPT-like sequences were assembled deeper in the phylogeny, assigned to two distinct branches (Group 1). A strongly supported cluster of three arabidopsis SPT sequences (100, 100, 100, 98%, ML, BI, MP, NJ) grouped together with one sequence from R. communis. In another cluster, one highly supported branch contained three V. vinifera sequences that clustered separately from the branch containing the peach sequence, strongly grouped with one Fragaria x ananassa SPT-like sequence. 3.4. Southern analysis Results from restriction enzyme digestions for the number of PPERSPT copies in peach were inconclusive since multiple bands 659 Fig. 3. Southern blot of genomic DNA from peach digested with Hind III (H) EcoRI (E) DraI (D), BamHI, (B), EcoRV (Ev) and probed with a PPERSPT-specific probe. were observed (Fig. 3). One band was hybridized after EcoRI and EcoRV digestion. The hybridization of two bands after HindIII and DraI digestion can be attributed to the presence of recognition sites of the aforementioned restriction enzymes inside the probe sequence. The hybridization of multiple bands after BamHI digestion could be ascribed to the presence of recognition sites of the restriction enzyme inside intron(s). However, existence of multiple copies for this gene cannot be excluded. 3.5. Expression analysis Initial RT-PCR experiments were carried out on cDNAs derived from leaves, sepals, petals, carpels and stamens of a fully developed flower. PPERSPT was present in all tissues examined although it is expressed at low levels in stamens (Fig. 4). The localization of the expression of PPERSPT in fully developed flowering buds was examined through in situ hybridization, too (Fig. 5). Expression of PPERSPT is most prominent in the ovary and the perianth, whereas its transcription is reduced to a weak signal in stamens (Fig. 5A1). Moreover, in the cross section of 5 weeks-old developing fruit signal is detected along the suture (margins) of the endocarp’s halves, starting from the union between the seed and the endocarp (Fig. 5B1). Examining the close up of the longitudinal section of 5 weeks-old fruit, PPERSPT transcripts are visible along the union of the two halves of the peach pericarp (Fig. 5C1). Strong expression of PPERSPT is detected in seeds in the longitudinal sections of 8 weeks-old fruit (Fig. 5C5, D2-4). A faint signal is also detected at the endocarp’s suture (Fig. 5D1). Finally PPERSPT expression is localized on the upper union of the two parts of peach endocarp (stone) as shown in Fig. 5C6. Real-time PCR provided a relative quantitative estimation of PPERSPT expression (Fig. 6). PPERSPT relative expression ratio was very low in both varieties during the first weeks of fruit development (2e3 weeks after full anthesis) and increased rapidly just before (5 weeks after anthesis) and during the period of pit hardening (7e10 weeks after full anthesis) (Fig. 6A, B). A significant augmentation of PPERSPT relative expression ratio in the susceptible variety ‘Andross’ compared to ‘Katherine’ was detected 4 Fig. 4. Expression analysis of PPERSPT by semi quantitative RT-PCR using ACTIN-2 as an internal control. Agarose gel, stained with ethidium bromide, showed a 178 bp PPERSPT fragment. The 402 bp ACTIN-2 fragment is shown at the bottom part of the figure. L: leaves, P: petals, S: sepals, ST: stamens, C: carpels. 660 E. Tani et al. / Plant Physiology and Biochemistry 49 (2011) 654e663 Fig. 5. Detection of the expression of PPERSPT in longitudinal sections of fully developed buds (A1). As a negative control, longitudinal sections (A2) were hybridized to sense RNA probes transcribed from pGEM-luc Vector (Promega, Madison, WI, U.S.A.). In this case, no hybridization signal was visible. Expression of PPERSPT is most prominent at the ovary (O), style (S) and perianth (P) (A1). No signal is detected at the stamens (St). Cross section (B1) and longitudinal section (C1) of 5 weeks-old developing fruit along with the respective negative controls (B3) and (C3). B2. Close up of the endocarp’s suture between two halves. Signal is detected along the suture beginning from the union between the seed and the endocarp. B4. No signal is detectable in the negative control. Examining the close up of the longitudinal section, PPERSPT transcripts are visible along the union of the two parts (hemispheres) of the peach pericarp (I) C4. Longitudinal sections of 8 weeks-old fruits (lower part of the fruit) with its negative control (C7). D1. Longitudinal sections of 8 weeksold fruits (upper part of the fruit) with its negative control (D5). PPERSPT expression is localized on the upper union of the two parts of peach endocarp (stone) (C6). No signal is visible in the negative control (C8). Strong expression of PPERSPT is detected in seeds (close ups C5, D2, D3) where no hybridization signal is visible in the negative controls (close ups C9, D6, 7, 8). A faint signal is also detected at the endocarp’s suture (D4). Size bars 2 mm (A1, 2), 8 mm (B1, 3 C1), 10 mm (C4, 7) and 12 mm (D1, 5). weeks after full anthesis (Fig. 6B,C). Interestingly, PPERSPT relative expression ratio was higher in variety ‘Andross’ than in ‘Katherine’ during the whole period of pit hardening exhibiting the highest peaks in differences just before the initiation of pit hardening (5 weeks after anthesis), as well as at the intermediate phase of pit hardening (8 weeks after full anthesis) (Fig. 6A,C). 4. Discussion In arabidopsis, a network of interacting factors that determine the proper fruit development and seed dispersal has been identified. Fruit opening occurs at the valve margins, which appear at the valve/replum border. The valve margins differentiate into narrow E. Tani et al. / Plant Physiology and Biochemistry 49 (2011) 654e663 661 Fig. 6. Quantitative analysis of PPERSPT expression in peach. Sampling took place from the beginning of fruit growth until the end of pit hardening (2e10 weeks after full anthesis). The relative amount of RNA in each sample was calculated by the 2eDDCT method using actin as internal control and the relative expression of ‘Andross’, two weeks after anthesis (A2) as calibrator (A,B). In Y axis, the numbers next to the letters indicate weeks after full anthesis (e.g., 2 ¼ week 2 after full anthesis). Comparison of ‘Andross’ versus ‘Katherine’ was made by dividing the 2 DDCT for both varieties (2 DDCT (A)/2 DDCT (K)) for each developmental stage(C). Again in Y axis, the numbers indicate weeks after full anthesis. A ¼ ‘Andross’, K ¼ ‘Katherine’. strips of cells consisting of a separation layer and a layer of lignified cells that both contribute to the process of fruit opening. The SHP1,2 genes are valve margin identity factors which promote expression of two genes encoding bHLH transcription factors, IND and ALC [9,12]. The genes RPL and FUL act upstream of the aforementioned genes, ensuring that their action is limited to the narrow strips where valve margin cells form [9,10,40]. While IND is necessary for specification of both the lignified layer and the separation layer of the valve margin, ALC is primarily involved in separation layer formation. IND belongs to the same subclade of the bHLH family as the HECATE1 (HEC1), HEC2 and HEC3, which are involved in transmitting tract formation [41], while the closest homologue of ALC is SPT [12,20,42]. In the present study we have cloned and characterized a SPATULA-like gene from peach, designated as PPERSPT. PPERSPT and the arabidopsis SPT gene share the same genomic organization. Both genes have seven exons interrupted by six introns: the bHLH domain is being translated by a part of the second exon, the third, fourth and part of the fifth exon. On the contrary, the arabidopsis ALC gene has a different genomic organization: it consists of five exons and four introns. Notably, the arabidopsis ALC protein is 210 a.a. long, significantly shorter than PPERSPT (386 a. a. long) and arabidopsis SPT (373 a.a. long). Thus, PPERSPT is more closely related to arabidopsis SPT and is more distant to arabidopsis ALC. The deduced PPERSPT sequence was aligned with homologous proteins from other plant species and categorization in subfamilies followed (Fig. 1). PPERSPT is placed in the same subfamily with arabidopsis SPT (subfamily 1) while other SPT-like proteins from monocotyledonous species like maize and rice are grouped separately constituting different subfamilies (subfamilies 2 and 3). The three protein isoforms of ALC from arabidopsis and one ALC from B. napus set up their own subfamily (subfamily 4). Similarly SPT-S proteins from Solanaceae species form another subfamily (subfamily 5) consistent with the previous findings by [15]. The recognition of potential specificity determining residues shows that subfamily 1, where PPERSPT and arabidopsis SPT belong and subfamily 4 of ALCs differ in significant a. a. such as the first and third a.a. of the bHLH domain (positions 220 and 222 of the alignment). Another two significant differences are spotted in positions 218 and 219 of the alignment, inside the region that includes the nuclear localization signal (NLS), according to [15]: the a.a. sequence of this region in the ALCs’ subfamily 4 is KRNI while in SPTs’ subfamily 1 is KRSR. From these differences inside the NLS region, the substitution of the last a.a., isoleukine in ALCs by an arginine in SPTs seems interesting since a hydrophobic I is replaced by a hydrophilic R. Deduced PPERSPT sequence and arabidopsis SPT share 59 (out of 62) conserved a.a. of the extended bHLH region (including the beta strand) (Fig. 1). The amphipathic helix and acidic domains were also identified in the PPERSPT protein. The amphipathic helix domain was not detected in rice and maize homologous proteins but a similar domain was identified in the group of SPT-S from Solanaceae and in ALCs. In agreement with [15], the acidic domain was detected only in SPTs from dicotyledonous species and not in the ALCs or in the other SPT-like from rice and maize. The amphipathic helix and the acidic domains were not detected in the other 14 members of the SPT subclade of bHLH proteins in arabidopsis in group VII of [20] and in subfamily 15 of [38], with the only 662 E. Tani et al. / Plant Physiology and Biochemistry 49 (2011) 654e663 exception of ALC protein, where an amphipathic helix could be identified between codons 17e27. The amphiphatic helix enhances the function of SPT in carpel growth while the presence of the acidic domain is necessary for it [15]. Phylogenetic analysis gave strong support to the assignment of the PPERSPT protein as a SPT rather than an ALC protein (Fig. 2). Specifically, the phylogeny created three groups of SPT-like sequences, one exclusively monocotyledonous (Group 4) and two other exclusively dicotyledonous (Groups 1 and 3). The PPERSPT protein was assigned together with Fragaria x ananassa, both belonging to the Rosaceae family, within a strongly supported group deep in the phylogenetic tree. The three ALC sequences formed Group 2 that assumed a more recent divergence from the monocotyledonous Group 4, while PPERSPT together with arabidopsis and other dicotyledonous sequences appear to evolve later. The classification of the PPERSPT protein in the same subfamily along with arabidopsis SPT and other SPT-like proteins from dicotyledonous species and the recognition of specific residues that differentiate this subfamily from the other subfamilies, including the ALC subfamily, providing further support that PPERSPT protein is a SPT-like protein rather than an ALC one. The presence of the acidic domain in the PPERSPT protein sequence and the phylogenetic analysis support this suggestion. In a recent paper by [43], qPCR reactions were performed for a number of genes involved in lignin and flavonoid biosynthesis, as well as other genes controlling stone formation in peach, including a gene designated as ALC-like from peach, during fruit development. The sequence of this putative ALC-like designated gene was not given and is not deposited in the gene sequence bank. Nevertheless, the ALC primers that the authors used to study this ALC-like designated gene matched our PPERSPT cDNA sequence 100%. This fact combined with our extensive bioinformatics analysis which gave no ALC-like hits in peach, suggest that the aforementioned ALC-like gene actually must be our PPERSPT gene. Thus taking into consideration the above data and the fact that no ALC genes were found outside the Brassicaceae family, the gene should be described as PPERSPT. In arabidopsis, many of the tissues in which SPT is expressed are actively growing tissues. Within the flower, SPT expression is present in proliferating cells within ovule primordia, in the lengthening funiculus as well as in elongating cells of the integuments. It is present in developing petals throughout their growth. Stamen primordia also, express SPT, as do growing sub-regions of the maturing anther, nevertheless SPT transcripts are apparently absent in sepals [39]. Based on results from [13] SPT promotes growth of carpel margins and the differentiation of specialised tissues from them. In the maturing silique, SPT expression becomes restricted to the edges of the valves in a pattern that resembles that of the SHP MADS genes; however valve dehiscence is not affected in spt mutants [39]. On the other hand, ALC is primarily involved in separation layer formation at the valve margins. In alc mutants, at the inner valve margin, the valves and the replum are held together by a lignified bridge preventing silique dehiscence which appeared due to the absence of a layer of nonlignified cells found in wild-type siliques between the lignified cells of the dehiscence zone and the cells of the replum. ALC is also expressed in other organs and developmental stages such as ovules at various stages of development, nectaries, the fruit pedicel branching point, and newly emerging leaves [12]. Expression analysis by both RT-PCR and in situ hybridization reveals that peach PPERSPT expression patterns resemble both ALC and SPT. RT-PCR experiments showed that it was present in all tissues examined (leaves, sepals, petals, stamens and carpels), although the localization of the expression of PPERSPT in fully developed buds by in situ hybridization showed that expression of PPERSPT is strong at the ovary and the perianth, whereas its transcription is reduced to a weak signal in stamens (Figs. 4 and 5A1). Moreover, in the cross-section of 5 weeks-old and 8 weeksold developing fruits (Fig. 5B1, D1) signal is detected along the endocarp suture of the two halves as well as along the union of the halves of the peach pericarp (Fig. 5C1). In the longitudinal sections of 8 weeks-old fruit, detection of PPERSPT transcript is most prominent in seeds (Fig. 5C5, D2-3). Interestingly, PPERSPT relative expression ratio was high both at 5 (just before the initiation of pit hardening) as well as at 8 weeks-old developing fruits (intermediate phase of pit hardening). Moreover at 5 weeks after full anthesis (approximately 35e37 days after full bloom) lignin deposition in peach endocarp was firstly detected [43]. Finally, as shown in Fig. 5C6, PPERSPT expression is localized on the upper part of the suture of peach endocarp close to the pedicel, where splitting of the peach fruit often occurs [4,44]. As previously described peach fruit undergoes four main stages of growth [6]. The first, rather rapid growth stage (stage I) is marked by cell division up to 7 weeks after full anthesis. This is followed by a slower fruit growth stage (stage II) where most of the dry matter is employed in pit hardening and seed and embryo growth (7e11 weeks after anthesis). The third stage (stage III) is characterized by more rapid growth where cell enlargement and elongation take place (12e15 weeks after anthesis). The last stage (stage IV) is the ripening phase (15e18 weeks after anthesis) characterized by a slight increase in fruit diameter [45]. During pit hardening and lignification, in clingstone varieties, the flesh is adhering to the endocarp. In the next stage of fruit growth, the outer surface of the endocarp becomes very furrowed and pitted while the flesh is still tightly attached to it. During the final stage of fruit growth, the flesh creates forces pulling out on the endocarp which will cause the pit to break in the weakest spot, which is located along the suture. In our previous study we presented results indicating that PPERFUL expression during the intermediate phase of pit hardening (8 weeks after full anthesis) was significantly lower in the susceptible variety (‘Andross’) compared to the resistant one, while PPERSHP expression was almost identical comparing the two varieties. In the present study we demonstrated that PPERSPT, a gene that possibly acts downstream of PPERSHP and is involved in endocarp margins development, is up-regulated in the susceptible variety compared to the resistant one just before the initiation of pit hardening (week 5 after full anthesis) until the intermediate phase of pit hardening (weeks 8 after full anthesis) (Fig. 6). [43] demonstrated that the ALC-like gene did not show considerable expression changes between 30 and 60 days after full bloom. Indeed, during this period (4e8 weeks after full anthesis) relative expression ratio of PPERSPT is constantly high whereas PPERSPT relative expression is low at the first stage of fruit growth as well as during the last stage that is pit hardening (Fig. 6). Finally, identification of PPERSPT gene adds one more component to the network of genes that control endocarp margins development in peach. Due to the fact that the arabidopsis SPT and ALC are apparently the products of recent gene duplication in an ancestor of the Brassicaceae family and no ALC-like gene has been isolated from the Rosaceae family, we can speculate that no such duplication took place in the peach genome. And most probably PPERSPT gene acts like an ALC-like gene too, in various aspects of peach fruit development, fulfilling both roles. Acknowledgments We thank Konstantinos Pasentsis for his invaluable help with RCA-RACE; the Institute of Pomology of the National Agricultural Research Foundation (NAGREF) and specifically Dr. Thomas Thomidis, and Venus Corporation for providing plant material. We also E. Tani et al. / Plant Physiology and Biochemistry 49 (2011) 654e663 acknowledge Ms. Laura Dadurian for critical reading the manuscript. Tsaballa A. holds a PhD scholarship from the “Alexander S. Onassis” Public Benefit Foundation This work was financially supported by the General Secretariat for Research and Technology (GSRT) of Greece. References [1] J.J. Giovannoni, Fruit ripening mutants yield insights into ripening control, Current Opinion in Plant Biology 10 (2007) 283e289. [2] M. Itkin, H. Seybold, D. Breitel, I. Rogachev, S. Meir, A. Aharoni, TOMATO AGAMOUS-LIKE 1 is a component of the fruit ripening regulatory network, Plant Journal 60 (2009) 1081e1095. [3] J. Vrebalov, I.L. Pan, A.J.M. Arroyo, R. McQuinn, M. Chung, M. Poole, J.K.C. Rose, G. Seymour, S. Grandillo, J. Giovannoni, V.F. Irish, Fleshy fruit expansion and ripening are regulated by the tomato SHATTERPROOF gene TAGL1, Plant Cell 21 (2009) 3041e3062. [4] C. O’Malley, J.T.A. Proctor, Split pits in Canadian peaches, Journal of the American Pomological Society 56 (2002) 72e75. [5] E. Tani, A.N. Polidoros, E. Flemetakis, C. Stedel, C. Kalloniati, K. Demetriou, P. Katinakis, A.S. Tsaftaris, Characterization and expression analysis of AGAMOUS-like, SEEDSTICK-like, and SEPALLATA-like MADS-box genes in peach (Prunus persica) fruit, Plant Physiology and Biochemistry 47 (2009) 690e700. [6] E. Tani, A.N. Polidoros, A.S. Tsaftaris, Characterization and expression analysis of FRUITFULL- and SHATTERPROOF-like genes from peach (Prunus persica) and their role in split-pit formation, Tree Physiology 27 (2007) 649e659. [7] H. Alonso-Cantabrana, J.J. Ripoll, I. Ochando, A. Vera, C. Ferrandiz, A. MartinezLaborda, Common regulatory networks in leaf and fruit patterning revealed by mutations in the Arabidopsis ASYMMETRIC LEAVES1 gene, Development 134 (2007) 2663e2671. [8] J.R. Dinneny, D. Weigel, M.F. Yanofsky, A genetic framework for fruit patterning in Arabidopsis thaliana, Development 132 (2005) 4687e4696. [9] S.J. Liljegren, A.H.K. Roeder, S.A. Kempin, K. Gremski, L. Ostergaard, S. Guimil, D.K. Reyes, M.F. Yanofsky, Control of fruit patterning in Arabidopsis by INDEHISCENT, Cell 116 (2004) 843e853. [10] Q. Gu, C. Ferrandiz, M.F. Yanofsky, R. Martienssen, The FRUITFULL MADS-box gene mediates cell differentiation during Arabidopsis fruit development, Development 125 (1998) 1509e1517. [11] A.H.K. Roeder, C. Ferrandiz, M.F. Yanofsky, The role of the REPLUMLESS homeodomain protein in patterning the Arabidopsis fruit, Current Biology 13 (2003) 1630e1635. [12] S. Rajani, V. Sundaresan, The Arabidopsis myc/bHLH gene ALCATRAZ enables cell separation in fruit dehiscence, Current Biology 11 (2001) 1914e1922. [13] J. Alvarez, D.R. Smyth, CRABS CLAW and SPATULA, two Arabidopsis genes that control carpel development in parallel with AGAMOUS, Development 126 (1999) 2377e2386. [14] J. Alvarez, D.R. Smyth, Crabs Claw and Spatula genes regulate growth and pattern formation during gynoecium development in Arabidopsis thaliana, International Journal of Plant Sciences 163 (2002) 17e41. [15] M. Groszmann, T. Paicu, D.R. Smyth, Functional domains of SPATULA, a bHLH transcription factor involved in carpel and fruit development in Arabidopsis, Plant Journal 55 (2008) 40e52. [16] C. Ferrandiz, S.J. Liljegren, M.F. Yanofsky, Negative regulation of the SHATTERPROOF genes by FRUITFULL during Arabidopsis fruit development, Science 289 (2000) 436e438. [17] C.A. Flanagan, Y. Hu, H. Ma, Specific expression of the AGL1 MADS-box gene suggests regulatory functions in Arabidopsis gynoecium and ovule development, Plant Journal 10 (1996) 343e353. [18] B. Savidge, S.D. Rounsley, M.F. Yanofsky, Temporal relationship between the transcription of 2 Arabidopsis MADS box genes and the floral organ identity genes, Plant Cell 7 (1995) 721e733. [19] A.N. Polidoros, K. Pasentsis, A.S. Tsaftaris, Rolling circle amplification-RACE: a method for simultaneous isolation of 50 and 30 cDNA ends from amplified cDNA templates, Biotechniques 41 (2006) 35. [20] M.A. Heim, M. Jakoby, M. Werber, C. Martin, B. Weisshaar, P.C. Bailey, The basic helix-loop-helix transcription factor family in plants: a genome-wide study of protein structure and functional diversity, Molecular Biology and Evolution 20 (2003) 735e747. 663 [21] K. Katoh, K.-i. Kuma, H. Toh, T. Miyata, MAFFT version 5: improvement in accuracy of multiple sequence alignment, Nucleic Acids Research 33 (2005) 511e518. [22] A.M. Waterhouse, J.B. Procter, D.M.A. Martin, M.l. Clamp, G.J. Barton, Jalview version 2 -a multiple sequence alignment editor and analysis workbench, Bioinformatics 25 (2009) 1189e1191. [23] D.P. Brown, N. Krishnamurthy, K. Sjölander, Automated protein subfamily identification and classification, PLoS Computational Biology 3 (2007) e160. [24] K. Ye, K. Anton Feenstra, J. Heringa, A.P. IJzerman, E. Marchiori, Multi-RELIEF: a method to recognize specificity determining residues from multiple sequence alignments using a machine-learning approach for feature weighting, Bioinformatics 24 (2008) 18e25. [25] P. Pérez-Rodríguez, D.M. Riaño-Pachón, L.G.G. Corrêa, S.A. Rensing, B. Kersten, B. Mueller-Roeber, PlnTFDB: updated content and new features of the plant transcription factor database, Nucleic Acids Research 38 (2009) D822eD827. [26] N. Pires, L. Dolan, Origin and diversification of basic-helix-loop-helix proteins in plants, Molecular Biology and Evolution 27 (2010) 862e874. [27] T. Keane, C. Creevey, M. Pentony, T. Naughton, J. Mclnerney, Assessment of methods for amino acid matrix selection and their use on empirical data shows that ad hoc assumptions for choice of matrix are not justified, BMC Evolutionary Biology 6 (2006) 29. [28] Felsenstein J., Phylip (Phylogeny inference package) version 3.6, Distributed by the author, Department of Genome Sciences, University of Washington, Seattle, 2004. [29] N. Saitou, M. Nei, The Neighbor-joining method e a new method for reconstructing phylogenetic trees, Molecular Biology and Evolution 4 (1987) 406e425. [30] D. Jones, W. Taylor, J. Thornton, The rapid generation of mutation data matrices from protein sequences, Comput Appl Biosci 8 (1992) 275e282. [31] R.V. Eck, M.O. Dayhoff, Atlas of Protein Sequence and Structure, NationalBiomedical Research Foundation. Maryland, Silver Spring, 1966. [32] L. Le Cam, Maximum likelihood e an introduction, ISI Review 58 (1990) 153e171. [33] F. Ronquist, J.P. Huelsenbeck, MrBayes 3: Bayesian phylogenetic inference under mixed models, Bioinformatics 19 (2003) 1572e1574. [34] T.B. Morrison, J.J. Weis, C.T. Wittwer, Quantification of low-copy transcripts by continuous SYBR (R) green I monitoring during amplification, Biotechniques 24 (1998) 954-þ. [35] K.J. Livak, T.D. Schmittgen, Analysis of relative gene expression data using real-time quantitative PCR and the 2(T)(-Delta Delta C) method, Methods 25 (2001) 402e408. [36] E. Flemetakis, R.C. Efrose, G. Desbrosses, M. Dimou, C. Delis, G. Aivalakis, M.K. Udvardi, P. Katinakis, Induction and spatial organization of polyamine biosynthesis during nodule development in Lotus japonicus, Molecular PlantMicrobe Interactions 17 (2004) 1283e1293. [37] O. Gandrillon, F. Solari, C. Legrand, P. Jurdic, J. Samarut, A rapid and convenient method to prepare DIG-labelled RNA probes for use in non-radioactive in situ hybridization, Molecular and Cellular Probes 10 (1996) 51e55. [38] G. Toledo-Ortiz, E. Huq, P.H. Quail, The Arabidopsis basic/helix-loop-helix transcription factor family, Plant Cell 15 (2003) 1749e1770. [39] M.G.B. Heisler, A. Atkinson, Y.H. Bylstra, R. Walsh, D.R. Smyth, SPATULA, a gene that controls development of carpel margin tissues in Arabidopsis, encodes a bHLH protein, Development 128 (2001) 1089e1098. [40] S.J. Liljegren, G.S. Ditta, H.Y. Eshed, B. Savidge, J.L. Bowman, M.F. Yanofsky, SHATTERPROOF MADS-box genes control seed dispersal in Arabidopsis, Nature 404 (2000) 766e770. [41] K. Gremski, G. Ditta, M.F. Yanofsky, The HECATE genes regulate female reproductive tract development in Arabidopsis thaliana, Development 134 (2007) 3593e3601. [42] L. Ostergaard, Don’t ’leaf’ now. The making of a fruit, Current Opinion in Plant Biology 12 (2009) 36e41. [43] C. Dardick, A. Callahan, R. Chiozzotto, R. Schaffer, M.C. Piagnani, R. Scorza, Stone formation in peach fruit exhibits spatial coordination of the lignin and flavonoid pathways and similarity to Arabidopsis dehiscence, BMC Biology 8 (2010) 13. [44] M. Nakano, M. Nakamura, Cracking and mechanical properties of the stone in peach cultivars after severe thinning, Acta Horticulturae 592 (2002) 531e536. [45] J. Gage, G. Stutte, Developmental indexes of peach e an anatomical framework, Hortscience 26 (1991) 459e463.