Introduction

Plant knotted-like genes (KNOX) encode homeodomain (HD) containing transcription factors (TFs) and are pivotal in regulating organ development (reviewed by Hake et al. 2004). They belong, together with BEL-like proteins, to the three amino acid loop extension protein cluster (Burglin 1997). KNOX were split into two classes on the basis of identity grade within the HD, intron position and expression patterns (Reiser et al. 2000). The third class, named M, was set after discovering KNOX lacking the HD (Kimura et al. 2008; Magnani and Hake 2008).

Class 1 KNOX are required for proper development of the shoot apical meristem (SAM) that act to maintain cells in an indeterminate state and are expressed in meristem cells (reviewed by Scofield and Murray 2006). Their functions have been assessed in A. thaliana during the diverse processes of vegetative and reproductive development (Hake et al. 2004; Scofield and Murray 2006; Truernit et al. 2006). The class 1 KNOX functions are mediated by interactions with cytokinins, auxin and gibberellins (reviewed by Hay et al. 2004), specific and combinatorial dimerization with BELL (synopsis by Hackbusch et al. 2005), mechanisms of gene hierarchy (Li et al. 2005; Guo et al. 2008).

Class 2 KNOX share HD identity ≤55% compared to that of class 1, are ubiquitously expressed, and typically contain an intron in the ELK domain vicinity (Reiser et al. 2000). Referring to dicot species, they have been characterised in Arabidopsis (http://arabidopsis.org/), tomato (Janssen et al. 1998), tobacco (Sentoku et al. 1998), barrel medic (Di Giacomo et al. 2008), pea (Hofer et al. 2001) and apple (Watillon et al. 1997). Organ wide-spread expression was reported for all these plants, though tissue-specific transcriptional preferences and alternative splicing events occur in monocots (Kerstetter et al. 1994; Morere-Le Paven et al. 2007; Tamaoki et al. 1995). The Arabidopsis KNAT3 has been the most deeply investigated, for which expression patterns were profiled in aerial organs (Serikawa et al. 1997) and roots, together with those of class 2 KNAT4 and 5 (Truernit et al. 2006). Phytohormones, such as cytokinin (Soucek et al. 2007; Di Giacomo et al. 2008; Truernit et al. 2006), abscissic acid and gibberellins (Morere-Le Paven et al. 2007) affect class 2 KNOX transcription in distinct organs. Light-mediated regulation of KNAT3 expression was observed in seedling as its promoter differentially responded to white, red and far-red spectra (Serikawa et al. 1996; Serikawa et al. 1997), whilst Knox7 was light-triggered in maize embryos (Morere-Le Paven et al. 2007). Class 2 proteins form heterodimers with KNOX and BEL members, for instance: KNAT3 binds to KNAT4 and nine BEL factors, whilst KNAT4 only binds to class 1 BP and BEL6 (Hackbusch et al. 2005). The sub-cellular localization of KNAT3 and 4 is regulated at the post-translational level (Truernit et al. 2006). The tobacco class 2 NTH201 was shown to localize at plasmodesmata (Yoshii et al. 2008) and is unable to self-traffic between cells, resembling the behaviour of KNAT3 (Kim et al. 2005).

The diversified expression patterns in many tissues at different times have made it difficult to propose single functions for class 2 KNOX. Moreover, role specificity is a result of spatio-temporal regulation of KNOX dimerization with other proteins. The functional redundancy amongst class 2 genes was evoked to explain the absence of phenotypes in silenced mutants (Serikawa et al. 1997); however, plastochron delay, leaf lamina shrink and petiole elongation were described in Arabidospsis knat3 null mutants (Chandler and Wolfgang 2004). The KNAT3 knock-out was found to reverse the eostre mutant and proposed to underlie both mega-gametogenesis and embryo sac development (Pagnussat et al. 2007), whilst KNAT7 was alleged to participate in secondary cell wall synthesis (Zhong et al. 2007). The disruption of a Physcomitrella class 2 member caused spore aberrancy (Singer and Ashton 2007), whilst that of tobacco NTH201 hampered both the accumulation of movement protein and the formation of replication complexes of the tobacco mosaic virus (Yoshii et al. 2008).

KNOX are proposed to play crucial roles in the control of tree agronomic traits (Groover et al. 2006; Testone et al. 2008), and to date the knowledge on class 2 members has been quite modest. Consequently, the peach KNOPE3 gene (phylogenetic ortholog of KNAT3) was cloned and mapped genetically on the Prunus reference map using cleavage amplified polymorphic sequence (CAPS) markers. Distinct mRNA localization patterns were observed in stem, flower buds, drupes and leaves, though the expression in sieves was common to all the organs, suggesting that KNOPE3 may participate in transport mechanisms of elaborated sap compounds. Generally, sugars are nutritive and signalling molecules able to control gene expression and plant development (Rolland et al. 2006); source-to-sink translocation of carbohydrates is essential for peach bud break and drupe growth (Maurel et al. 2004; Morandi et al. 2007). Consequently, the KNOPE3 expression was monitored in experiments causing sugar content variation in leaves. The gene was rapidly induced in sink leaves directly up-taking fructose, glucose and sucrose; it was also triggered in source leaves concomitantly with sucrose and sorbitol increase caused by transport block after stem girdling. Based on these results, a possible role of KNOPE3 in sugar translocation is proposed and discussed.

Materials and methods

Plant materials and growth conditions

Adult peach plants (Prunus persica cultivar ‘Chiripa’) grown in the IBBA-CNR fields belonged to two clonal lines (S1 and S2) obtained by micro-propagation of two seedlings derived from the open pollinated mother plant16 (‘Chiripa’). The clone 18 of the S1 line (named F0S1–18) was self-pollinated under controlled conditions and seeds were grown to generate F1S1–18 individuals (Testone et al. 2008), which provided most of the materials at distinct developmental stages for this work. Plants were grown in the greenhouse at 22–25°C, 16/8 h of light/dark with a light intensity of 100 μmol m−2 s−1 of photosynthetically active radiation (PAR). Several clones of KNOPE3 cDNA and genomic fragments were sequenced and we did not score any nucleotide differences amongst F1S1–18 sister plants, suggesting that KNOPE3 was tightly conserved (at least in the transcribed regions) and gene regulation mechanisms may be conserved.

Leaf developmental stages

One-year-old F1S1–18 plants were used during (a) vegetative resumption (March) for in situ hybridizations (see following paragraph) and (b) early shoot elongation (April) for gene expression analyses. In April, the shoots prolonging the main axes (trunks) were 28.3 ± 2.2 cm long and bore 8.8 ± 1.2 leaves. To mark leaf growth stages (Table 1), the leaf position on stem (from the apex to the basis), the expansion status and area of laminas and the chlorophyll content were determined. The apical tips (AT) included 1–4 leaflets; the leaves below (mainly the 5th leaf) were folded (F) or fully expanded (E). The same leaf types (n = 3) from distinct plants (n = 3) were pooled for RNA isolation and RT-PCR experiments, which were repeated for three distinct sampling. Lamina surfaces and chlorophyll content were measured, respectively, by ImageJ software (http://rsb.info.nih.gov/ij) and as described (Porra et al. 1989).

Table 1 Evolution of lamina area, chlorophyll (a + b) content, and relative expression of RUBCA and KNOPE3 genes during leaf development of 1-year-old peach plants after vegetative resumption (April)
Full size table

Isolation and sequence analysis of cDNA and genomic clones

The cDNA was synthesized from leaf RNA (see sub-paragraph of RT-PCR) of F1S1–18 plants and PCR-amplified by primers Kn3.1Fw (5′-CCATGGAAGCAGTGATGGC-3′) and Kn3.2Bw (5′-CTTCCTCAGTTGGATATGGCC-3′). These were designed based on the class 2 KNOX conserved amino acidic stretches AMEAVMA (in the KNOX2) and WPYPTEED (in the HD), respectively (see Fig. 1). The 390 base pairs (bp) fragment was cloned, sequenced and confirmed to share high identity with class 2 KNOX. The KNOPE3 full length cDNA was achieved into steps: (a) the 3′ region (PCR product of 1101 bp) was cloned by 3′RACE methodology (INVITROGEN), using the anchor primers provided by the manufacturer and the Kn3.1Fw primer; (b) the 5′ region was isolated using primers retrieved from KNAP3 of M. domestica (acc. n. Z71980), they were Kn3.3Fw (5′-GTATTCATGGCGTACCACAACC-3′) and Kn3.4Bw (5′-TAGCACCCGTGCCTTCACCGG-3′). PCR amplification produced a single fragment of 880 bp.

Fig. 1
figure 1

Features of KNOPE3 protein. a Alignment of KNOPE3 deduced amino acid sequence (bolded) with Malus domestica MdKN3 and Vitis vinifera KNAT3-like gene (VvCA067726). Gaps introduced for better alignment are shown by dashes, asterisks indicate strictly maintained residues. The typical KNOX1, KNOX2, ELK and the HD are indicated. b KNOPE3 (bolded) within the phylogenetic tree of class 2 KNOX proteins. Class 1 proteins were used to create an out-group. Bootstraps values (at the branching points) are given for major nodes and based on 1,000 replicates. Orthologous members of each subgroup are bordered and named referring to the respective Arabidopsis class 2 KNOX

Full size image

Introns were found by amplifying leaf genomic DNA (gDNA) with the following primer couples: Kn3.3Fw/Kn3.10Bw (5′-TTGTGCCTTGCGTTCTGCC-3′); Kn3.5Fw (5′-GCCCAGCTGGCTCAGTCGC-3′)/Kn3.2Bw and Kn3.1Fw/Kn3.6Bw (5′-GCAGCCTTTGAGCATACCTG-3′). The sequences of products from cDNA and gDNA were aligned by ClustalW (http://www.ebi.ac.uk.clustalw) and intron size and locations established.

The final PCR conditions were: 500 ng of gDNA or 200 ng of cDNA, 1 mM of each primer, 0.5 mM dNTPs, Taq DNA polymerase (TaqQUIA, Qiagen) 2.5 U, 1/10 of 10× Taq Buffer (Qiagen), 2.5 mM MgCl2, in a final volume of 50 μl. Thermocycle conditions included an initial cycle of 95°C for 5 min followed by 35 cycles of 95°C for 40 s, either 55°C (in cDNA based and RACE-PCR experiments) or 60°C (with gDNA) for 30–60 s and 72°C for 30–90 s, final extension at 72°C for 5 min. All PCR fragments were cloned into pGEM-T easy vector system (PROMEGA). The GenBank accession numbers of KNOPE3 cDNA and genomic sequences are DQ786755 and EU910092.

Alignments and phylogenetic analysis

The alignment of KNOPE3 with other class 2 KNOX proteins was performed by ClustalW (http://www.ebi.ac.uk.clustalw) and optimized by visual inspection (PILEUP program). Phylogenetic trees were constructed by MegaBlast2 (based on the minimum evolution criterion), using bootstrap values performed on 1,000 replicates and the 50% value was accepted as an indicative of a well-supported branch. The class 2 KNOX accession numbers were: PpKNOPE3 DQ786755; MtKNOX3 ABO33480.1; MtKNOX4 ABO33481.1; MtKNOX5 ABO33482.1; MdKNAP3 Z71980; NtH23 BAA25921; NtH201 AB333840; Osh45 BAA08552; LeT12 AAC49918; AtKNAT3 X92392; AtKNAT4 NP196667; AtKNAT5 NP194932; AtKNAT7 AF308451; VvCAO67726; VvCAO63645; VvCAO44300; ZmNP_001105852; PtABK92551; OsBAB55660; NtBAF95776. A few class 1 KNOX were used to create an out-group; acc. n.: AtKNAT1, U14174; AtSTM, U32344; AtKNAT2, U14175; AtKNAT6L, AB072362; AtKNAT6S, AB072361; PpKNOPE1 ABD52723.

Southern blot analysis

Extraction of gDNA and Southern blots were performed as previously described (Giannino et al. 2000). In brief, leaf gDNA (30 μg) was digested, fragments were size-fractionated on 0.8% agarose gel and transferred onto nylon membrane (Hybond N+, Amersham) according to Kingston et al. (2001). The probe (25 ng) was radio-labelled with to 5 μl of 32dCTP in 45 μl of TE buffer (10 mM Tris–HCl pH 8.0, 1 mM EDTA) and denatured in boiling water according to Rediprime™ II (Amersham) kit instructions. Filters were hybridized at 55°C (cDNA probe 1) and 60°C (genomic probe 2) overnight using a modified Church and Gilbert (1984) buffer (NaPi 0.5 M pH 7.2, SDS 5%, EDTA 10 mM), washed twice (2× and 1× SSC/0.1% SDS) at 55°C (cDNA probe 1) and 60°C (genomic probe 2) for 10 min and exposed to Biomax films (Kodak) for 4–12 h at −80°C. The cDNA probe 1 and the genomic probe 2 included the region between the primers Kn3.7Fw (5′-GTTTCCCCCGGTGAAGGCAC-3′)-Kn3.6Bw and were 735 and 1142 bp long, respectively. Hind III restriction enzyme cut twice (2108–2113 and 2139–2144, acc n. EU910092) in the exons of genomic probe.

Mapping

The F2 interspecific mapping population of the Prunus reference map (Joobeur et al. 1998; Dirlewanger et al. 2004) was used in this study. It was obtained by selfing a single F1 plant from the cross between the almond [Prunus dulcis (Mill) D. A. Webb; syn. P. amygdalus Batsch] cv ‘Texas’ and the peach cv. ‘Earlygold’ (abbreviation T × E). The tree population is maintained at CRA, Centro di Ricerca per la Frutticoltura, Ciampino, Rome. Leaf gDNA was isolated using DNeasy Plant Mini Kit (Qiagen). Primers derived from ‘Chiripa’ KNOPE3: Kn3.5Fw (see above) was in the exonic region, whilst Kn3I2.1Bw (5′-AACTTCGGTGCTCAAGAGCC-3′) fell in intron II. They were used to amplify gDNA from the parents, F1 and F2 progeny of the T × E population. The PCR conditions were: 30 ng of gDNA, 0.4 μM of each primer, 0.5 mM dNTPs, 1.25 U of high fidelity Platinum Taq (Invitrogen) for the parental genotypes, sequencing and Taq DNA polymerase (Qiagen) for segregation analyses, 1/10 of 10× Taq buffer, 2.5 mM MgCl2, in a final volume of 25 μl. Thermocycler conditions included an initial cycle at 95°C for 5 min followed by 35 cycles at 95°C for 30 s, 62°C for 60 s and 72°C for 90 s with a final extension at 72°C for 7 min. The sequences of KNOPE3 PCR fragments from E and T genotypes are reported in the supplemental data 1. The polymorphic profile generated by HincII digestion is described in “Results”. The PCR products were digested with 5 HincII units (New England Biolabs) in a final volume of 30 and 15 μl was electrophoresed onto 1.2% agarose gel. KNOPE3 CAPS was integrated in T × E map (Joobeur et al. 1998; Dirlewanger et al. 2004) using the software MAPMAKER (Lander et al. 1987; Lincoln et al. 1992). Initially, the “group” command assigned the locus to linkage groups (LOD threshold >7 and recombination fraction <10). The locus was then placed within the respective T × E linkage group using the “try” and “ripple” commands. After mapping, the “error detection” command was used and no candidate errors were detected for KNOPE3. Kosambi’s mapping function (Kosambi 1944) was used to convert recombination fractions into centimorgan (cM) map distances.

Semi-quantitative RT-PCR

Total RNA was extracted from various organ tissues (Giannino et al. 2000), DNAse-treated (RQ1, Promega), quantified by the spectrophotometer (NanoDrop technologies) and 3 μg was reverse transcribed at 55°C into a single strand cDNA by oligo-(dT)20 by Superscript III reverse transcriptase (Invitrogen), according to the manufacturer’s instructions. Table 2 contains details on the genes, primers and PCR conditions used in this work. PCR trials were performed at distinct cycles to assess the variation of transcript abundance before signal saturation, at fixed primer pair and reaction parameters. The Kn3.9fw and Kn3.8bw primers were designed in the 3′untranslated region (3′UTR) and yielded a 424 bp fragment. The expression of peach constitutive gene ACTIN 2 (same copy number per cell and stable expression in every cell, An et al. 1996) was monitored to check for equal transcript abundance. The PCR reaction was conducted in 50 μl total volume containing 100 ng of cDNA; the chemical parameters are explained in “Materials and methods”; 15 μl was electrophoresed in a 1% agarose gel.

Table 2 Genes, primers, PCR conditions and products in expression assays
Full size table

Real time RT-PCR (qRT-PCR)

All the qRT-PCR experiments included three repetitions of PCR reactions (technical triplicates) and each PCR sample consisted of cDNA retrieved from RNA of a leaf pool (n = 2–4 leaves, see details along the text). Total RNA was isolated from leaf pools (100 mg) by RNeasy Plant Mini Kit (Qiagen) and reverse-transcribed as described above, the cDNA (40 ng) was amplified in qRT-PCR reactions in 20 μl total volume containing: PCR reaction buffer 1×; dNTPs 50 μM; MgCl2 3.5 mM; d-(+)-Trehalose dihydrate (Fluka) 0.6 M; formamide 0.7 M; 0.1 μl of a 1:100 Sybr Green I dilution in dimethylsulfoxide (Invitrogen); Taq DNA polymerase (Qiagen) 0.2 U; 0.3 μM of each primer. Reactions were performed in triplicate at these conditions: 180 s at 95°C, 45 cycles of three-step amplification (30 s at 95°C, 1 min at 58°C and 1 min at 72°C). To test primer specificity, melting curve analysis (from 55 to 95°C with an increasing heat rate of 0.5°C s−1) was performed after amplifications. The threshold cycle (C t) value of each gene was normalized with that of ACTIN2 and compared with the C t of an untreated control using the \( 2^{{ - \Updelta \Updelta C_{\text{t}} }} \) method (Livak and Schmittgen 2001). In assays for leaf development, the KNOPE3 and RIBULOSE CARBOXYLASE ACTIVASE (RUBCA) expressions were, respectively, calibrated referring to the apical tip values, which were set as 1. The primers used for qRT-PCR are listed in Table 2.

In situ hybridization

One-year-old F1S1–18 plants were used at vegetative resumption (March 2006). Sections were made from growing shoots (n = 5) derived from the central axis. Excised tissues were fixed, dehydrated, embedded in paraffin, cut into 8 μM sections and hybridized (53°C) with the digoxygenin-labelled antisense RNA probe as described by Cañas et al. (1994). Probe 3 spanned the 89–536 bp cDNA stretch between primers Kn3.11Fw (5′-TTCAATCCGACCAACACGACC-3′) and Kn3.10Bw, and linearized by SpeI and NcoI. The anti-sense and sense RNA probes were in vitro synthesized by T7 and SP6 polymerases, respectively (Giannino et al. 2000).

Assays of imposed dark

Dark response was carried out in vitro using F1S1–18 seedlings. As for in vitro growth, the woody endocarp was removed mechanically by a nut cracker; seeds were sterilized in 70% ethanol for 1 min, followed by HClO 0.75% for 30 min, six washes in sterile water and a final immersion in sterile water for 12 h, and then dried with blotting paper under laminar flow cabinet. The brown coat was peeled gently by scalpel; seeds were germinated on MS0 (Murashige and Skoog 1962) supplied with 5 g l−1 agar (Micro Agar, DUCHEFA), 30 g l−1 sucrose to a final pH 5.7 in glass pots under irradiance of 80 μmol s−1 cm−2 with a photoperiod of 16/8 h light/dark cycle at 22°C. The germination occurred in ca. 4 days and seed vitality was ca. 55%; the seedlings were transferred on fresh agarised-MS0 11 days after sowing (das).

Dark was imposed on 15 das seedlings, which formed primary and secondary roots and bore 4–5 fully expanded leaves. The pots were covered with aluminium foil. The expanded leaves (n = 4) were excised in the dark and pooled. A separate plantlet was used for each different sampling time (0.5, 1, 2, 4, 8, 16 h). The dark period lasted 16 h so that treated samples could be compared to controls, which were under the normal range of 16/8 h light/dark. A longer dark time was not approached so as to minimize the interference of circadian effects. The experiment was repeated twice with a new set of seedlings (exceptionally, biological duplicates were performed). The expression of peach RUBCA and CATALASE2 (CAT2) genes were assayed as dark/light and stress responsive markers, respectively (Table 2 and supplemental data to Fig. 7). The diagrams of relative mRNA abundance represent actin-normalized values of treated samples with respect to those of untreated controls. Standard deviation for ratios was calculated and represented by bars.

Sugar assays

Sugar feeding was performed via the transpiration stream using the leaflets surrounding the shoot apex (AT, in Table 1) borne on the central axis of 1-year-old F1S1–18 plants (April 2006). Leaflets (n = 4) were excised at the junction between the petiole and subtending stem, and immersed into 2.2 ml tubes containing 1.8 ml of a 50 mM solution (separately: fructose, glucose, sucrose, sorbitol), maintained under irradiance of 50 μmol s−1 cm−2 to favour transpiration (Krapp et al. 1991), and sampled at 0, 3, 6 and 12 h. To check the occurrence of solution uptake, the overall volume of transpired water by the leaves (V T) in sorbitol feeding assays was measured by the formula: V Tn (%) = (V ET − V E) × 100 × (V 0)−1; V ET, evapo-transpired volume (μl) from tubes feeding the samples; V E, evaporated volume from tubes without samples; V 0, starting volume (1,800 μl); n, time of sampling. The V Tn (%) were: 0.00 ± 0.00, 2.37 ± 0.35, 5.23 ± 0.87, 9.27 ± 2.21 at 0, 3, 6 and 12 h post-treatment (hpt), respectively. The V Tn values of pure water and other sugar solutions (50 mM) did not vary significantly over time with respect to those for sorbitol, indicating that solution viscosity did not affect leaf uptake.

In April, 1-year F1S1–18 plants bore side branches with mature expanded leaves. The stem of the lateral branch closest to the soil was girdled (cortex and sieve removal). The bark ring was performed above and below the leaves laying at the branch basis, so as to hamper the transport towards the trunk and the upper sectors of the side branch. The leaves (n = 2) of the lateral branches of girdled and control plants were excised and pooled separately at 0 (10.00 a.m.), 6 (4.00 p.m.), and 30 (4.00 p.m.) hpt. Three repetitions of the experiments using distinct material (biological triplicates) were performed in sugar supply and girdling experiments and the relative gene expression was graphed by dots representing the mean values ± standard error.

Sugar quantification

The assay kits for sorbitol (cat. no. K-SORB) and sucrose, fructose and glucose and (K-SUFRG) were used (MEGAZYME International Ireland, Bray, Ireland), starting from 30 mg of leaf tissues and following the manufacturer’s instructions. The type and number of samples were the same as those used for the expression analyses of darkening and girdling assays. Sugar content fold change (Tables 3, 4; Figs. 7, 8) was graphed by dot representing the ratio between mean values from treated samples and those of untreated controls. Standard deviation for ratios was calculated and represented by bars. Student’s t test of treated versus control plants was performed and significant values (P < 0.05) were indicated by asterisks.

Table 3 Content of major transport sugars in expanded leaves of in vitro seedlings subjected to imposed darkening
Full size table
Table 4 Leaf sugar contents in girdling experiments
Full size table

Results

KNOPE3 belongs to the class 2 of KNOX

KNOPE3 full-length cDNA was 1894 bp long and encompassed an open reading frame of 1347 bp. The 5′UTR did not contain the typical Kozak and Shine-Dalgarno sequences, whereas the 3′UTR included a poly-adenylation tail (AATAAA, 1852-1857), a U-rich stretch (TTTTCT, 1811-1816) followed by a cleavage site (CA, 1823-1824) and terminal U-rich site (1825-1831), which are frequently located upstream the poly A tail of many plants (Graber et al. 1999).

KNOPE3 deduced protein consisted of 448 amino acids with a calculated weight of 50.05 kD; it contained the typical KNOX domains (Fig. 1a) and the highest overall identity was with Malus domestica KNAP3 (91%) followed by Medicago truncatula KNOX5 (77%), Arabidopsis thaliana KNAT3 and KNAT4 (69 and 70%), Vitis vinifera (CAO67726) and Populus thricocarpa (ABK92551) class 2 KNOX (64%). Referring to dicotyledonous trees, the identity grade of KNOPE3 C terminus (from the MEINOX domain to the protein end) was 98, 88 and 71%, respectively, compared to those of apple, western balsam poplar and grapevine class 2 KNOX, whilst the N terminus identity fell to 82, 23 and 16%. The KNOPE3 HD was 55% identical to that of class 1 Kn1 of maize and respected one of the three parameters telling class 1 from class 2 proteins (Reiser et al. 2000).

A phylogram (Fig. 1b) was constructed using various class 2 KNOX proteins from dicot and monocot species available in several databases. KNOPE3 was further confirmed to belong to the highly supported monophyletic class 2 proteins and to be closest to apple KNAP3.

Peach class 2 KNOX are a small family and KNOPE3 lays on linkage group 1

KNOPE3 harboured five introns (Fig. 2a, top scheme) maintaining the same positions as those of Arabidopsis KNAT3 (Fig. 2a, bottom scheme). Intron I and II fell in the KNOX2 sub-domain, intron III lay in the ELK domain, which is a feature that is specific to class 2 KNOX (Kerstetter et al. 1994). Intron IV exists in the HD and intron V was upstream of the stop codon. All the introns were AT rich (65% in average) and possessed GT/AG or GC/AG splice sites (Kitamura-Abe et al. 2004).

Fig. 2
figure 2

KNOPE3 genomic organization. a Scheme of peach KNOPE3 and Arabidopsis KNAT3 genes. The black triangles and numbers indicate intron positions and sizes (in base pairs, bp) with respect to the cDNA. Start and stop codons and polyadenylation tails are typed. The cDNA probe 1 spans the ELK-HD domains and a portion of the 3′UTR, whilst the genomic probe 2 includes introns III-V and 3′UTR. The cDNA probe 3 was used for in situ experiments. H, Hind III site. b Southern blot analysis. Molecular weights of a DNA marker are in kilo base pairs (kb) (left panel). Peach class 2 KNOX gene family: gDNA digested with EcoRV, XhoI, XbaI and hybridized with cDNA probe 1 (middle panel). Genomic probe 2 was used in combination with EcoRV, XbaI and HindIII, the latter cuts the probe twice (right panel). c KNOPE3 CAPS in T × E. Polymorphism profile of KNOPE3 CAPS in the parents (T, E), F1 and 9T × E F2 individuals (top panel). Map of linkage group 1 (G1) of the Prunus reference map and position of KNOPE3 (in bold) (bottom panel)

Full size image

Southern blot analysis (Fig. 2b) was first carried out with the KNOPE3 cDNA probe 1 (Fig. 2a) which encompassed the ELK to 3′UTR stretch and was remotely paralogous (11% nucleotide identity) to that of class 1 KNOPE1 (acc n. DQ358050). The signal pattern suggested that more than two members of class 2 occurred in the ‘Chiripa’ genome (Fig. 2b, middle panel). Subsequently, we used the genomic probe 2; spanning the stretch downstream of intron II to the 3′UTR and encompassed two Hind III sites in the exonic regions (H in Fig. 2a), lying 31 bp one from the other. A single band was observed (Fig. 2b, right panel) using EcoRV and XbaI, whilst two signals of ca. 2 kb and 450 bp appeared with Hind III. The ca. 450 bp size was consistent with that of 419 bp predicted by the restriction analysis (Fig. 2a), whilst the 31 bp band was not visible for it fell out of the electrophoretic run. Overall, this hybridization pattern indicated that KNOPE3 was a single copy per haploid genome.

KNOPE3 was mapped (Fig. 2c) scoring the segregation pattern of a CAPS marker in the Prunus T × E population (Joobeur et al. 1998; Dirlewanger et al. 2004). Primer pairs were designed based on the segment of ‘Chiripa’ KNOPE3 spanning intron I, and yielded 1113 and 1127 bp PCR products in E and T genotypes, respectively. They were sequenced and the differences consisted of 44 single nucleotide polymorphisms and ten insertion/deletion events (see “Material and methods” and supplemental data 1). The KNOPE3 alleles from E and T contained two and one HincII sites, respectively, at distinct positions. HincII digestion of PCR products generated a three band set of 529, 360, 224 bp in E, and a two band set of 903 and 2224 bp in T (Fig. 2c, top panel). The co-dominant polymorphic profile was confirmed in the F1, hence the allele segregation was determined in the F2 T × E population (Fig. 2c, top panel), processed by MAPMAKER EXP 3.0 and KNOPE3 was positioned at 66 cM of G1 (Fig. 2c, bottom panel), in co-segregation with the AC18 and AC23 markers.

KNOPE3 has a diversified pattern of expression in stem, flowers and fruits

KNOPE3 transcription was assayed in aerial vegetative and reproductive organs by semi-quantitative RT-PCR (Fig. 3) and in situ hybridization (Figs. 4, 5, 6) using the antisense dig-labelled probe 3 (Fig. 2a).

Fig. 3
figure 3

Detection of KNOPE3 expression by RT-PCR. Tissues are indicated above the lanes. The ACTIN2 was amplified to check for cDNA synthesis and equal loading. Transcript molecular sizes are in Table 2

Full size image
Fig. 4
figure 4

KNOPE3 transcript localization in stem The KNOPE3 mRNA is represented by the intense blue signal. a Sections at ca. 0.2 mm below the SAM. KNOPE3 mRNA was absent in the pith (p), spread in the cortex (cx) and visible in bundles of leaf traces (lt), marked by drawn circles. b Leaf trace magnification shows that the message localized to phloem and/or phloem-associated cells (p) and not to xylem (x). c Sections at ca. 5 cm from the apex. Leaf trace (lt-vsb) and stem vascular bundles (s-vsb) are circle-marked. The epidermis is also indicated (e). d Leaf trace magnification evidenced KNOPE3 signal in phloematic cells (ph) and not in the cambium (ca). e Close-up of stem vascular system showed an intense signal in phloematic cells, but not in the phloematic cap ( phc). f Control experiment, tissues were hybridized with a sense probe and no signal above background was detected (size bars a 100 μm, b 55 μm, c 115 μm, d, e 30 μm, f 80 μm)

Full size image
Fig. 5
figure 5

Transcript localization in flower and fruit. a–e Floral buds before anthesis (February) were used for the in situ experiments. a The KNOPE3 message occurred in the parenchymatous cells of receptacle (rc), and in the vascular bundles (vsb) of several flower components, such as anthers (a) filaments (fl), petals (pe) and sepals (se). b Magnification of a sepal (se) and a petal (pe) showed KNOPE3 mRNA stained sepals more intensely than petals. Intense signal occurred in vsb. c The message marked the anther endothecium (en) and tapetum region (tp), but was absent in perithecium (pr) and pollen grain (pg). d Deep blue signal marked vascular bundles (vsb) of anther filament (fl). e, f KNOPE3 was active in the vascular bundles (vsb) of stylus (st) and the stigma orifice (stg) and faintly in inner cell layers of the ovary (o). g Fruit at early development stage (S1, 3 mm of diameter). The transcript mainly localized in the vascular bundles (vsb) of mesocarp (msc); a slight signal was detected in funiculus tissue (fu) of seed (sd) and endocarp (enc) (size bars a 200 μm, b–d 40 μm, e, f 135 μm, g 500 μm)

Full size image
Fig. 6
figure 6

Transcript localization in vegetative SAM and leaves. a–d Shoots at the vegetative burst (late March) were analysed. a In SAM longitudinal section KNOPE3 message was absent in the central (CZ) and peripheral (PZ) zones, confined to the last cells of the rib zone (RZ), intense in provascular tissues (pv) underlying primordia (lp) and growing leaves (dl). b Shoot transversal section included growing leaves but not the apical meristem and primordia. Leaves were numbered arbitrarily 1–8 following the growth stage. Synoptically, the blue signal spread centripetally from inner (younger) towards outer (older) leaves. c Leaf 4 magnification pinpoints the mRNA in vascular bundles (vsb). d Close-up of leaf 8 shows the mRNA signal spread in the mesophyll (m), persisted in bundles (vsb), and was absent in the outermost layer of mesophyll and the epidermis (e). e Fully expanded leaf (main rib length 5 cm, sited 10 cm below the SAM, April). f Magnification of lamina showed the KNOPE3 expression in vascular bundles (vsb), palisade layer (pa) and aerenchyma (a). g Mid vein close-up evidenced specific transcript localization in the phloem (ph), absence in the xylem (x) and bundle sheath (bs). f In petiole transversal section the mRNA pattern resembled that of the leaf mid vein (size bars a 25 μm, b 280 μm, c, d 30 μm, e 110 μm, f 50 μm, g, h 25 μm)

Full size image

KNOPE3 was expressed in all organ tissues and the lowest transcript level occurred in stems and petals (Fig. 3). Stems of elongating shoots (April) were sectioned at 0.2 mm (Fig. 4a, b) and 5 cm (Fig. 4c–e) underneath the SAM. The transcript signal (blue stain, Fig. 4a) spread evenly in the cortex and was absent in the pith; it also marked the phloem and/or phloem-associated cells (also called hereafter sieve element/companion cell complex, acronym: SE-CCC) of leaf traces (Fig. 4b). At 5 cm, a strong signal featured in both the stem vascular system and leaf traces (Fig. 4c). As for the latter (Fig. 4d), KNOPE3 mRNA marked: (a) cell groups forming the SE-CCC residing underneath the phloem cap and (b) sporadic cells, likely to be the secondary phloem, which were located upon the cambium layer. As for stem vascular bundles (vsb), the message occurred in secondary phloem cells (Fig. 4e), but was absent in the intra-fascicular cambium, in the phloem cap and xylem cells.

In floral buds before anthesis (February), the gene transcript marked the receptacle, the sepals and the outer layers of anthers (Fig. 5a). More precisely, the signal diffused into the cortical region and peaked within the vsb of the receptacle (Fig. 5a); it occurred in the parenchimatous cells and vsb of sepals, whereas it only featured in vsb of petals (Fig. 5b). With regard to stamens, magnifications showed that the message localized to the endothecium and tapetum cells (Fig. 5c) and to the vsb of filaments connecting the anther lobes (Fig. 5d). KNOPE3 was not expressed in pollen grains and perithecia (Fig. 5c). As for gynoecium, a faint signal was observed in the outer cells of the stigma orifice, in the style vsb (Fig. 5e) and in a few inner cells of the ovary (Fig. 5f). Fruits of 3 mm diameter (Fig. 5g) were sectioned at early canonical stage S1 (Zanchin et al. 1994) and the transcript marked the vsb of mesocarp and seed funiculus.

KNOPE3 transcript localizes with different patterns during leaf growth

KNOPE3 expression was monitored in leaves at distinct growth stages borne on the main axis (April), following a top-down direction (Table 1). The lamina area, folding/expansion status, chlorophyll contents and RUBCA expression (Watillon et al. 1993) were measured to mark the growth stages and they increased with leaf maturity. The KNOPE3 transcript levels of fully expanded leaves at the shoot basis were significantly higher than those of upper folded leaves, evidencing that the expression raised in concomitance with growth.

In situ hybridization was performed on the apical meristems and surrounding leaves of shoots at the vegetative burst (late March). In SAM longitudinal sections, the KNOPE3 signal was absent in the central and peripheral zones, marked cells of the innermost layer in the rib zone, likely to be the vascular cordon, were intense in the pro-vascular strands at the basis of primordia and developing leaves, and absent in the pith (Fig. 6a). In transversal sections, the overall mRNA pattern showed a signal increase proceeding from the inner to the outer leaves (Fig. 6b). In younger leaves, it mainly localized to the vsb (Fig. 6c), whilst it extended in several group cells of older leaves (Fig. 6d), namely in the vsb and mesophyll cells, but not in sub-epidermal and epidermal layers. In expanded leaves (Fig. 6e), KNOPE3 message localized to the palisade layer, spongy parenchyma and vsb (Fig. 6f), but was absent in the bundle sheath of main (Fig. 6g) and secondary veins (not shown) and in the epidermis (Fig. 6f). Finally, KNOPE3 expression occurred in SE-CCC of leaf mid vein (Fig. 6g), which strongly resembled the pattern of stem leaf traces (Fig. 6h). Controls were performed using a sense probe, no signal above background was detected (not shown).

Leaf KNOPE3 rapidly responds to light variation and is triggered by sugars

The level of KNOPE3 transcript was assayed in leaves of darkened seedlings (Fig. 7). Light-exposed in vitro plants were obscured for 16 h and subsequently re-exposed to light. To characterise the physiological status of treated plants, the expression of peach RUBCA and CAT2 genes was monitored to mark dark/light (Watillon et al. 1993; Giannino et al. 2004) and stress events, respectively (supplemental material 2). The contents of major translocation sugars (Merlo and Passera 1991; Moing et al. 1997) were also determined (Table 3) and the fold variation of sorbitol and sucrose was graphed together with that of KNOPE3 transcription (Fig. 7). The latter augmented rapidly and climaxed with a ca. eightfold increase at 4 h post-darkness (hpd). An abrupt drop to the control levels occurred at 8 hpd, followed by an intense repression at 16 hpd (Fig. 7). The trend of sucrose content variation was similar to that of KNOPE3, suggesting that transport sugar may act on KNOPE3 expression. The RUBCA transcription was repressed from 0.5 hpd onwards and triggered at 1 h post-light restore. The CAT2 expression augmented significantly of ca. six times at 16 hpd, suggesting a late stress status (supplemental material 2).

Fig. 7
figure 7

KNOPE3 expression pattern and sugar leaf content variation in response to imposed darkening. Time-course of relative KNOPE3 mRNA levels by qRT-PCR and fold change variation of sorbitol and sucrose leaf contents ±1 h preceding the darkness and subsequent to light restoration, respectively. White and black circles indicate light-on and darkening, respectively. Bars indicate standard errors

Full size image

To assess whether KNOPE3 transcription is modulated by carbohydrates, shoot apical tips were fed with fructose, glucose, sucrose and sorbitol via the transpiration stream and gene expression monitored in a 12 h time lapse under illumination (Fig. 8a). The KNOPE3 transcription was induced differentially at 3 h post-treatment (hpt) by sucrose, glucose and fructose, maintaining the up-regulation state afterwards, whereas sorbitol feeding led to the transcriptional peak at 12 hpt. A 20- to 30-fold expression raise was observed after sucrose and glucose treatments, a less than tenfold induction was caused by fructose, whilst the sorbitol induction was ca. 15 times higher than controls (samples fed with water).

Fig. 8
figure 8

Time course of KNOPE3 expression in response to sugar variation. a The relative mRNA abundance of KNOPE3 determined by qRT-PCR in young (sink) leaves fed with 50 mM glucose (GLU, open square), fructose (FRU, open triangle), sorbitol (SORB, filled square) and sucrose (SUC, filled circle). b The KNOPE3 expression (filled square) was compared with that of SORBITOL DEHYDROGENASE (SDH, open square) in sorbitol uptake assay. c Stems bearing mature (source) leaves were girdled and the leaf KNOPE3 expression kinetics (filled square) was graphed together with sorbitol (SORB, open circle) and sucrose (SUC, filled circle) content variation

Full size image

The sorbitol kinetics was delayed with respect to the other sugars, suggesting that a sorbitol derivate may have mediated the gene triggering rather than the polyol itself. The latter is converted into fructose by the sorbitol dehydrogenase (SDH, EC 1.1.1.14), hence the peach SDH transcription was monitored upon sorbitol supply (Fig. 8b). The SDH up-regulation was confirmed (Zhou et al. 2006) and found to precede that of KNOPE3.

Stem girdling was performed to inhibit sugar transport from mature basal leaves. The KNOPE3 expression was significantly higher than controls (Fig. 8c) from 6 hpt afterwards, along with the raise of sorbitol and sucrose contents (Table 4) that followed the phloem removal. The down-regulation of the peach SORBITOL-6-PHOSPHATE DEHYDROGENASE gene (S6PDH), which encodes the enzyme synthesizing sorbitol (EC1.1.1.200), further suggested that sorbitol augment was caused by the translocation block rather than by a de novo synthesis (supplemental material 3).

The wounding effect (petiole cutting and stem cortex removal in sugar supply and girdling experiments, respectively) on KNOPE3 triggering was negligible, if not absent. This is viewable in the supporting material to Fig. 8a and c (named supplemental material 5), explicating KNOPE3 expression in control samples. To further check whether KNOPE3 was triggered by mechanical damages, gene transcription was measured in leaves after their blades were tweezer-injured (Supplemental material 4, Response to leaf injury). No significant variation of the message levels occurred along 6 h after wounding, whilst the expression of CATALASE gene stress markers (CAT1 and CAT2) increased from 3 hpt onwards. Hence, the KNOPE3 up-regulation appeared to be specifically and rapidly caused by glucose, fructose and sucrose.

Discussion

The peach KNOPE3 full length cDNA was isolated and assigned to the class 2 KNOX (Reiser et al. 2000). The gene consisted of six exons and five introns, maintaining a highly similar organization to that of Arabidopsis KNAT3. It was a single copy per haploid genome in the cultivar ‘Chiripa’ and belonged to a small family class 2 KNOX genes. In our lab, five class 1 KNOPE members were assessed in ‘Chiripa’, whilst the class 2 KNOPE4, distinct from KNOPE3, was retrieved by cDNA sequencing (unpublished). Hence, at least six KNOPE genes occur in the peach genome. KNOPE3 was positioned at 66 cM on G1 of the Prunus ‘Texas’ × ‘Earlygold’ genetic map and fell into a region where a QTL for leafing date was localized in almond (Sanchez Perez et al. 2007). The deduced protein KNOPE3 maintained the typical KNOX organization, including the N-terminal MEINOX, the ELK and the HD domains. The ELK domain of class 1 KNOX was proposed to be a nuclear localization signal (NLS, Reiser et al. 2000) and BEL dimerization was required for proper import to the nucleus (Cole et al. 2006). Class 2 KNOX are regulated at the post-translational level to reach the nucleus (Truernit et al. 2006), though the precise amino acidic signals are unknown. Interestingly, the C terminus of KNOPE3 contained (http://cubic.bioc.columbia.edu/predictNLS/) a predicted NLS (RKRNWHSNPSTSTVLKSKRKR, pos. 416–436, Fig. 1). Additionally, KNOPE3 harboured the same binding motif (WFIN pos. 411–414, Fig. 1) as that of class 1 KNOX, which recognize the TGACAGG/CT DNA stretch (Hake et al. 2004).

The KNOPE3 expression was widespread in several organs (Fig. 3), though the mRNA abundance varied in distinct tissues, in accordance with the behaviours of other class 2 KNOX (Reiser et al. 2000). In peach SAM, KNOPE3 was confined to pro-cambial strands underneath primordia and developing leaves (Fig. 4a). In the stem below the apex, the transcript featured in the cortex and vascular bundles, whereas it was confined to SE-CCC in stem distal sectors, where the secondary growth had started. Overall, the gene was repressed in the tunica of apical meristems and in cambia, pith, xylem and phloem cap of stems. The mRNA localization profile in undisclosed flower buds evidenced KNOPE3 action in the vsb of stamen threads, pistils, sepals, petals, receptacle and in tapetum/endothecium of anthers. The patterns were consistent with those reported for tomato Let12 in flower buds (Janssen et al. 1998). The KNOPE3 mRNA sequence is retrievable from EST databanks of drupe at S1–S4 stages (http://www.itb.cnr.it/estree/), proofing that the gene works during the whole fruit development. In situ hybridizations showed that the message localized to the vasculature innervating the mesocarp and the seed funiculus, suggesting that KNOPE3 maybe play a role in fruit nourishment.

With regard to leaf development, KNOPE3 mRNA abundance increased with growth, being lower in apical curled than basal expanded leaves (Table 1). The KNOPE3 mRNA was first seen in the vsb of younger leaves and then in SE-CCC and mesophyll of older leaves. In expanded leaves, it persisted in SE-CCC, palisade and spongy layers, but was absent in the epidermis, bundle sheaths and collenchyma. The message of Arabidopsis KNAT3 localized to vsb of leaf primordia, petioles of older leaves and hypocotyls of seedlings. Additionally, GUS assays showed that gene promoter activity declined from young to older leaves, being intense in veins and blade of the former, whilst strong in vascular net and diffuse in lamina of the latter. Hence, authors suggested that KNAT3 repression occurred with leaf development (Serikawa et al. 1997). The KNOPE3 mRNA loc patterns were partially consistent with those of Arabidopsis KNAT3; however, qRT-PCR experiments showed that KNOPE3 transcript increased with leaf growth. This was supported by in situ experiments showing that KNOPE3 domains expanded from vsb to mesophyll in parallel with leaf maturity. Hence, KNOPE3 and KNAT3 may not share the same functions in leaf development despite sharing high identity.

The KNOPE3 message occurs in pro-vascular strands and is associated to bundles of all aerial organs, suggesting that it may play a role throughout the functioning of vascular system. The intense localization signal inside the SE-CCC suggests that KNOPE3 mRNA may be either transported in the sap, similarly to class 1 messages (Kim et al. 2005), or produced directly by the nucleated companion cells of phloem. The class 1 KNOX protein localize to vasculature associated cells of leaves (Harrison et al. 2005; Hay and Tsiantis 2006) and are able to self-move, contrary to the class 2 KNAT3 (Kim et al. 2005) and NTH201, which localizes to plasmodesmata (Yoshii et al. 2008). Supposing that KNOPE3 maintains these skills, it might act at the level of connections between companion cells and adjacent sieve elements.

The KNOPE3 expression in green and photosynthetically active tissues, such as leaf mesophyll and sepal parenchyma (Vemmos and Goldwin 1994), suggests that gene activity is light regulated. In fact, obscuring the leaves caused a fast and intense expression increase, followed by a drastic down-regulation; the light restore triggered the gene immediately. In prolonged darkness, the KNOPE3 repression agrees with that of Arabidopsis KNAT3 in 6-day etiolated seedlings (Serikawa et al. 1996, 1997). However, in the latter case, multiple factors in addition to light absence can account for KNAT3 repression, whilst our experiments indicate a KNOPE3 fast sensing to both dark and light. Interestingly, the “high and low” transcription trend of KNOPE3 overlapped the “rise and fall” area of sucrose content. This is not unexpected since microarray analyses show that many genes are co-regulated by both light and sugar (Thum et al. 2008). In plants exposed to darkening, starch breakdown occurs and carbohydrate metabolism genes are reported to respond positively (reviewed by Smith et al. 2003; Rolland et al. 2006); moreover, senescence and stress defense mechanisms are also induced (Weaver and Amasino 2001; Lin and Wu 2004). The results from imposed obscurity suggest that the rapid increase of leaf KNOPE3 transcription may be caused by the quick variation of sugar content (mainly sucrose) together with the light turn-off. The role of light and carbohydrates is further supported by computational analyses (http://www.dna.affrc.go.jp/htdocs/PLACE) performed on KNOPE3 orthologues, namely Arabidopsis KNAT3 and grapevine CAO67726, which both hosted several light and sugar responsive motifs in the promoters (not shown).

Experiments of sugar uptake from sink leaflets showed that glucose, fructose and sucrose specifically triggered KNOPE3 transcription and more rapidly than sorbitol. The delayed sorbitol action and the clear difference between fructose and sucrose/glucose kinetic patterns (Fig. 8a) sustained a sugar-specific effect. The insignificant effect of wounding further corroborated the sugar peculiarity. The quick KNOPE3 up-regulation was consistent with that of several Arabidopsis homeobox TFs, including KNAT3, occurring at 3 h after glucose treatment of seedlings. Moreover, de novo protein synthesis was proposed to mediate the glucose induction (Price et al. 2004). In this context, the production of KNOPE3 antibodies would be auspicial to assess whether mRNA up-regulation associates to protein accumulation (as well as to confirm the mRNA and protein co-localization), considering that transcript and protein levels may not correlate as a result of multiple regulatory mechanisms (Bassett 2007; Schutze et al. 2008). KNOX alternative splicing is known to occur (reported in “Introduction”), but no specific data on post-translational modification have been available to our knowledge so far. The KNOPE3 late response to sorbitol suggests that sorbitol metabolites (and not the polyol itself) ignite the KNOPE3 up-regulation. The SDH induction preceded that of KNOPE3 after sorbitol supply, hence the fructose released by SDH might be a candidate of KNOPE3 triggering. Another candidate may be the glucose produced by the sorbitol oxidase (SOX, Yamaki 1980). SOX plays a negligible role in peach and apple sink leaflets (Lo Bianco et al. 1999; Zhou et al. 2006) and its expression was not tested because there has not been any sequence available in any plant database so far. Hence, it is not excluded that SOX-derived glucose may also mediate KNOPE3 response to sorbitol. Finally, leaf KNOPE3 up-regulation was observed to proceed along with the endogenous increase of transport sugars due to the block of source-to-sink trafficking. Class 2 KNOX may have a role in carbohydrate translocation by acting on genes encoding sugar carriers and/or enzymes of sugar metabolism. For instance, computational analyses show that the S6PDH genes of poplar and apple (acc n. P28475 and AF057134) harbour several KNOX binding motifs. This prompts our future investigation towards the elucidation of class 2 KNOX role in nutrient transport, which is pivotal for organ differentiation, growth and development.