Involvement of calcium signalling in dormancy release of grape buds

Abstract

Artificial induction of grape bud dormancy release by hydrogen cyanamide (HC) serves as a reliable model system to explore the events occurring shortly after the induction of dormancy release. Recently, a group of genes with remarkable differences in expression level between HC-treated and control buds was identified. The identification of several calcium signalling-related genes within that group raised the hypothesis of the involvement of Ca²⁺ signalling in grape bud dormancy release. Therefore, the effects of HC treatment on the expression profiles of several calcium sensors, the effect of the plasma membrane calcium channel blocker LaCl₃ and the calcium chelator EGTA on HC-induced and chilling-induced bud-break, and the effect of HC application on calcium-dependent protein phosphorylation activities in the bud tissue were studied. Here the HC-induced expression of Ca²⁺-ATPase is described, indicating that this treatment might evoke an increase in [Ca²⁺]cyt. Similar induction was confirmed for calmodulin, calmodulin-binding protein, and calcium-dependent protein kinase (CDPK). Both LaCl₃ and EGTA blocked the inducing effect of HC on bud-break, and their inhibitory effects were removed by supplying exogenous Ca²⁺. Calcium-dependent histone phosphorylation was up to 70% higher in HC-treated buds. Endogenous protein phosphorylation assays detected four proteins exhibiting increased phosphorylation following HC treatment, of which two were phosphorylated in a calcium-dependent manner. One of these, a 47 kDa protein, presented strong and Ca²⁺-dependent phosphorylation only in HC-treated buds. The potential role of CDPK in the phosphorylation of this protein was supported by an immunoprecipitation assay. The data suggest, for the first time, that calcium signalling is involved in the mechanism of bud dormancy release.

Bud dormancy, calcium signalling, heat shock, hydrogen cyanamide, Vitis

Introduction

The buds of temperate woody plants, including grapevine, undergo a dormancy cycle during the winter, induced by decreasing photoperiod and/or temperatures (Heide, 1974; Amling and Amling, 1980; Walser et al., 1981; Hummel et al., 1982; Scalaberilli and Couvillon, 1986; Fennel and Hoover, 1991; Rodriguez et al., 1994), and ending when plants have received adequate chilling temperatures and are exposed to favourable growing conditions (Samish, 1954; Brown et al., 1967; Saure, 1985; Dennis, 1986).

In warm-winter regions, prolonged dormancy is a major obstacle for commercial production of temperate fruits, including grapes, which are widely distributed in subtropical regions (Shulman et al., 1983; Saure, 1985; Erez, 1987). At present, dormancy release needs to be controlled by the use of artificial dormancy-breaking compounds to compensate for the lack of natural chilling, and this practice is indispensable for maintaining economic production of table grapes in these regions (Erez, 1985, 1987, 1995). However, the currently available effective compounds are both costly and entail a risk of bud damage due to their phytotoxicity (Erez, 1987, 1995; Or et al., 1999).

Understanding the biological mechanisms involved in bud dormancy release is crucial for the manipulation of bud-break timing. Although extensive studies have been performed on various physiological aspects of dormancy (Nitsch, 1957; Fuchigami et al., 1982; Saure, 1985; Dennis, 1986; Powell, 1987; Fuchigami and Nee, 1987; Martin, 1991; Lang, 1994; Borchert, 2000), the complex network of biochemical and cellular processes responsible for the regulation and execution of bud dormancy release has not yet been characterized (Seeley, 1994).

Hydrogen cyanamide (HC) is the most useful dormancy-breaking compound for grapevine (Shulman et al., 1983; Erez, 1987, 1995; Henzell et al., 1991). Although the mechanism by which HC exerts its dormancy-breaking effect is not clear, its application provides a uniform and effective response, as well as controlled induction timing. HC-based induction of dormancy release may, therefore, serve as a reliable model system to study early events occurring shortly after the induction of dormancy release, by providing an accurate and clear-cut definition of the actual timing of dormancy release induction (Or et al., 1999, 2000a, b, 2002). This advantage could then be used to identify the biochemical components making up the signal transduction cascade, which is activated following application of an environmental and/or chemical trigger.

To identify such components, an analysis of alterations in gene expression during early stages of HC-induced dormancy release in grape buds was initiated in our laboratory. This analysis showed that catalase expression and activity are both inhibited in grape buds shortly after HC application, leading to the accumulation of hydrogen peroxides (Nir et al., 1986; Shulman et al., 1986; Or et al., 2000b). It was also shown that GDBRPK, a transcript for a sucrose non-fermenting (SNF)-like protein kinase, is up-regulated following chemical induction of dormancy release by HC. Several studies suggest that SNF-like protein kinases function as stress pseudoreceptors in yeast, mammals, and plants (Anderberg et al., 1992; Jelinsky and Samon, 1999). It was therefore suggested that this protein kinase might be involved in the perception of a stress signal induced by HC (Or et al., 2000b).

The simultaneous and remarkable transient induction of transcripts encoding pyruvate decarboxylase (PDC) and alcohol dehydrogenase (ADH) shortly after HC application was then described. Based on the above, it was speculated that HC application leads to transient respiratory stress, maybe through the development of oxidative stress, which probably results in a temporary increase in the AMP/ATP ratio (Or et al., 2000b). Since AMP is known as a stress signal that is sensed by SNF-like kinases, it was suggested that the SNF-like GDBRPK could serve as the sensor of this signal (Or et al., 2000b, 2002). However, the identification of such changes in expression patterns of single genes provides only partial clues for the possible involvement of particular metabolic pathways in the dormancy process.

To gain a broader perspective on the complex biochemical network responsible for the regulation and execution of the dormancy process, much more detailed insight is needed into the coordinated induction (or repression) of metabolic cassettes that act together during dormancy release of grape buds. As an initial step in the application of genomic approaches in our research, two cDNA libraries were constructed from cDNA originated from control and HC-treated buds, respectively, and a total of 2352 and 2400 clones from the respective libraries were randomly selected and sequenced (T Keilin et al., unpublished results).

Since the abundance of a particular transcript in an expressed sequence tag (EST) collection from a non-normalized cDNA library can be used to estimate its expression level (Fei et al., 2004; Ma et al., 2004), differential gene expression was computed based on a digital expression profile for the HC-treated bud library compared with the control one. This analysis identified remarkable differences in the abundance of clones for a few genes between the two libraries (T Keilin et al., unpublished results). Among these were several calcium signalling-related genes, including a calmodulin-binding protein (CBP), calmodulin (CaM), and Ca²⁺-ATPase.

These data suggest the involvement of calcium signalling in dormancy release. This notion may be supported by recent findings which show that oxidative stress induces calcium signalling in plant cells (Price et al., 1994; Pei et al., 2000; Foreman et al., 2003; Rentel and Knight, 2004). Further support may stem from the involvement of calcium in the regulation of the activities of several SNF-like proteins (Hotta et al., 1998).

Calcium is a prominent second messenger in signal transduction cascades in plants which are involved in a wide variety of growth and developmental processes, including plant responses to various environmental stimuli (Hepler and Wayne, 1985; Bush, 1995; Trewavas, 1997). The involvement of Ca²⁺/CaM-dependent NAD⁺ kinase (Gallais et al., 2000, 2001), calcineurin B-like protein kinase (Kim et al., 2003; Pandey et al., 2004), and calcium-dependent protein kinase (CDPK; Anil et al., 2000) in seed germination was documented, and suggested the involvement of Ca²⁺ signalling in seed germination/seed dormancy release. However, involvement of calcium signalling in bud dormancy release has never been documented.

To analyse further the hypothesis that calcium signalling is involved in bud dormancy release, the effect of HC treatment on the expression profiles of several calcium sensors, the effects of the plasma membrane calcium channel blocker LaCl₃ and the calcium chelator EGTA on HC-induced and chilling-induced bud break, and the effect of HC application on calcium-dependent protein phosphorylation activities in the bud tissue were studied.

Materials and methods

Plant material

The experiments were conducted with mature, cordon-trained grapevines (Vitis vinifera cv. Perlette) from commercial vineyards in the central Jordan Valley. All plants were subjected to the cultural practices commonly used in commercial vineyards in this region.

Chemical induction of dormancy release

Dormancy-breaking treatments were applied in mid-December. Vines were pruned to three-node spurs and the detached canes, each containing nine buds (in positions 4–12) were transferred to the laboratory. Canes were cut into single-node cuttings that were randomly mixed and placed in vases, with their bases immersed in water, at 22 °C overnight. The next day, the cuttings were separated into two groups. Cuttings of one group were sprayed with 3% ‘Dormex’ (SKW, Trostberg, Germany), a commercial formulation containing 49% HC, and those of the other group were sprayed with water. Both ‘Dormex’ and water were sprayed with the addition of 0.02% Triton X-100 as a wetting agent. Vases with sprayed cuttings were then placed in a growth chamber at 22 °C under a 14 h/10 h light/dark regime. Buds were harvested from both HC-treated and control cuttings at selected time points after HC application (6, 12, and 24 h, 2 d, and 4 d), immediately transferred to liquid nitrogen, and then stored at –80 °C.

Ten groups of 10 single-node sprayed cuttings were prepared from control and HC-treated cuttings and then forced, under the conditions described above, for 30 d. The bud-break percentages were used to assess the effect of the treatment on dormancy release.

Analyses of the influence of calcium-blocking and recovery treatments on the inducing effect of HC on dormancy release

The calcium channel blocker LaCl₃ and the calcium chelator EGTA [ethylene glycol-bis(β-aminoethylether)-N,N,N′,N′-tetraacetic acid, Sigma] were used in calcium-blocking treatments. Recovery treatments were then carried out by application of either Ca²⁺ or Mg²⁺ solution after removal from LaCl₃ or EGTA solution.

Single-node cuttings were prepared as described above and were randomly mixed. Nine groups of 10 single-node cuttings were prepared for each treatment. For the blocking treatments, the bases of the cuttings were dipped in different concentrations of LaCl₃ or EGTA for 9 h, and then washed with double-distilled water (ddH₂O) and dipped in ddH₂O for another 9 h. For the recovery treatment, the bases of the cuttings were first dipped in different concentrations of LaCl₃ or EGTA for 9 h, then washed with ddH₂O and dipped in different concentrations of CaCl₂ or MgCl₂ solutions for another 9 h. An orthogonal design was used to study the blocking effect of 10, 20, and 30 mM LaCl₃, and the recovery effect of 10, 20, and 30 mM Ca²⁺ or Mg²⁺. A similar design was used to study the blocking effect of 10, 20, and 30 mM EGTA, and the recovery effect of 10, 20, and 30 mM Ca²⁺. After 18 h, all the cuttings from both the blocking and recovery treatments were washed with ddH₂O, and then sprayed with 3% Dormex, with the addition of 0.02% Triton X-100. The sprayed cuttings were placed in ddH₂O-containing vases in a growth chamber under the above-described conditions. Bud-break percentages were used to assess the effect of the treatment on dormancy release.

Analyses of the influence of calcium-blocking and recovery treatments on the inducing effect of controlled chilling on dormancy release

Controlled chilling treatments were applied at the beginning of December. Detached canes were divided into six groups that were each wrapped with wet cloth and stored at 4 °C for 0, 200, 400, 500, 600, and 800 h. At each time point, one group was removed from storage and cut into single-node cuttings. Ten groups of 10 cuttings were placed in ddH₂O-containing vases in a growth chamber under the above-described conditions. Bud-break percentages were used to assess the effects of various chilling durations on dormancy release.

For the blocking treatments, the bases of single-node cuttings were dipped in 30 mM EGTA for 12 h, then washed with ddH₂O and dipped in ddH₂O for another 12 h. For the recovery treatment, cuttings were first treated with EGTA as described above, then washed with ddH₂O and dipped in 10 mM CaCl₂ solution for another 12 h. Cuttings that were dipped in ddH₂O for 24 h served as controls. All of the dipping procedures were carried out at 4 °C. The cuttings were then removed from solution, stored at 4 °C for 800 h, and placed under forcing conditions. All other details are as described above.

RNA extraction and northern analysis

Total RNA was extracted from frozen grape buds using a modified cetyltrimethylammonium bromide (CTAB)–LiCl protocol published by Chang et al. (1993). Total RNA was denatured, fractionated by formaldehyde/agarose gel electrophoresis, and transferred to a nylon membrane (Hybond N⁺, Amersham) according to Sambrook et al. (1989). Hybridization was performed with ³²P-labelled probes. The probes of CBP (TIGR Grape Unigene TC 34227), CaM (TIGR Grape Unigene TC 39147), Ca²⁺-ATPase (TIGR Grape Unigene TC 48993), and CDPK (TIGR Grape Unigene TC 33887) were prepared by PCR amplification of the corresponding pTripleEx2-based clones, using the insert screening primers LDF and LDR according to the manufacturer's instructions (Clontech). Hybridization buffer containing 5× SSC and 10× Denhardt's was prepared according to Sambrook et al. (1989). The hybridization temperature was 55 °C. The membranes were washed twice with 2× SSC/0.1% SDS for 20 min at room temperature, followed by a 15 min wash with 1× SSC/0.1% SDS and a 30 min wash with 0.1× SSC/0.1% SDS, both at 65 °C.

Protein extraction

Frozen buds (2 g) were ground in liquid nitrogen. The powder was then suspended in 3 ml of extraction buffer [20 mM TRIS, pH 7.2, 74 mM NaCl, 1 mM EDTA, 0.5% (v/v) NP-40, 30 μl of complete protease inhibitor (one tablet of complete protease inhibitor from Roche Diagnostics GmbH, Mannheim, Germany was dissolved in 1 ml of ddH₂O), 5 mM dithiothreitol (DTT), 2% (w/v) polyvinylpyrrolidone (PVP), 8 mM imidazole, 1 mM Na₂MoO₄, 100 M Na₃VO₄, 20 mM sodium L-tartrate]. The crude extract was homogenized on ice for 1 min, and then centrifuged at 12 000 g, 4 °C for 30 min. The supernatant, containing the soluble proteins, was filtered through two layers of microcloth (Calbiochem). After precipitation by 25% saturation of ammonium sulphate and centrifugation at 13 000 g, 4 °C for 15 min, the supernatant was dialysed against 10 mM TRIS pH 7.2 at 4 °C overnight. Protein concentration was determined by the Bradford reagent (BioRad) according to the manufacturer's instructions.

Kinase activity assays

Substrate phosphorylation assay:

Protein kinase activity was determined by measuring the incorporation of ³²P from [γ-³²P]ATP into the exogenous substrate histone III-S according to the method of Anil et al. (2000). In a total volume of 30 μl, the assay mixture contained Ca²⁺/EGTA kinase buffer (25 mM TRIS-HCl, pH 7.2, 10 mM MgCl₂, 0.3 mM EGTA, 25 mM β-glycerolphosphate, 50 μM vanadate, with or without 0.1 mM CaCl₂), 0.43 μg μl⁻¹ histone III-S, and 10 μg of the protein sample. The reaction was initiated with the addition of 0.5 μl of [γ-³²P]ATP (10 μCi μl⁻¹, 3.3 μM, PerkinElmer). After incubation of the reaction at 30 °C for 25 min, the reaction was terminated by the addition of 10 μl of 3× Laemmli's sample buffer (Laemmli, 1970) and incubation at 95 °C for 5 min. Proteins were resolved on a 10% SDS–polyacrylamide gel at 20 V overnight. The gels were dried and exposed to X-ray film (Fuji) for 2 h.

Autophosphorylation and/or endogenous substrate phosphorylation assay:

Soluble protein extracts were also assayed to determine Ca²⁺-dependent autophosphorylation and/or endogenous substrate phosphorylation activities. The reaction was carried out basically as described above but here histone III-S was omitted and 20 μg of the protein sample used. The dried gels were exposed to biomax-sensitive X-ray film with intensifying screen (Kodak) for 3 h at –80 °C.

Immunodetection of CDPK using polyclonal anti-soybean CDPK

Soluble protein extracts from frozen grape buds which were sampled at different time points were resolved by 10% SDS–PAGE and electroblotted onto a nitrocellulose membrane (Amersham) following standard conditions (Sambrook et al., 1989). Immunoblots were performed according to Trebitsh et al. (2000). Membranes were incubated with polyclonal antibodies directed against the CaM-like domain of soybean CDPK (Bachmann et al., 1996; kindly supplied by Professor Harmon, University of Florida) and visualized by the enhanced chemiluminescence technique (SuperSignal, Pierce).

Immunoprecipitation

For the immunoprecipitation (IP) studies, 100 μl of protein extract obtained by the procedure described above was incubated with 0.04 μg μl⁻¹ rabbit polyclonal CDPK antibodies at 4 °C for 6 h. To monitor the efficiency of the IP, an identical protein sample was incubated without antibodies, or with unrelated rabbit polyclonal anti-NF-κB-inducing kinase (NIK) of human origin (Santa Cruz Biotechnology, SC-7211). Immunocomplexes were precipitated by incubation for 2 h at 4 °C with 10 μl of 1:2 (v/v) protein A–agarose beads (Pharmacia) as precipitating agent. After centrifugation at 20 000 g for 2 min, the supernatant was assayed for remaining kinase activity under the reaction conditions described above. Remaining CDPK protein was detected by western analysis as described above.

Results

Effect of HC treatment on the expression profiles of calcium signalling-related genes

To study temporal reprogramming of gene expression during HC-induced grape bud dormancy release, two cDNA libraries were constructed from cDNA originated from control and HC-treated buds, respectively (T Keilin et al., unpublished results). A total of 2352 and 2400 clones from the respective libraries were randomly selected and sequenced.

Since the abundance of a certain transcript in an EST collection from a non-normalized cDNA library can be used to estimate its expression level (Fei et al., 2004; Ma et al., 2004), differential gene expression was computed based on a digital expression profile for the HC-treated bud library compared with the control one. This analysis identified remarkable differences in the abundance of clones for several genes between the HC and control libraries. Among these were a few calcium signalling-related genes, including a CBP, CaM, and Ca²⁺-ATPase.

The EST which was annotated as CaM (581 bp) showed 99% identity to TC 39147 in the TIGR Grape Unigene collection (ftp://ftp.tigr.org/pub/data/tgi/Vitis_vinifera), and contained a complete open reading frame (ORF) for a CaM protein. The EST that was annotated as CBP (531 bp) showed 95% identity to TC 34227 in the same database. A partial ORF of 37 amino acids was detected in the 5′ region of this EST by the ORF finder of NCBI (http://www.ncbi.nlm.nih.gov/gorf/). Based on BLASTP analysis against the NR database (http://www.ncbi.nlm.nih.gov/Ftp/index.html), the deduced amino acid sequence of this EST shared 88% identity with a calmodulin-binding domain (CBD) located at the C-terminal end of an Arabidopsis CBP (ACBP60_like NP_973527, Reddy et al., 2002). The third EST (1029 bp), annotated as a plasma membrane Ca²⁺-ATPase, showed 83% identity to TC 48993 in the TIGR Grape Unigene collection. A partial ORF of 188 amino acids was detected in the 5′ region of this EST by the ORF finder of NCBI. Based on BLASTP analysis against NR, the deduced amino acid sequence of this EST showed 87% identity to a plasma membrane Ca²⁺-ATPase from soybean (Glycine max, AAG28435.1; Chung et al., 2000), and the EST appeared to be a partial cDNA clone that covers the 3′-untranslated region (UTR) and part of the cation transport domain at the C-terminal end of the grape Ca²⁺-ATPase.

To confirm that the increased frequency of these ESTs in the HC library indeed reflects an increased transcript level following induction of dormancy release, northern analyses were carried out using the described ESTs as hybridization probes. Transcript levels of CBP, CaM, and Ca²⁺-ATPase were indeed markedly higher in the HC-treated buds compared with controls (Fig. 1), thus confirming the results from the frequency analysis. The induction of Ca²⁺-ATPase expression by HC treatment first became evident 12 h after treatment, peaked markedly at 24 h, and decreased prominently thereafter. By contrast, Ca²⁺-ATPase mRNA was almost undetectable in control buds at all time points analysed. A similar expression pattern was observed for CBP. The expression pattern of CaM differed: transcript was present in both HC and control buds; nevertheless, it was present at significantly higher levels in HC-treated buds at all time points analysed, except 6 h.

Fig. 1.

The effect of HC application on the transcript level of several calcium signalling-related genes in grape buds. Equal amounts of total RNA (20 μg) originated from HC-treated (HC) and control buds (C) that were sampled at several time points after treatment. The same blot was sequentially probed with radiolabelled PCR products amplified from clones representing the following ESTs: calmodulin (CaM; TC 39147); Ca²⁺-ATPase (TC 48993); and calmodulin-binding protein (CBP; TC34227). The autoradiograms and the methylene blue-stained membrane are shown.

Effect of LaCl₃ on HC-induced grape bud dormancy release

Consistent with previous results (Or et al., 2000b), HC treatment markedly induced dormancy release of grape buds compared with controls (Fig. 2). To evaluate the role of calcium signalling during HC-induced dormancy release, the buds were treated with various concentrations of LaCl₃, a plasma membrane calcium channel blocker, prior to HC treatment. As shown in Fig. 2A, all three LaCl₃ treatments markedly reduced HC's inducing effect on bud-break percentage 14 d after treatment. However, after 16 d, only the higher concentrations of LaCl₃ (20 mM and 30 mM) showed a significant inhibitory effect. Overall, the higher the LaCl₃ concentration, the longer and stronger was its inhibitory effect. Nevertheless, about 1 month after the treatments, bud-break percentage reached above 80% for all the treatments, indicating that LaCl₃ did not affect bud viability.

Fig. 2.

The inhibitory effect of LaCl₃ on HC-induced grape bud dormancy release and the recovery effect of Ca²⁺ and Mg²⁺ on this inhibition. (A) Inhibiting effect of LaCl₃: single-bud cuttings were dipped in 10, 20, or 30 mM LaCl₃ for 9 h, transferred to ddH₂O for another 9 h, and then sprayed with 3% Dormex. (B) Recovery effect of Ca²⁺ and Mg²⁺: single-bud cuttings were dipped in 30 mM LaCl₃ for 9 h, transferred to 10 mM Ca²⁺ or 10 mM Mg²⁺ solution for another 9 h, and then sprayed with 3% Dormex. Control buds were dipped and sprayed in ddH₂O in both (A) and (B). HC-treated buds were dipped in ddH₂O for 18 h and sprayed with 3% Dormex in both (A) and (B). All the treated cuttings were placed with their bases immersed in water at 22 °C under 14 h/10 h light/dark conditions. Bud-break (%) was calculated at several time points during the forcing period. The values are the averages of nine replications, 10 cuttings each. Bars represent the standard deviation.

Effective recovery from the inhibitory effect of LaCl₃ (30 mM) was achieved by exogenous application of a low concentration of Ca²⁺ (10 mM): 80% and 92% recovery was evident after 16 d and 19 d, respectively. By contrast, application of 10 mM Mg²⁺ led to significantly less effective recovery (Fig. 2B)—46% and 64% after 16 d and 19 d, respectively.

These results indicate that the induction of bud dormancy release by HC treatment could be interrupted by blocking the plasma membrane calcium channels. The better recovery demonstrated following exogenous Ca²⁺ supply after LaCl₃ treatment, as opposed to Mg²⁺, suggests that the inhibitory effect of LaCl₃ on HC treatment is due to Ca²⁺ blockage. These results suggest a functional link between the inducing effect of HC on bud-break rate and Ca²⁺ availability.

Effect of EGTA on HC-induced grape bud dormancy release

To support further the assumption of a potential role for calcium signalling during HC-induced dormancy release, the buds were treated with EGTA, a specific chelator of extracellular free calcium, prior to HC treatment. Similar to the effect of LaCl₃, EGTA treatment (30 mM) markedly reduced HC's inducing effect on bud-break rate, and its inhibitory effect could be effectively reversed by a low concentration of Ca²⁺ (10 mM), which led to improved bud-break compared with HC-treated buds (Fig. 3). As in the LaCl₃ experiment, bud-break reached about 95% for all treatments after 1 month, indicating that EGTA treatment does not affect bud viability and that HC has an inducing effect on bud-break rate rather than on the total bud-break percentage under the present conditions. The inhibitory effect of the Ca²⁺-specific chelator EGTA, and the recovery from this inhibition upon exogenous Ca²⁺ supply after EGTA treatment, suggest that HC's inducing effect on bud-break rate may be mediated by calcium signals.

Fig. 3

The inhibitory effect of EGTA on HC-induced grape bud dormancy release and the recovery effect of Ca²⁺ on this inhibition. Two groups of single-bud cuttings were dipped in 30 mM EGTA for 9 h. One group was then transferred to 10 mM CaCl₂ for another 9 h while the other group was transferred to water for the same period. The treated buds were then sprayed with 3% Dormex. All the other details are as described in Fig. 2.

Effect of EGTA on chilling-induced grape bud dormancy release

First, the effect of different durations of chilling on grape bud dormancy release was evaluated. Controlled chilling for 500, 600, and 800 h significantly enhanced dormancy release, with a higher bud-break rate following a longer chilling duration. Shorter chilling periods (400 h and 200 h) did not show any significant inducing effect on bud-break rate compared with non-chilled controls (Fig. 4A). To determine the role of Ca²⁺ signalling on chilling-induced dormancy release, single-node cuttings were treated with 30 mM EGTA prior to the 800 h of chilling treatment. As shown in Fig. 4B, the inducing effect of 800 h chilling on bud-break rate was negated by the EGTA treatment: the presented bud-break rate was similar to that of the non-chilled control. The inhibitory effect of EGTA was eliminated when 10 mM Ca²⁺ was supplied following EGTA treatment. Based on the above, it is speculated that Ca²⁺ is involved in the transduction of both natural and artificial signals leading to bud-break.

Fig. 4.

The inhibitory effect of EGTA on chilling-induced bud dormancy release and the recovery effect of Ca²⁺ on this inhibition. (A) Effect of chilling duration on bud dormancy release. Detached canes were divided into six groups that were stored at 4 °C for 0, 200, 400, 500, 600, and 800 h. At each time point, one group was cut into single-node cuttings and transferred to a growth chamber under forcing conditions. (B) Effect of EGTA on chilling-induced dormancy release. Single-node cuttings were divided into three groups. The first group was dipped in ddH₂O for 24 h. The second and third groups were dipped in 30 mM EGTA for 12 h, and then dipped for another 12 h in ddH₂O or 10 mM CaCl₂, respectively. All dipping procedures were carried out at 4 °C. The cuttings were removed from solution, stored at 4 °C for 800 h, and then placed under forcing conditions. All other details are as described in Fig. 2.

Possible involvement of calcium-dependent protein phosphorylation in the response to HC treatment during grape bud dormancy release

To gain more information about the involvement of calcium signalling in HC-induced bud break, the transcript and protein levels of a CDPK, and calcium-dependent protein phosphorylation activities were investigated.

Within the EST collection a partial cDNA clone (550 bp) was found which was annotated as CDPK, presenting 91% identity to TC 33887 in the TIGR Grape Unigene collection (ftp://ftp.tigr.org/pub/data/tgi/Vitis_vinifera/). The deduced amino acid sequence of this EST shared 73% identity with CDPK19 (CPK8) of Arabidopsis (NP_197446.1). Based on alignment with TC 33887, the clone, which originated from the HC library, contains the 3′-UTR and part of the C-terminal EF-hand of the Ca²⁺-binding domain. Amplified product from this EST was used as a probe for northern analysis (Fig. 5). Marked induction of CDPK expression by HC treatment was first evident 12 h after treatment, peaked at 24 h, and then declined slightly. The CDPK transcript level in the control buds was much lower throughout the time-course experiment.

Fig. 5.

The effect of HC-induced bud dormancy release on transcript level of a calcium-dependent protein kinase (CDPK). Equal amounts of total RNA (20 μg) originated from HC-treated (HC) and control buds (C) that were sampled at several time points after treatment. The northern blot was probed with radiolabelled PCR product amplified from a clone representing a CDPK EST (TC 33887). The autoradiogram and the methylene blue-stained membrane are shown.

Calcium-dependent histone phosphorylation activities in grape buds during HC-induced dormancy release

Using histone III-S as a substrate, protein phosphorylation activity was characterized in protein extract originated from HC-treated buds harvested 24 h after the treatments (Fig. 6A, B). To verify whether histone phosphorylation activity is calcium dependent, increasing concentrations of EGTA were added to the assay buffer. Addition of 0.3 mM EGTA resulted in a significant decrease in histone III-S phosphorylation activity. However, a further increase in EGTA concentration did not result in any further decrease in phosphorylation activity (Fig. 6A). These results suggest that the recorded histone phosphorylation activity is calcium dependent, and imply that as little as 0.3 mM EGTA is needed to chelate the endogenous free calcium present in the bud extract. Accordingly, the effect of increasing concentrations of CaCl₂ on histone III-S phosphorylation activity in the presence of 0.3 mM EGTA in the assay buffer was investigated (Fig. 6B). Addition of as little as 10 μM CaCl₂ resulted in 4.3-fold activation of phosphorylation. Activation increased with increasing Ca²⁺ concentration, to a maximal activation level at 40 μM CaCl₂. The activation of calcium-dependent protein phosphorylation by micromolar amounts of Ca²⁺ is in agreement with the activation of CDPKs found in other plant species (Lee et al., 1998; Anil et al., 2000; Harmon et al., 2000). These results suggest the existence of CDPK activity in grape buds during dormancy release.

Fig. 6.

Characterization and time-course of histone phosphorylation activities in grape buds during dormancy release. The activities are presented as the intensities of the signals on the autoradiogram of the SDS–polyacrylamide gel. (A) The influence of 0–0.6 mM EGTA on phosphorylation of histone III-S by bud protein extract (10 μg) originated from HC-treated buds harvested 24 h after treatment. (B) The influence of 0–640 μM CaCl₂ on phosphorylation of histone III-S by the bud protein extract (10 μg) described in (A), in the presence of 0.3 mM EGTA in the reaction buffer. (C) Phosphorylation of histone III-S by protein extracts (10 μg) from control and HC-treated buds at 24, 48, and 96 h, in the presence of 0.3 mM EGTA and 100 μM CaCl₂ in the reaction buffer. (D) Phosphorylation of histone III-S by protein extracts (10 μg) from control and HC-treated buds at 24 h and 48 h, in the presence of 0.3 mM EGTA and in the presence (+) or absence (–) of 100 μM CaCl₂ in the reaction buffer. The intensity of the bands was quantified by Fuji ImageGouge software (version 4.0). The experiments were repeated three times with similar results.

The histone phosphorylation activites in control and HC-treated buds, sampled at different time points, were compared. Histone phosphorylation activity was detected in both control and HC-treated buds. While similar activity was detected 24 h after HC treatment, around 70% higher activities were detected in HC-treated buds at 48 h and 96 h, compared with activites in the control buds (Fig. 6C). These activities were markedly lower in the absence of calcium, in both control and HC-treated buds (Fig. 6D). These results indicate that histone phosphorylation activities in the buds are in large part calcium dependent.

Calcium-dependent endogenous protein phosphorylation activities in grape buds during HC-induced dormancy release

Phosphorylation of protein extracts from control or HC-treated buds was also performed in the absence of exogenous substrate to determine endogenous substrate phosphorylation and/or autophosphorylation activities. Overall higher phosphorylation was observed in HC-treated bud extracts relative to control bud extracts at all time points. Four proteins with estimated sizes of 35, 47, 60, and 120 kDa (Fig. 7A) were phosphorylated. The 35, 60, and 120 kDa proteins were phosphorylated in both control and HC-treated samples. While a much higher level of phosphorylation was visible for the 35 kDa and 120 kDa proteins in the HC-treated buds (especially between 12 h and 48 h after treatment), the difference was less clear for the 60 kDa protein. However, quantification of the intensity of the 60 kDa bands (ImageGouge, Fuji) showed 3.3-, 3.7-, 4.4-, and 1.2-fold higher degrees of phosphorylation in HC-treated samples at 12, 24, 48, and 96 h, respectively. Phosphorylation of the 47 kDa protein was only evident in HC samples, at all time points.

Fig. 7.

Time-course and calcium dependency of endogenous protein phosphorylation activities in grape buds. The activities are presented as the intensities of the signals on the autoradiogram of the SDS–polyacrylamide gel. The panels below the autoradiogram images represent the same gels after Comassie staining. (A) Phosphorylation activities in protein extracts (20 μg) from control and HC-treated buds harvested at different time points, in the presence of 0.3 mM EGTA and 100 μM CaCl₂ in the reaction buffer. (B) Phosphorylation activities in protein extracts (20 μg) from control or HC-treated buds harvested 24 h and 48 h after treatment, in the presence of 0.3 mM EGTA and in the presence (+) or absence (–) of 100 μM CaCl₂ in the reaction buffer. The four phosphorylated proteins are identified by arrows, and their relative positions highlighted on the Comassie-stained gels in the lower panels. The pre-stained molecular mass markers used were: myosin (206 kDa), β-galactosidase (124 kDa), bovine serum albumin (80 kDa), carbonic anhydrase (49.1 kDa), soybean trypsin inhibitor (30.6 kDa), lysozyme (17.8 kDa), and aprotinin (6.9 kDa). Band intensity was quantified as described in Fig. 6.

The possible calcium-dependent nature of the endogenous substrate phosphorylation and/or autophosphorylation activities was investigated by comparing the phosphorylation signal of the same protein sample in the presence and absence of Ca²⁺ (Fig. 7B). Whereas phosphorylation of the 35 kDa and 120 kDa proteins turned out to be Ca²⁺ independent, the 47 kDa and 60 kDa proteins exhibited a weaker phosphorylation signal in the absence of calcium, indicating that their phosphorylation is calcium dependent. Based on the above, it appears that phosphorylation of the 60 kDa and 47 kDa proteins is both HC induced and Ca²⁺ dependent.

Using a polyclonal anti-CDPK antibody prepared against the CaM-like domain of soybean CDPK (Bachmann et al., 1996), a polypeptide was detected in the size range of ∼60 kDa in HC-treated and control bud extracts at several time points. This size range coincides with the size of the putative bud-expressed CDPK protein presented above, which is 60 kDa. Quantification of signal intensities, using ImageGouge software, revealed a 3-fold higher level of the putative CDPK protein in HC-treated buds 48 h and 96 h after HC application (Fig. 8A). These results agree with the increased CDPK transcript level and increased level of calcium-dependent phosphorylation activities in HC-treated buds (Figs 5–⁶7), and support the assumption that the protein detected by the polyclonal antibody, which has previously recognized different CDPK isoforms in many plant species (Anil et al., 2000; Martin and Busconi, 2001), is also a CDPK.

Fig. 8.

CDPK protein levels and endogenous protein phosphorylation activities after immunoprecipitation. (A) Western blot analysis of equal amounts of protein (50 μg) extracted from control or HC-treated buds, using polyclonal anti-soybean CDPK. (B) Protein extract from HC-treated buds at 96 h was incubated with 40 μg ml⁻¹ polyclonal anti-soybean CDPK, or with 40 μg ml⁻¹ polyclonal anti-human NIK. Antibody complexes were immunoprecipitated, and the supernatant was assayed for remaining endogenous protein phosphorylation activity. The same protein extract which was not subjected to immunoprecipitation served as a control. The 47 kDa protein is identified by an arrow. (C) The supernatant obtained in (B) was assayed for remaining CDPK protein by western analysis. Band intensity was quantified as described in Fig. 6.

To investigate the possible involvement of CDPK in the HC-induced and Ca²⁺-dependent phosphorylation of the 47 kDa protein, protein extract from HC-treated buds harvested 96 h after the treatment was subjected to IP by the anti-soybean CDPK in order to remove CDPK from the extract. The supernatant after IP was then assayed for remaining endogenous protein phosphorylation activity. The phosphorylation signal of the 47 kDa protein was 60% lower after the IP than its intensity in the same extract not subjected to IP (Fig. 8B). Parallel IP of the same extract with anti-human NIK was used to demonstrate the specific effect of the anti-soybean CDPK. Accordingly, the former did not reduce the phosphorylation signal, which exhibited an intensity similar to that in extract that had not been subjected to IP. The phosphorylation was further confirmed to be Ca²⁺ dependent based on weaker phosphorylation in the absence of calcium in a reaction with the same extract that was not subjected to IP.

To monitor the efficiency and specificity of the IP, western analysis was carried out using anti-soybean CDPK to detect CDPK protein remaining in the supernatant after IP. The CDPK protein level was significantly lower in the supernatant of the extract subjected to IP by anti-soybean CDPK, compared with supernatants from extract that was not subjected to IP, and from extract that was subjected to IP by anti-human NIK (Fig. 8C). These results suggest that the reduction in the intensity of the phosphorylation signal for the 47 kDa protein is due to the removal of CDPK activity.

Based on the above, it is assumed that the 47 kDa protein is a CDPK target and that the interaction between the two proteins is HC induced and calcium dependent.

Discussion

In warm-winter regions, which lack sufficient chilling, artificial dormancy-breaking compounds such as HC are needed to trigger dormancy release of deciduous fruit tree buds (Erez, 1987, 1995; Shulman et al., 1983; Henzell et al., 1991). It is not clear how either natural or artificial inducers trigger the signal cascades inside the buds, or what biochemical events construct these cascades, although there are some indications of oxidative stress being involved in the process (Or et al., 2000b, 2002). Based on the high redundancy of clones coding for several Ca²⁺ signalling-related genes in the EST collection originated from HC-treated buds, the hypothesis that Ca²⁺ signalling might be involved in grape bud dormancy release was raised and investigated. The results of the current study add support to this hypothesis and suggest that such Ca²⁺ signalling may be mediated by calcium-dependent protein phosphorylation.

To confirm critically the involvement of Ca²⁺ in a particular stimulus response, Sanders et al. (1999) suggested that three experimental criteria should be met: (i) a stimulus should evoke a change in [Ca²⁺]cyt; (ii) blocking of the increase in [Ca²⁺]cyt should block the downstream response; and (iii) appropriate Ca²⁺ sensors should be present. Measurement of an elevation in [Ca²⁺]cyt following either HC or chilling stimulus within the bud would certainly give direct proof of the involvement of a calcium signal in the response. However, such measurements are currently impossible in woody bud tissue. Therefore, experiments were conducted that would meet the other two criteria and demonstrate the involvement of Ca²⁺ in grape bud dormancy release.

While direct evidence of an elevation in [Ca²⁺]cyt following HC treatment is not available, the pronounced activation of transcription from a Ca²⁺-ATPase gene following HC application could serve as an indirect indicator of this elevation, due to the crucial role of Ca²⁺-ATPase in attenuating and modulating the stimulus-induced Ca²⁺ rise (Geisler et al., 2000). Such an induction of Ca²⁺-ATPase transcript (CAP1) was demonstrated in maize seedlings following the elevation of [Ca²⁺]cyt evoked in response to anoxic conditions (Subbaiah et al., 1994; Subbaiah and Sachs, 2000). Similarly, increased [Ca²⁺]cyt elicited dose- and time-dependent up-regulation of Ca²⁺-ATPase2b in human lens cells, at both the mRNA and protein levels (Liu et al., 2002). In addition to the activation of transcription from the Ca²⁺-ATPase gene reported in this study, HC-mediated induction from a Ca²⁺/H⁺ exchanging protein gene and an additional three isoforms of Ca²⁺-ATPase genes were recently detected, using a bud cDNA array that was produced from the above-mentioned EST collection (X Pang et al., unpublished results). These preliminary findings regarding HC-induced up-regulation of several Ca²⁺ transporters may add support to the assumption that HC stimulated an elevation in [Ca²⁺]cyt.

The fact that oxidative stress leads to a transient increase in [Ca²⁺]cyt in several systems, via activation of plasma membrane Ca²⁺ channels (Pei et al., 2000; Foreman et al., 2003), may add support to the assumption that [Ca²⁺]cyt is elevated following HC application, since HC evidently induces a transient oxidative stress within the treated buds following the inactivation of catalase (Or et al., 2000b, 2002; T Halaly et al., unpublished results).

The plasma membrane Ca²⁺ channel blocker La³⁺ and the specific Ca²⁺ chelator EGTA have been used to block Ca²⁺ influx from the apoplast in many plant systems, as a means of proving the need for a Ca²⁺ signal in a cascade of events leading from a certain stimulus to its downstream response (Philsoph-Hadas et al., 1996; Gallardo et al., 1999; Kenton et al., 1999; Liu et al., 2003; Zhao et al., 2003). In the present study, application of both blocking agents led to a significant decrease in the inducing effect of HC on bud-break rate, and exogenous application of Ca²⁺ (10 mM), following the application of LaCl₃ or EGTA, resulted in 80% and 100% recovery of the inducing effect of HC on bud-break, respectively. Application of the identical concentration of another divalent ion, Mg²⁺, was significantly less effective, showing only 46% recovery after removal of LaCl₃. These data suggest that the inhibitory effect of both a Ca²⁺ channel blocker and a Ca²⁺-specific chelator on the inducing effect of HC is due specifically to the blockage of [Ca²⁺]cyt influx which may be evoked by HC and may then serve as a secondary messenger in the cascade that leads to HC-induced bud break. Future study of the effect of the blockers on the expression from HC-induced Ca²⁺ sensor genes will add support to this assumption.

Compared with many other deciduous fruit crops, grapevine buds have low chilling requirements. Nevertheless, the absence or an inadequate degree of chilling will result in limited, uneven, and delayed bud-burst (Lavee and May, 1997), and increased chilling duration increases the rate of bud-break (Kliewer and Soleimani, 1972; Dokoozlian, 1999). Accordingly, the current study revealed that 500 or more chilling hours increase the rate of bud dormancy release relative to non-chilled buds. The inducing effect of such chilling duration on bud dormancy release was negated by EGTA treatment, while addition of Ca²⁺ following the EGTA treatment fully restored the inducing effect, indicating that Ca²⁺ signal might also be involved in chilling-induced bud dormancy release. It has been reported that there is a marked increase in unsaturated fatty acid (C_18:3) when the chilling requirement is close to being satisfied in dormant apple buds (Wang and Faust, 1990). Such unsaturation may increase membrane depolarization (Penzo et al., 2002), which in turn leads to increased Ca²⁺ uptake via depolarization-activated calcium channels (Miedema et al., 2001). Such a scenario could explain the inhibitory effect of EGTA on chilling-induced bud dormancy release. The involvement of Ca²⁺ in the transduction of both natural and artificial dormancy release signals implies that similar mechanisms are activated following the application of these signals and similar events/pathways/activities may belong to core functions within the mechanism of dormancy release. Future study of the effect of chilling on the expression from the HC-induced Ca²⁺ signalling candidate genes presented here may add further support to this assumption.

Elevations of [Ca²⁺]cyt in response to signals, which is the initial step in Ca²⁺ signal transduction cascades, could be due to influx of Ca²⁺ from the apoplast and/or Ca²⁺ release from intracellular stores (endoplasmic reticulum, vacuoles, mitochondria, chloroplasts, and nucleus) (Hetherington and Brownlee, 2004). The results presented here suggest that HC and chilling might lead to an elevation in [Ca²⁺]cyt by inducing an influx of Ca²⁺ from the apoplast, since EGTA is a specific chelator of extracellular free calcium. This may coincide with the finding in poplar buds that Ca²⁺ is located mainly in the cell walls and intercellular spaces after dormancy induction by short-day exposure (Chen et al., 1997).

Evidence has been supplied that elevated Ca²⁺ levels control diverse cellular processes via Ca²⁺ sensors (Sanders et al., 1999; Reddy, 2001). CaM- and Ca²⁺-regulated protein kinases are two major groups of such Ca²⁺ sensors (Reddy, 2001). CaM is a ubiquitous Ca²⁺ receptor in eukaryotes which interacts with various CBPs and modulates their activity (Snedden and Fromm, 2001). In the present study, transcript levels of both a CaM gene and a CBP gene increased in response to HC application, which is indicative of the involvement of a Ca²⁺/CaM pathway in the HC-induced bud-break in grape. The high homology of the reported Ca²⁺-ATPase to soybean SCA1, which is a plasma membrane CaM-activated Ca²⁺-ATPase (Chung et al., 2000), may add further support to this assumption.

The deduced amino acid sequence of CBP shows high similarity to the CBDs from a CBP60 family found in maize (Reddy et al., 1993), tobacco (Lu and Harrington, 1994), Arabidopsis (Reddy et al., 2002), and bean (Ali et al., 2003). Other than the presence of CBD and the binding to CaM in a Ca²⁺-dependent manner, these CBPs do not show sequence similarity to any other known domains, and therefore no clue is provided as to their function (Reddy et al., 2002). Interestingly, one isoform from the CBP family in bean, to which our CBP shows high similarity, was found to be strongly induced by H₂O₂ and other elicitors (Ali et al., 2003). Such a link between the development of oxidative stress and elevation in [Ca²⁺]cyt may exist in the buds’ response to HC application.

To gain more information on the involvement of calcium signalling in HC-induced bud-break, a potential response of Ca²⁺-regulated protein kinases was looked for. Currently, most of the known calcium-stimulated protein kinase activities in plants are associated with CDPKs, which present both protein kinase and CaM-like domains in a single polypeptide and therefore bear a unique advantage that cannot be provided by other Ca²⁺ sensors (Harmon et al., 2000; Cheng et al., 2002; Hrabak et al., 2003). A marked increase in the mRNA level of a CDPK isoform expressed in the buds following HC application was detected, and the involvement of CDPK in mediating Ca²⁺ signalling in HC-induced bud-break was further investigated. Western analysis using an anti-soybean CDPK revealed a 3-fold higher level of the putative CDPK protein in HC-treated buds, 48 h and 96 h after HC application (Fig. 8A). Based on the ability of this polyclonal antibody to recognize different CDPK isoforms in many plant species (Anil et al., 2000; Martin and Busconi, 2001), it is assumed that the detected grapevine protein is also a CDPK. These results agree with the increased CDPK transcript level and increased level of calcium-dependent phosphorylation activities in HC-treated buds (Figs 5–⁶7). Increased CDPK activity is generally ascribed to activation of its transcription and, consequently, accumulation of the protein (Anil et al., 2000; Abbasi et al., 2004), although post-translational stimulation of CDPK has also been found (Abo-El-Saad and Wu, 1995; Martin and Busconi, 2001).

Detection of Ca²⁺-dependent protein phosphorylation activities in grape buds by in vitro assay, using histone III as the substrate, showed similar levels between control and HC-treated buds at 24 h, and around 70% higher activity in HC-treated buds at 48 h and 96 h. While such an assay might not reflect in vivo activities, since Ca²⁺ supply in the in vitro assay could not represent the situation in vivo (Harper et al., 2004), it certainly demonstrates the increased level of potential Ca²⁺-dependent phosphorylating agents. Moreover, when these data are combined with increased transcription of an indirect indicator of [Ca²⁺]cyt elevation, such as Ca²⁺-ATPase, it might be reasonable to infer from the in vitro studies the situation in vivo that might lead to induced calcium-dependent protein phosphorylation activity in HC-treated buds.

While the present data support the involvement of CDPK in HC-induced Ca²⁺-dependent phosphorylation activity, the possibility of a contribution of SNF-like and CBL-interacting protein kinases to the Ca²⁺-dependent histone phosphorylation activity detected in the present study should not be excluded, since these kinases have overlapping substrate specificity with CDPKs (Bachmann et al., 1996; Huber and Huber, 1996; Sugden et al., 1999; Kim et al., 2003). Actually, in a previous study, a higher transcript level of an SNF-like protein kinase (GDBRPK) was detected in HC-treated buds compared with control buds (Or et al., 2000b). On the other hand, recent microarray analysis indicates that two isoforms of CBL-interacting protein kinase were down-regulated by HC treatment. The detection of transcripts for various kinases that can phosphorylate histone in a calcium-dependent manner in the mature bud suggests that the level of calcium-dependent histone phosphorylation reflects the overall activity of these kinases. This insight can help to settle the inconsistency between the 3-fold increase in CDPK protein level (Fig. 8) and a near 2-fold increase in histone phosphorylation activity (Fig. 6) by the assumption that the final level of activity increase is the result of both induction of CDPK and reduction of CBL-interacting protein kinase activity.

Using endogenous protein as a substrate, much higher phosphorylation activity was observed in HC-treated samples than in controls. Two proteins were detected that were phosphorylated in a calcium-dependent manner. Phosphorylation of one of these proteins, around 60 kDa in size, was evident in controls, but was induced by HC application. By contrast, phosphorylation of the other, 47 kDa protein, was HC dependent and did not appear in control buds. This 47 kDa protein appears to be phosphorylated by CDPK, based on the results of IP assay. Since CDPK is available in the tissue in both control and HC-treated buds, it is assumed that the detection of its phosphorylation only in HC-treated buds stems from different availability of the 47 kDa substrate in HC-treated versus control tissues. This coincides with massive activation of the transcription of various genes following HC treatment (T Keilin et al., unpublished results), some of which have been reported to be substrates of CDPKs, such as sucrose synthase (Anguenot et al., 2006) and actin depolymerization factor (Allwood et al., 2001). Alternatively, detection of the 47 kDa protein's phosphorylation only in HC buds may stem from HC-induced conformational changes in this protein, which might facilitate its phosphorylation. Induction of phosphorylation of light-harvesting complex II by light is regulated by just such a substrate conformational change mechanism (Zer et al., 1999). Identification of the 47 kDa substrate may uncover the next step in the calcium signal cascade (Cheng et al., 2002; Harper et al., 2004) and help track the cascade that leads to bud-break.

Since CDPKs can phosphorylate themselves (autophosphorylate) in a Ca²⁺-dependent manner (Saha and Singh, 1995; Chaudhuri et al., 1999; Harmon et al., 2000), the possibility is raised that the ∼60 kDa protein, which is phosphorylated in a Ca²⁺-dependent manner, represents the CDPK itself. However, unequivocal data to support this claim cannot yet be supplied, since it was not possible to carry out in-gel phosphorylation activity assays successfully.

To our knowledge, this is the first systematic investigation of the involvement of calcium signalling in bud dormancy release. Overall, the study supports the assumption that Ca²⁺ signalling is involved in the HC- and chilling-induced mechanism that leads to release of grape bud dormancy.

The authors would like to thank Professor Shimon Lavee for support and encouragement. This research was supported by Research Grant Award No. IS-3340-02 from BARD, The United States–Israel Binational Agricultural Research and Development Fund.