The above-described experimental system led to the conclusion that gene expression is affected by the induction of dormancy release, and exposed changes in the expression of several genes that appear to be relevant to dormancy release, based on the existing horticultural and physiological knowledge (Or et al. 2000b). However, to gain a broad perspective of the complex biochemical network responsible for the regulation and execution of the dormancy-release process, more detailed insight into the coordinated induction/repression of metabolic cassettes that act together during dormancy release is required. High-throughput analysis of the expression profiles of large subsets of genes may be valuable in this endeavor.
5.2.1. Detection of changes in the trancriptome during a natural dormancy cycle
The first attempt to carry out a high-throughput study was reported by Pacey-Miller et al. (2003). Buds were collected from a vineyard at weekly intervals, beginning 9 weeks prior to bud burst. One cDNA library representing all sampling points was generated, from which 4270 clones were randomly selected, sequenced and printed on nylon membranes. RNA from each sampling point was P33-labeled and used to probe the membranes.
While the scope of the study was wide, the dormancy status of the buds at the different sampling points was not analyzed (i.e. a dormancy curve was not established), making the interpretation of the results somewhat questionable. From the indicators supplied in that study—bud swell at time point 6 and bud burst at time point 9—only partial and indirect information could be obtained. Since bud swelling represents the beginning of active growth, which starts only after ED is released, and bud burst commonly occurs no earlier than 3 weeks from the time of ED release under winter field conditions, a valid hypothesis could be that ED was released at no later than time point 5. However, if an ECD effect was involved, the situation would be more complicated: ED release could occur even earlier, and bud swelling could be delayed until the environmental conditions become satisfactory. Hence, it is difficult to properly differentiate between changes related to ED, ECD and meristem growth in the analyzed population. Accordingly, only expression data obtained from buds sampled at time points 1-5 will be considered here, as they most likely represent ED buds during their transition towards dormancy release.
The ESTs with identified functions (65%) were divided into sub-categories according to their cellular role, and the outcome led the authors to conclude that because the whole metabolism of the bud is being assayed, it is difficult to clearly distinguish between the different stages of the developmental process. The diffuculty to reach a conclusion from these data may partially stem from the mixed analysis of ED, ECD and active growth stages. As active growth may represent at least 40% of the biomass used for the library pool, it can account for the high percentage of genes related to cell structure and plant growth. A comparison to the size and content of similar sub-divisions in non-dormant buds or actively growing organs would likely be instrumental in revealing potentially unique features of the mature bud during late stages of ED. Nevertheless, there are indications of a high percentage of defense/stress-related genes, which support the results of other studies (Or et al. 2000b, 2002, Keilin et al. 2007, Mazzitelli et al. 2007, Halaly et al. 2008, Mathiason et al. 2008, Perez et al. 2008).
The authors presented the array-based expression patterns of a few candi date genes that were either associated to the dormancy-release process in previous studies, or represented related functions. An increase in ADH at time point 2 was followed by induction of members of the antioxidative machinery, such as ascorbate peroxidase (APX), glutathione reductase (GR), glutathione peroxidase (GPX) and glutathione S-transferase (GST), at the time points 3 and 4, which may reflect the timing of dormancy release. These findings support the aforementioned hypothesis regarding the development of respiratory and oxidative stress during dormancy release (Or et al. 2000b) and suggest that the antioxidant machinery is induced to cope with the increased H2O2 level in the bud during natural dormancy release. Significant induction of phospholipid hydroperoxide glutathione peroxidase and the stress-induced gene OSMOTIN at time point 3 is in line with this suggestion. A clear and stable decrease in the transcript level of pathogenesis-related genes and DEHYDRIN was also reported, which will be discussed further on.
While this study provided some valuable insight, the high-throughput data could not be further exploited; the number of unigenes in the EST collection was not reported, the ESTs were not made available to the public for further analyses, and expression profiles of the entire collection were not supplied. In addition, the lack of validations for the array-based profiles made it difficult to assess the reliability of the arrays and the quality of the profiles, particularly those that fluctuate redundantly between time points, as in the case of CAT and SN^-related kinase ESTs.
5.2.2. Expression profiling of grape-bud EST collections from dormant and active buds
The next attempt at taking a high-throughput approach was presented by Keilin and co-workers (2007), who generated a publicly available large-scale grape-bud EST collection. Clones for this collection were selected from libraries from buds at various developmental stages: (i) mature woody buds sampled in the vineyard during the dormancy cycle; (ii) young green buds sampled during shoot development; (iii) mature buds sampled at different time points after ED release by HC; and (iv) untreated buds sampled in parallel to the HC-treated ED buds. The collection included 5516 consensus sequences, of which 59% were not included in the Vitis TIGR collection at the time of analysis. About 22% of these transcripts showed no similarity to any known plant transcript.
When mature and young bud libraries were compared, it appeared that the functional distance between the young and mature buds was reflected by both quantitative and qualitative differences within the clone populations in each of the functional categories analyzed. The young bud library presented a much larger variety of functions which were related to active differentiation and growth. Overall, the data indicated that mature buds are biochemically active, but exhibit a different repertoire of biochemical activities. On the other hand, when the HC-treated and control libraries were compared, changes in several functional categories reflected mainly changes in the expressions of single genes. This indicates a significant effect of the treatment on the expression of certain potentially key genes. Among other findings, a comparison between the HC-treated library and its control revealed significant differences in ubiquitin ligase complex, ribonucleoprotein complex, plasma membrane, cell division, cell cycle and biopolymer metabolism functional categories. The categories of response to stress and response to biotic stimuli appeared to be significantly more highly represented in the mature bud library, confirming data presented previously (Pacey-Miller et al. 2003). On the other hand, signal transduction, carbon utilization, photosynthesis, regulation of metabolism, secondary metabolism, cell cycle and cell division were more intensively represented in the library prepared from young buds.
The abundance of a particular transcript in an EST collection from a non-normalized cDNA library can be used to estimate its expression level in the tissue from which the library originated (Fei et al. 2004). Following this rationale, Keilin and co-workers (2007) detected significant differences in the number of clones between HC and control libraries for a collection of unigenes. Detailed expression analysis was reported for seven selected genes with increased expression and two genes with decreased expression at 6, 12, 24, 48 and 96 h post-HC treatment, validating the 'virtual' blot and tracking changes at the early stages of the bud response.
DEHYDRIN expression was shut down following application of HC, in agreement with the profile detected during the natural dormancy cycle (Pacey-Miller et al. 2003). Dehydrins are normally synthesized during cellular dehydration; they are present in mature seeds and buds exposed to chilling and they disappear once growth resumes (Lang. 1994, Muthalif and Rowland 1994, Salzman et al. 1996, Faust et al. 1997, Rinne et al. 1999, Welling et al. 2002). The observed reduction in transcription level supports the putative effect of HC on growth resumption as early as 48 h after its application, and suggests that DEHYDRIN transcript level is controlled by dormancy-status regulators that might be induced by stimuli other than chilling temperature (Keilin et al. 2007).
Two oxidative-stress-related genes, thioredoxin h (TRXH) and GST, were markedly induced following HC application, supporting the hypothesis that HC application leads to the development of oxidative stress. Thioredoxins play a role in redox regulation of target enzymes and transcription factors. They also serve as hydrogen donors to peroxiredoxins which reduce hydroperoxides to water (Jacquot et al. 2002, Sweetlove et al. 2002, Marchand et al. 2004, Serrato et al. 2004). Accordingly, induction of TRXH expression has been reported in response to oxidative stress (Laloi et al. 2004, Serrato et al. 2004, Rey et al.
2005) and TRXH has been shown to localize to the mitochondria and regulate the activity of the alternative oxidase (AOX; Gelhaye et al. 2004).
Analysis of the expression profile of GST supplied solid and controlled confirmation of data reported by Pacey-Miller et al. (2003). Among other functions, GSTs serve as GPXs, reducing the organic hydroperoxides produced during oxidative stress. They also function as dehydroascorbate reductases (DHARs), reducing dehydroascorbate to ascorbic acid to allow proper function of the ascorbate-glutathione cycle, a major player in the antioxidant defense system (Edwards and Dixon 2005). Accordingly, expression of GST1 transcript in Arabidopsis is induced by oxidative stress (Reuber et al. 1998, Kliebenstein et al. 1999), and import of the major cytosolic GST into mitochondria was demonstrated in mouse cells following oxidative stress (Raza et al. 2002). Evidence from mammalian systems strongly suggests that GSTs play an important physiological role in the regulation of cellular signalling processes (Awasthi et al. 2005).
Among the other genes that were significantly induced were those encoding for Rab GTP-binding protein, actin depolymerization factor (ADF), ADP-ribosylation factor (ARF), calmodulin-binding protein (CBP) and polyubiquitin (Keilin et al. 2007). Interestingly, analysis of the impact of oxidative stress on the proteome of plant mitochondria revealed a similar combination of induced genes, including those encoding for disulfide isomerase, peroxiredoxin, GST, ADF and the endoplasmic reticulum-based Ca2+-sequestering protein cal-reticulin (Sweetlove et al. 2002). Differential expression of ADF has also been associated with cold hardiness in wheat (Ouellet et al. 2001). The combined induction of ADF, ARF, and small GTPases led Keilin and coworkers (2007) to raise the possibility that the intracellular vesicle trafficking system plays a role in dormancy release. Interestingly, involvement of small GTPase in the regulation of actin dynamics and in the modulation of hydroperoxide production has been previously reported during polar cell growth and in response to abiotic stress (Mazel et al. 2004, Molendijk et al. 2004).
Another gene that was clearly induced in HC-treated buds was annotated as sucrose synthase (SuSy), catalyzing sucrose cleavage to fructose and UDP-glucose. The functional significance of SuSy is particularly important under low oxygen conditions and its requirement for survival under such conditions was shown using SuSy-deficient mutants (Subbaiah and Sachs 2003, Koch 2004). SuSy up-regulation is also evident in response to cold shock, water deprivation, anoxia and salt stress (Marana et al. 1990, Baud et al. 2004, Subbaiah and Sachs 2003, Gu et al. 2004). A potential role for SuSy in dormancy release may be inferred from findings connecting sugar-sensing systems and SuSy to regulation of plant-hormone biosynthesis and sensing, and to meristem development. In this context, it is interesting to note that the SuSy gene is one of the first genes induced as leaf primordia differentiate from the apical meristem (Pien et al. 2001, Koch 2004).
5.2.3. The potential of transcriptome profiling to reveal factors potentially involved in dormancy release: possible role of calcium signalling in bud dormancy release
Under stress conditions upregulation of SuSy is accompanied by calmodulin upregulation and potentially mediated by Ca2+ (Subbaiah and Sachs 2003, Gu et al. 2004). Interestingly, a Ca2+-binding protein (CBP) that interacts with calmodulin in a Ca2+-dependent manner was markedly induced following HC application (Keilin et al. 2007). Thus, the possibility that Ca2+-signalling is involved in grape-bud dormancy release and that cytosolic Ca2+ might be a transducer of the HC signal was proposed. This assumption may be supported by the findings that oxidative stress induces Ca2+ signalling in plant cells (Price et al. 1994, Pei et al. 2000, Foreman et al. 2003, Rentel and Knight 2004), and that cytosolic Ca2+ is a transducer of low oxygen signals in maize seedlings, leading to changes in the expression of various genes including SuSy (Subbaiah and Sachs 2003).
The involvement of Ca2+/calmodulin-dependent NAD+ kinase (Gallais et al. 2000, 2001), calcineurin B-like protein kinase (Kim et al. 2003, Pandey et al. 2004) and Ca2+-dependent protein kinase (CDPK) (Anil et al. 2000) in seed germination has been documented, suggesting a mechanistic role for Ca2+-signalling in seed germination/dormancy release. However, evidence for the involvement of Ca2+ signalling in bud dormancy release was not reported until recently. In their report, Pang and co-workers (2007) showed HC-induced expression of Ca2+-ATPase, which in turn could be an indication that HC treatment evokes an increase in Ca2+cyt. Similar induction was confirmed for calmodulin, CBP and CDPK. Both the Ca2+-channel blocker LaCl3 and the calcium chelator EGTA blocked the inductive effect of HC, and their inhibitory effect was removed by supplying exogenous Ca2+. Calcium-dependent histone phosphoryla-tion was up to 70% higher in HC-treated buds. Endogenous protein-phosphorylation assays detected a 47-kD protein that presented strong and Ca2+-dependent phosphorylation only in HC-treated buds. The potential role of a CDPK in the phosphorylation of this protein was supported by an immunopre-cipitation assay. Based on these data, the authors suggested that calcium signalling is involved in the mechanism of bud dormancy release (Pang et al. 2007).
Was this article helpful?