Mitochondria Change Dynamics and Morphology during Grapevine Leaf Senescence

Leaf senescence is the last stage of development of an organ and is aimed to its ordered disassembly and nutrient reallocation. Whereas chlorophyll gradually degrades during senescence in leaves, mitochondria need to maintain active to sustain the energy demands of senescing cells. Here we analysed the motility and morphology of mitochondria in different stages of senescence in leaves of grapevine (Vitis vinifera), by stably expressing a GFP (green fluorescent protein) reporter targeted to these organelles. Results show that mitochondria were less dynamic and markedly changed morphology during senescence, passing from the elongated, branched structures found in mature leaves to enlarged and sparse organelles in senescent leaves. Progression of senescence in leaves was not synchronous, since changes in mitochondria from stomata were delayed. Mitochondrial morphology was also analysed in grapevine cell cultures. Mitochondria from cells at the end of their growth curve resembled those from senescing leaves, suggesting that cell cultures might represent a useful model system for senescence. Additionally, senescence-associated mitochondrial changes were observed in plants treated with high concentrations of cytokinins. Overall, morphology and dynamics of mitochondria might represent a reliable senescence marker for plant cells.


Introduction
Senescence defines the final stage of development of any organ. In leaves, senescence is particularly spectacular since it is characterized by chlorophyll degradation and the display of other pigments. Senescence is aimed to an ordered disassembly of the cellular components and recycling of nutrients that are exported to other plant organs. Eventually, senescence terminates with a genetically programmed cell death (PCD) [1]. Senescence is a complex physiological process whose onset and regulation is under control of several internal (hormones, plant stage and organ age) and external (light, nutrient availability, biotic and abiotic stresses) stimuli. The complexity of the process is revealed by the change in expression of thousand genes and the activation of many signalling pathways [2].
Since senescence is an active and slow process, cells require energy to stay alive until the onset of PCD. Early in the senescence process, the photosynthetic capacity of leaves drops, due to chlorophyll degradation. The energy demand of senescing cells therefore relies completely on mitochondria [3]. Moreover, besides their central role in energy and carbon metabolism [4], mitochondria appear to play a significant role as stress sensors and signal dispatchers during PCD events. In animals, but also in plants, mitochondria regulate PCD by releasing pro-apoptotic factors such as cytochrome c and possibly also through their dynamics and morphology alterations [5]. Although mitochondria are often portrayed as static, oval or rod-shaped organelles, recent studies have demonstrated that their shape and distribution are variable [6]. Changes in the distribution pattern of mitochondria and their movement appear to be critical in executing their cellular functions. A link between PCD and plant mitochondrial dynamics and morphology alterations has been already reported in protoplasts of Arabidopsis [7], during dark-induced senescence in Arabidopsis leaves [3] and in cell cultures of Medicago truncatula [8] experiencing starvation or treatment with high levels of cytokinins, treatments able to trigger senescence-like processes [9][10][11][12][13]. However, despite their utility as a simplified model, protoplasts and cell cultures might not cover all the aspects of a complex developmental process such as leaf senescence.
Here we investigated whether the shape and dynamics of mitochondria in leaves of grapevine (Vitis vinifera) change during senescence. Grapevine is an important crop species and a model for deciduous perennial plants, characterized by a full seasonal senescing syndrome and whose genome has been fully sequenced and annotated [14][15][16]. We also extended our observations to leaves treated with high concentrations of cytokinins and to starved cell cultures, in order to test whether these senescence-like processes resembled spontaneous leaf senescence in terms of mitochondrial modifications.

Materials and Methods
Cloning For the expression of the GFP targeted to mitochondria, the b-GFP coding sequence was subcloned from the b-GFP plasmid [17,18] into the pBI121 binary vector (Clontech Laboratories, USA), by replacing the b-glucuronidase coding sequence into the BglII/SacI restriction sites as described by Zottini et al. [19]. The pBI121 binary vector harbouring the neomycin phosphotransferase II coding sequence allowed the selection of kanamycin resistant transgenic cells.
Competent cells of Agrobacterium tumefaciens GV3101 strain (resistant to rifampicin and gentamycin) were prepared according to Sambrook et al. [20] and the pBI121 b-GFP binary vector was introduced by electroporation as reported by Zottini et al. [19]. Bacterial growth was optimized by using YEP medium (10 g/L Bacto-Trypton, 10 g/L yeast extract, 5 g/L NaCl; pH 7.0). The medium was supplemented with appropriate antibiotic selection (100 mg/mL rifampicin, 50 mg/mL gentamycin, 50 mg/mL kanamycin).
The transformation procedure was performed as described below. Cell suspensions were initiated from habituated embryogenic cultures by transferring 1 g of pre-embryogenic masses collected from solid MS-to Erlenmeyer flasks (250 mL) filled with 50 mL liquid MS-medium. The flasks were cultured for 3 days on an orbital shaker at 80 rpm and incubated at 25uC in the dark. Before transformation with A. tumefaciens, 500 mg of pre-embryogenic masses and embryos were transferred into a Petri dish contained 1 mL of induction liquid medium (ILM, i.e. NN medium supplemented with 58 mM sucrose and 2.26 mM 2,4-D) and incised with a sharp razor blade.
Bacteria suspension was prepared as in Zottini et al. [19]. A. tumefaciens suspension was diluted to OD 550 0.5 in ILM and added to the dissected embryos previously transferred to bacteria-free ILM. Embryos were incubated at room temperature in the dark for 10 min, after which the cultures were washed 5 times with ILM (3 min each). Infected embryos were blotted dry on sterile filter paper and then plated on NN solid medium and incubated at 25uC in the dark. Two days later, the cultures were transferred to NN solid medium supplemented with 300 mg/mL cefotaxime and maintained at 25uC in the dark. After 10 days the cultures were transferred on NN solid medium supplemented with 20 mg/mL kanamycin and 300 mg/mL cefotaxime. After 20 days the cultures were transferred on NN solid medium supplemented with specific antibiotics (40 mg/mL kanamycin and 300 mg/mL cefotaxime) and subcultured at 20 day-intervals. Embryo clusters were collected and transferred to NN hormone free solid medium for germination. Each somatic embryo was transferred to Microbox Containers (Duchefa) on solid (8 g/L plant agar, Duchefa) half strength MS medium, supplemented with 44 mM sucrose.

Cytokinin treatments
Stem cuttings, containing an internode, a node and a still expanding leaf, were dissected from hydroponic grown plants. Stem cuttings were transferred into plastic tubes containing 6 mL hydroponic solution and maintained in a climate growth chamber at 25uC under a 16 h light photoperiod. Leaves were analysed 12 days after treatment with 100 mM BA.

Grapevine suspension cell cultures
Establishment and maintenance of suspension cell cultures are described in Zottini et al. [19]. Briefly, grapevine cell lines were obtained from leaf disk explants incubated in solid B5F medium. After several subculture cycles, aliquots of callus were inoculated into liquid B5F medium. Every week, 2 mL of suspension cell cultures were transferred to Erlenmeyer flasks (250 mL) filled with 50 mL liquid B5F medium. The suspension cultures were maintained in a climate growth chamber at 2561uC on an orbital shaker (80 rpm) under a long photoperiod (16 h light and 8 h dark). To determine fresh weight, intact cells were separated from culture medium and cell debris through a vacuum filtration unit (Sartorius, Florence, Italy).

DNA analysis
DNA was isolated from V. vinifera cell cultures as described in [9]. For DNA fragmentation analysis, 10 mg of each sample DNA were electrophoresed on a 1% (w/v) agarose gel containing TAE buffer (40 mM Tris-acetate, 1 mM EDTA) and stained with ethidium bromide.

Analyses of chlorophyll content and photosynthetic efficiency
Chlorophylls were extracted from leaf tissues by N,N-dimethylformamide and analysed in a double-beam spectrophotometer (GBC UV/VIS 918). Pigment concentrations were calculated using the extinction coefficients of Porra et al. [25].
In vivo, chlorophyll fluorescence was measured at room temperature with a Dual PAM-100 (Walz) fluorometer, with a saturating light at 3000 mE m 22 s 21 . Before measurements, plants were dark-adapted for 40 min at room temperature. The parameter Fv/Fm was calculated as the variable (Fm-Fo) over the maximum fluorescence (Fm) [26]. Measurements were replicated at least 3 times.

Analyses of mitochondrial morphology, volume and number in leaves
Transformed leaves were analysed using a Nikon PCM2000 laser scanning confocal imaging system. For GFP detection, excitation was at 488 nm and emission between 515/530 nm. For the quantification of the number and volume of mitochondria, confocal Z-sections along the leaf thickness were acquired. These analyses were performed on the leaf tips at the different senescence stages (M mature leaves; S1, S2, S3 subsequent senescence leaf stages; 10 replicates for each stage). Image analysis was performed by ImageJ bundle software (http://rsb.info.nih.gov/ij/). Threedimensional reconstructions of confocal microscope laser scanning images of leaf tissue were obtained by the Volocity 3D Image Analysis Software (PerkinElmer, UK).

Mitochondria tracking
Time-lapse images of leaf tissues expressing the mitochondriatargeted GFP were acquired every 1.6 s. Size, shape, and intensity parameters were specified so that most organelles would be defined by the software. Tracks combining 12-20 sequential images were calculated by the shortest path model, and mean velocity was calculated for each track. Organelles were tracked using the Volocity 6.0 (PerkinElmer, UK) track utility, and the velocity was calculated for 100-300 organelles/experiment. At least three biological replicates were analysed.

Analysis of mitochondrial morphology in cell culture
A Nikon PCM2000 laser scanning confocal microscope was used to analyse mitochondrial morphology. The tetramethylrhodamine methyl ester (TMRM) dye (Molecular Probes, Leiden, the Netherlands), a mitochondrial membrane potential sensor, was used to visualise mitochondria in cell cultures as described by Zottini et al. [8]. Cell suspensions (300 mL) were collected at different times during their growth cycle, and incubated for 15 min on a rotary shaker in 700 mL B5F medium containing 1 mM TMRM. Cells were centrifuged for 3 min at 10,0006g, the supernatant was discarded and the pellet washed twice with 700 mL B5F. Cells were then resuspended in 500 mL B5F. For microscopy analysis, 100 mL cell suspensions were placed on a microscope slide and visualised under a confocal microscope (excitation 548 nm, emission 573 nm). Images were processed using ImageJ bundle software (http://rsb.info.nih.gov/ij/). For mitochondrial morphology experiments, a randomized complete block design was used with three replicates (individual Erlenmeyer flasks). Each experiment was repeated three times.

Cell viability assay
Cell viability was determined by the fluorescein diacetate (FDA, SIGMA) method according to Amano et al. [27]. Immediately before each assay, a stock solution of FDA (0.5% w/v in acetone) was diluted with distilled water to create a fresh 0.01% w/v FDA working solution kept in the dark at 4uC. Grapevine cell suspensions were aliquoted in 2 mL cell culture package on Poly-Prep chromatography columns (BioRad) and then diluted 1/ 10 with PBS (2.7 mM KCl, 137 mM NaCl, 1.8 mM KH 2 PO 4 , 4.0 mM Na 2 HPO 4 ). One hundred mL of this solution were mixed by gentle stirring with 0.01% w/v FDA in a quartz cuvette. A spectrophotofluorimeter (Perkin Elmer, UK) equipped with a stirrer was employed. Excitation and emission wavelengths were selected at 493 and 510 nm, respectively. The increase in fluorescence was recorded over a 120 s time period. The slope

Statistical analysis
Data are presented as mean and standard deviation of the biological and technical replicates. Differences were tested by unpaired, two-tailed T test. Significant differences by p,5% or p,1% are indicated by one or two asterisks, respectively.

Physiological and molecular characterization of leaf senescence
Leaf senescence was assessed for three months in grapevine plants grown in vitro. Following a common classification for senescent leaves [28,29], we identified four different stages, based on the progressive chlorophyll loss. Expanded green leaves (M) had the maximum chlorophyll content, whereas in the subsequent senescent stages S1 (30 day-old), S2 (60 day-old) and S3 (90 dayold) the chlorophyll loss was 17, 47 and 90%, respectively ( Figure 1A, B). Chlorophyll degradation was accompanied by a decrease in the photosynthetic capacity, another parameter useful in identifying senescence stages [30]. Leaves remain active during the first stages of senescence, as evidenced by the stable values of photosynthetic efficiency (Fv/Fm) of photosystem II (PSII) until stage S2. In S3, photosynthetic efficiency dropped, in agreement with the low chlorophyll content measured in this stage ( Figure 1B).
For the molecular characterization, the transcriptional profiles of known molecular markers of senescence, namely VvNAM, VvSAG12 and VvSAG13 genes [31], normalised against the expression of the housekeeping actin gene, were analysed. During senescence, VvNAM and VvSAG12 expression levels increased from stage M to S2, and then declined. The transcript level of VvSAG13, instead, constantly increased until stage S3 ( Figure 1C).

Analyses of mitochondrial morphology and dynamics during leaf senescence
Mitochondria were visualized by stable expression of the fluorescent reporter GFP carrying the sequence tag for mitochondrial localization (b-GFP) (see Figure S1 for details on transformation procedures). At stage M, mitochondria appeared abundant and inter-connected in long, branched structures, homogeneously distributed throughout the cytoplasm (Figure 2A). In senescent leaves (S2 stage), mitochondria were scarce, isolated and had a different morphology, characterized by a large, round shape ( Figure 2B). Since senescence seems to follow different routes in mesophyll and stomata cells [32,33,3], we analysed mitochondria morphology in details in these cell types ( Figure 3A,B). The decrease in the mitochondria number during senescence was more pronounced in mesophyll cells (decrease of 25% at stage S3 as compared to M, Figure 3C) than in stomata (decrease of 50%, at stage S3 as compared to M, Figure 3D). In both cell types, the mean volume of mitochondria increased at S2 and then decreased again during S3 ( Figure 3E,F). If we consider the total mitochondrial volume, measured in the different stages of leaf development, we observed that is was substantially unchanged up to the S2, while in the S3 stage the total volume and the total number of mitochondria drastically decreased.
In order to determine whether mitochondrial movement was different in mature and senescent epidermal leaf tissues, dynamics of mitochondria was analysed in several time-lapse movies (Movie S1; Movie S2). Figure 4 (A) shows the distribution of mitochondria velocity (mm/sec), measured in both mature and senescent leaves. It is noteworthy that the majority of senescent leaf mitochondria were slow (54% with #0.2 mm/sec), while a broad distribution of mitochondria velocity was observed in mature leaves (in mature leaves the average and median of mitochondria velocity were 0,64 and 0.51 mm/sec respectively, while in senescent leaves average and median were 0,32 and 0.22 mm/sec respectively). Figure 4 (B,C) describes the movement patterns of mitochondria populations in mature (B) and senescent (C) leaves, reported as displacement rates. Also in this case, two different behaviours are evident, with mitochondria from senescent leaves less dynamic than those from mature leaves.

Effects of high levels of the cytokinin on leaf senescence
Senescence is a slow process, under the influence of several endogenous and environmental stimuli. Under certain conditions, young and mature leaves can be induced to undergo a process of accelerated senescence and PCD. We had previously reported that high concentrations of cytokinins are able to induce a senescencelike process both in plants and cell cultures [9,10]. In order to check whether mitochondria alterations observed during spontaneous senescence were also detectable when senescence was artificially induced, grapevine stem cuttings treated with high levels of cytokinins were analysed. Leaf treatment with 100 mM BA affected chlorophyll content (32% loss) and photosynthetic efficiency (16% decrease of Fv/Fm) as compared to untreated leaves ( Figure 5A,B), and induced a strong increase of the senescence markers VvSAG12 and VvSAG13 expression ( Figure 5C). Altogether, these results show that high concentrations of the cytokinin BA induced a senescence-like process in grapevine leaves. Mitochondria abundance and morphology were then analysed both in stomata and mesophyll cells ( Figure 5D-G). As expected, mitochondria were abundant, elongated and forming dynamic branched structures in untreated leaves ( Figure 5D,E; Movie S3). Cytokinin treatment resulted in dramatic alterations in mitochondrial morphology and dynamics. In presence of 100 mM BA, the branched structures disappeared, and mitochondria appeared enlarged and round-shaped (Figure 5D,E,G). In parallel, cytokinin treatment caused a significant reduction in the mitochondrial number in the mesophyll cells (Fig 5F). Mitochondria dynamics was also analysed in leaf epidermal cells upon cytokinin treatment and a strong delay in velocity was detected and measured as shown in Figure 5H and Movie S4 (the average and median of mitochondria velocity in untreated leaves were 0.34 and 0.18 mm/sec, respectively; while in senescent leaves were 0.22 and 0.10 mm/sec, respectively).

Analyses of mitochondrial morphology in suspension cell cultures
Senescence is a complex phenomenon, involving nutrient recycling between separate plant organs. Even within a single leaf, cell types are affected differently during senescence, as shown for stomata and mesophyll cells. In order to reduce this source of variability, it is convenient to have a study system with a reduced complexity. Cell cultures present several advantages as model system, since they are homogeneous, fast-growing and chemical treatments are easily applied. In order to test whether the changes associated with leaf senescence could be detectable in cell cultures, we established and characterized a cell line from grapevine leaves. Suspension cell cultures actively proliferated during the first 11 days after a subculture ( Figure 6A) and then progressively declined and died due to nutrient depletion ( Figure 6B). We previously reported that starvation triggers a senescence-like PCD process in Arabidopsis cell cultures [10,11]. In order to test whether grapevine cell cultures had features similar to leaf senescence, we evaluated the expression pattern of VvNAM, VvSAG12 and VvSAG13. The senescence gene markers were expressed at low levels in cells at mid-exponential growth phase (7 days after culture initiation), but strongly increased in the decline phase ( Figure 6C; 18 days), confirming that cell cultures are a suitable model to study certain aspects of leaf senescence.
In addition, the cell DNA integrity was measured as a marker of PCD. As reported in Fig. 6D, the typical DNA laddering pattern was observed in DNA from 18 day old cells, when the percentage of cell death was 55%. This result is in agreement with previous data reported in Arabidopsis cultured cells, where it has been shown that DNA laddering was detectable only when the level of cell death reached 40% [11].
Mitochondrial morphology was analysed at different times of the sub-culture cycle through tetramethylrhodamine methyl ester (TMRM) staining ( Figure 6E,F). At the initial phase of the growth curve (4 days after culture initiation), mitochondria were punctiform, abundant and homogeneously distributed into the cytoplasm ( Figure 6E,F; 4 days). At the mid-exponential phase of growth (7 days), elongated and branched mitochondrial structures were detected ( Figure 6E,F). At 18 days, in advanced senescence phase (55% of cells were already dead, Figure 6B), mitochondria were scarce and appeared as giant spherical or ring-shaped organelles ( Figure 6F; 18 days). Quantification of mitochondria number and volume in senescing cells showed a ten-fold reduction of mitochondria number while the volume increased more than four-fold ( Figure 6G, H).

Discussion
Senescence is a fundamental phase in plant organ development and, as such, has strong influence on crop quality, productivity and fruit shelf-life. Understanding the molecular mechanisms underlying senescence could provide new tools for enhanced crop management and breeding. Control of senescence might result in increased photosynthetic lifetime of leaves and augmented biomass. For fruit crops, senescence affects source-sink relationship through nutrient recycling. Moreover, fruit ripening itself can be considered a type of senescence. V. vinifera, a perennial fruit crop with a full seasonal leaf senescence syndrome, and a sequenced, annotated genome, represents therefore an optimal model species for studying senescence.
In the present work, we first characterized leaf senescence in grapevine by physiological (chlorophyll content, photosynthetic efficiency) and molecular (pattern of expression of senescence marker genes) analyses. We then focused on the alterations of mitochondria morphology and dynamics during senescence. Transgenic grapevine plants harbouring mitochondria-targeted GFP represent a new, useful tool for studying the role of these organelles in woody plants. Stable expression lines avoided the artefacts induced by agroinfiltration procedure, due to wounding and heterogeneous transformation events. Moreover, it allowed the long-term assessment of mitochondrial pattern evolution during the slow process of spontaneous leaf senescence.
Our analyses revealed that mitochondria underwent dramatic changes during leaf senescence. In mature leaves, mitochondria appeared as long, branched structures formed by inter-connected mitochondria. These structures gradually evolved, along senescence, toward enlarged, round-shaped mitochondria, and in the meantime the total number of organelles significantly decreased. These alterations were particularly evident in the S2 senescence phase, when leaves had lost about 50% of their chlorophyll content and the expression of senescence marker genes, coding for proteins involved in nutrient recycling [31], highly increased.
Basing on our previous studies, we also show that treatment of grapevine cuttings with high levels of the cytokinin BA induced an accelerated process of leaf senescence and affected mitochondrial morphology similarly to spontaneous senescence. This experimental approach provides therefore a rapid, reproducible system in which senescence is induced in a short and defined period of time. It is worth mentioning that the induction of senescence by cytokinins was stage dependent. In fact, BA was effective on expanding leaves but not on mature, completely expanded ones (results non reported), suggesting that in leaves, as previously shown in cell cultures [9], a sort of cell/tissue competence dependent on the developmental stage is required.
An interesting observation of the present work is that mitochondria behave slightly different in stomata and mesophyll cells during senescence. In particular, mitochondria appeared more structured in stomata than in mesophyll cells (Figure 2 and  3), possibly because in stomata, senescence is not as orderly paced as in mesophyll cells [33]. The apparent delay in the onset of senescence in guard cells might be instrumental in ensuring functional stomata and gas exchange until the senescence process undergoing in the other leaves cells is completed.
Our observations indicate that timing and progression of senescence differ among cell types in leaves. This heterogeneity, and the experimental difficulties that characterize the studies on perennial woody species, make cell cultures an attractive system. Being homogeneous, fast growing, easy to handle and readily accessible to treatments, cell cultures might represent an ideal study system. Previous observations from our laboratory had shown that mitochondria undergo modifications in their morphology along the growth curve of a cell culture of M. truncatula [8]. This prompted us to look for a correlation between leaf senescence and nutrient depletion in cell cultures of grapevine. Analysis of the expression pattern of three senescence marker genes, detected in the final phase of cell growth curve, confirmed that senescence of cell cultures mimics leaf senescence also at a molecular level. Similar alterations in mitochondrial morphology and dynamics were observed in senescing leaves and cell cultures. We foresee that cell cultures can be used as a first, convenient study system, to test the effect of different treatments on senescence.
An important finding of this study is that spontaneous as well cytokinin-induced leaf senescence were accompanied by a general and strong decrease in mitochondria motility, about 40-60% lower than mature and untreated stem cuttings leaves. This effect might depend on the progressive degradation of the cytoskeleton [34]. It is currently unclear whether mitochondria motility controls, or is controlled by, the physiological state of cells. Recent reports suggest that mitochondria motility is correlated to calcium signalling, both in animal [35] and plants [36]. The use of new tools for measuring in vivo mitochondrial calcium concentration and dynamics, such as genetically encoded sensors [37], might contribute to clarify the role of calcium in the regulation of the senescence process. A growing number of genetically encoded sensors for a variety of ions and metabolites have been successfully employed in mitochondria [38]. Since senescence in plants involves nitrogen recycling, the recently developed activity sensors for ammonium transport might provide an interesting tool [39].
In conclusion, here we report alterations in mitochondria number, morphology and dynamics during spontaneous and cytokinin-induced senescence of grapevine leaves and cell cultures. Long, branched mitochondria characterizing mature and early senescent leaves progress into enlarged, less motile mitochondria in the late senescent stages. In this context, mitochondria morphologies represent useful markers to identify different physiological and developmental stages of senescence. Supporting Information Figure S1 Transgenic grapevine embryogenic cell lines stably expressing b-GFP targeted to mitochondria. (A) Epi-fluorescence images of embryogenic callus: starting material for transformation procedure (t 0 ); 10 days after transformation (t 1 ); 3 months after transformation (t 2 ; arrows indicate transformed globular embryos); 4 months after transformation (t 3 ; arrows indicate torpedo embryos). (B) Epi-fluorescence (1, 2) and confocal (3,4)  Movie S1 Movement of b-GFP-labeled mitochondria in leaf at mature M stage. The leaf was analysed by means of confocal laser scanning microscope. The movie was recorded at 1 frame per 1.6 seconds. Playback is at 7.5 frames/second (12x normal speed), so the movie represents 48 seconds of real time. (AVI) Movie S2 Movement of b-GFP-labeled mitochondria in leaf at the senescent S2 stage. The movie was recorded at 1 frame per 1.6 seconds and is playing at 7.5 frames/second, representing 48 seconds of real time (12x normal speed). (AVI) Movie S3 Movement of b-GFP-labeled mitochondria in control leaf of hydroponic grown plant. The movie was recorded at 1 frame per 1.6 seconds and is playing at 7.5 frames/second, representing 48 seconds of real time (12x normal speed). (AVI) Movie S4 Movement of b-GFP-labeled mitochondria in leaf of hydroponic grown grapevine plants treated with 100 mM BA. The movie was recorded at 1 frame per 1.6 seconds and is playing at 7.5 frames/second, representing 48 seconds of real time (12x normal speed). (AVI)