Symplastic connection is required for bud outgrowth following dormancy in potato (Solanum tuberosum L.) tubers



    1. Unit of Plant Biochemistry and
    Search for more papers by this author
    • *

      Present address: Istituto Agrario San Michele all'Adige (IASMA), Via E. Mach, S. Michele all'Adige, I38010, Trento, Italy.


    1. Unit of Plant Biochemistry and
    Search for more papers by this author
    • Present address: Groupe de Génomique Fonctionnelle des Plantes, EA3900 Université de Picardie 33, Rue St Leu, 80039 Amiens Cedex, France.


    1. Unit of Plant Biochemistry and
    Search for more papers by this author
    • Present address: Horticultural Production Chains Group, Wageningen University, Marijkeweg 22, 6709 PG Wageningen, the Netherlands.


    1. Programme of Plant Pathology, Scottish Crop Research Institute, Invergowrie, Dundee DD2 5DA, UK
    Search for more papers by this author
    • §

      Present address: Centre for Integrative Physiology, School of Biomedical Sciences, University of Edinburgh, Hugh Robson Building, George Square, Edinburgh EH8 9XD, UK.


    1. Unit of Plant Biochemistry and
    Search for more papers by this author

    1. Programme of Plant Pathology, Scottish Crop Research Institute, Invergowrie, Dundee DD2 5DA, UK
    Search for more papers by this author

    Corresponding author
    1. Unit of Plant Biochemistry and
      R. D. Hancock. Fax: +44 1382 568 575; e-mail:
    Search for more papers by this author

R. D. Hancock. Fax: +44 1382 568 575; e-mail:


To gain greater insight into the mechanism of dormancy release in the potato tuber, an investigation into physiological and biochemical changes in tuber and bud tissues during the transition from bud dormancy (immediately after harvest) to active bud growth was undertaken. Within the tuber, a rapid shift from storage metabolism (starch synthesis) to reserve mobilization within days of detachment from the mother plant suggested transition from sink to source. Over the same period, a shift in the pattern of [U-14C]sucrose uptake by tuber discs from diffuse to punctate accumulation was consistent with a transition from phloem unloading to phloem loading within the tuber parenchyma. There were no gross differences in metabolic capacity between resting and actively growing tuber buds as determined by [U-14C]glucose labelling. However, marked differences in metabolite pools were observed with large increases in starch and sucrose, and the accumulation of several organic acids in growing buds. Carboxyfluorescein labelling of tubers clearly demonstrated strong symplastic connection in actively growing buds and symplastic isolation in resting buds. It is proposed that potato tubers rapidly undergo metabolic transitions consistent with bud outgrowth; however, growth is initially prevented by substrate limitation mediated via symplastic isolation.


The potato plant propagates vegetatively via tubers, storage organs that form at the tips of underground stems (stolons). The process of tuberization involves the cessation of growth in the stolon apical meristem and the induction of first longitudinal and later random cell division and expansion in the subapical region (Xu, Vreugdenhil & van Lammeren 1998). This is accompanied by massive deposition of storage carbohydrates (Visser et al. 1994) and proteins (Shewry 2003) in the tubers. After harvest, tuber buds are generally dormant and will not grow even if the tubers are placed under optimal conditions for sprouting (i.e. warm temperature, darkness, high humidity). Following a period of ‘rest’ of between 1 and 15 weeks (Wiltshire & Cobb 1996), dormancy is broken and buds start to grow (conventionally, a tuber is defined as sprouting if it contains one or more buds of 2 mm length or greater). The duration of the dormancy period is primarily dependent on the genotype, but other factors such as the growth conditions of the crop and storage conditions after tuber harvest are also important (Turnbull & Hanke 1985; Wiltshire & Cobb 1996; Fernie & Willmitzer 2001; Sonnewald 2001).

Although potato tuber dormancy has been studied extensively, there is still a poor understanding of the mechanisms controlling the process. For example, it is not clear whether the growth status of buds is controlled from within the buds (endodormancy) or by the rest of the tuber (paradormancy). Excision of buds from dormant tubers inevitably leads to bud growth (Rappaport et al. 1965), and several investigators have suggested that tuber dormancy is regulated by changes in the content of individual growth hormones (Turnbull & Hanke 1985; Suttle 2004a) or by their relative balance (Hemberg 1985) within the tubers. However, the situation is complicated by the production of growth-promoting substances [e.g. gibberellins (Rappaport & Sachs 1967)] in buds following excision, and it is known that gibberellins and other hormones induce dormancy break even when applied exogenously (Rappaport, Lippert & Timm 1957). Interpretation is further complicated by the observation that potato tissues have a differential sensitivity to hormones dependent on time from harvest (Suttle 2004b). Despite difficulties associated with study of the hormonal control of tuber dormancy, consensus is starting to emerge regarding the role of individual hormones (Suttle 2004a). Recent analysis of temporal changes in endogenous hormone levels provides strong evidence for a role of abscisic acid in maintaining dormancy (Destefano-Beltrán et al. 2006), while experiments with transgenic tubers over-expressing an enzyme for cytokinin synthesis suggest a role for this compound in dormancy release (Zubko et al. 2005). Gibberellins appear to be of little importance in dormancy release, but play a role in the control of subsequent sprout growth (Suttle 2004b).

In addition to the synthesis or breakdown of growth regulators, dormancy may also result from the absence of specific physiological or biochemical competences in the tubers that are required for bud growth or phytohormone perception (Sonnewald 2001). During its life cycle, the tuber undergoes a functional transition from an active sink for assimilates to a source of nutrients for the developing sprouts. During tuber development, the storage parenchyma actively converts incoming soluble assimilates (e.g. sucrose, amino acids) into polymeric reserves (starch and protein) (Prat et al. 1990; Visser et al. 1994; Fernie & Willmitzer 2001). At maturity, over 70% of the tuber carbohydrate reserves are sequestered as starch, which must be converted into transport-compatible solutes for sustaining sprout growth. The development of functional competences for the mobilization and transport of carbohydrates and other nutrients from the storage parenchyma into the growing buds is requisite for the sink–source transition in the tuber. Sink activation (bud growth) may precede or follow the development of functional source competence within the tuber. However, complete dormancy break can be defined as the establishment of sink–source relationships between the bud (sink) and the storage parenchyma (source), and completion of the systemic transition within the tuber.

The available information on physiological and biochemical properties of individual tuber tissues and organs during the sink–source transition is very scarce. In dormant tubers, cell division in the buds is absent and most cells in the meristem are arrested primarily in the G1 state (Campbell, Suttle & Sell 1996). In dormant buds, DNA synthesis is virtually absent (MacDonald & Osborne 1988), while RNA and protein biosynthesis progress at a highly reduced rate (Korableva & Ladyzhenskaya 1995). However, there is no information on metabolic properties of tuber buds at different developmental stages, and no biochemical markers of sink activity within the buds have been identified although some studies have examined gene expression during bud meristem activation (Bachem et al. 2000) and isolated potential molecular markers (Faivre-Rampant et al. 2004). There is also little information on the development of source competence in the storage parenchyma as, in particular, the enzymology of starch degradation in this tissue is still ill defined (Smith, Zeeman & Smith 2005). Potato tubers are known to contain phosphorolytic and hydrolytic starch-degrading enzymes throughout their life cycle (Davies & Ross 1987), and total phosphorolytic and amylolytic activities actually decline during potato sprouting. It is clear that some net starch degradation occurs in stored tubers for sustaining maintenance metabolism and sugar accumulation at low temperatures (Wiltshire & Cobb 1996), but the key players in this process have not yet been unequivocally identified although the recent identification of quantitative trait loci (QTLs) associated with low temperature sweetening combined with a candidate gene approach should allow rapid progress (Menéndez et al. 2002). Other metabolic changes are known to occur in the tuber storage parenchyma during the tuber sink–source transition. For example, storage parenchyma cells of developing tubers possess a sucrose uptake capacity (functionally, an apoplastic sucrose retrieval mechanism), which is carrier mediated and turgor sensitive (Oparka & Wright 1988). On the other hand, sucrose uptake by storage parenchyma of sprouting tubers is essentially diffusional, largely unaffected by both the sulphydryl reagent p-chloromercuribenzene sulphonic acid (PCMBS) and the protonophore carbonylcyanide m-chlorophenylhydrazone (CCCP), and insensitive to turgor (Wright & Oparka 1989). This functional change is assumed to allow sucrose diffusion into the apoplast in source tubers, thus making it available for phloem uptake and transport to the buds. However, active phloem uptake of sucrose in source tubers has not been demonstrated.

In this study, physiological and biochemical evidence for a rapid transition from sink to source in the potato tuber parenchyma following detachment of the tuber from the mother plant is presented. Furthermore, evidence is provided that dormant tuber buds are fully metabolically competent but growth is substrate limited. The data presented suggest that substrate limitation is the result of symplastic isolation in the tuber bud, and bud outgrowth is associated with symplastic reconnection. Taken together with previous data regarding a shift from apoplastic to symplastic phloem unloading during tuber development (Viola et al. 2001), these data highlight the fundamental importance of plasmodesmatal gating in the potato tuber life cycle.


Plant material

Tubers of Solanum tuberosum L. cv. Desiree were planted in the field in early April. Tuberization initiated approximately 50 d after planting. Experiments were undertaken on developing tubers approximately 70 d after planting. Once plants had reached maturity, tubers were harvested, and tubers of similar size [100 ± 10 g fresh weight (FW)] were stored in the dark at 4 °C. Dormant tubers were classed as those that were less than 4 weeks from harvest and visible bud growth was not observed prior to 8 weeks storage. To study changes in carbohydrate composition and amylolytic enzymes during sprouting, tubers that had been stored for 10 weeks at 4 °C but that showed no signs of visible sprouting were transferred to moist compost and left to sprout at 15 °C in the dark. Tubers were removed at designated intervals, sprout dry weight (DW) was recorded and tuber starch and sugar content and amylase activities were determined as outlined as follows.

Extraction and quantification of metabolites

For sugar and starch analysis in tuber parenchyma, 1 mm slices were prepared from fresh tuber tissue and discs were rapidly cut using a 5 mm cork borer prior to immediate freezing in liquid N2. For the analysis of bud metabolites, buds were excised from tubers and immediately frozen in liquid N2. Tissues were ground to a fine powder under liquid N2, resuspended in hot (80 °C) ethanol (20:1 v/w) and extraction was continued at 80 °C with gentle shaking for 1 h. Samples were centrifuged (10 000 g, 10 min, 1 °C), the supernatant was decanted and the extraction was repeated twice. The three supernatant fractions were combined, reduced to the aqueous phase by evaporation at 80 °C, lyophilized and resuspended in 10 volumes of distilled H2O. This fraction was subsequently used for analysis of glucose, fructose and sucrose according to the method of Viola & Davies (1992) with the modifications described by Mitchell et al. (1998). The cell pellet was resuspended in 9 volumes of distilled water, and starch was gelatinized by vigorous boiling for 60 min. The suspension was cooled to room temperature; 1 volume of 1 m acetate buffer pH 5.2 containing 100 U mL−1 amyloglucosidase (Megazyme, Wicklow, Ireland) was added and incubation was continued at 37 °C for 18 h. After centrifugation, glucose in the supernatant (released from starch) was quantified spectrophotometrically as described (Viola & Davies 1992).

For the extraction of organic acids, tissues were prepared as described earlier and ground in liquid N2. Samples were extracted in 5% (w/v) metaphosphoric acid, and cell debris was removed by centrifugation (16 000 g, 5 min, 1 °C). Supernatants were filtered through 0.45 µm filters, and 20 µL was used for organic acid quantification by high-performance liquid chromatography (HPLC) as previously described (Hancock, Galpin & Viola 2000). Under the conditions used, retention times for oxalate, citrate, malate and succinate were 6.7, 9.8, 11.8 and 14.2 min, respectively.

Determination of β-amylase activity by native gel electrophoresis

β-Amylase activity was determined as described by Hill et al. (1996). Tuber discs were ground to a powder in liquid N2 and extracted 2:1 (v/w) in 100 mmN-2-hydroxyethylpiperazine-N′-2-ethanesulphonic acid (HEPES)–KOH (pH 7.6), 20 mm MgCl2, 2 mm CaCl2, 3% (w/v) polyethyleneglycol 8000 and 14.3 mm 2-mercaptoethanol. Cell debris was removed by centrifugation, and the supernatant was prepared for gel loading by the addition of glycerol to 0.125% (v/v) and bromophenol blue to 0.1% (w/v). Twenty-five microlitres of each sample was loaded onto polyacrylamide gels (30:0.8 acrylamide : bisacrylamide) that consisted of a stacking gel containing 3% (w/v) polyacrylamide and 0.2% (w/v) potato amylopectin in 6.3 mm tris(hydroxymethyl)aminomethane (Tris)–HCl (pH 6.8) and a resolving gel containing 7.5% (w/v) polyacrylamide and 0.2% (w/v) potato amylopectin in 37.5 mm Tris–HCl (pH 8.8). Gels were run at constant current with 25 mm Tris–glycine (pH 8.9) running buffer. After running, gels were equilibrated in 100 mm sodium acetate buffer (pH 6.0), 1 mm CaCl2 and 5 mm dithiothreitol for 3 h, and stained with iodine to detect amylolytic enzyme activities.

Metabolism of d -[U-14C]glucose by potato tuber tissues

Tuber discs were prepared from 1 mm slices using a 5 mm cork borer, and dormant or open buds were excised from appropriate tubers. Tissues were incubated as previously described (Davies et al. 2005) with 0.148 MBq d-[U-14C]glucose for 3 h and extracted in ethanol as described earlier. The soluble fraction was further fractioned into neutral and ionic components by strong anion exchange as previously described (Souleyre et al. 2004). Radioactivity in the CO2, anionic, starch, other insoluble and neutral fractions was determined by liquid scintillation counting after dilution into ScintLogic HiCount cocktail (Lablogic Systems Ltd, Sheffield, UK) using a Packard Tri-Carb 3100TR counter (Packard Instrument Co., Meriden, CT, USA). Radioactivity in sucrose, glucose and fructose was determined by HPLC with flow scintillation detection as described (Davies et al. 2005).

Distribution of [U-14C]sucrose in tuber parenchyma

Slices (1 mm) were prepared from tubers and washed six times in 50 mm methanesulphonic acid (MES)–KOH (pH 6.5), 300 mm mannitol to remove enzymes released by cellular damage. Tissues were then incubated for 2 h in the same buffer containing 0.037 MBq mL−1[U-14C]sucrose (specific activity 17.1 MBq mmol−1), washed three times (5 min) in fresh buffer without sucrose and patted dry with paper towels. Samples were then placed between photographic paper (gloss side towards tissue) and gently pressed between two aluminium plates held in place with a G-clamp. Samples were rapidly frozen in liquid N2 and lyophilized. After lyophilization, samples were gently removed from the paper and exposed to CS-molecular imaging screens (Bio-Rad Laboratories Ltd., Hemel Hempstead, UK) for up to 48 h. Radioactivity was imaged using a Bio-Rad GS-525 Molecular Imager with 100 µm resolution.

Carboxyfluorescein diacetate (CFDA) labelling

To image phloem transport into tuber buds, CFDA (Molecular Probes, Eugene, OR, USA) was loaded into potato tubers. Tubers were cut in half, ensuring a minimum of 3 cm of parenchyma tissue between the cut surface and the apical bud; the cut surface was immersed, to a depth of 3 mm, in 20 mg mL−1 aqueous CFDA solution. CFDA is membrane soluble, but endogenous esterases cleave off the acetate moieties to generate the fluorescent, membrane impermeable carboxyfluorescein (CF) molecule. CF then acts as a marker of phloem transport and provides evidence for symplastic unloading (Roberts et al. 1997). After loading, tubers were allowed to translocate CF for 14 h prior to examination of buds by confocal microscopy using a Leica SP2 microscope (Leica Microsystems, Mannheim, Germany) equipped with an argon laser. Tissue was excited at 488 nm, and emission was collected at 517–552 nm for the CF signal. To allow an outline of the buds and tuber tissue to be imaged, autofluorescence from the cell walls and tuber surface was simultaneously collected at 680–800 nm to provide a background signal. Test samples of tuber were also incubated in Texas Red dye (Molecular Probes) to ensure that the CF was being transported via the phloem and not the xylem (Roberts et al. 1997). Furthermore, prior to excision of the apical region for imaging, loaded tuber halves were hand sectioned from the cut surface upwards at 5 mm intervals to ensure that only phloem transport (and not parenchyma transport) was responsible for dye delivery to the apical region.


Sink–source transition in tuber tissues

Figure 1 shows the relative recovery of radioactivity from [U-14C]glucose in starch and sucrose in the storage parenchyma of tubers at different developmental stages. Approximately 55% of metabolized radioactivity was recovered in starch and 10% in sucrose in discs excised from developing tubers of approximately 5 cm diameter. The proportion of radioactivity recovered in starch progressively declined following harvest, and in tubers bearing sprouts totalling 4 g it was reduced to 5%. Conversely, the recovery of radioactivity in sucrose progressively increased and reached 70% of the label incorporated in sprouting tubers. The partitioning of [U-14C]glucose to sucrose in discs excised from the storage parenchyma of developing tubers was markedly increased (to levels similar to those observed in sprouting tubers) by preincubation of the tissue for 1 h with NaF or following 24 h detachment from the mother plant.

Figure 1.

Distribution of radioactivity in metabolic pools following the incubation of discs excised from tubers at various developmental stages with d-[U-14C]glucose. Discs were prepared as described in the text and incubated for 3 h with d-[U-14C]glucose after which discs were washed and radioactivity in the sucrose (▪) and starch (□) pools quantified as described. Developmental stages selected were 1, developing tubers (5 cm diameter); 2, resting (dormant) tubers (5 d from harvest from naturally senesced plants); 3, sprouting tubers with 2 mm buds; 4, sprouting tubers with 1 g sprouts; 5, sprouting tubers with 4 g sprouts; 1a, developing tubers preincubated with 5 mm NaF for 1 h; and 1b, developing tubers 24 h after detachment from mother plant. The data are given as mean ± SE, n = 3. Percentage of metabolized label indicates the label recovered in any particular pool as percentage of label metabolized (total label recovered less that recovered as [14C]glucose). Total radioactivity metabolized was 3.08 ± 0.22 KBq g−1 FW, and there were no significant differences between treatments.

Figure 2 shows the distribution of radioactivity in slices of tubers at different stages of development after incubation with [U-14C]sucrose. Autoradiographs of slices from developing tubers revealed uniform uptake of label across all tuber tissues (Fig. 2a). On the other hand, autoradiographs of slices of dormant (5 d from harvest from naturally senesced plants; Fig. 2b) or sprouting (1 cm sprout; Fig. 2c) tubers showed concentration of label in discrete islets (arrows), corresponding to the phloem strands internal and external to the vascular ring. Evidence for phloem-localized sucrose uptake was also found in autoradiographs of developing tubers following detachment from the mother plant for a period of 5 d (Fig. 2d). Phloem-localized sucrose uptake was not observed in slices of sprouting tubers preincubated with the highly effective sucrose–proton cotransport inhibitor p-chloromercuribenzene sulphonate (Turgeon 1996) (Fig. 2e).

Figure 2.

Autoradiographs of tuber slices at various stages of development following incubation with [14C]-sucrose for 2 h. Tubers slices were prepared, incubated and imaged as described in the text. Developmental stages selected were: (a) developing tuber (5 cm diameter), (b) resting (dormant) tuber (5 d from harvest from naturally senesced plants), (c) sprouting tuber with 1 cm sprout length, (d) developing tuber after 5 d detachment from mother plant and (e) sprouting tuber preincubated with the sucrose transport inhibitor p-chloromercuribenzene sulphonic acid (PCMBS) (2 mm) for 1 h. Scale bar in (a) = 1 cm.

Changes in the carbohydrate composition of tubers during sprouting in compost are shown in Fig. 3a. Uniform size tubers (100 ± 10 g FW) were selected and stored for 10 weeks at 4 °C prior to transfer to compost at 15 °C. At this stage, the tubers contained approximately 2.4 g of soluble sugars. After initial transfer to compost and stimulation of sprout growth, tuber sugar content declined, and this decline continued until the sprouts reached approximately 1 g dry matter. The inverse relationship between sink weight and tuber sugar content is consistent with utilization of these sugars for sustaining sprout growth. Subsequently, the rate of dry matter accumulation in sprouts increased substantially, and sugar accumulation was observed in the tubers. This was accompanied by a decline in the tuber starch content suggesting that starch hydrolysis was responsible for sugar accumulation. In support of this hypothesis, zymograms of tuber protein extracts showed the appearance of two novel amylase isoforms at the onset of starch mobilization (Fig. 3b).

Figure 3.

Changes in carbohydrate composition in tuber during sprouting and identification of β-amylase isozyme induction. (a) Uniform tubers [100 ± 10 g fresh weight (FW)] that had been cold stored for 10 weeks were transferred to moist compost and left to sprout at 15 °C in the dark. Total soluble sugar content (▪), sprout dry weight (DW) (□) and starch content (●) of tubers at various developmental stages (days after transfer) were recorded and are expressed on the basis of 100 g FW tuber on day 0. The data are given as mean ± SE, n = 3. (b) Zymogram of β-amylase activity of total protein extracts from tubers at the developmental stages presented in (a). The arrow marks the β-amylase couplet induced at the onset of starch mobilization.

Changes in bud metabolism associated with dormancy release

Figure 4 shows the partitioning of label in dormant (<0.5 mm) or open (1.5–2.0 mm) buds after incubation with d-[U-14C]glucose. Although there were minor differences in label partitioning, with more radioactivity recovered in the insoluble and sucrose fractions in open buds, these differences were statistically insignificant (P > 0.05) as determined using the Student's t-test. Despite the similarity in metabolic competency between dormant and open buds, analysis of metabolite pools showed striking differences. Figure 5 shows changes in the content of selected organic acids in buds at different stages of development. Only minor differences were observed in the concentration of oxalate (Fig. 5a) and malate (Fig. 5b) as the buds transitioned from the resting to growing state. Oxalate concentrations subsequently declined slightly as growth continued, while malate increased as the buds grew to 1.5–2 mm in length. Conversely, citrate (Fig. 5c) and succinate (Fig. 5d) showed dramatic concentration increases as the buds transitioned from rest to growth after which the citrate concentration remained relatively stable while there was a further increase in succinate concentration as buds grew larger.

Figure 4.

Radioactive partitioning following the incubation of dormant or open buds with d-[U-14C]glucose. Dormant (<0.5 mm length, ▪) or open (1.5–2 mm length, □) buds were incubated with d-[U14C]glucose for 3 h. Buds were processed, and radioactivity in each metabolic fraction was determined as described in the text. The results are expressed as a percentage of metabolized label incorporated into the appropriate fraction and are presented as mean ± SE, n = 3. Total radioactivity metabolized was 392 ± 20 KBq g−1 FW in dormant buds and 344 ± 15 KBq g−1 FW in open buds.

Figure 5.

Changes in organic acid content of tuber buds at various stages of development. Apical buds were sampled and organic acid concentrations were analysed as described in the text. (a) Oxalate, (b) malate, (c) citrate and (d) succinate. The x-axis shows the size of the sampled buds (mm) and the y-axis shows the organic acid concentration. The data are given as mean ± SE, n = 3.

Similarly, there were marked changes in carbohydrate pools as tuber buds developed (Fig. 6). Resting buds contained approximately 15 mg g−1 FW starch but very little soluble sugar. After bud burst and the resumption of bud growth, there was a sharp increase in both starch and soluble sugars, particularly sucrose. As buds developed further, the starch and sucrose contents gradually declined (Fig. 6a). In older sprouts (10 mm), starch was undetectable in the top and middle sections of the sprout and throughout the length of the sprout; hexoses were much more abundant than sucrose with glucose as the most abundant sugar (Fig. 6b).

Figure 6.

Changes in sugar content of apical tuber buds at various stages of development. Buds from potatoes maintained at 4 °C in potato stores were sampled at the appropriate size as described in the text. (a) Early developing buds up to 2 mm in length. The x-axis shows the length of the sampled buds (mm) and the y-axis shows the concentration of glucose (▪), fructose (□), sucrose (bsl00008) or starch (inline image). (b) Equal top, middle and bottom sections from 1 cm sprouts. Symbols as describedfor (a). The data are given as mean ± SE, n = 3.

Unloading of CF in tuber buds

More than 90% of the potato tubers loaded with CFDA showed uptake and trapping of the dye in the phloem, and subsequent transport through the vascular strands of the tuber. These data suggest that the phloem is active and that solutes are mobile in the sieve tubes even in dormant or resting tubers. The phloem is capable not only of short-distance uptake or retrieval of solutes from nearby parenchyma cells, but also of mobilizing those solutes over long distances using mass flow. However, although phloem connectivity in the main rings of transport phloem through the cortex of the tuber is active, dormant buds remain disconnected at early developmental stages soon after tuber harvest.

At 1 and 3 weeks postharvest, functional phloem connections (as indicated by transport of CF dye into the region) were not seen in any of the apical buds (Fig. 7a,b). Although dye was sometimes visible in the transport phloem subtending the bud (e.g. P in Fig. 7a), there were no phloem strands carrying dye up into the bud and no unloading of dye into cells close to the apical meristem (Fig. 7e). From week 5 postharvest onwards, an increasing number of apical buds showed some degree of phloem connectivity until at week 7; approximately 50% of the tubers studied showed some phloem connectivity into the apical bud (Fig. 7c). At this stage, the CF was contained primarily in the phloem strands in and around the bud, but by 9 weeks postharvest, both the amount of dye being transported through the phloem and the amount being unloaded into the apical meristem and central cells had increased (Fig. 7d,f).

Figure 7.

Carboxyfluorescein (CF) transport into apical buds. Potato tubers were imaged at weekly or fortnightly intervals from 1 to 11 weeks postharvest. (a–d) Representative images of apical buds at 1, 3, 7 and 9 weeks postharvest, respectively. The CF signal is shown in yellow, and the autofluorescence from the tissue is shown in blue. Some autofluorescence from compounds in the cuticle (epidermis) and from starch granules (ground tissues) also occurs in the detection channel used to collect the CF signal and therefore also appears yellow. Only buds that showed CF transport in the phloem subtending or close to the apical buds were imaged, although dye was not always visible immediately below the bud [e.g. (b)]. At 1 and 3 weeks postharvest, no CF was found in the immediate vicinity of the bud or unloaded into the region of the apical meristem. Some tubers showed CF in the phloem strands (P) subtending the buds (a), but no functional phloem connectivity existed into the bud. By week 7 postharvest, approximately 50% of the tubers imaged showed dye in functional phloem strands that extended up into the bud (c; arrows), and CF transport and unloading occurred into the apical bud and meristem region (c; arrowheads). At week 9 postharvest, the majority of tubers showed strong transport of CF in the phloem strands both subtending and into the growing buds, and extensive dye unloading into the region of the apical mersistem. (e and f) Show only the CF signal of enlarged areas (shown inside the dotted lines) from (a and d), respectively. The difference between the lack of dye unloading into the region of the apical meristem (M) of dormant buds (e), and CF unloading into the meristem of sprouting buds (f) can clearly be seen inside the dotted areas. Scale bar in (a) = 500 µm for (a–d). Scale bar in (e) = 1 mm for (e) and (f).


Sink–source transition in the storage parenchyma

In potato tubers, both starch and sucrose are synthesized from the same pool of metabolic intermediates, namely hexoses, hexose phosphates and sugar nucleotides. Examination of metabolic fluxes in excised storage parenchyma during the sink–source transition was achieved by the supply of [U-14C]glucose, which is rapidly converted into these metabolites (Viola 1996). During tuber maturation, the starch-synthesizing capacity in the storage parenchyma declined and this was accompanied by a progressive increase in net glucose conversion into sucrose (Fig. 1). This change is in keeping with the developmentally regulated metabolic sink–source transition within the tissue. In the storage parenchyma of developing tubers, limited sucrose-synthesizing capacity was detected. However, sucrose synthesis in immature tubers could be induced very rapidly (to the levels observed in sprouting tubers) by treatments that lead to inhibition of starch synthesis such as preincubation of tuber tissue with NaF (Viola & Davies 1991) or detachment of tubers from the mother plant (Fig. 1) (Geigenberger et al. 1994). The capacity for de novo sucrose synthesis by storage parenchyma of developing tubers has been detected both in excised tissue (Viola 1996) and in intact tubers (Geigenberger et al. 1999), and has been attributed to the presence of sucrose phosphate synthase (SPS) activity (Geigenberger & Stitt 1993). This enzyme is allosterically modulated (Reimholz, Geigenberger & Stitt 1994), and the enhanced sucrose synthesis in tissues with decreased starch-synthesizing capacity is likely to be attributed to the rapid accumulation of glucose-6-phosphate, which occurs in tissues treated with NaF (Viola & Davies 1991) or following tuber detachment (Geigenberger et al. 1994). From these observations, it is apparent that the storage parenchyma of sink tubers possesses the metabolic competence for sucrose synthesis, and this capacity can be rapidly up-regulated in response to changing metabolic status. However, it is clear that increased sucrose synthesis in storage parenchyma is not per se conducive to bud outgrowth. Transgenic tubers with reduced starch-synthesizing capacity accumulate large amounts of sucrose via adaptive up-regulation of SPS activity, but show no modification of dormancy parameters (Trethewey et al. 1999). In addition, storage of tubers at low temperature induces soluble sugar accumulation, but bud outgrowth is actually delayed (Wiltshire & Cobb 1996).

Sucrose synthesis appears as a dominant anabolic pathway in the storage parenchyma of dormant and sprouting tubers. However, in dormant tubers the absence of a sink for the utilization of sucrose produced results in its hydrolysis, presumably in the vacuole (Isla, Vattuone & Sampietro 1998), with the ensuing production of hexoses (Zrenner, Schuler & Sonnewald 1996) which can accumulate to levels 10-fold or higher than that of sucrose in cold-stored tubers (Amir, Kahn & Unterman 1977). It is unclear whether soluble sugar accumulation in tubers stored at low temperature results from enhanced starch degradation (Hill et al. 1996) or from reduced utilization of metabolic intermediates (Viola & Davies 1994). However, it is clear that sucrose synthesis precedes hexose accumulation in stored tubers (Duplessis, Marangoni & Yada 1996). In the present work, sugars were shown to be readily utilized for bud growth upon dormancy break in cold-stored tubers (Fig. 3). The high sucrose-synthesizing capacity in dormant tubers ensures a rapid conversion of hexoses (or hexose–phosphates) into sucrose, which is available for transport into the growing buds. In the initial phases of bud growth (up to 2 mm length), sucrose is the most abundant sugar in the tissue (Fig. 6). Under our conditions, sustained starch mobilization in the storage parenchyma appeared to be induced at an advanced sprouting stage upon depletion of the tuber soluble carbohydrate reserves. Starch mobilization appeared accompanied by soluble sugar accumulation in the sprouts and the appearance of a new β-amylase couplet on zymograms (Fig. 3). These isoforms showed similar electrophoretic mobility to the β-amylase induced by low temperature in Desiree reported by Hill et al. (1996) and characterized by Nielsen, Deiting & Stitt (1997). The similarities between these amylases are intriguing and may indicate that the isoform transiently induced upon cold storage has the physiological function of provision of metabolites to sprouting buds. Although outside the scope of the current work, this issue merits further investigation.

Taken together, the results presented here indicate that induction of starch mobilization in the storage parenchyma is not a prerequisite for bud outgrowth. Upon dormancy break, bud growth will be sustained in the first instance by soluble reserves in the storage parenchyma, where available. On the other hand, the accumulation of soluble sugars in the storage parenchyma per se is not conducive to bud growth. This strongly suggests that metabolic changes in the storage parenchyma have little or no control over bud dormancy.

Sink–source transition in the sieve element companion cell complex (SECC)

From a metabolic perspective, the SECC represents a constitutive sink for carbohydrates. However, from a physiological perspective, the SECC must acquire new functional competences as the tuber develops, to fulfil its role in the source tuber. Autoradiographs of tuber slices incubated with [U-14C]sucrose show uniform sucrose uptake by the storage parenchyma of developing tubers, while in slices of sprouting tubers there is clear evidence of sucrose uptake localized to discrete regions, corresponding to phloem bundles. The absence of sucrose uptake by the phloem of developing tubers (Fig. 2) is in agreement with our previous work demonstrating that the delivery of sucrose from the SECC into the parenchyma cells of developing tubers occurs symplastically (Viola et al. 2001). The sucrose uptake capacity by parenchyma cells of developing tubers apparently serves the purpose of retrieving sucrose ‘leaked’ into the apoplast (Wright & Oparka 1989). On the other hand, the evidence of active phloem sucrose uptake in sprouting tubers (Fig. 2) together with the previously demonstrated absence of active sucrose uptake by the storage parenchyma cells (Wright & Oparka 1989) suggest that sucrose freely diffuses from the storage parenchyma cells into the apoplast where it is actively retrieved by the SECC for transport to the buds. Although the development of active sucrose uptake by the SECC appears as an essential functional requirement for the sink–source transition in potato tubers, there was no correlation between the development of this capacity and dormancy release. Dormant tubers also showed a capacity for active sucrose uptake by the SECC. Indeed, this functional capacity appears to develop rapidly (within days) following tuber detachment. It is likely that following detachment when the tuber becomes a self-organized system, the process for physiological maturation is accelerated. Markers for these developmental changes are the increase in sucrose-synthesizing capacity (decline in starch-synthesizing capacity) in the storage parenchyma and the development of active sucrose uptake by the SECC. These tissues appear ‘primed’ for sustaining sprout growth in mature (dormant) tubers, but these physiological competencies are insufficient to trigger dormancy release.

Metabolic changes in apical tuber buds during the tuber sink–source transition

Unlike the tuber tissues, metabolic activities within the buds are thought to be very restricted during dormancy (Suttle 2004a). However, uptake and metabolism of [U-14C]glucose by excised resting or growing apical buds were remarkably similar (Fig. 4), suggesting that they possess similar metabolic competence for sugar metabolism. Although a large proportion of radioactivity was recovered in products of respiratory metabolism in buds supplied with radio-labelled sugars, the organic acid concentration was low in resting buds, and a marked increase in the concentration of both citrate and succinate was observed on transition to growth (Fig. 5). Similarly, a large increase of carbohydrate concentration in the apical buds coinciding with the onset of bud growth was observed (Fig. 6). The correlation between growth activation and carbohydrate and organic acid accumulation in the tuber apical buds indicates that carbon availability may be a limiting factor for bud growth. This suggests that outgrowth of the bud may occur as a result of (the induction of) sugar availability to the buds. Data presented here demonstrate that the functional capacity for sucrose uptake in the phloem is induced early during the resting period, prior to bud break, and that the transport phloem within the tuber is capable of long-distance solute transport throughout the tuber by mass flow. Thus, sucrose unloading into the buds appears to be a key prerequisite for bud outgrowth. We have previously shown that tuberization is characterized by the induction of symplastic unloading in the swelling region of the tuber, while the apical region, corresponding to the apical bud remained symplastically isolated (Viola et al. 2001). Thus, the apical bud corresponds to a discrete cell domain characterized by symplastic isolation from the rest of the tuber. Using CF as a marker of symplastic continuity, evidence for the termination of symplastic isolation of the apical bud at the onset of bud growth was obtained (Fig. 7). This physiological shift correlates with carbohydrate accumulation in the bud and subsequent bud outgrowth. It is not clear what governs the initiation of bud outgrowth, and where the point of control exists. It is possible that developmental processes which begin within the buds lead to subsequent phloem connectivity which in turn brings sufficient metabolites to fuel bud outgrowth, or that signals from outwith the buds initiate phloem connectivity which in turn stimulates bud development and outgrowth. The data presented here do not show whether mature phloem strands exist between the transport phloem of the cortex and the region of the apical meristem, but are incapable of solute transport, or whether lack of vascular tissue differentiation limits transport; these data would require a detailed anatomical study using electron microscopy. However, CF transport within the sieve elements can be seen as thin strands of dye within the tissues, as shown in Fig. 7c (arrows), while unloading from the phloem appears as a more diffuse signal from the cytoplasm of parenchyma cells. Given that there is a developmental progression from phloem transport (Fig. 7c) to phloem unloading (Fig. 7d), it is possible to speculate that some degree of physical, or anatomical, barrier is present at tuber harvest. What is clear from this study is that functional phloem connectivity and symplastic transport into the bud are commensurate with bud development and outgrowth. In the light of the changes in phloem unloading during tuber formation, the identification of the mechanisms involved in the regulation of phloem unloading in potato is of paramount importance for understanding the control of developmental processes in the tuber.

Symplastic gating and the potato tuber life cycle

In a previous publication, we demonstrated that induction of tuberization in elongating stolons was associated with: (1) a switch from apoplastic to symplastic phloem unloading in the subapical region behind the meristem; and (2) a reduction in label deposition in the stolon apical meristem from 14CO2 supplied to whole plants (Viola et al. 2001). Therefore, stolon elongation is associated with an active apical meristem supported by the symplastic delivery of metabolites, and tuberization is associated with a dormant apical meristem that is symplastically isolated. In the present work, it has been demonstrated that the apical tuber bud meristem remains dormant despite: (1) a switch from starch to sucrose synthesis in the storage parenchyma upon detachment; (2) the development of an active sucrose uptake system by the tuber phloem; and (3) no detectable metabolic incompetencies in the apical bud. In developing and freshly harvested tubers, the apical bud remains symplastically isolated; however, symplastic connection was reestablished in growing buds. Therefore, control of meristem activity is achieved in part by the supply of metabolites, which is modulated by symplastic gating; this mechanism should be considered in future studies of the tuber life cycle. The next challenge is to determine what environmental, genetic and hormonal controls are involved in switching symplastic gating within the potato.


We wish to thank Ralph Wilson (Scottish Crop Research Institute) and Dr. Luigi Tedone (University of Bari) for supplying potatoes. This work was funded by the Scottish Executive Environment and Rural Affairs Department.