Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
As reported in a previous paper (Plant, Cell and Environment 24, 357–365, 2001), introduction of sucrose phosphorylase into the cytosol of potato results in increased respiration, an inhibition of starch accumulation and decreased tuber yield. Herein a more detailed investigation into the effect of sucrose phosphorylase expression on tuber metabolism, in order to understand why storage and growth are impaired is described. (1) Although the activity of the introduced sucrose phosphorylase was low and accounted for less than 10% of that of sucrose synthase its expression led to a decrease in the activities of enzymes of starch synthesis relative to enzymes of glycolysis and relative to total amylolytic activity. (2) Incubation of tuber discs in [14C]glucose revealed that the transformants display a two-fold increase of the unidirectional rate of sucrose breakdown. However this was largely compensated by a large stimulation of sucrose re-synthesis and therefore the net rate of sucrose breakdown was not greatly affected. Despite this fact major shifts in tuber metabolism, including depletion of sucrose to very low levels, higher rates of glycolysis, and larger pools of amino acids were observed in these lines. (3) Expression of sucrose phosphorylase led to a decrease of the cellular ATP/ADP ratio and energy charge in intact growing tubers. It was estimated that at least 30% of the ATP formed during respiration is consumed as a result of the large acceleration of the cycle of sucrose breakdown and re-synthesis in the transformants. Although the absolute rate of starch synthesis in short-term labelling experiments with discs rose, starch synthesis fell relative to other fluxes including respiration, and the overall starch content of the tubers was lower than in wild-type tubers. (4) External supply of amino acids to replace sucrose as an osmoticum led to a feed-back inhibition of glycolysis, but did not restore allocation to starch. (5) However, an external supply of the non-metabolizable sucrose analogue palatinose – but not sucrose itself – stimulated flux to starch in the transformants. (6) The results indicate that the impaired performance of sucrose phosphorylase-expressing tubers is attributable to decreased levels of sucrose and increased energy consumption during sucrose futile cycling, and imply that sucrose degradation via sucrose synthase is important to maintain a relatively large sucrose pool and to minimize the ATP consumption required for normal metabolic function in the wild type.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Conversion of sucrose to starch in growing potato tubers is well characterized with respect to the pathway and the location of the individual steps (Kruger 1997) but much less is known about the regulation of this important pathway. It has been demonstrated that ADP-glucose pyrophosphorylase (AGPase), plastidial phosphoglucomutase (PGM) and import of ATP into the plastid via the amyloplast adenylate translocator exert influence over the control of starch synthesis in potato tubers (Tjaden et al. 1998; Geigenberger, Müller-Röber & Stitt 1999a; Fernie et al. 2001b). AGPase catalyses the first irreversible step that is unique to the pathway of starch synthesis, and is activated by 3-phosphoglycerate (3PGA) and inhibited by Pi (inorganic phosphate; Sowokinos & Preiss 1982; Hnilo & Okita 1989). These allosteric properties allow the rate of starch synthesis to respond to changes in levels of glycolytic metabolites that are generated when the balance between the rate of sucrose breakdown and flux into respiratory pathways changes (Geigenberger et al. 1997; Geigenberger, Geiger & Stitt 1998a). However, on their own, they are not sufficient to explain other metabolic scenarios in which the rate of starch synthesis is altered; in particular when the sucrose level is altered. In these conditions, the rate of starch synthesis changes reciprocally to the levels of the glycolytic intermediates (Geigenberger et al. 1994; Geiger, Stitt & Geigenberger 1998).
Genetic manipulation of cytosolic sucrose mobilization within the tuber by the ectopic expression of yeast invertase resulted in a massive metabolic shift, involving induction both of glycolysis and a rapid cycle of sucrose breakdown and re-synthesis, whereas starch accumulation was decreased (Sonnewald et al. 1997; Trethewey et al. 1998, 1999a, b). Similarly, the expression of sucrose phosphorylase led to an elevation in the activities of several glycolytic enzymes and the levels of glycolytic intermediates, and an even more dramatic decrease of tuber starch content (Trethewey et al. 2001). In both cases the lower starch content was intriguing because the tubers contained elevated levels of hexose-phosphates and 3-PGA, the substrate and activator of the AGPase reaction, respectively.
The role of sugars in the regulation of metabolic and developmental processes in plants has recently received much attention (see Smeekens 2000) and there is a growing body of evidence that sugar-mediated regulation contributes to the regulation of metabolism in potato tubers (see Fernie, Roessner & Geigenberger 2001a). Furthermore, several independent studies (Borisjuk et al. 1998; Geiger et al. 1998; Weber et al. 1998; Grimmer, Bachfischer & Komor 1999) indicate that the sucrose concentration may be a key determinant in the regulation of carbon partitioning in heterotrophic tissues. The invertase and sucrose phosphorylase lines are both characterized by a dramatic reduction of sucrose. Sucrose might act as a signal or, alternatively, influence the osmotic relations. Sucrose represents a major osmotic component of the cytosol, and decreased levels of this metabolite are often accompanied by major changes in the levels of other osmotically active compounds, including amino acids (Hare, Cress & Van Staden 1998).
Intriguingly, the general metabolic phenotype resulting from expression of a sucrose phosphorylase in potato tubers resembles that following expression of invertase, even though the introduced activity was far lower in the case of sucrose phosphorylase (Trethewey et al. 2001). Further, the decrease in starch accumulation was larger in the sucrose phosphorylase lines than the invertase lines (Trethewey et al. 2001). Taken together, these results indicate that the metabolic consequences following expression of heterologous sucrose degrading activities are not due to an increased rate of sucrose degradation per se. To test this hypothesis directly we have now measured the rates of sucrose degradation and re-synthesis, and the overall activities of sucrose synthase and sucrose-phosphate synthase in sucrose phosphorylase-expressing tubers. To evaluate the alternative possibility that the metabolic shifts are due to differences between properties of the endogenous sucrose synthase and the introduced sucrose phosphorylase, we analysed whether sucrose phosphorylase led to an accelerated energy consumption in tuber tissue, and whether external feeding of sucrose, amino acids or the sucrose analogue palatinose to discs of transgenic tubers were able to increase the rate of starch synthesis.
Materials and methods
Transformants and wild-type potato plants (Solanum tuberosum L. cv. Desirée, Saatzucht Fritz Lange, Bad Schwartau, Germany) were grown in soil (3 L pots) supplemented with Hakaphos grün (100 g per 230 L soil; BASF, Ludwigshafen, Germany) in a growth chamber (350 µmol photons m−2 s−1 irradiance, 14 h/10 h day/night regime, 20 °C, 50% relative humidity). Growing tubers from 9-week-old, daily-watered plants with high activities of sucrose synthase, which is taken as an indicator for rapidly growing tubers (Merlo et al. 1993), were used for the experiments. The generation of transgenic plants has been described previously (Trethewey et al. 2001).
Analysis of enzyme activities
Unless described here enzyme analyses were performed as described in Geigenberger et al. (1998b). In the case of sucrose synthase, two different assay conditions were used: 150 mm sucrose and 4 mm UDP (Vmax), and 15 mm sucrose and 0·4 mm UDP (Vsel). Sucrose phosphorylase activity was determined using the extraction buffer as detailed by Geigenberger & Stitt (1993) and an assay modified from Birkenberg & Brenner (1984) in that the reaction was buffered at pH 7·0 using Hepes instead of phosphate buffer. Starch synthases, branching enzymes and the specific activities of the cytosolic and plastidic isoforms of phosphoglucomutase were determined as described in Fernie et al. (2001b).
Tuber discs (diameter 8 mm, thickness 1–2 mm) were cut directly from growing tubers attached to the fully photosynthesizing mother plant, washed three times with 10 mm 2-(N-morpholino) ethane sulfonic acid (MES) (pH 6·5; KOH) and then incubated (eight discs in a volume of 4 mL in a 100 mL Erlenmeyer flask shaken at 90 r.p.m) for 25 min or 2 h in 10 mm Mes-KOH (pH 6·5) containing 2 mm glucose including 1·4 KBq µmol−1[U-14C]-glucose (Amersham-Buchler, Freiburg, Germany) supplemented with asparagine, mannitol, palatinose, proline or sucrose at various concentrations. Then the discs were harvested, washed three times in buffer (100 mL per eight discs), and frozen in liquid nitrogen.
Fractionation of 14C labelled components
Tuber discs were extracted with 80% (v/v) ethanol at 80 °C (1 mL per two discs), re-extracted in two subsequent steps with 50% (v/v) ethanol (1 mL per two discs at each step), the combined supernatants dried under an air stream at 40 °C, taken up in 1 mL H2O, and labelled separated by ion-exchange chromatography and thin-layer chromatography exactly as described in Geigenberger et al. (1997). The insoluble material left after ethanol extraction was analysed for label in starch as in Merlo et al. (1993). Specific activities of the hexose phosphate pool were estimated by dividing the label retained in the phospho-ester fraction by the summed carbon in this fraction as detailed in Geigenberger et al. (1997). In the case of the 25 min feeding experiments the specific activities were divided by two in order to get the average value during incubation.
Where differences are described in the text as significant, a t-test was performed using the algorithm incorporated into Microsoft Excel 7·0 (Microsoft Corp., Seattle, WA, USA) that yielded a value below 5% (P < 0·05).
Determination of key enzyme activities mediating the sucrose–starch transition
Developing tubers of the transformants contained sucrose phosphorylase at an overall activity that was less than 10% of that of sucrose synthase (Table 1). The transformants contained significantly increased activity of sucrose synthase, when the latter was assayed using saturating substrates. However, under non-saturating assay conditions which are closer to the in-vivo situation, sucrose synthase activity remained unchanged compared with the wild type, indicating that the increase in overall activity was largely compensated in planta, possibly via reversible protein phosphorylation (Huber et al. 1996). This contrasts with sucrose-phosphate synthase (SPS), in which the maximal catalytic activity was not significantly changed, but the activity assayed under non-saturating substrate concentrations (Vsel) was markedly increased in all transgenic lines. Potato tuber SPS is regulated by reversible protein phosphorylation which changes the kinetic properties of the enzyme rather than the maximal catalytic activity, and can be tracked as a change in the Vsel/Vmax activity ratio (see Trethewey et al. 1999b). There was an up to three-fold increase in the Vsel/Vmax ratio in all transgenic lines, in comparison with the wild type, indicating that the activation state of SPS was increased.
Table 1. Sucrose levels, starch levels, and enzyme activities in growing tubers of sucrose phosphorylase-expressing transgenic plants. Sucrose synthase (SuSy) activity was assayed in the presence of 150 mm sucrose and 4 mm UDP (termed Vmax), or in the presence of 15 mm sucrose and 0·4 mm UDP (termed Vsel). Sucrose phosphate synthase (SPS) activity was assayed in the presence of 12 mm Fru-6-P, 36 mm Glc-6-P and 6 mm UDPGlc (termed Vmax), or in the presence of 2 mm Fru-6-P, 6 mm Glc-6-P, 6 mm UDPGlc and 5 mm Pi (termed Vsel). The activation state is calculated as Vsel/Vmax ratio and given as a percentage
The data presented are the mean ± SE (n = 4–5 tubers from different plants). GB, granule bound. aValues that were determined by the t-test to be significantly different from the wild type (P < 0·05).φ, branching enzyme is expressed as fold-stimulation of glycogen phosphorylase; n.m.; not measured; n.d., not detectable.
The expression of sucrose phosphorylase was accompanied by an up to a 50% decrease in the activity of AGPase in lines SP-29 and SP-12, by a reduction in the activity of soluble starch synthase in line SP-29, and by a decrease in starch phosphorylase activity in lines SP-2, SP-29 and SP-12 (Table 1). There were no clear changes of the activities of other enzymes of starch metabolism. No significant changes were observed in the activities of hexokinases, fructokinases and invertases compared to the wild type, whereas glycolytic activities were increased (data not shown, see also Trethewey et al. 2001). The activity of sucrose phosphorylase was also accompanied by a significant decrease in sucrose and starch levels (with the exception of the starch level in line SP-11; Table 1) similar to that previously observed (Trethewey et al. 2001; Roessner et al. 2001).
Metabolism of labelled glucose in tuber discs of the transgenic lines
In order to analyse the rate of sucrose degradation and its re-synthesis, freshly cut slices of growing potato tubers were incubated with 2 mm[U-14C]-glucose in the presence of 25 mm mannitol (Table 2). Labelled glucose was preferred to sucrose due to the large changes in the internal sucrose levels observed in the transformants. Furthermore, this choice of radio-labelled substrate also provided the opportunity to perform experiments under a range of sucrose concentrations and to measure the rate of sucrose re-synthesis under these conditions. In both wild-type and transgenic tissue, 14C glucose uptake rose slightly when 25 or 100 mm sucrose was included in the medium. The metabolism of label differed markedly between wild-type discs and discs from sucrose phosphorylase-expressing tubers. A markedly and significantly smaller proportion of the label was incorporated into insolubles (of which between 90 and 93% of the label was consistently recovered in starch regardless of the genotype or incubation conditions) in the transformants. Label incorporation into the intermediates and products of glycolysis (phospho-esters, organic and amino acids) also decreased in the transformants, but less markedly than label incorporation in starch. Moreover a markedly and significantly larger proportion of the label was recovered in sucrose in the transformants than in wild-type discs (approximately 60% compared with 30%), reflecting the increased activation state of SPS (see above). Intriguingly, whereas inclusion of high concentrations of sucrose in the incubation medium significantly stimulated label incorporation into insoluble compounds (starch) and inhibited label incorporation into sucrose and phospho-esters in wild-type discs, there was no marked or consistent effect in discs from the transformants
Table 2. Redistribution of radiolabel within tuber discs excised from wild-type and sucrose phosphorylase-expressing lines following incubation in 2 mm[U-14C]glucose and different sugar supplements. Discs were cut from developing tubers of 10-week-old plants, washed three times in buffer, and then pre-incubated for 20 min in 10 mm MES-KOH (pH 6·5) supplemented as detailed. Incubations then received [U-14C]glucose to a final specific activity of 5625 dpm nmol−1. The discs were then incubated for a further 25 min, washed three times, extracted and analysed for radiolabel in insolubles, sucrose, P-ester, organic acids and amino acids
Uptake (dpm g FW−1 25 min−1)
Distribution of radiolabel, % of total metabolized recovered in:
Data presented are the mean ± SE, n = 4. aValues that were determined by the t-test to be significantly different from wild type (P < 0·05). bValues that were determined by the t-test to be significantly different in response to sugar supplied (P < 0·05).
Estimation of rates of sucrose synthesis and breakdown, glycolysis and starch synthesis
In short-term labelling experiments the relative distribution of label does not always reflect the actual metabolic fluxes due to isotopic dilution (for details see Geigenberger et al. 1997). We divided label retention in the phospho-ester fraction after 25 min (Table 2) by the summed carbon in the hexose phosphate pool (data not shown), and then divided again by two to estimate the average specific activity of the hexose phosphate pool during the 25 min incubation. The endogenous pools of hexose-phosphates (approximately 150–500 nmol g FW−1, data not shown) and sugars (i.e. sucrose 2000–10000 nmol g FW−1, see Table 1) were much larger than the amount of external (labelled) glucose that is taken up in 25 min (approximately 30–40 nmol g FW−1, calculated from Table 2). As a result the specific activity of the phospho-ester pool will rise gradually during incubation, and the value obtained after the end of the 25 min incubation period will overestimate the average value during incubation by a factor of two.
Incubation of discs in 100 mm sucrose significantly decreased the proportion of radiolabel in the phospho-ester fraction by about 30% in discs from wild-type tubers, compared with discs incubated with 25 mm sucrose or mannitol, resulting in a decrease of the specific activity of the hexose phosphate pool (Table 2). This presumably reflects an increased rate of sucrose mobilization when sucrose is entering the discs from the medium. The specific activity of the hexose phosphate pool was much (and significantly) lower in the transformants than in wild-type tuber discs, probably reflecting an increased rate of carbohydrate mobilization in these lines (Table 3).
Table 3. Absolute rates of starch synthesis, glycolysis and sucrose re-synthesis within tuber discs excised from wild-type and sucrose phosphorylase overexpressors following incubation as in Table 2. Absolute rates of starch synthesis, glycolysis and sucrose re-synthesis were calculated from the label incorporation data using the specific activity of the hexose phosphate pool in order to account for isotopic dilution factors. The specific activity of the hexose phosphate pool was estimated by dividing the label released by the action of acid phosphatase by the sum of carbon in phosphorylated sugars measured in the same samples at the end of the incubation. Values were corrected by dividing by two, to get the mean specific activity during the course of the 25 min incubation
HP specific activity (dpm nmol−1)
Metabolic flux (nmol hexose equivalents g FW−1 25min−1)
Data presented are the mean ±SE, n = 4. aValues that were determined by the t-test to be significantly different from wild type (P < 0·05). bValues that were determined by the t-test to be significantly different in response to sugar supplied (P < 0·05).
The specific activities were then used to estimate absolute fluxes within the tuber discs (Table 3). When discs incubated in 25 mm mannitol were compared, the expression of sucrose phosphorylase led to a 2·6- to 3·3-fold significant increase in glycolytic flux, a 1·4- to 2·5-fold increase in the absolute rate of starch synthesis (only significant for line SP-2), and also a 6·5- to 10-fold significant increase in the unidirectional rate of sucrose re-synthesis. Thus, the transformants partitioned less label from 14C glucose to starch, but the calculated absolute carbon flux towards starch was, surprisingly, higher than in the wild type.
The net rate of sucrose degradation was measured directly by subtracting the sucrose levels in the discs before and at the end of the incubation (Table 4). This approach is only valid in the case in which no external sucrose is supplied, and we therefore only present data for discs incubated in 25 mm mannitol. Unexpectedly, the expression of sucrose phosphorylase did not increase the net rate of sucrose degradation, with the rate not being significantly altered in SP-2 and SP-29, and decreased in SP-11. The unidirectional rate of sucrose breakdown was calculated as the difference between the net rate of sucrose degradation and the rate of unidirectional sucrose synthesis. Expression of sucrose phosphorylase led to 1·6- to 2·3-fold increase in the unidirectional rate of sucrose breakdown (significant for SP-2 and SP-29).
Table 4. Direct measurement of the net rate of sucrose degradation and estimation of unidirectional rates of sucrose synthesis and degradation within tuber discs excised from wild-type and sucrose phosphorylase-expressing lines following incubation in 25 mm mannitol as in Table 2. Net sucrose degradation was measured as the difference of the sucrose content of the discs before incubation (see Table 1) minus the sucrose content of the discs at the end of the incubation (see Fig. 2). Unidirectional (UD) sucrose degradation was estimated as the sum of net sucrose degradation plus UD sucrose synthesis (see Table 3)
Fluxes (nmol hexoses g FW−1 25 min−1)
UD sucrose degradation
Data presented are the mean ±SE, n = 4. aValues that were determined by the t-test to be significantly different from wild type (P < 0·05).
Estimation of ATP consumption during sucrose cycling and determination of adenylate levels
Although a cycle of sucrose synthesis and degradation allows sensitive regulation of sucrose breakdown, it also represents a ‘futile cycle’ (Dancer, Hatzfeld & Stitt 1990; Geigenberger & Stitt 1991) leading to consumption of ATP. We used the data in Table 3 to calculate the wastage of ATP during sucrose cycling. In the case of wild-type tubers we assumed that the cycle involves operation of sucrose synthase and SPS and that one molecule of ATP is required per molecule of sucrose, which is degraded and re-synthesized (to phosphorylate the fructose formed by sucrose synthase). In the case of the transformants expressing sucrose phosphorylase we assumed that two molecules of ATP are required per molecule of sucrose during the cycle (as no UTP is generated when sucrose is converted to Glc-1-P by sucrose phosphorylase). Table 5 summarizes the loss of ATP during the cycle as well as the maximum rate of ATP production, assuming that (1) flux into glycolysis equals respiration, that (2) each glucose-equivalent entering glycolysis is fully oxidized, and that (3) the total yield of a fully oxidized glucose is 36 ATP (for the latter see Stryer 1988).
Table 5. Estimation of the ATP demand of sucrose cycling. The wastage of ATP during sucrose cycling was calculated for wild type and transformants by using the metabolic flux data of Table 3. In the case of the wild type we assumed that the cycle involves operation of sucrose synthase and SPS and that one molecule of ATP is required per molecule of sucrose which is degraded and re-synthesized (to phosphorylate the fructose formed by sucrose synthase). In the case of the transformants expressing sucrose phosphorylase we assumed that two molecules of ATP are required per molecule of sucrose during the cycle (as no UTP is synthesized when sucrose is converted to Glc-1-P by sucrose phosphorylase). The rate of sucrose degradation was estimated as the sum of the rates of starch synthesis, glycolysis and sucrose re-synthesis. The total gain of ATP was estimated from the glycolytic flux assuming that (i) flux into glycolysis equals respiration, that (ii) each glucose-equivalent entering glycolysis is fully oxidized, and that (iii) the total yield of a fully oxidized glucose is 36 ATP (for the latter see Stryer 1988)
Estimated rates of ATP production and consumption [µmol (g FW)−1 (25min)−1]
Total ATP produced
ATP required for sucrose cycling
ATP required for sucrose cycling (in % of total produced)
In wild-type tubers, ATP production was estimated to be 5·4–8·7 µmol ATP (g FW)−1 (25 min)−1, whereas 0·20–0·26 µmol ATP (g FW)−1 (25 min)−1 were consumed during sucrose cycling (Table 5). Therefore only 3–4% of the total ATP produced was wasted in the cycle, which is similar to the proportion estimated in germinating Ricinus seedlings (Geigenberger & Stitt 1991). A different picture emerged in the transgenic lines in which ATP production and ATP wastage during cycling was estimated to be 14–19 and 3–6·3 µmol (g FW)−1 (25 min)−1, respectively, revealing that 20–35% of the total ATP produced in respiration is being wasted during operation of this cycle. This represents a minimum estimate of the proportion of the total ATP that is consumed in this cycle, because not every glucose molecule that enters glycolysis will be respired, and it is unlikely that respiratory electron transport and ATP production are perfectly coupled. These calculations demonstrate that expression of sucrose phosphorylase leads to a substantial loss of cellular energy in potato tubers. This is consistent with the significant decrease of the ATP/ADP ratio and the adenylate energy charge in intact tubers following sucrose phosphorylase expression (Fig. 1) and furthermore also provides an explanation for the elevated rate of respiration in these tubers.
Influence of high external sucrose on the metabolism of labelled glucose
In agreement with the results of previous studies (Geiger et al. 1998; Loef, Stitt & Geigenberger 2001) there was a 2·5-fold significant increase in the rate of starch synthesis and a smaller increase in the rate of sucrose synthesis and glycolysis when external sucrose was supplied to wild-type discs. In contrast, starch synthesis was not significantly changed in response to high external sucrose in discs from transgenic tubers (Table 3). This might be because the transformants have lost the capacity to respond to sucrose, or because the introduced sucrose cleaving activity mobilizes sucrose so efficiently that it does not accumulate in the discs, despite being supplied at high concentrations in the medium. To distinguish between these possibilities we determined the internal sucrose level in the discs at the end of the incubation (Fig. 2). The discs were washed three times to remove sucrose from the apoplast. A separate experiment showed that 94% of the labelled sucrose was removed on the first wash, another 5% on the second and up to 1% by the third, and that the fourth to tenth washes only yielded trace levels of label. The sucrose content in wild-type discs doubled following incubation in 100 mm sucrose. The level of sucrose within the transgenic lines was very low, and although it doubled when 100 mm sucrose was included in the medium it remained far below wild-type levels. These results show that supplying exogenous sucrose does not lead to a substantial enough increase of the internal sucrose pool in the transgenic discs to allow complementation of the metabolic phenotype (see below for further experiments).
External amino acid supply to tuber discs leads to feed-back inhibition of glycolysis, without influencing starch synthesis
Expression of sucrose phosphorylase led to a significant and up to two-fold increase in the level of asparagine, which is the major amino acid in potato tubers (Table 6). There were also significant changes in the levels of minor amino acids, including a significant increase in alanine, tyrosine, arginine and serine, whereas GABA, valine and isoleucine decreased. The total level of amino acids increased, but this was only significant in line SP-2 (Table 6). This accumulation of amino acids is consistent with the increased rate of glycolysis in the transgenic tuber discs. Similar results were obtained in invertase and invertase/glucokinase overexpressing lines (Trethewey et al. 1998). The higher amino acid content in tubers expressing sucrose phosphorylase or invertase and containing decreased levels of sucrose is consistent with the proposal that sucrose plays a role in the maintenance of the cellular osmotic status (Winter, Robinson & Heldt 1993; Hare et al. 1998; Geigenberger et al. 1997, 1999b). Interestingly, transgenic lines with decreased expression of AGPase or plastidial PGM that contain increased levels of sucrose are characterized by a reduction in amino acid levels (Trethewey et al. 1999a; Fernie et al. 2001b).
Table 6. Amino acid contents of transgenic tubers. Amino acid contents were determined in samples from developing tubers
The data presented are the mean ±SE for determinations on five individual plants per line, and are given as µmol g FW−1. Gaba = γ-amino-butyric acid. aValues that were determined by the t-test to be significantly different from wild type (P < 0·05).
One possible explanation for the lower starch content in the sucrose phosphorylase-expressing lines could be that there is increased allocation of carbon towards glycolysis and amino acid biosynthesis, to compensate for the low sucrose and meet the requirement for cellular osmotica. To investigate this possible interplay between amino acid and starch synthesis, we investigated whether externally supplied amino acids influence metabolic fluxes in tuber discs. In this experiment (Fig. 3) and also in the following experiment of Fig. 4 we used only two transgenic lines SP-11 and SP-29. These lines have been shown to be representative lines in the present study (see above) and during extensive investigations in a previous study (Trethewey et al. 2001).
In an initial experiment the relative uptake rates of proline, asparagine and sucrose were compared by incubating wild-type tuber discs in 50 mm of [U-14C] labelled substrate (100 mm in the case of asparagine). Asparagine was chosen because it is the major amino acid representing a fairly large pool within the tuber (see Table 6) and is mainly located in the cytosol (A. Tiessen and P. Geigenberger, unpubl. results). Proline was chosen as it often increases inside cells in response to osmotic stress (Roosens et al. 1999). The uptake rates of sucrose, proline and asparagine were equivalent in all cases. Fractionation revealed that the amino acids were relatively inert, with 15–20% of the label being recovered in protein after 2 h of incubation, and the remainder staying in the amino acid pool (data not shown).
Tuber discs were incubated in tracer levels of labelled glucose with and without supplementary 50 mm proline or 100 mm asparagine for a 2 h period (Fig. 3). When discs incubated without external amino acids were compared, expression of sucrose phosphorylase led to an increase in the proportion of label incorporation into sucrose (Fig. 3c) whereas the proportion of label incorporated into starch (Fig. 3b), phosphate-ester (Fig. 3d), organic acid (Fig. 3e) and especially that in amino acids (Fig. 3f) decreased. Furthermore, the specific activities of the phospho-ester pool of the transgenic lines decreased sharply with respect to the wild type (Fig. 3g). These results resemble those already presented above. In discs from wild-type tubers external amino acids led to decreased labelling of organic acids and amino acids from 14C-glucose (Fig. 3e & f, changes were not significant at the 5% level, but a consistent trend is observed), whereas labelling of starch remained unchanged (Fig. 3b). The same trend was observed after feeding proline or asparagine to discs of tubers expressing sucrose phosphorylase (significant in three cases with respect to the decrease in organic acids, and in one case with respect to amino acids). The sum of label in organic acids and amino acids provides an estimate of the glycolytic flux. In wild-type and transformant discs, label entering glycolysis decreased after feeding asparagine or proline. Label in glycolytic products decreased to 64 ± 8*, 86 ± 11 and 69 ± 9%* of control levels after feeding proline and 79 ± 9, 60 ± 8* and 47 ± 4%* of control levels after feeding asparagine to wild type, SP-11 and SP-29, respectively (* denotes values that are significantly different from the control). The results indicate that amino acid biosynthesis and glycolysis are feed-back inhibited when amino acids are supplied. Crucially, feed-back inhibition of glycolysis did not lead to a corresponding increase in the flux to starch. Further, similar changes were seen for discs from wild-type tubers and transformant lines. It is therefore unlikely that the decreased partitioning to starch in response to sucrose phosphorylase expression is a direct consequence of changes in glycolytic flux.
The partitioning of carbon toward starch in the transgenic lines can be considerably increased on incubation with palatinose
In a second approach to investigate whether the low sucrose levels per se restrict starch synthesis in the sucrose phosphorylase-expressing tubers, we supplied the sucrose analogue palatinose. This approach was taken because addition of external sucrose did not lead to an increase of the internal sucrose content of discs from the transformants (see above), and sucrose feeding would anyway not distinguish between possible effects due to signalling, sucrose supply and osmotic status. Palatinose cannot be cleaved by SuSy, invertase or any other sucrose-degrading activity present in wild-type tubers (Fernie et al. 2001a), however, there are no reports in the literature of whether palatinose influences the activity of sucrose phosphorylase. For this reason we assayed sucrose phosphorylase activity in a desalted enzyme extract from a sucrose phosphorylase-expressing tuber in the presence of various palatinose and sucrose concentrations. Palatinose could not be metabolized by sucrose phosphorylase but it was a weak competitive inhibitor of the enzyme: sucrose phosphorylase exhibited a Ki of 37 mm (data not shown). Since the estimated cellular concentration of palatinose following a 2 h incubation with 5–100 mm of the analogue is 0·1–1 mm (see Fernie et al. 2001a) it is approximately three orders of magnitude lower than estimated concentrations of sucrose and is therefore far too low to inhibit sucrose phosphorylase in vivo.
In both wild-type and transgenic discs, the rates of 14C-glucose uptake and metabolism were slightly increased by including 100 mm palatinose in the medium (Fig. 4a & b). Palatinose led to a change in the metabolism of labelled glucose in both wild type and transformants. In both cases, there was a consistent trend of palatinose to increased partitioning of radiolabel into starch (Fig. 4c). When absolute fluxes are calculated it becomes apparent that palatinose stimulates the absolute rate of starch synthesis in transgenic as well as wild-type discs (Fig. 4k, significant at the 5% level only for the wild type, in SP-29 significant at the 10% level). The relative stimulation of starch synthesis was similar in wild-type and transgenic lines. The ratio between starch synthesis and glycolysis increased significantly from 3·2 to 4·8 when palatinose was supplied to wild-type discs, from 2·8 to 3·7 when palatinose was supplied to SP-11, and significantly from 3·8 to 5·9 when palatinose was supplied to SP-29 (Fig. 4n). These results show that the capacity for the transgenics to respond to the sucrose analogue palatinose is not impaired. As in previous experiments, the absolute rates of metabolic fluxes are higher in the transformants than in wild-type discs.
Labelling of starch degradation products was also analysed in the experiment of Fig. 4. In discs incubated with 14C-glucose and mannitol, label in maltose (Fig. 4h) was not detectable in the wild type but represented 5 and 10% of the total 14C absorbed in SP-11 and SP-29, respectively. Label in soluble glucans (Fig. 4i) was not detectable in wild type and SP-11, but represented approximately 2% of the total label in SP-29. Similar results were obtained when palatinose was included in the incubation medium, except in the case of soluble glucans where palatinose feeding led to a decrease in the labelling of soluble glucans in SP-29. The results indicate that expression of sucrose phosphorylase leads to a stimulation of starch degradation.
Introduction of sucrose phosphorylase leads to depletion of sucrose and increased energy consumption by only a small increase of sucrose breakdown
Introduction of sucrose phosphorylase has widespread consequences for many aspects of tuber metabolism, including depletion of sucrose by only a moderate increase in the rate of sucrose breakdown, higher rates of glycolysis and respiration and an activation of SPS and consequent increase in the rate of sucrose re-synthesis. These changes are coupled to reduction in the starch content and the activities of starch synthesizing enzymes (Table 1) and larger pools of amino acids (Table 6). These marked changes in flux and tuber composition occur even though the unidirectional rate of sucrose breakdown is only increased about two-fold, and this increase is largely offset by a large stimulation of sucrose re-synthesis, with the result that the net rate of sucrose mobilization is not markedly stimulated (Table 4). This relatively small stimulation of sucrose breakdown is consistent with the relatively low activity of sucrose phosphorylase in the transformants, which was 10-fold lower than the activity of sucrose synthase (Table 1).
This comparison indicates that the changes we observe in tuber growth and metabolism may be due to specific features of sucrose phosphorylase, or the reaction which it catalyses, rather than a major change in the overall rate of sucrose breakdown per se. One difference between sucrose phosphorylase and sucrose synthase is that the differences in their kinetic properties mean that sucrose phosphorylase will allow it to drive sucrose concentrations much lower than sucrose synthase. Sucrose phosphorylase has a much higher affinity for sucrose (Km(sucrose) approximates 1 mm, see Silverstein et al. 1967) than sucrose synthase (Km(sucrose) 40–200 mm, see Avigad 1982), which is also subject to inhibition by fructose and glucose (Doehlert 1987; Dancer et al. 1990). Thermodynamic factors will also contribute to the depletion of sucrose, as the equilibrium constant of the reaction is further displaced towards sucrose degradation for sucrose phosphorylase than sucrose synthase, and the Pi concentration in the cytosol of tubers is about 100-fold higher than the UDP concentration (Loef, Stitt & Geigenberger 1999; and data not shown). Another major difference is that sucrose phosphorylase leads directly to formation of Glc-1-P, and does not require the reaction catalysed by UGPase, which will remove the requirement for PPi and uridine nucleotides and increase the energy requirement for sucrose mobilization. Expression of sucrose phosphorylase indeed resulted in a dramatic decrease of the steady-state sucrose content (Table 1), and in increased ATP consumption (Table 5) and a decreased ATP/ADP ratio and adenylate energy charge (Fig. 1).
These results imply that the metabolite levels maintained in growing tubers as a consequence of sucrose degradation via sucrose synthase are important for normal metabolic function. These include the maintenance of a relatively large sucrose pool which has functions in signalling and/or in cellular osmotic regulation, and the minimization of ATP consumption.
Introduction of sucrose phosphorylase had a complex effect on starch metabolism, leading to contradicting changes in intact tubers and tuber discs
Sucrose phosphorylase had a complex effect on starch synthesis. On the one hand, the higher rates of sucrose breakdown and elevated levels of glycolytic intermediates should allow allosteric activation of AGPase and stimulate starch synthesis. This was indeed observed in short-term labelling experiments with tuber discs. Nevertheless, allocation to starch was decreased relative to other major fluxes (Table 3), and the starch content of intact tubers was reduced in comparison with wild type (Table 1). Clearly, the introduction of sucrose phosphorylase leads to a series of changes in metabolism that interact to counteract the effects of high levels of hexose-phosphates and 3PGA, resulting in reduced rather than increased levels of starch in intact tubers.
First, consumption of ATP in the sucrose substrate cycle leads to a decrease of the cellular energy status (Fig. 1), which may restrict starch synthesis. There is mounting evidence that the rate of starch synthesis is restricted by the ATP supply, even in wild-type tubers. Studies with transgenic potato tubers with decreased and increased expression of the plastidic adenylate transporter indicate that starch synthesis is limited by the availability of ATP in the plastid (Tjaden et al. 1998; Geigenberger et al. 2001), and incubation of wild-type tuber discs with adenine to increase ATP levels led to a stimulation of starch synthesis from sucrose (Loef et al. 2001).
Second, sucrose phosphorylase leads to a dramatic decrease of the cellular sucrose concentration. It has been shown elsewhere that sucrose stimulates starch synthesis via a mechanism that operates independently of changes in the levels of the substrates and effectors of AGPase (Geigenberger et al. 1994; Geiger et al. 1998). Incubation of transgenic tuber discs in high external sucrose did not result in increased partitioning of label to starch (Table 3). Analysis of the cellular sucrose content at the end of the experiment revealed, however, that this is probably because the introduced enzyme activity prevents sucrose from accumulating even when high concentrations are supplied in the medium. A further set of experiments was therefore performed using the non-metabolizable sucrose analogue palatinose which stimulates allocation to starch in wild-type tuber discs (Fernie et al. 2001a). This also led to increased partitioning to starch in the transformants, relative to the rate of glycolysis (Fig. 4). The selective stimulation of starch synthesis after feeding palatinose indicates that low sucrose is one of the factors leading to the decreased allocation of carbon to starch in the transformants.
Third, several enzymes of starch synthesis show a decrease of activity in sucrose phosphorylase-expressing tubers. Studies with transgenic tubers with decreased expression of AGPase (Geigenberger et al. 1999a) or starch synthase (Marshall et al. 1996) document that small changes of the activity of individual enzymes do not alter starch accumulation significantly. It is possible, however, that a parallel change of both AGPase and starch synthase activities allow a more substantial alteration in starch accumulation. The mechanisms leading to this coordinated response in the transgenic tubers are unknown, but one possible explanation is that they too are linked to changes in the levels of sucrose.
Fourth, labelling studies indicate that starch degradation may be increased in these transformants, leading to an inhibition of net starch accumulation. Labelling of starch degradation products, like maltose and soluble glucans increased in the transformants (Fig. 4) and it is clear from the specific activities of the hexose phosphate pool that the transgenics have increased rates of isotopic dilution. It is not immediately clear why starch degradation should be increased, although an attractive explanation would be that it is connected with the low level of sucrose in the transformants.
The results of the present paper therefore indicate that a combination of factors decrease allocation to starch synthesis in intact tubers that overexpress sucrose phosphorylase. These include a shift in the relative activities of enzymes in the starch synthesis pathway and glycolysis, a greatly accelerated sucrose cycling leading to increased consumption of ATP, and depletion of sucrose to low levels which may lead to an inhibition of starch synthesis by a novel mechanism that has not yet been characterized. However, the absolute flux to starch in short-term feeding experiments with tuber discs was nevertheless still higher in the transformants than in wild-type tissue, indicating that further factors will also be required to explain the reduced starch content of the transgenic tubers. One possible explanation for this could be that growing potato tubers are essentially hypoxic (Geigenberger et al. 2000), whereas oxygen supply is not limiting in discs. Further studies are therefore required to determine the regulation of the dominant fluxes of carbohydrate metabolism in the transformants under the oxygen concentrations found inside growing tubers.
This work was supported by the Deutsche Forschungsgemeinschaft (Ge 878/1-1, grant to P.G. and A.T.).
Received 25 February 2002;received inrevised form 13 May 2002;accepted for publication 15 May 2002