Metabolite export and chlorophyll fluorescence
Triose phosphates and 3-PGA are the main carbohydrates released from mesophyll cell chloroplasts in the light (Lilley et al., 1977; Heber & Heldt, 1981). We tested whether this was also true for guard cell chloroplasts. The content of triose phosphates in the medium increased nearly linearly within 20 min in the light, as is shown in Fig. 1. By contrast, the content of 3-PGA remained constant at a low level. Even in darkness, 3-PGA was not exported, and triose phosphates were released in darkness only at a one-eighth of the rate measured in the light (data not shown).
Figure 1. Time dependent release of C3 metabolites of illuminated guard cell chloroplasts from Vicia faba. Immediately before the start of the illumination 0.5 mm Pi was added to the plastid suspension. The results of three experiments using separate chloroplast preparations are shown. Values are calculated for 100% intactness of the plastids. 3-PGA, 3-phosphoglycerate.
Download figure to PowerPoint
In the experiments leading to Fig. 1, NaHCO3 was omitted from the medium because, in initial investigations, we had found that in the presence of 5 mm NaHCO3, no significant export of 3-PGA had occurred and, surprisingly, triose phosphates were exported in the light and in the presence of NaHCO3 at only 77% of the rates measured in the absence of NaHCO3 (ranging from 56 to 114%, n = 4). The cause for this high variability is not known. Obviously, NaHCO3 was not required for maximum export of triose phosphates. Possibly, due to the suspension of a small plastid volume in a relatively large assay volume and due to the contamination of the chloroplasts with mitochondria, the CO2 content of the assay medium had sufficed to support CO2 fixation by Rubisco. Alternatively, triose phosphates were derived predominantly from the breakdown of starch and not from the carboxylation of ribulose-1,5-bisphosphate.
In a second series of experiments we directed our investigation to a possible export of hexose phosphates, glucose and maltose. Neuhaus & Schulte (1996) reported that isolated mesophyll cell chloroplasts of Mesembryanthemum crystallinum released these substances during the breakdown of starch. In addition, we examined effects of orthophosphate (Pi) on the export of phosphorylated and nonphosphorylated sugars. Because commercial sorbitol is significantly contaminated with glucose, we used mannitol as osmoticum in these experiments; the intactness of the plastids was unaffected by this modification.
In addition to triose phospates, guard cell chloroplasts exported hexose phosphates, glucose and maltose in considerable amounts (Table 2). Because starch from broken plastids is likely to be present in the suspension and the formation of neutral sugars does not depend on ATP or NADPH, we tested whether or not the appearance of glucose and maltose in the medium was related to the intactness of the plastids. Suspensions of untreated guard cell chloroplasts as well as guard cell chloroplasts lysed by addition of 0.025% Triton-X-100, were incubated for 30 min in the light. The increase in the contents of neutral sugars in the medium surrounding the untreated plastids was set to 100%. Compared with this control the contents of glucose and maltose in the suspension of lysed plastids increased by 18 and 6%, respectively. Similar results were obtained if the plastids were ruptured by osmotic shock (data not shown). We concluded that the release of glucose and maltose was due to export and not due to an artefact of the isolation procedure.
Table 2. Influence of Pi on metabolite release in the light
| ||Without Pi µmol (mg Chl)−1 h−1||µatom C (mg Chl)−1 h−1||0.5 mm Pi µmol (mg Chl)−1 h−1||µatom C (mg Chl)−1 h−1|
|Triose phosphates||2.3 ± 2.4|| 6.9||39.0 ± 11.3||117.0|
|Hexose phosphates||1.8 ± 2.5|| 10.8||12.9 ± 3.1|| 77.4|
|Glucose||9.6 ± 3.6|| 57.6||21.2 ± 5.0||127.2|
|Maltose||3.3 ± 1.4|| 39.6||13.7 ± 4.9||164.4|
|Total|| ||115|| ||486|
The transport of triose phosphates and hexose phosphates is accomplished by the triose phosphate-phosphate translocator and the glucose-6-phosphate-phosphate translocator. These two transporters catalyze the counter exchange of orthophosphate with the respective sugar phosphates (Flügge, 1998). Consequently, in the absence of orthophosphate in the medium, export of both triose phosphates and hexose phosphates was low, if not insignificant (Table 2). Remarkably, even the efflux of glucose and maltose decreased by factors of 2 and 4, respectively, if orthophosphate was lacking.
The mechanism underlying the stimulating effect of orthophosphate on the release of maltose and glucose remains to be identified. A similar observation, however, far less pronounced, was made by Stitt & Heldt (1981) when they investigated starch breakdown in illuminated mesophyll cell chloroplats of spinach. For pea mesophyll cell chloroplasts, Kruger & ap Rees (1983) reported an approximately three-fold increase in maltose production during starch breakdown in darkness if phosphate was added. They suggested that maltose phosphorylase was involved. However, to our knowledge maltose phosphorylase has never been purified from plant extracts and a gene encoding to a corresponding activity in plants was not reported so far.
With respect to the total flux of carbon, export of neutral sugars was more important than export of phosphorylated sugars (Table 2). The sugars most likely originated from hydrolytic starch breakdown. Transport of glucose across the membrane of mesophyll cell chloroplasts has been studied in detail by Servaites & Geiger (2002), and a putative plastidic glucose transporter has been cloned (Weber et al., 2000). Transport of maltose (and also glucose) across the envelope was documented for mesophyll chloroplasts (Neuhaus & Schulte, 1996; Rost et al., 1996) and amyloplasts of cauliflower buds (Neuhaus et al., 1995), however, no cDNA encoding for a plastidic maltose transporter has been identified so far. The release of neutral sugars from guard cell chloroplasts in considerable amounts is remarkable, not only with respect to the carbon supply for osmotica production during stomatal opening, but also in terms of the transport properties of the chloroplast envelope. Related to chlorophyll content, the rates of glucose and maltose export of guard cell chloroplasts were higher by factors of 85 and 37, respectively, than those of mesophyll cell chloroplasts of Mesembryanthemum crystallinum during starch degradation at physiological rates (as calculated from data in Table 3 of Neuhaus & Schulte, 1996).
The amounts of carbohydrates released by guard cell chloroplasts in vitro we report here are comparable in magnitude with in vivo values. We determined an export of metabolites from isolated guard cell chloroplasts in the light and in the presence of Pi that was equivalent to 486 µatom C (mg Chl)−1 h−1. In stomata opening within 1.5 h, Outlaw & Manchester (1979) observed a decrease in starch content of 72 mmol anhydroglucosyl equivalents per kg d. wt. With a dry mass of 6 ng per guard cell pair (Outlaw & Lowry, 1977) and a chlorophyll content of 1.8 pg per guard cell (this report) this value corresponds to starch breakdown of 480 µatom C (mg Chl)−1 h−1. Although the close numeric identity is possibly fortuitous, this result demonstrates that isolated plastids are able to release carbon at physiological rates. This conclusion is confirmed by a comparison of the export of carbohydrates from guard cell chloroplasts with the solute requirement for stomatal opening. In order to simplify the estimation we assume that only potassium malate served as osmoticum. During an opening movement lasting 2 h a guard cell of Vicia faba accumulates about 1300 fosmol h−1 (Reckmann et al., 1990), which corresponds to a required malate production of 540 fmol h−1 (activity coefficient of potassium malate = 0.8). Total carbon release from isolated guard cell chloroplasts was 486 µatom C (mg Chl)−1 h−1. After CO2 fixation by PEPcarboxylase, this allows a formation of malate at 162 µmol (mg Chl)−1 h−1 or 292 fmol cell−1 h−1, respectively, which corresponds to 54% of the demand. Because the plastids we isolated did not display their full photosynthetic activity, we assume that, in vivo, guard cell chloroplasts could make a higher contribution to the synthesis of malate than our estimates show.
What is the origin of the exported carbon? First we consider the possibility that the triose phosphates from the guard cell plastids were derived from 3-PGA produced by the Calvin cycle. The in vivo activity of Rubisco in guard cell protoplasts of Vicia faba was calculated to be 22 µmol (mg Chl)−1 h−1 (Reckmann et al., 1990, based on data of Gotow et al., 1988). This activity would suffice to produce triose phosphates at a rate of 7.3 µmol (mg Chl)−1 h−1; this is not more than about one-fifth of the measured average export of 39 µmol (mg Chl)−1 h−1 (Table 2). However, the in vitro activity of Rubisco from lysed guard cell protoplasts of Vicia faba was found to have been much higher than the in vivo activity: Shimazaki (1989) determined a Rubisco activity of 140 µmol (mg Chl)−1 h−1. Such a rate would correspond to a production of triose phosphates at a rate of 47 µmol (mg Chl)−1 h−1; this is slightly higher than the average of the rates of export we measured. But if all of the triose phosphates originating from Rubisco activity were exported, the Calvin cycle would be depleted of intermediates and stop. Nevertheless, release of triose phosphates continued linearly over 20 min (Fig. 1). We conclude that triose phosphates were mainly derived from starch breakdown. Robinson & Preiss (1987) analysed carbohydrate metabolizing enzymes in guard cells of Commelina communis. The enzymes necessary for phosphorolytic starch breakdown were present in the guard cell chloroplasts. By contrast, Robinson & Preiss (1987) could not detect activity of hexokinase (in both plastid fraction and nonplastid fraction). The authors stated, however, that this was possibly due to low activity or nonoptimal assay conditions. Hexokinase activity has been detected in preparations of guard cell chloroplasts of pea (c. 17 µmol mg Chl−1 h−1, S. Overlach & K. Raschke, unpublished), pea root plastids (Borchert et al., 1993), and inside cauliflower bud plastids (Journet & Douce, 1985). Thus, in guard cell chloroplasts, glucose-6-phosphate could be produced from starch via the phosphorolytic path and possibly also via the hydrolytic path and hexokinase. Triose phosphates could then be produced via glycolysis.
Alternatively, starch degradation products could have been used to prevent the depletion of Calvin cycle intermediates. Fructose-6-phosphate could have been produced via phosphoglucoisomerase from starch derived glucose-6-phosphate and then fed into the Calvin Cycle leading to the regeneration of the CO2 acceptor ribulose-1,5-bisphosphate. Also, glucose-6-phosphate could have entered the oxdative pentose phosphate path and provided a substrate for the production of ribulose-5-phosphate and ribulose-1,5-bisphosphate. (Unpublished results of P. Müller and K. Raschke showed that the dehydrogenases of the oxidative pentose phosphate path in guard cell chloroplasts were active in the light at rates equal in magnitude to that of Rubisco).
These are attractive possibilities because, in these cases, the generation of ribulose-1,5-bisphosphate would not depend on the activity of plastidic fructose-1,6-bisphosphatase (FBPase). Rather low activities of this enzyme in guard cells were reported by Robinson & Preiss (1987) and Shimazaki et al. (1989), and in our laboratory, presence of plastidic FBPase activity could not be verified in guard cells of both, Pisum sativum and Vicia faba (Hedrich et al., 1985, P. Müller & K. Raschke, unpublished). It appears that the functional plasticity of guard cells, as recognized by Outlaw & De Vlieghere-He (2001), Olsen et al. (2002), and Zeiger et al. (2002) among several causes, is based on a choice of available metabolic pathways.
By contrast to the low Rubisco activity (Reckmann et al., 1990) the activities of NADP-malate dehydrogenase and NADP-GAP-DH were especially high in guard cell chloroplasts (Shimazaki et al., 1989; Scheibe et al., 1990). It was therefore proposed that most of the redox power derived from photosynthetic electron transport in guard cell chloroplasts was used for the reduction of OAA and 3-PGA imported from the cytosol (Gotow et al., 1985; Shimazaki et al., 1989). If this was true, the addition of OAA and 3-PGA to isolated guard cell chloroplasts (lacking the surrounding cytosol) should stimulate photosynthetic electron transport. Indeed, the analysis of the chlorophyll fluorescence of isolated guard cell chloroplasts indicates that this was the case (Fig. 2). Following addition of OAA to the medium, photochemical quenching qQ = ([Fm–Fv]/Fm) increased from 0.31 to 0.42 within 20 s (calculated from the recording shown in Fig. 2a). Similarly, addition of 3-PGA led to an increase of qQ from 0.32 to 0.45 within 1 min (Fig. 2b). Obviously, OAA and 3-PGA were taken up by the plastid, photosynthetic electron transport was enhanced, and the necessity of energy dissipation by chlorophyll fluorescence was reduced. Triose phosphates produced by the reduction of 3-PGA were available for export. In fact, if 0.5 mm 3-PGA was present in the medium, export of triose phosphates increased to about 167% (109–222%, n = 3) of that of the control not supplied with 3-PGA. This finding supports the operation of a 3-PGA-triose phosphate shuttle as proposed by Shimazaki et al. (1989). Likewise, if OAA was provided and taken up by guard cell chloroplasts, malate was exported. In one experiment, the malate content of the surrounding medium increased by 12 µmol (mg Chl)−1 within 20 min. This effect was strictly light dependent.
Figure 2. Demonstration of photosynthetic reduction of oxaloacetate (OAA) and 3-phosphoglycerate (3-PGA) by reduction of chlorophyll fluorescence of guard cell chloroplasts from Vicia faba. Guard cell chloroplasts equivalent to 1 µg Chl were suspended in 1 ml chloroplast medium, supplemented with 0.5 mm Pi. Short saturating light pulses (2800 µmol m−2 s−1) were given every 40 s to determine maximal fluorescence (Fm). During the pulses, the modulation frequency of the measuring beam was increased from 1.6 kHz to 100 kHz. Start and end of illumination with actinic light (120 µmol m−2 s−1) is indicated by arrowheads pointing upwards or downwards, respectively. At 2–3 min after the start of illumination (a), 1.5 mm OAA or (b), 1 mm 3-PGA were applied. F0, fluorescence intensity of the dark adapted sample, measuring beam of negligible actinic intensity. Fv, variable fluorescence, observed upon illumination with actinic light.
Download figure to PowerPoint
In conclusion, in illuminated guard cell chloroplasts, starch degradation and photosynthetic electron transport can occur simultaneously. The breakdown products of starch are exported. Redox power and ATP produced in conjunction with photosynthetic electron transport is used for the conversion of 3-PGA (derived from Rubisco activity or imported from the cytosol) to triose phosphates, for the reduction of imported oxaloacetate to malate (to be returned to the cytoplasm), and for the phosphorylation of products of starch breakdown. Isolated guard cell chloroplasts exported reduced carbon at a rate one-half of that required for the malate production during stomatal opening.