Acclimation of plants to an increase in atmospheric carbon dioxide concentration is a well described phenomenon. It is characterized by an increase in leaf carbohydrates and a degradation of ribulose 1, 5-bisphosphate carboxylase protein (Rubisco) leading in the long term to a lower rate of CO2 assimilation than expected from the kinetic constants of Rubisco. This article summarizes studies with transgenic plants grown in elevated pCO2 which are modified in their capacity of CO2 fixation, of sucrose and starch synthesis, of triosephosphate and sucrose transport and of sink metabolism of sucrose. These studies show that a feedback accumulation of carbohydrates in leaves play only a minor role in acclimation, because leaf starch synthesis functions as an efficient buffer for photoassimilates. There is some evidence that in elevated pCO2, plants grow faster and senescence is induced earlier.
A large number of studies have shown that short-term elevation of atmospheric CO2 partial pressure (pCO2) leads to a higher rate of CO2 assimilation (for review see Drake, Gonzàlez-Meler & Long 1997). This observation agrees with the known kinetic properties of ribulose 1, 5-bisphosphate carboxylase protein (Rubisco), i.e. (i) at the current atmospheric CO2 concentration Rubisco is not CO2 saturated (Woodrow & Berry 1988, Andrews & Lorimer 1987); and (ii) oxygen competes with CO2 at the catalytic site of Rubisco and photorespiration decreases the efficiency of CO2 assimilation by about 20–30%. Therefore, a doubling of the atmospheric pCO2 increases carboxylation and inhibits the oxygenation reaction of Rubisco.
When plants were grown for an extended period of time in high pCO2 under conditions where light and nitrogen supply were not limiting, acclimation effects sometimes occurred and the stimulation of CO2 assimilation disappeared or was even reversed (Sage, Sharkey & Seemann 1989). Several studies describe this acclimation, which often is accompanied by higher carbohydrate content and lower levels of soluble proteins, especially Rubisco. The degree of acclimation differs between plant species (Sage et al. 1989) and in the same species differences were found when the rooting volume was varied (Arp 1991; for review Drake et al. 1997). This reduction in Rubisco protein was interpreted as a possible consequence of the disturbed source/sink balance (Stitt 1991). The surplus carbohydrates produced that cannot be exported and metabolized in sink tissues, leads to a feedback accumulation in source leaves and reduces the expression of photosynthetic genes by a mechanism involving sugar sensing (Krapp et al. 1993; Jang & Sheen 1997). Other observations indicate that the growth of plants in elevated pCO2 accelerated ontogenesis of the plant (Miller et al. 1997).
This article describes studies which were carried out to prove possible explanations for the acclimation to enhanced pCO2 by the use of transgenic plants. These plants are altered in different steps of CO2-assimilation, assimilate translocation or sink metabolism. It has been shown several times, that increasing or decreasing the activity of key enzymes, which are catalysing possible regulatory steps, by overexpression or antisense repression is an excellent way to analyse the basis of the acclimation process and to determine which steps are involved in the feedback accumulation of carbohydrates. Despite several publications describing the growth of transgenic plants under elevated CO2-concentrations, this article will consider only those which are suitable for understanding of the influence on one of these steps on the acclimation.
The response of transgenic plants grown in ambient and elevated pCO2 in the atmosphere were considered and compared with wild-type plants grown in the same conditions for the following: (i) altered Rubisco and Rubisco-activase activity (Sicher, Kremer & Rodermel 1994; He et al. 1997) (ii) altered triosephosphate translocator activity (Kauder et al. unpublished) (iii) increased sucrose phoshate synthase activity (Micallef et al. 1995) (iv) reduced capacity of phloem loading (Kauder et al. unpublished) or starch synthesis capacity (Ludewig et al. 1998) or (v) with reduced sink capacity in potato tubers (Ludewig et al. unpublished).
GROWTH CHARACTERISTICS OF WILD-TYPE POTATO PLANTS
Most of the work described in this article was carried out with transgenic potato plants. For an understanding of the results of the genetic transformations, some growth characteristics of wild-type potato plants in ambient and elevated pCO2 (100 Pa) are summarized. The determination of the growth rates of shoot and tubers of potato plants revealed that plants grown in elevated pCO2 grew faster (Ludewig et al. in preparation, Kauder et al. unpublished). Between the second and the third week after planting the increase in shoot fresh weight was doubled in plants grown in elevated pCO2, and after 5 weeks fresh weight decreased faster than in ambient pCO2 (Fig. 1a). After 3 weeks, the tuber yield was also increased and was about 50% higher after 7 weeks (Fig. 1b). When leaf metabolism in both growth conditions was compared after 7 weeks, only a few differences were found (Table 1). The content of sucrose was slightly increased whereas the starch content was dramatically higher than in the ambient controls. No changes were found in the content of Rubisco, a finding which is consistent with the expectation that in the absence of sink limitation no acclimation occur. It is clearly different from the results of most other experiments in elevated pCO2 (for review see Drake et al. 1997). However, the observations described here confirm those found with tobacco plants (Miller et al. 1997). Miller et al. (1997) have recently shown that the development of photosynthetic CO2 assimilation in high pCO2-grown tobacco plants was shifted temporarilly to an earlier maximum and subsequent senescent decline. From these studies it is more likely that the decline of Rubisco observed in several other studies is the result of accelerated senescence rather than a direct effect of elevated pCO2 on the expression of Rubisco.
Table 1. . CO2 assimilation and metabolite content of potato leaves grown in ambient (350 p.p.m.) or elevated (1000 p.p.m.) pCO2 for 7 weeks in a climatic chamber with a photon flux of 600 μmol m–2 s–1 in a 12 h light/12 h dark period. Leaf samples were harvested from fully developed source leaves at the end of the light period
ANTISENSE REPRESSION OF RUBISCO AND RUBISCO ACTIVASE
For a detailed inspection of the role of Rubisco in acclimation, Sicher et al. (1994) grew wild-type and Rubisco-antisense tobacco plants in ambient pCO2 and after 6 weeks transferred them into a growth chamber with 70 Pa pCO2. Nine days after transfer, the leaf Rubisco content had decreased by about 20% (Sicher et al. 1994). Similar observations were made when transgenic tobacco plants retaining 50% of their Rubisco activity were analysed (Sicher et al. 1994). The reduction of Rubisco protein in the transgenic lines only moderately altered the assimilation rates, carbonic anhydrase activity, chlorophyll and soluble protein content. These findings suggest that Rubisco is not a limiting step for plant productivity when grown in elevated pCO2. However, the amount of Rubisco protein varied, when the plants were subjected to an altered environment for a short period. Such a change leads to a new whole plant adjustment which requires a redistribution of all resources. This has to be separated from a direct CO2 effect. One possible response is the induction of senescence in mature leaves, which is characterized by a decreased Rubisco content. This mechanism is obviously induced in both control and Rubisco-antisense plants. A conclusive study allowing the separation of direct and indirect effects of CO2 is still missing and to this end, Rubisco-antisense plants should be grown under constant conditions of elevated pCO2.
A second approach to the analysis of the role of Rubisco during acclimation of tobacco plants has been descibed by He et al. (1997). Rubisco-activase-antisense expressing plants were grown in elevated pCO2 (150 Pa) and analysed for developmental changes of the old and youngest source leaves. In this study the growth of transgenic plants was found to be retarded. In old leaves the catalytic turnover of Rubisco was reduced as a consequence of senescence, and elevated pCO2 was not involved in the adjustment of Rubisco protein. He et al. (1997) speculated that the observed effects may be explained by an age-dependent decrease in the activase to Rubisco ratio, but this does not explain the age-dependent reduction in catalytic activity of Rubisco. Additionally, He et al. (1997) have compared the Rubisco content of Rubisco-activase-antisense plants with NADP-glyceraldehyd-3-phosphate dehydrogenase (NADP-GAPDH)-antisense plants to determine whether sugar signals might be involved in the regulation of Rubisco content. Both lines were impaired in the rate of photosynthesis to the similar extent. The age-dependent delay was only found in Rubisco-activase-antisense plants in Rubisco degradation, but not in the NADP-GAPDH-antisense plants. From this observation He et al. (1997) excluded the involvement of a general sugar signal for the regulation of the Rubisco content.
In summary, modification of the amount of Rubisco protein or reducing the enzyme activation state have shown that Rubisco degradation, often observed when plants were submitted to elevated pCO2, is possibly a result of the induction of senescence in fully developed leaves, when transferred from ambient into elevated pCO2 rather than a problem of nitrogen redistribution. Rubisco activase interferes with this induction of senescence in an unknown way.
RESTRICTION OF TRIOSEPHOSPHATE EXPORT FROM THE LEAF CHLOROPLAST
Triosephosphates are the products of the CO2-fixation of the chloroplast. Their export into the cytosol is catalyzed by the triosephosphate translocator (TPT) in counter-exchange with inorganic phosphate. The TPT has previously been shown to constitute one of the key steps in controlling the partitioning of carbon between starch and sucrose (Riesmeier et al. 1993a; Heineke et al. 1994). In these growth conditions of ambient pCO2 and a photon flux of 300 μmol m–2 s–1, 80% of the photoassimilates accumulated in the chloroplasts as starch, with no differences in pCO2 assimilation and tuber yield. From these findings Heineke et al. (1994) concluded that potato leaves have a sufficiently large capacity for starch synthesis to transiently store most of the carbon assimilated during the light period. In the dark period the starch which was accumulated in the light was degraded and possibly exported out of the chloroplast via the hexose transporter (Schäfer, Heber & Heldt 1977). This alternative route of carbon export out of the chloroplast bypasses the deficiency in the triosephosphate translocator protein without reducing the rate of CO2 assimilation or tuber yield.
To test if this bypass is sufficient under conditions of much higher carbon fluxes, TPT-antisense plants were grown in elevated pCO2 of 100 Pa and a photon flux of about 600 μmol m–2 s–1. No obvious phenotypic differences to wild-type plants were found and the rates of CO2 assimilation, chlorophyll and Rubisco content were unaltered (Table 2). As already described, starch, which was low in the leaves of wild-type plants in ambient pCO2, increased in elevated pCO2. In addition, the TPT-antisense plants accumulate much higher starch levels than wild-type plants under ambient pCO2, as already discussed but under elevated pCO2 this difference disappeared. A lower sucrose level was found in the leaves of TPT plants under both growth conditions and the tuber yield of TPT plants was slightly reduced. These observations are slightly different from those in earlier experiments (Heineke et al. 1994), possibly as the consequence of the higher photon flux which resulted in a four-fold increase in CO2 assimilation. These observations show that starch metabolism can nearly compensate for the reduced TPT-activity even under high flux rates. From these observations it seems to be unlikely that in leaves of wild-type plants starch and sucrose synthesis are limiting for the rate of CO2-fixation.
Table 2. . CO2 assimilation and metabolite content of leaves from potato plants expressing an antisense gene of the chloroplastic triosephosphate translocator. Growth conditions are described in Table 1
INCREASE OF SUCROSE SYNTHESIS BY EXPRESSING MAIZE SUCROSE-PHOSPHATE SYNTHASE IN TRANSGENIC TOMATO PLANTS
In the cytosol, triosephosphates are converted to sucrose. One key enzyme of sucrose synthesis is the enzyme sucrose-phosphate synthase. Tomato plants overexpressing maize sucrose-phosphate synthase (SPS) are characterized by an increased sucrose to starch ratio when grown in ambient pCO2, but the assimilation rates of the leaves were unchanged (Galtier et al. 1993). These authors mentioned that, when SPS is catalyzing a limiting step, one might expect these transgenic tomato plants to respond to elevated pCO2 with higher CO2 assimilation and/or higher yield. Micallef et al. (1995) have found that the assimilation rates of SPS-overexpressing plants were higher when grown in an elevated pCO2 of 65 Pa. They concluded that the limitation of photosynthesis imposed by end-product synthesis was reduced. The yield of SPS transformants grown in elevated pCO2 was identical with that of control plants. This could reflect the determinate nature of growth of these tomato lines (Micallef et al. 1995) Unfortunately, in this article no data for sucrose and starch contents in elevated pCO2 were shown, but from 14C partitioning studies there is evidence for a stimulation of sucrose synthesis in maize SPS-expressing plants. Surprisingly, the Rubisco activity of both control and maize SPS-expressing plants was not reduced in elevated pCO2, when the plants were 40-day-old. These results are in agreement with those of the Rubisco-activase-antisense plants described above. There is no evidence for the model of sink limitation of photosynthesis in elevated pCO2, involving an increase in leaf carbohydrate content and a downregulation of photosynthetic genes (Drake et al. 1997).
REDUCTION OF PHLOEM LOADING OF SUCROSE BY NONSPECIFIC OR PHLOEM-SPECIFIC-ANTISENSE REPRESSION OF THE SUCROSE TRANSPORTER IN TRANSGENIC POTATO PLANTS
Sucrose synthesized in the mesophyll cells of source leaves is exported into the phloem to supply sink tissues. In potato, apoplastic sucrose loading into the sieve tube/companion cell complex is an essential step of photoassimilate allocation. Restriction of phloem loading by ectopic targeting of a yeast invertase into the apoplast or by antisense repression of the sucrose transporter lead to high accumulation of soluble sugars and to an inhibition of CO2 assimilation (von Schaewen et al. 1990; Heineke et al. 1992; Riesmeier, Hirner & Frommer 1993b). Transgenic plants with reduced sucrose export capacity are suitable tools to further test the hypothesis of sugar-dependent down regulation of photosynthetic enzymes in elevated pCO2. Two sets of sucrose-transporter-antisense plants were grown under elevated pCO2. In one set of plants expression of the SUT antisense gene was driven by the constitutive 35S CaMV promoter (lines αSUT1–35S 17, 34 and 43, Riesmeier et al. 1993b). In a second set the RolC promoter was used (lines αSUT1-P 39, 44 and 50, Kühn et al. 1996). There were no obvious phenotypic differences between both types of transgenic plants. Therefore, only data from one set (αSUT1-P plants) are shown in Table 3. When transformants were grown in elevated pCO2 of 100 Pa, the phenotypic effects observed in ambient pCO2 (leaf curling, chlorosis, retardation of development, increased shoot-to-root ratio, reduced root and tuber production) were reinforced and leaf carbohydrate content increased (Table 3). In elevated pCO2, chlorophyll and Rubisco content were slightly reduced in wild-type plants (Table 1) and the transformant line αSUT1-P 39 (Table 3) and in both lines this reduction was independent from the degree of hexose accumulation. The line αSUT1-P 44 had similar Rubisco and chlorophyll contents in ambient pCO2 as wild-type leaves and in elevated pCO2 the reduction of both contents was accompanied by an accumulation of hexoses. In αSUT1-P 50 chlorophyll and Rubisco were lower in ambient pCO2 and even more reduced in elevated pCO2. In both conditions high hexose contents were found. These observations show that a lower chlorophyll and Rubisco content in elevated pCO2 are not strongly correlated with the increase in hexose levels in the leaves. Therefore, these experiments confirm the conclusion of the other sections that hexose accumulation does not seem to be involved in downregulation of photosynthesis.
Table 3. . CO2 assimilation and metabolite content of leaves of potato plants expressing an antisense gene of the sucrose transporter in the phloem tissue. Growth conditions are described in Table 1
A second point of this experiment should briefly be addressed. In all antisense plants the rates of photosynthesis in elevated pCO2 were lower than in ambient pCO2. Surprisingly, tuber yield in elevated pCO2 was increased by about 30% as compared with controls. This observation argues against the role of sugar accumulation itself for the reduction of Rubisco content. Possibly the reduced Rubisco content is the result of an acceleration of ontogeny of transgenic plants in elevated pCO2, which is modulated by sugar-dependent signals.
REDUCTION OF THE CAPACITY OF STARCH SYNTHESIS IN LEAVES
Leaves of many crop plants accumulate starch during the light period, which is mostly degraded during the following night. Starch functions as a buffer for carbohydrates to allow a continous export of assimilates even during the dark period. The extent of starch accumulation is under genetic and environmental control. One of the most obvious changes in leaf metabolism of wild-type potato plants resulting from long-term exposure to elevated pCO2 is the dramatic increase in leaf starch content (Table 1). In ambient pCO2 the starch that accumulated during the light period was mainly degraded during the following dark period. In fully expanded leaves of plants grown in elevated pCO2 a high proportion of the starch is not degraded during the night. Only 10% of the starch found at the end of the light period was degraded during the night. The remaining starch accumulated during leaf development and might be the result of an imbalance between CO2 assimilation and carbohydrate export. The role of this starch buffer was analysed by growing potato plants in elevated pCO2 of 100 Pa, in which the activity of the key enzyme of starch synthesis, the ADP-glucose pyrophosphorylase (AGPase), was inhibited by leaf-specific antisense repression (Ludewig et al. 1998). It has previously been shown that in ambient pCO2 a reduction of leaf starch synthesis by antisense repression of AGPase reduces the chloroplast starch content, without having any effect on CO2-assimilation or tuber yield (Leidreiter et al. 1995). When AGPase-antisense transformants were grown in elevated pCO2, the starch accumulation was slightly increased when compared with ambient pCO2, but it was much lower than in wild-type plants grown in elevated pCO2. The inability to store carbon as starch resulted in a reduced rate of CO2 assimilation, and an inverse correlation was found between the ability to accumulate starch and the reduction in CO2 assimilation (Ludewig et al. 1998). This observation clearly shows that the chloroplastic starch pool is necessary to allow a constant rate of CO2 assimilation. These transgenic potato plants obviously are limited by end-product synthesis capacity.
TUBER SPECIFIC MODULATION OF SINK CAPACITY BY ANTISENSE REPRESSION OF SUCROSE SYNTHASE OR OF ADP-GLUCOSE PYROPHOSPHORYLASE
Stitt (1991) has postulated that acclimation of plants to elevated pCO2 is induced by an insufficient sink demand. It was earlier shown that the removal of sink organs in plants led to an increase in leaf carbohydrate levels and a reduction of CO2 assimilation (for review see Herold 1980). For the experimental proof of this ‘sink limitation’ in elevated pCO2 of 100 Pa two sets of transgenic potato plants, which were reduced in their sink strength by tuber-specific antisense repression of sucrose synthase or ADP-glucose pyrophosphorylase were selected. Both sets of transformants do not show any differences in leaf phenotype or metabolism, when grown in ambient pCO2 (Zrenner et al. 1995; Ludewig et al. unpublished). Tuber yield was reduced and the composition changed, whereas shoot fresh weight increased, compared with wild-type plants. These results demonstrate that reduced sink strength of the tuber can be compensated by stimulated growth of alternative sinks, i.e. the shoot. On the basis of this flexibility, carbon demand can be kept constant, resulting in unchanged rates of CO2 assimilation. Under elevated pCO2 both sets of transformants show the same tendencies of acclimation as wild-type plants: unchanged rates of CO2 assimilation, a decrease in the shoot : tuber ratio and an increase in tuber yield. The shoot fresh weight of transformants was increased in elevated pCO2 as it was in ambient pCO2 (Sonnewald, unpublished).
SUMMARY AND FURTHER ASPECTS
This article summarizes studies on transgenic plants with altered enzymes or transporters involved in carbohydrate metabolism in elevated pCO2. These studies are discussed in order to prove two hypotheses: (i) does the ability to export excess carbon into storage sinks determine the capacity of plant species to cope with elevated atmospheric pCO2 and (ii) does the acclimation of leaf metabolism to elevated pCO2 result from a limiting sink capacity? Stitt (1991) has speculated that limiting sink capacity increase leaf sugar content, which causes the ‘sugar-mediated’ downregulation of expression of photosynthetis.
With respect to the first question, from the studies of the last section it can be concluded that the ability of plants to create sinks might be an important factor of acclimation, but these sinks do not necessarily need to be storage sinks. Differences between species may be a consequence of flexibility. The present results do no support ‘sugar-mediated’ downregulation as the reason for leaf acclimation to elevated pCO2. They indicate that the growth of plants in elevated pCO2 is accelerated and senescence is induced earlier. These observations imply that the reduction in the Rubisco content that is often observed possibly reflects the developmental stage of the investigated leaves rather than the environmental conditions, i.e. elevated pCO2. In the period between the second and the third week, the potato plants were found to grow faster, and in this period the rate of CO2 assimilation was increased in elevated pCO2. The studies with transgenic plants with antisense repression of TPT or AGPase indicate that during this period the capacity of sucrose and starch synthesis possibly limits the CO2 assimilation. This might be caused by the limitation on end-product synthesis. Further studies are needed to verify this limitation more precisely.
The work of the authors was supported by the Deutsche Forschungsgemeinschaft (SPP ‘Stoffwechsel und Wachstum der Pflanze unter erhöhter CO2-Konzentration’).