Coordination of the cell-specific distribution of the four subunits of glycine decarboxylase and of serine hydroxymethyltransferase in leaves of C3-C4 intermediate species from different genera

The cell-specific distribution of the four subunit proteins (P, L, T and H) of glycine decarboxylase (GDC) and of serine hydroxymethyltransferase (SHMT) has been studied in the leaves of C₃-C₄ intermediate and C₄ species of three genera (Flaveria, Moricandia and Panicum) using immunogold localization. Antibodies raised against these proteins from pea leaf mitochondria were used to probe Western blots of total leaf proteins of F. linearis Lag., M. arvensis (L.) DC and P. milioides Nees ex Trin. (C₃-C₄), and F. trinervia (Spring.) Mohr and P. miliaceum (L.) (C₄). For all species, each antibody recognised specifically a protein of similar molecular weight to that in pea leaves. In leaves of M. arvensis the P protein was present in the mitochondria of the bundle-sheath cells but was undetectable in those of the mesophyll, whereas the L, T and H proteins and SHMT were present in both cell types. The density of immunogold labelling of SHMT on the mitochondria of mesophyll cells was less than that on those of the bundle-sheath cells, which correlates with the relative activities of SHMT in these cell types. These data reveal that the lack of functional GDC in the mesophyll cells of M. arvensis, which is the principal biochemical reason for reduced photorespiration in this species, is due to the loss of a single subunit protein. This lack of coordinate expression of the subunit proteins of GDC within a photosynthetic cell represents a clear difference between M. arvensis and other C₃ and C₃-C₄ species. None of the GDC proteins was detectable in the mesophyll cells of the C₃-C₄ and C₄ Flaveria and Panicum species but all were present in the bundle-sheath cells. The differences in the distribution of the GDC proteins in leaves of the C₃-C₄ species studied are discussed in relation to the evolution of photosynthetic mechanisms.     

Leaf anatomy of c(3)-c(4) species as related to evolution of c(4) photosynthesis

This study was undertaken to examine the degree of Kranz anatomy development in the species intermediate to C₃ and C₄ types (C₃-C₄) in Panicum, Neurachne, Flaveria, and Moricandia. In each genus, C₃ and/or C₄ species were used for comparison. Leaf transections from each species were examined by light and transmission electron microscopy. The percentages of leaf photosynthetic cell profiles partitioned to bundle sheaths were higher in C₄ than in C₃ species, while C₃-C₄ species tended to be in between. However, percentages for C₃-C₄ species in Moricandia and some C₃-C₄Flaveria species were not greater than C₃. When expressed on a cell profile area basis, C₃-C₄ species partitioned more photosynthetic tissue to bundle sheaths than C₃ species in Moricandia, but not in Flaveria. Neurachne minor S. T. Blake (C₃-C₄) partitioned a very small portion of cell profile area to the inner bundle sheaths (5%) compared to Neurachne munroi F. Muell (C₄) (21%). The percentage of organelles partitioned to bundle sheaths was much greater in C₃-C₄ than in C₃ species. The average C₃ percentages for mitochondria plus peroxisomes were 19, 8, and 19.5% for Neurachne, Flaveria, and Moricandia, respectively, compared to 41, 29, and 46.5% for the C₃-C₄ species. The CO₂ compensation concentration was negatively related to the partitioning of tissue to bundle sheaths and to the percentage of organelles in bundle sheaths. It is concluded that all of the C₃-C₄ species examined have developed some degree of Kranz anatomy and that this altered anatomy is involved in their reduced apparent photorespiration.     

Molecular phylogeny of Flaveria as deduced from the analysis of nucleotide sequences encoding the H-protein of the glycine cleavage system

Sequence comparisons have shown that nucleotide sequences of the H-protein, a component of the glycine cleavage system, are only moderately conserved and can be used as molecular markers for intrageneric phylogenetic studies. We have analysed the respective cDNA sequences from 12 species of Flaveria, and a more limited set of gdcsH upstream regions. These data are discussed with respect to a phylogenetic reconstruction of Flaveria, a small genus which includes species of different photo-synthetic types, namely C₃, C₃-C₄, C₄-like and C₄. Our analysis essentially supports an earlier hypothesis, based on morphological and eco-geographical data, of the evolution of Flaveria (Powell 1978). This close agreement shows the usefulness of H-protein nucleotide sequences at a low taxonomic level. Our analysis independently confirms that C₄ photosynthesis has evolved two times in different lineages of Flaveria. Most remarkably, the C₄ taxa of Flaveria appear as derived relative to the C₃-C₄ intermediate taxa, i.e. they probably have common direct predecessors. This is the first direct evidence for a phylo-genetically intermediate position of C₃-C₄ intermediate photosynthesis. Our data also confirm the antiquity of C₃ photosynthesis in Flaveria but suggest that the collection of F.pringlei used in our experiments, although clearly of C₃ photosynthetic metabolism, possibly originated from hybridization with a more advanced taxon.     

Reassimilation of carbon dioxide by Flaveria (Asteraceae) species representing different types of photosynthesis

The capability to reassimilate CO₂ originating from intracellular decarboxylating processes connected with the photorespiratory glycolate pathway and-or decarboxylation of C₄ acids during C₄ photosynthesis has been investigated with four species of the genus Flaveria (Asteraceae). The C₃-C₄ intermediate species F. pubescens and F. anomala reassimilated CO₂ much more efficiently than the C₃ species F. cronquistii and, with respect to this feature, behaved similarly to the C₄ species F. trinervia. Therefore, under atmospheric conditions the intermediate species photorespired with rates only between 10–20% of that measured with F. cronquistii. At low oxygen concentrations (1,5%) the reassimilation potential of F. anomala approached that of F. trinervia and was distinct from that found with F. pubescens. The data are discussed with respect to a possible sequence of events during evolution of C₄ photosynthesis. If compared with related data for C₃-C₄ intermediate species from other genera they support the hypothesis that, during evolution of C₄ photosynthesis, an efficient capacity for CO₂ reassimilation evolved prior to a CO₂-concentrating mechanism.Abbreviations C₃, C₄ assimilated CO₂ initially found in 3-phosphoglycerate (C₃) or malate and aspartate (C₄) - D reassimilation coefficient - R _n , R _t net, total CO₂ evolution as measured with 0.03 and 3% CO₂, respectively - RuBP ribulose-1,5-bisphosphate - TPS true photosynthesis     

The relative contributions of reduced photorespiration,and improved water-and nitrogen-use efficiencies,to the advantages of C3−C4 intermediate photosynthesis in Flaveria

Summary Analyses of carbon-assimilation patterns in response to intercellular CO₂ concentrations, and the photosynthetic water-and nitrogen-use efficiencies, were conducted for a C₃, a C₄, and three C₃–C₄ species in the genus Flaveria in order to determine some of the advantages and disadvantages of C₃–C₄ intermediate photosynthesis. Operational intercellular CO₂ partial pressures (pi), determined when the atmospheric CO₂ partial pressure (pa) was approximately 330 bar, in the C₃–C₄ species were generally equal to, or greater than, those observed in the C₃ species under well-watered or water-stressed conditions. This reflects equal, or lower, water-use efficiencies (WUEs) in the C₃–C₄ species. The only case in which higher WUEs were observed in the C₃–C₄ species, compared to the C₃ species, was when photosynthesis rates were limited by available nitrogen and were less than 12.5 mol CO₂ m^-2s^-1. At higher photosynthesis rates, the C₃–C₄ species exhibited lower values of photosynthesis rate for equal values of stomatal conductance (lower WUE), compared to the C₃ species. Comparing slopes for the linear regions of the relationship between leaf nitrogen content and net photosynthesis rate (taken as an index of photosynthetic nitrogen-use efficiency, NUE), the C₄ species exhibited the highest NUE, followed by the C₃–C₄ species, F. ramosissima, with the other two C₃–C₄ species and the C₃ species being equal and exhibiting the lowest NUEs. The lack of consistent advantages in NUE and WUE in the C₃–C₄ species F. pubescens and F. floridana suggest that in some C₃–C₄ Flaveria species C₄-like anatomy and biochemistry do not provide the same gas exchange advantages that we typically attribute to the CO₂-concentrating mechanism of fully-expressed C₄ plants.     

Photosynthesis research on yellowtops: Macroevolutionn in progress

The vast majority of angiosperms, including most of the agronomically important crop plants (wheat, etc.), assimilate CO₂ through the inefficient C₃ pathway of photosynthesis. Under ambient conditions these organisms loose about 1/3 of fixed carbon via photorespiration, an energetically wasteful process. Plants with C₄ photosynthesis (such as maize) eliminate photorespiration via a biochemical CO₂-pump and thus have a larger rate of carbon gain. The genus Flaveria (yellowtops, Asteraceae) contains not only C₃ and C₄ species, but also many C₃-C₄ intermediates, which have been interpreted as evolving from C₃ to fully expressed C₄ metabolism. However, the evolutionary significance of C₃-C₄Flaveria-intermediates has long been a matter of debate. A well-resolved phylogeny of nearly all Flaveria species has recently been published. Here, we review pertinent background information and combine this novel phylogeny with physiological data. We conclude that the Flaveria species complex provides a robust model system for the study of the transition from C₃ to C₄ photosynthesis, which is arguably a macroevolutionary event. We conclude with comments relevant to the current Intelligent Design debate.     

Effects of polyploidy on photosynthesis

In polyploid plants the photosynthetic rate per cell is correlated with the amount of DNA per cell. The photosynthetic rate per unit leaf area is the product of the rate per cell times the number of photosynthetic cells per unit area. Therefore, the photosynthetic rate per unit leaf area will increase if there is a less than proportional increase in cell volume at higher ploidal levels, or if cell packing is altered to allow more cells per unit leaf area. In autopolyploids (Medicago sativa, C₃ species, and Pennisetum americanum, C₄ species) there is a doubling of photosynthesis per cell and of cell volume in the tetraploid compared to the diploid. However, there is a proportional decrease in number of cells per unit leaf area with this increase in ploidy such that the rate of photosynthesis per leaf area does not change. There is more diversity in the relationship between ploidal level (gene dosage) and photosynthetic rates per unit leaf area in allopolyploids. This is likely to reflect the effects of natural selection on leaf anatomy, and novel genetic interactions from contributed genomes which can occur with allopolyploidy. In allopolyploid wheat (C₃ species) a higher cell volume per unit DNA at the higher ploidal level is negatively correlated with photosynthesis rate per unit leaf area. Although photosynthesis per cell increases with ploidy, photosynthesis per leaf area decreases, being lowest in the allohexaploid, cultivated bread wheat (Triticum aestivum). Alternatively, doubling of photosynthetic rate per cell with doubling of DNA, with apparent natural selection for decreased cell volume per unit DNA, results in higher rates of photosynthesis per leaf area in octaploid compared to tetraploid Panicum virgatum (C₄) which may be a case of allopolyploidy. Similar responses probably occur in Festuca arundinacea. Therefore, in some systems anatomical factors affecting photosynthesis are also affected by ploidal level. It is important to evaluate that component as well as determining the effect on biochemical processes. Current information on polyploidy and photosynthesis in several species is discussed with respect to anatomy, biochemistry and bases for expressing photosynthetic rates.Abbreviations Chl chlorophyll - RuBPC ribulose-1,5-bisphosphate carboxylase     

Molecular properties of phosphoenolpyruvate carboxylase from C3, C3−C4 intermediate,and C4 Flaveria species

Phosphoenolpyruvate carboxylase (PEPCase; EC 4.1.1.31) from Flaveria trinervia Mohr (C₄), F. floridana Johnston (C₃–C₄), and F. cronquistii Powell (C₃) leaves were compared by electrotransfer blotting/enzyme-linked immunoassay (Western-blot analysis), mobility of the native enzyme in polyacrylamide gels and in isoelectric focusing (IEF) gels, peptide mapping, and in-vitro translation of RNA isolated from each plant. The PEPCases from the C₃ and C₃–C₄ plants were very similar to each other in terms of electrophoretic mobilities on gels and isoenzyme patterns on IEF gels, and identical in peptide mapping. Quantitative differences were noted, however, in that the C₃–C₄ intermediate plant contained more PEPCase overall and that the relative activity of individual isoenzymes shifted between the C₃ and C₃–C₄ intermediate PEPCases. The PEPCase from the C₄ plant had a different isoenzyme pattern, a different peptide map, and was far more abundant than the other two enzymes. Western blot analysis demonstrated the cross-reactivity of PEPCases from all three Flaveria species with antibody raised against maize PEPCase. The results provide evidence, at the molecular level, that supports the view of C₃–C₄ intermediate species as C₃-like plants with some C₄-like photosynthetic characteristics, but there are differences from the C₃ plant in the quantity and properties of the PEPCase from the C₃–C₄ intermediate plant.Abbreviations IEF isoelectric focusing - kDa kilodalton - PEPCase phosphoenolpyruvate carboxylase - Rubisco Ribulose-1,5-bisphosphate carboxylase/oxygenase     

Glycine decarboxylase is confined to the bundle-sheath cells of leaves of C3−C4 intermediate species

Immunogold labelling has been used to determine the cellular distribution of glycine decarboxylase in leaves of C₃, C₃–C₄ intermediate and C₄ species in the genera Moricandia, Panicum, Flaveria and Mollugo. In the C₃ species Moricandia foleyi and Panicum laxum, glycine decarboxylase was present in the mitochondria of both mesophyll and bundle-sheath cells. However, in all the C₃–C₄ intermediate (M. arvensis var. garamatum, M. nitens, M. sinaica, M. spinosa, M. suffruticosa, P. milioides, Flaveria floridana, F. linearis, Mollugo verticillata) and C₄ (P. prionitis, F. trinervia) species studied glycine decarboxylase was present in the mitochondria of only the bundle-sheath cells. The bundle-sheath cells of all the C₃–C₄ intermediate species have on their centripetal faces numerous mitochondria which are larger in profile area than those in mesophyll cells and are in close association with chloroplasts and peroxisomes. Confinement of glycine decarboxylase to the bundle-sheath cells is likely to improve the potential for recapture of photorespired CO₂ via the Calvin cycle and could account for the low rate of photorespiration in all C₃–C₄ intermediate species.Abbreviation and symbol kDa kilodaltons - CO₂ compensation point     

Co-function of C3-and C4-photosynthetic pathways in C3, C4 and C3-C4 intermediate Flaveria species

The potential for C₄ photosynthesis was investigated in five C₃-C₄ intermediate species, one C₃ species, and one C₄ species in the genus Flaveria, using ¹⁴CO₂ pulse-¹²CO₂ chase techniques and quantum-yield measurements. All five intermediate species were capable of incorporating ¹⁴CO₂ into the C₄ acids malate and aspartate, following an 8-s pulse. The proportion of ¹⁴C label in these C₄ products ranged from 50–55% to 20–26% in the C₃-C₄ intermediates F. floridana Johnston and F. linearis Lag. respectively. All of the intermediate species incorporated as much, or more, ¹⁴CO₂ into aspartate as into malate. Generally, about 5–15% of the initial label in these species appeared as other organic acids. There was variation in the capacity for C₄ photosynthesis among the intermediate species based on the apparent rate of conversion of ¹⁴C label from the C₄ cycle to the C₃ cycle. In intermediate species such as F. pubescens Rydb., F. ramosissima Klatt., and F. floridana we observed a substantial decrease in label of C₄-cycle products and an increase in percentage label in C₃-cycle products during chase periods with ¹²CO₂, although the rate of change was slower than in the C₄ species, F. palmeri. In these C₃-C₄ intermediates both sucrose and fumarate were predominant products after a 20-min chase period. In the C₃-C₄ intermediates, F. anomala Robinson and f. linearis we observed no significant decrease in the label of C₄-cycle products during a 3-min chase period and a slow turnover during a 20-min chase, indicating a lower level of functional integration between the C₄ and C₃ cycles in these species, relative to the other intermediates. Although F. cronquistii Powell was previously identified as a C₃ species, 7–18% of the initial label was in malate+aspartate. However, only 40–50% of this label was in the C-4 position, indicating C₄-acid formation as secondary products of photosynthesis in F. cronquistii. In 21% O₂, the absorbed quantum yields for CO₂ uptake (in mol CO₂·[mol quanta]^-1) averaged 0.053 in F. cronquistii (C₃), 0.051 in F. trinervia (Spreng.) Mohr (C₄), 0.052 in F. ramosissima (C₃-C₄), 0.051 in F. anomala (C₃-C₄), 0.050 in F. linearis (C₃-C₄), 0.046 in F. floridana (C₃-C₄), and 0.044 in F. pubescens (C₃-C₄). In 2% O₂ an enhancement of the quantum yield was observed in all of the C₃-C₄ intermediate species, ranging from 21% in F. ramosissima to 43% in F. pubescens. In all intermediates the quantum yields in 2% O₂ were intermediate in value to the C₃ and C₄ species, indicating a co-function of the C₃ and C₄ cycles in CO₂ assimilation. The low quantum-yield values for F. pubescens and F. floridana in 21% O₂ presumably reflect an ineffcient transfer of carbon from the C₄ to the C₃ cycle. The response of the quantum yield to four increasing O₂ concentrations (2–35%) showed lower levels of O₂ inhibition in the C₃-C₄ intermediate F. ramosissima, relative to the C₃ species. This indicates that the co-function of the C₃ and C₄ cycles in this intermediate species leads to an increased CO₂ concentration at the site of ribulose-1,5-bisphosphate carboxylase/oxygenase and a concomitant decrease in the competitive inhibition by O₂.Abbreviations PEP phosphoenolpyruvate - PGA 3-phosphoglycerate - RuBP ribulose-1,5-bisphosphate     

Photosynthetic Characteristics of C(3)-C(4) Intermediate Flaveria Species : I. Leaf Anatomy, Photosynthetic Responses to O(2) and CO(2), and Activities of Key Enzymes in the C(3) and C(4) Pathways

Four species of the genus Flaveria, namely F. anomala, F. linearis, F. pubescens, and F. ramosissima, were identified as intermediate C₃-C₄ plants based on leaf anatomy, photosynthetic CO₂ compensation point, O₂ inhibition of photosynthesis, and activities of C₄ enzymes. F. anomala and F. ramosissima exhibit a distinct Kranz-like leaf anatomy, similar to that of the C₄ species F. trinervia, while the other C₃-C₄ intermediate Flaveria species possess a less differentiated Kranz-like leaf anatomy. Photosynthetic CO₂ compensation points of these intermediates at 30°C were very low relative to those of C₃ plants, ranging from 7 to 14 microliters per liter. In contrast to C₃ plants, net photosynthesis by the intermediates was not sensitive to O₂ concentrations below 5% and decreased relatively slowly with increasing O₂ concentration. Under similar conditions, the percentage inhibition of photosynthesis by 21% O₂ varied from 20% to 25% in the intermediates compared with 28% in Lycopersicon esculentum, a typical C₃ species. The inhibition of carboxylation efficiency by 21% O₂ varied from 17% for F. ramosissima to 46% for F. anomala and were intermediate between the C₄ (2% for F. trinervia) and C₃ (53% for L. esculentum) values. The intermediate Flaveria species, especially F. ramosissima, have substantial activities of the C₄ enzymes, phosphoenolpyruvate carboxylase, pyruvate, orthophosphate dikinase, NADP-malic enzyme, and NADP-malate dehydrogenase, indicating potential for C₄ photosynthesis. It appears that these Flaveria species may be true biochemical C₃-C₄ intermediates.     

Molecular phylogenies disprove a hypothesized C4 reversion in Eragrostis walteri (Poaceae)

Background and Aims

The main assemblage of the grass subfamily Chloridoideae is the largest known clade of C₄ plant species, with the notable exception of Eragrostis walteri Pilg., whose leaf anatomy has been described as typical of C₃ plants. Eragrostis walteri is therefore classically hypothesized to represent an exceptional example of evolutionary reversion from C₄ to C₃ photosynthesis. Here this hypothesis is tested by verifying the photosynthetic type of E. walteri and its classification.

Methods

Carbon isotope analyses were used to determine the photosynthetic pathway of several E. walteri accessions, and phylogenetic analyses of plastid rbcL and ndhF and nuclear internal transcribed spacer DNA sequences were used to establish the phylogenetic position of the species.

Results

Carbon isotope analyses confirmed that E. walteri is a C₃ plant. However, phylogenetic analyses demonstrate that this species has been misclassified, showing that E. walteri is positioned outside Chloridoideae in Arundinoideae, a subfamily comprised entirely of C₃ species.

Conclusions

The long-standing hypothesis of C₄ to C₃ reversion in E. walteri is rejected, and the classification of this species needs to be re-evaluated.     

The occurrence of both C3 and C4 photosynthetic characteristics in a single Zea mays plant

The activities of the carboxylating enzymes ribulose-1,5-biphosphate (RuBP) carboxylase and phosphoenolpyruvate (PEP) carboxylase in leaves of three-week old Zea mays plants grown under phytotron conditions were found to vary according to leaf position. In the lower leaves the activity of PEP carboxylase was lower than that of RuBP carboxylase, while the upper leaves exhibited high levels of PEP carboxylase. Carbon dioxide compensation points and net photosynthetic rates also differed in the lower and upper leaves. Differences in the fine structure of the lowermost and uppermost leaves are shown. The existence of both the C₃ and C₄ photosynthetic pathways in the same plant, in this and other species, is discussed.Abbreviations PEP phosphoenolpyruvate - RuBP ribulose-1,5-biphosphate     

Effects of irradiance and temperature on photosynthesis in C3, C4 and C3/C4 Panicum species

Species in the Laxa and Grandia groups of the genus Panicum are adapted to low, wet areas of tropical and subtropical America. Panicum milioides is a species with C₃ photosynthesis and low apparent photorespiration and has been classified as a C₃/C₄ intermediate. Other species in the Laxa group are C₃ with normal photorespiration. Panicum prionitis is a C₄ species in the Grandia group. Since P. milioides has some leaf characteristics intermediate to C₃ and C₄ species, its photosynthetic response to irradiance and temperature was compared to the closely related C₃ species, P. laxum and P. boliviense and to P. prionitis. The response of apparent photosynthesis to irradiance and temperature was similar to that of P. laxum and P. boliviense, with saturation at a photosynthetic photo flux density of about 1 mmol m^-2 s^-1 at 30°C and temperature optimum near 30°C. In contrast, P. prionitis showed no light saturation up to 2 mmol m^-2 s^-1 and an optimum temperature near 40°C. P. milioides exhibited low CO₂ loss into CO₂-free air in the light and this loss was nearly insensitive to temperature. Loss of CO₂ in the light in the C₃ species, P. laxum and P. boliviense, was several-fold higher than in P. milioides and increased 2- to 5-fold with increases in temperature from 10 to 40°C. The level of dark respiration and its response to temperature were similar in all four Panicum species examined. It is concluded that the low apparent photorespiration in P. milioides does not influence its response of apparent photosynthesis to irradiance and temperature in comparison to closely related C₃ Panicum species.Abbreviations AP apparent photosynthesis - I CO₂ compensation point - gl leaf conductance; gm, mesophyll conductance - PPFD photosynthetic photon flux density - PR apparent photorespiration rate - RuBPC sibulose bisphosphate carboxylase     

Primary partitioning and storage of photosynthate in sucrose and starch in leaves of C4 plants

A procedure involving pulse labelling of leaves with ¹⁴CO₂ was developed to measure the primary (initial) partitioning of photosynthate between sucrose and starch. Partitioning of photosynthate into sucrose and starch was determined in leaves of C₄ plants and compared with the patterns of storage of carbon in these products during the light period. The ratio of primary partitioning into sucrose and starch varied from about 0.5 in those species that accumulated mostly starch in the leaves (Amaranthus edulis L., Atriplex spongiosa F. Muell. and Flaveria trinervia (Spreng.) C. Mohr) to about 8 in Eleusine indica (L.) Gaertn., which accumulated mostly sucrose. No label was detected in free glucose or fructose. Generally there was a reasonable link between the primary partitioning of photosynthate and the type of carbohydrate stored in the leaf during the day. However, the ratio of carbon initially partitioned into sucrose versus starch was about 3 to 4 times higher in leaves of NADP-malic enzyme-type monocotyledonous species compared with phosphoenolpyruvate carboxykinase-type species, although the ratio of sucrose to starch accumulated in leaves during the day was very similar in the two groups. Sucrose and starch were the principal carbohydrates accumulated in leaves during the day. None of the species examined contained significant amounts of fructan and only one species, Atriplex spongiosa, contained substantial amounts of hexose sugars. In most of the species studied, the proportion of photosynthate partitioned into starch was greater at the end of the day than at the beginning. With the exception of Flaveria trinervia, the rate of CO₂ assimilation did not decline during the day, showing that, under our conditions, accumulation of carbohydrate in the leaves did not lead to feedback inhibition of photosynthesis in these C₄ species.Abbreviations Chl chlorophyll - NAD-ME NAD-malic enzyme - NADP-ME NADP-malic enzyme - PCK phosphoenolpyruvate carboxykinase We thank Prof. H.W. Heldt (Pflanzenphysiologisches Institut, Universität Göttingen) for discussions and advice during the course of this work.