Manipulating PEPC levels in plants - Semantic Scholar

4 downloads 0 Views 227KB Size Report
Abstract. This review examines the current understanding of the structural, functional and regulatory properties of. C4 and C3 forms of higher plant ...
Journal of Experimental Botany, Vol. 53, No. 376, pp. 1837±1845, September 2002 DOI: 10.1093/jxb/erf061

REVIEW ARTICLE

Manipulating PEPC levels in plants M. Jeanneau1, J. Vidal1,3, A. Gousset-Dupont1, B. Lebouteiller1, M. Hodges1, D. Gerentes2 and P. Perez2 1

Institut de Biotechnologie des Plantes, UMR CNRS 8618, Universite de Paris Sud, F-91405 Orsay cedex, France 2 Biogemma, Campus Universitaire des CeÂzeaux, F-63170 AubieÁre, France Received 17 April 2002; Accepted 21 June 2002

Abstract This review examines the current understanding of the structural, functional and regulatory properties of C4 and C3 forms of higher plant phosphoenolpyruvate carboxylase. The emphasis is on the interactive metabolic and post-translational controls acting on the enzyme in the physiological context of C4 photosynthesis and the anaplerotic pathway. A brief overview is given concerning the recent developments of PEPC-based genetic engineering of C3 plants with the aim of improving photosynthetic performance in normal and limiting environmental conditions. So far, in spite of achieving a considerable increase in PEPC levels, more work needs to be done with respect to the correct dosage and location before that goal is reached. Some unpublished results on the transformation of maize with a sorghum C4 PEPC cDNA are also presented. They show that it is possible to increase photosynthetic PEPC levels in this C4 plant and that the modi®cation in enzyme content has a pleiotropic physiological impact and, notably, an improved water use ef®ciency when water is limited. Key words: Genetic engineering, carboxylase, signal transduction.

phosphoenolpyruvate

Introduction Higher plant phosphoenolpyruvate carboxylase (PEPC; E.C. 4 1 1 31) is a multifaceted enzyme involved in various physiological contexts (Andreo et al., 1987). In C4 and CAM photosynthesis, a speci®c form of the enzyme initiates the pathway (Lepiniec et al., 1994), while in all plant types, a C3 PEPC is branched to glycolysis to replenish Krebs cycle intermediates (the so-called ana3

plerotic function), thereby providing precursors for amino acid synthesis (Stitt, 1999). Furthermore, in lipid-rich seeds, malate has been reported to be the best precursor for fatty acid synthesis in leucoplasts and a high PEPC activity coincides with seed maturation (Smith et al., 1992). A wealth of data has been gathered about the multigenic family encoding this enzyme, including the structure and function of genes and promoters and phylogenetic relationships (Chollet et al., 1996; Lepiniec et al., 1994). Since the discovery of PEPC by Bandurski and Greiner (1953), many studies have established the allosteric nature of this homotetrameric enzyme that is subject to an opposite and antagonistic effect of metabolites, for example, malate (negative feedback) and sugar-P (positive feedback). Interest concerning this enzyme has been revitalized after it was discovered to be subjected to post-translational control (Nimmo et al., 1984; Chollet et al., 1996; Vidal and Chollet, 1997). All plant PEPCs studied so far contain an N-terminal consensus domain that is phosphorylated on a regulatory serine by a dedicated, Ca2+-independent PEPC kinase (PEPCk). Pharmacological and molecular-based approaches have led to the dissection of the highly complex signalling cascade controlling PEPCk levels in C4 and CAM plants. Now, there are emerging data to support the idea that a similar cascade is also involved in the regulation of C3 plant PEPCk (Vidal and Chollet, 1997; J Vidal, unpublished data). Phosphorylation modulates the metabolic regulation of photosynthetic PEPC in a way that ensures protection against the feedback effect exerted by malate. Recently, the 3-D structure of the E. coli enzyme has been determined by X-ray diffraction thereby making clear the topography of active and inhibitor sites as well as the mechanism of inhibition (Kai et al., 1999). Genetic manipulations have been undertaken to modify PEPC levels in plants with the goal of clarifying the role of the enzyme in various physiological contexts and to

To whom correspondence should be addressed. Fax: +33 1 69 15 34 23. E-mail: [email protected]

ã Society for Experimental Biology 2002

1838 Jeanneau et al.

Fig. 1. The pathway of C4 photosynthesis in malate formers. The enzymes of the C4 cycle (phosphoenolpyruvate carboxylase, PEPC; NADPmalate dehydrogenase, NADP-MDH; pyruvate-Pi-dikinase, PPDK; NADP-malic enzyme, NADP-ME) provide CO2 to the ribulose-bisphosphate carboxylase (Rubisco) in the bundle sheath cell chloroplasts. In addition, 3-phosphoglyceric acid (3-PGA) moves to the mesophyll chloroplasts for reduction to triose-P (phosphoglycerate kinase, PGK; NADP-glyceraldehyde phosphate dehydrogenase, GPDH).

improve plant productivity. One major approach has been to introduce this key biochemical component of the C4 photosynthetic pathway into C3 plants (Ku et al., 1999; Matsuoka et al., 2001). In this regard, preliminary work has been devoted to solving the problems of targeting the PEPC to the desired location and to achieving high expression levels in the host plants. This report is designed to summarize data related to the regulatory and functional properties of both the photosynthetic (C4 type) and anaplerotic (C3 type) forms of PEPC. Then, a brief survey of what has been learnt from PEPC genetic manipulations in both C3 and C4 plants will be presented. Regulatory properties of the photosynthetic PEPC C4 plants display a concentric organization of photosynthetic leaf tissues (mesophyll and bundle sheath) in which enzymes of the photosynthetic pathway, C4 cycle and Benson±Calvin cycle, are distributed (Fig. 1). In the mesophyll cell cytosol, the C4 PEPC is subject to both light-dependent transcriptional and post-translational control (SchaÈffner and Sheen, 1992; Chollet et al., 1996). The expression of the photosynthetic gene and the accumulation of the corresponding C4 PEPC form in the cytosol of sorghum mesophyll cells are mediated by phytochrome and the regulatory phosphorylation of trans-acting factors

during leaf greening (Rydz et al., 2000). These aspects of PEPc regulation are beyond the scope of this review and will not be developed further. Regulation of C4 PEPC in the mesophyll cell cytosol involves photosynthesis-related metabolite effectors like glucose-6-phosphate (G-6-P; activator) and malate (feedback inhibitor) and a highly complex, light-dependent, reversible phosphorylation process (Chollet et al., 1996). This latter modi®cation changes the functional properties of C4 PEPC, increasing Vm and Ka for G-6-P, and decreasing its sensitivity to malate when measured at suboptimal pH (7.3) and PEP concentration (2.5 mM). The recent determination of the 3-D-structure of E. coli PEPC has shed some light on the molecular mechanism of malate inhibition and how phosphorylation can relieve this effect (Kai et al., 1999). In the E. coli PEPC, arginine 587, located in a highly conserved glycine-rich loop, is shared by the aspartate binding site and the active site. Upon binding of this effector (equivalent to L-malate in the plant enzyme), the loop is removed from the catalytic site thus perturbing substrate binding and causing a loss of catalytic activity (Kai et al., 1999). It has been suggested that the plant invariant phosphorylated N-terminus (acid-base-XX-S-I-D-A-Q-L-R) of the enzyme moves closer to the entry of the inhibitor site thus impeding access of malate. Recent mutational analyses based on the structural features of the C4 PEPC from maize have identi®ed putative residues (R183 and R184) involved in the allosteric

Manipulating PEPC levels in plants 1839

Fig. 2. Schematic model for the spatio-temporal organization of the transduction cascade in the C4 leaf. An increase in 3-posphoglyceric acid (3PGA) ensures the intercellular coupling via pHc changes in the mesophyll cell cytosol. This is followed by activation of a mesophyll cell phospholipase C (PI-PLC) and transient production of the second messengers inositol-1,4,5-trisphosphate (InsP3) and diacylglycerol (DAG). InsP3 causes tonoplast calcium channels to open and ef¯ux of calcium into the cytosol while DAG and calcium activate a PKC-like activity. The rapid synthesis of the calcium-independent PEPC kinase leads to the phosphorylation of C4 PEPC. Question marks indicate the steps that remain to be elucidated.

activator site binding G-6-P (Terada et al., 2001). Moreover, a study of the enzyme chimera between C4 and C3 PEPC isoforms from Flaveria species and sitedirected mutagenesis have revealed the crucial role of the residue at position 774 (serine in C4 and alanine in C3 enzymes) as a major determinant for the speci®c properties of each enzyme form (BlaÈsing et al., 2000). In vitro, C4 PEPC activity, phosphorylation and metabolic control are highly sensitive to pH (Echevarria et al., 1994; Gao and Woo, 1996). The increase in pH that has been proposed to occur in the mesophyll cell cytosol upon illumination of the C4 leaf (between pH 7 and 7.5, see below) is expected to activate the C4 PEPC and, partially, to promote desensitization of the enzyme towards the effectors, notably, L-malate. Therefore, most of the interactive players acting on the regulation of PEPC (i.e. pHc, G-6-P, phosphorylation) are opposed to the negative feedback exerted by malate. The light-transduction chain leading to C4 PEPC phosphorylation by its requisite, Ca2+-independent protein kinase has been studied by ¯ow cytometry, confocal microscopy, molecular biology, and cellular pharmacology techniques using mesophyll protoplasts from crabgrass (Digitaria sanguinalis) and pharmacological reagents (Giglioli-Guivarc'h et al., 1996). This has led to a model of the spatio-temporal organization of the cascade

in the C4 leaf as shown in Fig. 2. In this scenario, 3phosphoglyceric acid, generated during photosynthesis in bundle sheath cell chloroplasts represents the intercellular metabolic message that diffuses into mesophyll cells. In these cells, its subsequent transport into the chloroplasts under the protonated 2± form is expected to cause alkalization of pHc. This early cascade event is followed by activation of a mesophyll cell phospholipase C (PIPLC) and transient production of the second messenger inositol-1,4,5-trisphosphate (InsP3). The cloning, immunological-based identi®cation and biochemical characterization of a Digitaria PI-PLC has been performed in this laboratory showing that this component of the phosphoinositide signalling pathway is a d-type enzyme present in mesophyll cells (Coursol et al., 2000). However, the mechanism underlying its activation remains to be clari®ed. The PI-PLC reaction product InsP3 modulates tonoplast calcium channels, thereby increasing the ¯ux of calcium into the cytosol. This results in the activation of a calcium-dependent protein kinase that is awaiting identi®cation, although recent pharmacological and biochemical data strongly support the hypothesis that this missing link is a PKC-like enzyme (J Vidal, unpublished data). This observation makes sense since PI-PLC catalysed phosphoatidylinositol-4,5-bisphosphate (PIP2) cleavage produces both InsP3 and the PKC activator 1,2-

1840 Jeanneau et al.

Fig. 3. The anaplerotic pathway leading to amino acid synthesis. PEPC replenishes the Krebs cycle intermediates that are depleted when a-ketoglutarate (a-KG) is used for amino acid synthesis in the chloroplasts via the glutamine synthetase (GS)/glutamate-aketoglutarate amino transferase (GOGAT) cycle.

diacylglycerol (DAG). The ultimate step in the cascade implicates the nucleus and the up-regulation of a PEPCk gene. One consistent hypothesis might be that the Ca2+dependent protein kinase phosphorylates a transcription factor that increases PEPCk gene expression. This highly complex, light-dependent cascade involves the contribution of various components (pHc, ion channels and signalling enzymes), cell types (bundle sheath and mesophyll) and subcellular compartments (chloroplast, vacuole, nucleus). In CAM plants, up-regulation of the PEPCk and PEPC phosphorylation occurs during the night and is dependent on a circadian oscillator (Nimmo, 2000). In the CAM species Mesembryanthemum crystallinum, it has been shown that a cascade similar to that found in C4 plants operates in mesophyll cells during the night (Bakrim et al., 2001). The Ca2+-independent PEPCk gene/cDNA has been cloned recently from CAM, C3 and C4 plants (Nimmo, 2000; Taybi et al., 2000; Tsuchida et al., 2001). This unique Ser/Thr protein kinase exhibits several interesting features. It is the smallest protein kinase known so far with a predicted molecular mass around 31 kDa (274, 279 and 284 amino acids in the case of the KalanchoeÈ fedtschenkoi, Mesembryanthemum crystallinum and Arabidopsis thaliana enzyme, respectively). Indeed, it is made up of a catalytic domain with minimal or no additions. Although it belongs to the Ca2+/calmodulin-regulated group of protein kinases, it lacks the regulatory auto-inhibitory region and

the EF-hands. It displays an alkaline pHo of around 8 (using the recombinant enzyme from Mesembryanthemum crystallinum) and, in reconstituted assays, it speci®cally phosphorylates the N-terminal regulatory serine of the target PEPC, thereby decreasing the malate sensitivity of the enzyme, as expected. Its activity is not modulated directly by second messengers (such as Ca2+/calmodulin or cyclic nucleotides) or by phosphorylation/dephosphorylation processes, but rather through rapid changes in its turnover rate. From a physiological point of view, the blocking of PEPCk synthesis in the illuminated C4 leaf led to a marked inhibition of CO2 assimilation (Bakrim et al., 1993). Therefore C4 PEPC phosphorylation appears to be a critical event governing carbon assimilation in the C4 photosynthetic pathway. These complex mechanisms involving the light-dependent control of C4 PEPC at both the transcriptional and post-translational levels allow the adjustment of intercellular carbon ¯ow according to the demand of the Calvin cycle and thus ensure ef®cient functioning and homeostasis of C4 photosynthesis. Regulatory phosphorylation of PEPC in the anaplerotic pathway linked to amino acid synthesis In the C3 plant leaf, PEPC is no longer involved in photosynthesis, but ful®ls a variety of physiological roles. In the anaplerotic pathway (Fig. 3), that also occurs in C4 plants, the enzyme contributes to the replenishment of Krebs cycle intermediates when organic acids are directed towards other metabolic pathways such as amino acid (via the GS/GOGAT cycle) and protein synthesis (Stitt, 1999; Champigny and Foyer, 1992). In this respect, PEPC can be considered as a branch of the glycolytic pathway. In relation to the anaplerotic function, PEPC activity also contributes to the homeostasis of cell cytosolic pH (Sakano, 1998) and the chloroplastic OAA/malate shuttle that provides the cytosol with reducing power required by the nitrate reductase (NR) (Oaks, 1994). As nitrogen assimilation proceeds, primary metabolism is reset so that more carbon is diverted to respiratory metabolism by means of a complex co-ordinated regulation of many enzymes and transporters, including signalling networks and metabolites. Intuitively, the concept that PEPC must be protected against malate (via a phosphorylation process) as proposed in C4 and CAM plant photosynthesis, should apply to any system in which the concentration of this metabolite increases, such as anaplerotic C-¯ow and its interaction with N-metabolism. Indeed, regulatory phosphorylation of C3 PEPC is supported by a number of facts. (1) The presence of the N-terminal phosphorylation domain in all plant PEPCs sequenced so far, whatever the physiological type. (2) The presence of a Ca2+-independent PEPC-kinase in leaves of C3 plants (Vidal and Chollet,

Manipulating PEPC levels in plants 1841

1997) and the isolation of C3-plant PEPCk cDNAs and genes (Taybi et al., 2000; Nimmo, 2000). (3) The induction of a PEPCk activity in illuminated C3 leaves and protoplasts that is blocked by cycloheximide (CHX), like the corresponding C4 and CAM enzymes, suggesting that protein turnover is involved in the up-regulation of the PEPCk in C3 plants (Vidal and Chollet, 1997). Collectively, these data support the hypothesis that upregulation of PEPCk in a C3-plant context is via a transduction cascade similar to that operating in C4 and CAM plants. However, whether the upstream signalling elements identi®ed in C4 mesophyll cells are also key players in the C3 cascade remains poorly documented. Experiments using barley leaf protoplasts have suggested that, while PEPC phosphorylation occurs in situ in the light and is modulated by protein synthesis and calcium, the mechanism leading to up-regulation of the corresponding PEPCk might differ from that found in C4 mesophyll protoplasts (Smith et al., 1996). However, recent results (J Vidal, unpublished data) suggest that, in addition to the PEPCk, the central component of the C4 cascade (namely, a C3 PI-PLC) is present in the C3 plant signalling system. This ®nding may also indicate that alkalization of pHc is a requisite step of the C3 cascade. In support of this hypothesis are experimental data showing that mesophyll cell pHc increases upon illumination of the C3 plant leaf (Yin et al., 1990). However, this point needs to be characterized further. The rate of C-¯ux through the anaplerotic PEPC is also modulated by NO3± and/or amino acids via a change in PEPC phosphorylation status (Li et al., 1996; Van Quy and Champigny, 1992; Champigny and Foyer, 1992). Collectively, the data indicate that the C3 leaf PEPCk activity increases in the light and that leaf N-status can in¯uence the regulatory phosphorylation of C3 PEPC. Transgenic plants Recent developments in plant genetic engineering have allowed the production of C3 and C4 transgenic plants that are aimed at studying the function of PEPC gene promoters and the impact of over- and under-expression of the enzyme on the physiology and growth of the transformed plants. C3 plants

C3 photosynthesis suffers from O2 inhibition due to the fact that Rubisco is a bifunctional enzyme with competing oxygenase and carboxylase activities. Oxygenation of RubP initiates the photorespiratory pathway that subsequently leads to a considerable loss of carbon (and perhaps N). By contrast, C4 plants have evolved a CO2 concentrating mechanism (the C4 cycle) to overcome this wasteful process (see below and Fig. 1). The acquisition of this new photosynthetic strategy by a wide variety of plant species

indicates that it has originated independently and on many separate occasions during the evolution of ¯owering plants (Lepiniec et al., 1994). The important question is: can evolution be mimicked by taking advantage of genetic manipulations to import some C4 photosynthetic characteristics into C3 species? One key enzyme of the C4 cycle is the C4 PEPC that occurs at very high levels in the mesophyll cell cytosol where it very ef®ciently assimilates CO2 (as the hydrated HCO3± form). Over the last three years, several attempts have been made to transfer C4 photosynthetic traits, with emphasis on PEPC, into C3 plants in order to improve CO2 assimilation rates. It has been found that the C4 PEPC gene promoter (from maize) can drive high level expression of a reporter gene in transgenic rice plants in an organ-speci®c, mesophyll-cellspeci®c and light-dependent manner (see Matsuoka et al., 2001, for a review). The next experiments used plasmid constructs containing a full-length PEPC cDNA fused to the C4 PEPC gene promoter. In tobacco, a 2±5-fold increase in leaf PEPC activity (maize C4 PEPC) was accompanied by a corresponding increase in malate content. However, these biochemical differences did not produce any signi®cant physiological changes with respect to photosynthetic CO2 assimilation rate and CO2 compensation point (Matsuoka et al., 2001). In transgenic potato overexpressing a bacterial PEPC gene, the CO2 compensation point was found to be lowered, together with an increase in respiration and glucose and starch contents (Gehlen et al., 1996; HaÈusler et al., 1999). Therefore, these ®rst attempts were successful in terms of PEPC ectopic expression in C3 plants, however, the overexpression levels were generally low. More recently, the use of the Agrobacterium-mediated transformation of rice with an intact C4 PEPC gene of maize has allowed this problem to be solved. Indeed, a tremendous increase in mesophyll cell PEPC was attained that accounted for about 12% of total leaf soluble protein, a level that exceeds by 2±3-fold the enzyme concentration in C4 leaves (Ku et al., 1999). These plants exhibited reduced O2 inhibition of photosynthesis but photosynthetic rates were comparable to those of untransformed plants. Moreover, the maize PEPC in transgenic rice leaves remained in its dephosphorylated and less active form during illumination (Matsuoka et al., 2001). Therefore, although progress is being made in manipulating PEPC levels in C3 plants, ef®cient strategies to obtain the desirable phenotypes are yet to be formulated. As mentioned above, the maize C4 PEPC is not phosphorylated in the leaves of transformed rice and, therefore, it is expected to be weakly active. This raises the question of whether the exogenous enzyme is a poor target for the C3 leaf PEPCk, or alternatively, whether a compensation mechanism is induced to buffer the impact of the genetic modi®cation. Another potential problem is the availability of PEP in the mesophyll cells of C3 plants. Too much

1842 Jeanneau et al.

PEPC is expected to decrease the content of this substrate to a level that would perturb plant metabolism. In a similar strategy, transgenic plants have been produced using genes or cDNAs encoding pyruvate-Pi dikinase (PPDK) and NADP-malic enzyme (NADP ME) and it has been shown that it is possible to overexpress the corresponding enzymes in C3 leaf mesophyll cells. However, it was found that PPDK overexpression had no effect on carbon metabolism while NADP-ME had a detrimental effect (Matsuoka et al., 2001). Ideally, the C4 pathway genes should be expressed at the correct level and ratios in the correct cell compartments of leaf tissues that lack the Kranz anatomy. Interestingly, a recent report has described the existence of the biochemical characteristics of C4 photosynthesis in the stems and petioles of C3 plants (Hibberd and Quick, 2002). The presence of the correct biochemical pathways (e.g. decarboxylating enzymes, NAD and NADP-ME; PEP regeneration, PPDK) in the photosynthetic cells surrounding the vascular bundles indicates that the essential biochemical components and regulatory elements controlling the cell speci®c gene expression required for C4 photosynthesis are already present in C3 plants. This may explain the polyphyletic evolution of C4 plants. Understanding how this system has evolved in leaves of C3 plants to give rise to C4 plants during evolution may give important clues that would help to manipulate C3 plants more ef®ciently. For example, the promoter of an NADP-ME gene from bean has been shown to direct expression in cells around the vascular systems (Schaaf et al., 1995). C4 plants In photosynthetic tissues of the illuminated C4 leaf, the C4 cycle acts as a pump increasing CO2 levels in the vicinity of Rubisco. As a consequence of the elevated ratio of CO2/ O2, the oxygenase activity of this enzyme and, thus, photorespiration, are markedly reduced with respect to C3 plants. When grown in their natural environment, often consisting of hot climates with sporadic rainfalls, C4 plants have a selective advantage over C3 plants. They show higher photosynthetic rates, biomass productivity and water and nitrogen use ef®ciencies. Within the last century, the use of classical plant breeding methods has led to a doubling of the yield of maize. This was essentially due to an increase in the leaf surface and not to an improved performance of the photosynthetic apparatus. Because the CO2 pumping system overcomes O2 inhibition of photosynthesis and causes the CO2 compensation point to be very low, C4 photosynthesis is near saturated at atmospheric CO2 levels. Therefore, it is believed that engineering higher PEPC levels in C4 plants should not lead to a signi®cant improvement in photosynthetic ef®ciency. In addition, the metabolic and post-translational regulation of the C4 PEPC may afford compensation effects counteracting the quantitative modi®cation.

However, when maize is subjected to a water stress, photosynthesis is depressed, although C4 PEPC content is slightly increased (Rodriguez-Penagos and Munoz-Clares, 1999). The impact of water stress is quite complex and, in essence, pleiotropic, notably altering photosynthesis following an increase in stomatal closure (mild progressive stress) (Cornic, 2000) and/or irreversible damages to the photosynthetic apparatus (e.g. ATP-synthase) (Tezara et al., 1999) leading to lower ATP contents and CO2 assimilation rates. However, within the range of relative water content occurring in natural conditions, it is likely that stomatal closure plays the major role in the decrease in leaf photosynthesis while the photosynthetic machinery remains intact (Cornic, 2000). As a consequence of stomatal closure, a consistent hypothesis is that both the internal and chloroplastic CO2 molar ratio decline, and CO2 is progressively replaced by O2 thus favouring the oxygenase activity of Rubisco. Maize is particularly sensitive to water stress during its reproductive stage which leads to a strong impairment of grain ®lling. Under a short-term, moderate water stress induced by sorbitol, the activity of C4 PEPC showed a 50% increase in maize leaf discs. This increase was correlated with enhanced proline levels, thus suggesting that C4 PEPC could be involved in plant stress adaptation (Rodriguez-Penagos and Munoz-Clares, 1999). Although the C4 PEPC of Amaranthus edulis has an estimated ¯uxcontrol coef®cient over the CO2 assimilation rate of 0.26 for plants grown at ambient CO2, this becomes much higher (0.68) when plants are grown at low CO2. Therefore, the enzyme is expected to contribute signi®cantly to the control of C ¯ux in the photosynthetic pathway under the limited internal CO2 conditions promoted by drought (Dever et al., 1997). Based on these data, it can be hypothesized that increasing the C4 PEPC content by genetic engineering may positively impact on (1) osmotic adjustment, providing precursors for proline biosynthesis and (2) the capacity to ®x CO2 at lower gas conductance. Enhancing the performance of the CO2 trapping system would ensure both higher CO2 levels in the vicinity of Rubisco and lower rates of photorespiration, and may contribute to a better tolerance of C4 plants to water de®cit. To test this hypothesis, maize plants were engineered by Agrobacterium-mediated infection or biolistic techniques with a plasmid containing a full length C4 PEPC (from sorghum) cDNA under the control of the C4 gene promoter (leaf speci®c). Transgenic maize overexpressing (2.2-fold) or underexpressing (0.2-fold by a co-suppression effect) this C4 PEPC were produced and characterized at the molecular and physiological levels. Sorghum C4 PEPC was expressed in the mesophyll cell cytosol and found to undergo phosphorylation in the light. However, the phosphorylation status of the enzyme was signi®cantly lower in overexpressor plant leaves subjected to non-

Manipulating PEPC levels in plants 1843

saturating light. This was not due to a decrease in PEPCk levels and down-regulation of the phosphorylation cascade. In PEPC-de®cient mutants of the C4 plant Amaranthus edulis, a compensatory mechanism increasing the C4 PEPC phosphorylation status has been described (Dever et al., 1997). Therefore, the plant can respond to the modi®cation of PEPC levels by modulating the phosphorylation status of the enzyme. Physiological studies showed that ectopic expression of C4 PEPC has a pleiotropic effect on maize plants. In particular, the engineered plants displayed a better capacity to ®x CO2 in water-suf®cient conditions. Under moderate drought, most of the characteristics investigated (CO2 ®xation rate, CO2 compensation point, fresh and dry weight, leaf surface, and stomatal density) were improved in the overexpressors and depressed in the underexpressors. Furthermore, in addition to the above-mentioned characteristics, the overexpressor showed a signi®cant (+30%) improvement in the intrinsic water use ef®ciency (WUE). Elevated values of WUE can be due to the better capacity of the plants to ®x CO2 when the gas conductance is reduced by water-limiting conditions. Contributing to the limitation in gas conductance is the observed reduction in leaf stomatal density of the overexpressor. Recent results have shown that the stomatal density of developing leaves depends on light intensity and CO2 concentration sensed by mature leaves (Brownlee, 2001). Studies on Arabidopsis mutants altered in stomatal density response and patterning suggest that HIC and SDD1 genes are involved in favouring diffusion of long-distance systemic signals and processing of signalling proteins. Along these lines, it is interesting to note that the stomatal density of transgenic maize leaves (4th leaf) varied in an opposite way with respect to C4 PEPC amount. Thus, the lower gas conductance, due to the reduction of stomatal aperture (under stress) and density, and high C4 PEPC activity, that maintain higher CO2 ®xation levels, may account for the improved WUE and biomass production of the C4 PEPC overexpressor. As mentioned above, PEPC has been suggested to be implicated in the lipid ®lling of castor oil seeds during maturation (Smith et al., 1992). The highest enzyme content coincided with the most active phase of storage oil accumulation. Furthermore, it has been demonstrated that malate supports very high rates of fatty acid synthesis in isolated leucoplasts from developing seeds (Smith et al., 1992). In another set of experiments, Agrobacteriummediated and biolistic transformation techniques were used to alter the PEPC content of maize seeds with DNA constructs containing the C4 PEPC cDNA from sorghum in fusion with the High-Molecular-Weight Glutelin (HMWG) promoter from wheat. Biochemical analyses showed that there was no signi®cant change in the lipid content of transformed seeds compared to wild-type seeds (although the transgenics contained up to 10-fold more

PEPC than the wild-type seeds). By contrast, a weak but statistically signi®cant positive correlation was detected with protein, starch and soluble sugar levels. Several hypotheses can account for these observations. Perhaps the PEPC is not contributing signi®cantly to ¯uxes through the different metabolic pathways leading to storage seed components. Maybe the functional and regulatory properties of the C4 PEPC (e.g. Km for PEP, regulatory phosphorylation and malate sensitivity) are not well adapted to the physiological context of seeds. It is also possible that the HMWG promoter activity does not coincide with the phase of seed ®lling in which PEPC plays a role. Concluding remarks It is now possible to produce transgenic C3 and C4 plants that ectopically express foreign PEPC cDNA and genes. These techniques pave the way to studies devoted to engineering crop plants with PEPC in an attempt to improve photosynthetic rates, yield, biomass, and tolerance to abiotic stress. However, future advances could be hindered by the fragmented state of the current understanding of plant metabolism and the growing gap between an ability to clone, study and manipulate individual genes and proteins and an understanding of how they are integrated into and impact on the complex metabolic networks in plants (DellaPenna, 2001). Indeed, this approach is still in its infancy and suffers from several technical and conceptual limitations. From a general point of view, the control analysis theory states that control over a metabolic ¯ux is generally shared by several enzymes of a pathway, and it is expected that `large increases in ¯ux cannot be generated by activation of single enzymes, but can be by activation at several sites along the pathway' (Fell and Thomas, 1999). This has been observed in a number of cases and, to overcome this problem, it will be necessary simultaneously to increase, together with PEPC, those enzymes which exert a control over the ¯ux pathway, and perhaps also regulatory genes impacting on their activity. Furthermore, in the case of highly regulated enzymes, compensatory mechanisms to offset the effect of the genetic modi®cation do occur. Finally, for the recombinant enzyme in a given physiological context, the substrate may be limiting, or the increased amount of enzyme may have deleterious effects on the subtle balance of related metabolites. It is now known that many metabolites trigger signalling processes impacting on target gene expression and metabolic controls of several enzymes. Collectively, these data emphasize the point that a correct dosage and location of the introduced enzymes must be achieved. However, how is it possible to install the Kranz anatomy with its division of labour between specialized photosynthetic cells in a C3 leaf? Assessment of the single cell-C4 system present in some aquatic plants

1844 Jeanneau et al.

like Hydrilla verticillata as an alternative approach is currently under investigation (Leegood, 2002; HaÈusler et al., 2002), yet it is thought that a certain degree of compartmentation will be required to achieve the goal (Leegood, 2002). In spite of the wealth of data currently available to guide genetic manipulation, it is still mandatory to develop knowledge about the physiological context in which the enzyme (or gene) of interest is to be inserted and the functional and regulatory mechanisms modulating its activity. Even when this knowledge becomes available, the introduction of a new gene may have pleiotropic effects on the host plant, making the result of the transgenic experiment somewhat unpredictable and implying an unavoidable touch of empiricism. This is the case, for example, of the transgenic maize overexpressing the C4 PEPC, for which both growth an developmental processes have been altered by the single gene transformation. In addition, this work has revealed that improved performances induced by the recombinant protein might only be detected following exposure of the plant to limiting environmental conditions. References Andreo CS, Gonzalez DH, Iglesias AA. 1987. Higher plant phosphoenolpyruvate carboxylase: structure and regulation. FEBS Letters 213, 1±8. Bakrim N, Prioul J-L, Deleens E, Rocher J-P, Arrio-Dupont M, Vidal J, Gadal P, Chollet R. 1993. Regulatory phosphorylation of C4 phosphoenolpyruvate carboxylase. A cardinal event in¯uencing the photosynthesis rate in sorghum and maize. Plant Physiology 101, 891±897. Bakrim N, Brulfert J, Vidal J, Chollet R. 2001. Phosphoenolpyruvate carboxylase kinase is controlled by a similar signaling cascade in CAM and C4 plants. Biochemical and Biophysical Research Communications 286, 1158±1162. Bandurski RS, Greiner CM. 1953. The enzymatic synthesis of oxalacetate from phosphenolpyruvate and carbon dioxide. Journal of Biological Chemistry 204, 781±786. BlaÈsing OE, Westhoff P, Svensson P. 2000. Evolution of C4 phosphoenolpyruvate carboxylase in Flaveria, a conserved serine residue in the carboxyl-terminal part of the enzyme is a major determinant for C4-speci®c characteristics. Journal of Biological Chemistry 275, 27917±27923. Brownlee C. 2001. The long and short of stomatal density signals. Trends in Plant Science 6, 441±442. Champigny M-L, Foyer C. 1992. Nitrate activation of cytosolic protein kinases diverts photosynthetic carbon from sucrose to amino acid biosynthesis. Plant Physiology 100, 7±12. Chollet R, Vidal J, O'Leary MH. 1996. Phosphoenolpyruvate carboxylase: a ubiquitous, highly regulated enzyme in plants. Annual Review of Plant Physiology and Plant Molecular Biology 47, 273±298. Coursol S, Giglioli-Guivarc'h N, Vidal J, Pierre J-N. 2000. An increase in phosphoinositide-speci®c phospholipase C activity precedes induction of C4 phosphoenolpyruvate carboxylase phosphorylation in illuminated and NH4Cl-treated protoplasts from Digitaria sanguinalis. The Plant Journal 23, 497±506. Cornic G. 2000. Drought stress inhibits photosynthesis by decreasing stomatal apertureÐnot by affecting ATP synthesis. Trends in Plant Science 5, 187±188.

Dellapenna D. 2001. Plant metabolic engineering. Plant Physiology 125, 160±163. Dever LV, Bailey KJ, Leegood RC, Lea PJ. 1997. Control of photosynthesis in Amaranthus edulis mutants with reduced amounts of PEP carboxylase. Australian Journal of Plant Physiology 24, 469±476. Echevarria C, Pacquit V, Bakrim N, Osuna L, Delgado B, Arrio-Dupont M, Vidal J. 1994. The effect of pH on the covalent and metabolic control of C4 phosphoenolpyruvate carboxylase from Sorghum leaf. Archives of Biochemistry and Biophysics 315, 425±430. Fell DA, Thomas S. 1999. Increasing the ¯ux in a metabolic pathway: a metabolic control analysis perspective. In: Kruger, Hill, Ratcliffe G, eds. Regulation of primary metabolic pathways in plants. Proceedings of the Phytochemical Society of Europe. Kluwer Academic Publishers, 257±273. Gao Y, Woo KC. 1996. Regulation of phosphoenolpyruvate carboxylase in Zea mays by protein phosphorylation and metabolites and their roles in photosynthesis. Australian Journal of Plant Physiology 23, 25±32. Gehlen J, Pastruga R, Smets H, Merkelbach S, Kleines M, Porsch P, Fladung M, Becker I, Rademacher T, HaÈusler E, Hirsch H-J. 1996. Effects of altered phosphoenolpyruvate carboxylase activities on transgenic C3 plant Solanum tuberosum. Plant Molecular Biology 32, 831±848. Giglioli-Guivarc'h N, Pierre JN, Brown S, Chollet R, Vidal J, Gadal P. 1996. The light-dependent transduction pathway controlling the regulatory phosphorylation of C4 phosphoenolpyruvate carboxylase in protoplasts from Digitaria sanguinalis. The Plant Cell 8, 573±586. HaÈusler RE, Hirsch H-J, Kreuzaler F, PeterhaÈnsel C. 2002. Overexpression of C4-cycle enzymes in transgenic C3 plants: a biotechnological approach to improve C3 photosynthesis. Plant and Cell Physiology 53, 591±607 HaÈusler RE, Kleines M, Uhrig H, Hirsch H-J, Smets H. 1999. Overexpression of phosphoenolpyruvate carboxylase from Corynebacterium glutamicum lowers the CO2 compensation point (T) and enhances dark and light respiration in transgenic potato. Journal of Experimental Botany 50, 1231±1242. Hibberd JM, Quick WP. 2002. Characteristics of C4 photosynthesis in stems and petioles of C3 ¯owering plants. Nature 415, 451±454. Kai Y, Matsumura H, Inoue T, Terada K, Nagara Y, Yoshinaga T, Kihara A, Tsumura K, Izui K. 1999. Three-dimensional structure of phosphoenolpyruvate carboxylase: a proposed mechanism for allosteric inhibition. Proceedings of the National Academy of Sciences, USA 96, 823±828. Ku MSB, Agarie S, Nomura M, Fukayama H, Tsuchida H, Ono K, Hirose S, Toki S, Miyao M, Matsuoka M. 1999. High-level expression of maize phosphoenolpyruvate carboxylase in transgenic rice plants. Nature Biotechnology 17, 76±80. Leegood RC. 2002. C4 photosynthesis: principles of CO2 concentration and prospects for its introduction into C3 plants. Plant and Cell Physiology 53, 581±590. Lepiniec L, Vidal J, Chollet R, Gadal P, Cretin C. 1994. Phosphoenolpyruvate carboxylase: structure, regulation and evolution. Plant Sciences 99, 111±124. Li B, Zhang X-Q, Chollet R. 1996. Phosphoenolpyruvate carboxylase kinase in tobacco leaves is activated by light in a similar but not identical way as in maize. Plant Physiology 111, 497±505. Matsuoka M, Furbank R, Fukayama H, Miyao M. 2001. Molecular engineering of C4 photosynthesis. Annual Review of Plant Physiology and Plant Molecular Biology 52, 297±314. Nimmo GA, Nimmo HG, Fewson CA, Wilkins MB. 1984. Diurnal changes in the properties of phosphoenolpyruvate

Manipulating PEPC levels in plants 1845 carboxylase in Bryophyllum leaves: a possible covalent modi®cation. FEBS Letters 178, 199±203. Nimmo HG. 2000. The regulation of phosphoenolpyruvate carboxylase in CAM plants. Trends in Plant Science 5, 75±80. Oaks A. 1994. Ef®ciency of nitrogen utilization in C3 and C4 cereals. Plant Physiology 106, 407±414. Rodriguez-Penagos M, Munoz-Clares R. 1999. Response of phosphoenolpyruvate carboxylase from maize leaves to moderate water de®cit. Journal of Plant Physiology 155, 631±638. Rydz KS, Prieto JL, Rychter AM, Vidal J. 2000. A DNA-binding activity for the promoter of the gene encoding C4 phosphoenolpyruvate carboxylase is modulated by phosphorylation during greening of the sorghum leaf. Plant Science 159, 65±73. Sakano K. 1998. Revision of biochemical pH-stat: involvement of alternative pathway metabolisms. Plant and Cell Physiology 39, 467±473. Schaaf J, Walter MH, Hess D. 1995. Primary metabolism in plant defence. Plant Physiology 108, 949±960. SchaÈffner AR, Sheen J. 1992. Maize C4 photosynthesis involves differential regulation of phosphoenolpyruvate carboxylase genes. The Plant Journal 2, 221±232. Smith LH, Lillo C, Nimmo HG, Wilkins MB. 1996. Light regulation of phosphoenolpyruvate carboxylase in barley mesophyll protoplasts is modulated by protein synthesis and calcium, and not necessarily correlated with phosphoenolpyruvate carboxylase kinase activity. Planta 200, 174±180. Smith RG, Gauthier DA, Dennis DT, Turpin DH. 1992. Malateand pyruvate-dependent fatty acid synthesis in leucoplasts from developing castor endosperm. Plant Physiology 98, 1233±1238.

Stitt M. 1999. Nitrate regulation of metabolism and growth. Current Opinion in Plant Biology 2, 178±186. Taybi T, Patil S, Chollet R, Cushman JC. 2000. A minimal serine/threonine protein kinase circadianly regulates phosphoenolpyruvate carboxylase activity in crassulacean acid metabolism-induced leaves of the common ice plant. Plant Physiology 123, 1471±1481. Terada A, Kotera M, Tsumura K, Furumoto T, Matsumura H, Kai Y, Izui K. 2001. Phosphoenolpyruvate carboxylase (PEPC): mutational analysis of a ¯exible loop and a putative binding site for an allosteric activator, glucose 6-phosphate (G6P). In: Proceedings of the 12th international congress of photosynthesis. Brisbane, Australia, S17±029. Tezara W, Mitchell VJ, Discoll SD, Lawlor DW. 1999. Water stress inhibits plant photosynthesis by decreasing coupling factor and ATP. Nature 401, 914±917. Tsuchida Y, Furumoto T, Izumida A, Hata S, Izui K. 2001. Phosphoenolpyruvate carboxylase kinase involved in C4 photosynthesis in Flaveria trinervia: cDNA cloning and characterization. FEBS Letters 507, 318±322. Van Quy L, Champigny M-L. 1992. NO3± enhances the kinase activity for phosphorylation of phosphoenolpyruvate carboxylase and sucrose phosphate synthase proteins in wheat leaves. Plant Physiology 99, 344±347. Vidal J, Chollet R. 1997. Regulatory phosphorylation of C4 PEP carboxylase. Trends in Plant Science 2, 230±237. Yin Z-H, Neimanis S, Wagner U, Heber U. 1990. Lightdependent pH changes in leaves of C3 plants. 1. Recording pH changes in various cellular compartments by ¯uorescence probes. Planta 182, 244±252.