C4 photosynthesis: principles of CO2 concentration and prospects for its introduction into C3 plants (2024)

Search for other works by this author on:

Article history

Abstract

C₄ photosynthesis has a number of distinct properties that enable the capture of CO₂ and its concentration in the vicinity of Rubisco, so as to reduce the oxygenase activity of Rubisco, and hence the rate of photorespiration. The aim of this review is to discuss the properties of this CO₂‐concentrating mechanism, and thus to indicate the minimum requirements of any genetically‐engineered system. In particular, the Kranz leaf anatomy of C₄ photosynthesis and the division of the C₄‐cycle between two cell types involves intercellular co‐operation that requires modifications in regulation and transport to make C₄ photosynthesis work. Some examples of these modifications are discussed. Comparisons are made with the C₄‐type photosynthesis found in single‐celled C₄‐type CO₂‐concentrating mechanisms, such as that found in the aquatic plant, Hydrilla verticillata and the single‐celled C₄ system found in the terrestrial chenopod Borszczowia aralocaspica. The outcome of recent attempts to engineer C₄ enzymes into C₃ plants is discussed.

Introduction

The discovery of C₄ photosynthesis in the 1960s and awareness of the high photosynthetic capacity of C₄, as compared with C₃, plants created an upsurge of interest in the impact of photorespiration on the carbon economy of C₃ plants, its origins in the Rubisco oxygenase reaction, and in the ecophysiological relationships between C₃ and C₄ plants. This was recorded strikingly in the proceedings of a conference held in Canberra in 1970 (Hatch et al., 1971). Later, Burris and Black wrote, ‘A continuing interest in plant productivity has been quickened in recent years primarily by a growing concern for feeding a rapid expanding world population and by the discovery of the C₄ pathway of photosynthetic carbon assimilation. Population growth has served as a goad to research, and the low photorespiration and high efficiency of the C₄ pathway have suggested a promising route to aid in achieving increased plant yields’ (Burris and Black, 1976). At the time, attempts were made to transfer the advantages of C₄ photosynthesis (reduced photorespiration, and increased nitrogen, water efficiency and, under certain conditions, increased light use efficiency) into C₃ plants by breeding. Although C₃ and C₄ species of Atriplex were hybridized, the independent inheritance of the genes conferring C₄ characteristics, such as Kranz anatomy, elevated PEP carboxylase and low CO₂ compensation point, indicated that plant breeding was unlikely to be a successful route (Björkman et al., 1971; Björkman, 1976). However, the advent of genetic engineering in crop plants has recently resulted in attempts to introduce C₄ characteristics into C₃ plants, such as tobacco, potato and rice, by genetic manipulation (Matsuoka et al., 2001). Clearly, it is easier to over‐express enzymes of the C₄ pathway either individually or in concert, than it is to introduce the complex Kranz leaf anatomy of C₄ photosynthesis, with its division of labour between thin‐walled mesophyll cells surrounding the thick‐walled chlorenchymatous bundle‐sheath. For this reason, considerable interest has focused on engineering single‐celled C₄‐type CO₂‐concentrating mechanisms, such as that found in the aquatic plant, Hydrilla verticillata. The aim of this review is to discuss the requirements of a CO₂‐mechanism as found in C₄ plants, and thus to indicate the minimum requirements of any genetically‐engineered system.

What are the requirements for a CO₂‐concentrating mechanism in C₄ plants?

C₄ photosynthesis has a number of features characteristic of, and essential to, CO₂‐concentrating mechanisms in general. These components have been defined earlier (Badger and Spalding, 2000) (Fig. 1). The number for each component discussed below corresponds to those in Fig. 1.

(1) An active, photosynthetically‐driven, CO₂ capture system

In C₄ plants this role is taken by PEP carboxylase operating in the cytosol of the specialized mesophyll cells. A specialized C₄ isoform of PEP carboxylase is expressed in very high amounts (c. 20 times that of the housekeeping ‘C₃’ isoform), under the control of a promoter that directs expression specifically in the mesophyll cells (Stockhaus et al., 1997). The properties of the C₄ PEP carboxylase are slightly modified compared with the C₃ form (Chollet et al., 1996). Mutants that lack the C₄ form, but which retain the housekeeping form of PEP carboxylase, are unable either to concentrate CO₂ or to assimilate CO₂ diffusing from the atmosphere directly in the bundle‐sheath (Dever et al., 1995).

An aspect which is of crucial importance, but which is frequently overlooked, is the role of carbonic anhydrase in C₄ photosynthesis. PEP carboxylase utilizes bicarbonate, rather than CO₂. This means that CO₂ entering the mesophyll cells from the atmosphere must be rapidly converted to bicarbonate, catalysed by carbonic anhydrase (Hatch and Burnell, 1990; Badger and Price, 1994). Carbonic anhydrase in the leaves of C₄ plants is probably confined to the mesophyll cytosol, albeit at activities which are only just sufficient to ensure that the conversion of CO₂ to bicarbonate does not limit photosynthesis. This is largely attributable to the low concentration of CO₂ in the mesophyll (∼4 μM) compared with the high K_m(CO₂) of carbonic anhydrase (between 0.8 and 2.8 mM in a range of C₄ species) (Hatch and Burnell, 1990). In C₃ plants, on the other hand, carbonic anhydrase is largely confined to the chloroplast (Badger and Price, 1994).

(2) A supply of photosynthetic energy

In C₄ plants ATP is used to drive the C₄ cycle. In the simplest form of C₄ photosynthesis, found in NADP‐malic enzyme species, such as sugar cane (Fig. 2), photosynthetically‐generated ATP (2ATP) is used to regenerate PEP from pyruvate, catalysed by pyruvate, Pi dikinase, and 1 NADPH is used to reduce oxaloacetate to malate in the mesophyll. In the bundle‐sheath chloroplast 1 NADPH is generated by the action of NADP‐malic enzyme. The C₄ cycle is thus an ATP‐driven CO₂ pump, utilizing 2ATP per CO₂ transferred from the mesophyll to the bundle‐sheath.

(3) An intermediate pool of captured CO₂

Oxaloacetate, the immediate product of PEP carboxylation is highly reactive and does not accumulate to high concentrations in maize (an NADP‐malic enzyme species), though it is present in significant quantities in Amaranthus edulis, an NAD‐malic enzyme species (Leegood and von Caemmerer, 1988). Generally oxaloacetate is converted to C4 acids, such as malate and aspartate, which act as transient stores of fixed CO₂ (Hatch, 1971). The relative proportions of malate and aspartate formed from oxaloacetate are dependent on the C₄ sub‐type (little aspartate is formed in pure NADP‐malic enzyme species), and probably other factors such as light intensity and nitrogen nutrition (Khamis et al., 1992).

(4) A mechanism to release CO₂ from the intermediate pool

In C₄ plants this role is taken by specific enzymes decarboxylating C4 acids in the bundle‐sheath. These decarboxylases (NAD‐ and NADP‐malic enzymes and PEP carboxykinase) all release CO₂ (the substrate of Rubisco) rather than bicarbonate, although this release occurs in the chloroplast in the case of NADP‐malic enzyme, in the cytosol in the case of PEPCK, and in the mitochondria in the case of NAD‐malic enzyme. Although C₄ plants have traditionally been classified into three subgroups based on the presence of one of these decarboxylases, it is now becoming apparent that there is much more diversity. In PEPCK types, PEPCK works in tandem with NAD‐malic enzyme (Burnell and Hatch, 1988b). Many NADP‐malic enzyme species, including maize, also have PEPCK as an auxiliary decarboxylase, decarboxylating aspartate (Gutierrez et al., 1974; Wingler et al., 1999), and other NADP‐malic enzyme types exhibit appreciable decarboxylation of aspartate as well as malate (Meister et al., 1996). This diversity probably reflects the multiple evolutionary origins of C₄ photosynthesis. The advantages and disadvantages of particular pathways of decarboxylation are unclear, but their presence is a testament to the flexibility of plant metabolism.

(5) A compartment in which to concentrate CO₂ around Rubisco

In the majority of C₄ plants the photosynthetic cells within the leaf are organized in two concentric cylinders. The outer cylinder comprises thin‐walled mesophyll cells with large intercellular spaces which radiate from the inner cylinder of thick‐walled bundle sheath cells. The thickened cell wall is a major barrier to the diffusion of solutes and gases in all C₄ plants and the bundle‐sheath is the compartment in which CO₂ is concentrated. Direct measurements of the inorganic carbon pool show that it is about some 10–20 times the mesophyll CO₂ concentration (Hatch, 1999). Although many C₃ plants possess a bundle‐sheath, it is seldom green to the degree found in C₄ plants. In rice or barley, for example, the bundle‐sheath contains only a few chloroplasts. In most C₃ plants, the parenchymatous bundle‐sheath may be important in the loading of assimilates into the vasculature (Koroleva et al., 2000) rather than playing any major role in CO₂ fixation. It has also been suggested that increased bundle‐sheath size may be important in maintaining hydraulic integrity in hot, arid environments (Sage, 2001) and could, therefore, be preadaptive. Factors that are involved in the development of the bundle‐sheath have yet to be elucidated at the molecular level (Taylor, 2000; Dengler and Nelson, 1999).

(6) A means to reduce leakage of CO₂ from the site of CO₂ elevation

Three factors are important here. First, the C₄ structure provides a long liquid diffusion pathway from the bundle‐sheath organelles to the intercellular spaces surrounding the mesophyll cells, particularly if the chloroplasts are centripetally arranged in the bundle‐sheath, as, for example, in NAD‐malic enzyme and PEP carboxykinase type C₄ species. Second, the bundle sheath has thickened walls, which reduce its permeability to CO₂ (Table 1). One factor that may affect leakage of CO₂ from the bundle‐sheath is the occurrence of a suberized lamella. It is absent in dicotyledonous species, and in grasses is present only in species with either an uneven bundle sheath outline or with centrifugally located chloroplasts. In those species with uneven cell outlines the suberized lamella may be important in restricting CO₂ leakage through the high surface area of the bundle sheath/mesophyll interface (Hattersley, 1992). Third, the absence of carbonic anhydrase in bundle sheath cells is necessary for the effective concentration of CO₂ in the bundle‐sheath. This is critically important because the substrate for Rubisco is CO₂, not bicarbonate, and carbonic anhydrase activity would lead to the accumulation of bicarbonate rather than CO₂ as the predominant inorganic carbon species. For example, at equilibrium at pH 8 the bicarbonate concentration would be 50 times that of CO₂ (Furbank and Hatch, 1987; Burnell and Hatch, 1988a).

Detailed modelling of the compartmentation of the inorganic carbon pool in mesophyll and bundle sheath has shown that the efflux of

\({\mathrm{HCO}_{3}^{{-}}}\)

via plasmodesmata is insignificant compared to the flux of C4 acids (Furbank and Hatch, 1987). It has, therefore, been suggested that leakage of

\({\mathrm{HCO}_{3}^{{-}}}\)

via the plasmodesmata is not likely to be a serious problem, nor is the leakage of CO₂, because the diffusion coefficients of gases in solution are 10⁴ times less than in air. Of course, if substantial CO₂ leakage occurred, this would reduce the efficiency of C₄ photosynthesis by acting as a futile cycle and increasing its quantum (light) requirement (Hatch et al., 1995).

(7) Modification of the kinetic properties of Rubisco

In C₄ plants, Rubisco has an intermediate affinity for CO₂. The K_m(CO₂) of C₄ grasses ranges between 28–60 μM, compared to C₃ grasses with a range of 13–26 μM (Yeoh et al., 1980), a high S_rel (the relative specificity factor for Rubisco, relating the carboxylase to oxygenase reaction kinetics) comparable to that in C₃ plants, and a catalytic turnover rate about double that of C₃ Rubisco (Seemann et al., 1984). An increased catalytic turnover of Rubisco is always associated with a reduced affinity for CO₂, i.e. there is a trade‐off between the two (von Caemmerer and Quick, 2000; Badger and Spalding, 2000). Since Rubisco in C₄ plants operates at high CO₂ concentrations (well above the K_m for CO₂), the effects of the reduced affinity are negligible, and the increased catalytic turnover results in improved nitrogen use efficiency in that C₄ plants maintain a higher photosynthetic rate for a lower investment in Rubisco, which can form up to 50% of the soluble protein in the leaves of C₃ plants.

Specific requirements of C₄ photosynthesis

In addition to these general requirements of a CO₂‐concentrating mechanism, C₄ plants, by virtue of having the C₄ cycle separated into two distinct cell types, have specialized arrangements for both intracellular and intercellular transport of metabolites, and differences in the regulation, that are crucial to its operation.

One vital aspect of C₄ photosynthesis is metabolite movement between the bundle sheath and the mesophyll. This is sustained by diffusion via numerous plasmodesmata, driven by gradients in their concentrations (Leegood, 1985; Stitt and Heldt, 1985). The C₄ plasmodesmatal frequency can be 3–5 times the C₃ frequency at the mesophyll/bundle‐sheath interface (Botha, 1992). The necessity for metabolite transport between the mesophyll and bundle sheath also sets limits on the amount of mesophyll tissue that can be functionally associated with bundle sheath tissue. Close contact between the two is required, dictating the proximity of mesophyll and bundle sheath cells and, therefore, influencing leaf structure. For this reason, the leaf thickness is limited in C₄ plants and the interveinal distance (i.e. the number of mesophyll cells between adjacent bundle sheaths) is usually smaller than in the leaves of C₃ plants (Hattersley, 1992). Thus C₃ plants engineered to express C₄ photosynthesis between the mesophyll and bundle sheath would be compromised with regard to metabolite transport if the interveinal distance were large.

The intracellular exchange of metabolites occurs at a much greater rate than in C₃ plants and C₄ plants possess translocators in the chloroplasts and mitochondria with unique, or considerably altered, kinetic properties that can transport key metabolites of the C₄ cycle, such as the C4 acids, malate and aspartate, as well as pyruvate, oxaloacetate and PEP (Leegood, 2000). For example, the exchange of inorganic phosphate, PEP, glycerate‐3‐P, and triose‐P occurs in chloroplasts of C₃ and C₄ plants via the phosphate translocator. The affinity for PEP of this translocator is 55‐fold higher than that of the phosphate translocator in C₃ chloroplasts (Gross et al., 1990).

Again, in C₃ and C₄ plants, the dicarboxylates (oxaloacetate, malate, 2‐oxoglutarate, glutamate, and aspartate) undergo counter‐exchange across the chloroplast envelope on the dicarboxylate translocator (Day and Hatch, 1981). However, in malate‐forming C₄ plants, such as maize, oxaloacetate uptake would not occur when oxaloacetate concentrations are several orders of magnitude less than malate concentrations. There is, therefore, an additional high affinity oxaloacetate carrier in maize mesophyll chloroplasts which is little affected by malate. Although this carrier is also present in chloroplasts of the C₃ plant, spinach, it has a very much lower capacity (Hatch et al., 1984). Clearly, limitations imposed by C₃ transporters on transport of intermediates of the C₄ cycle between the cytosol and the chloroplast could easily arise. Thus it would seem desirable to introduce specific transporters in any programme to engineer a CO₂‐concentrating mechanism.

Specialization of the functions of the mesophyll and bundle‐sheath has required modification of their processes. For example, specialization of electron transport pathways in the bundle‐sheath chloroplasts of NADP‐malic enzyme plants has led to a predominance of photosystem 1‐dependent cyclic electron transport leading to ATP formation (Chapman et al., 1980), and a low photosystem 2 activity. This means that NADPH is only generated by NADP‐malic enzyme, which is sufficient to reduce only 50% of the glycerate‐3‐P generated by Rubisco in the bundle sheath. All C₄ sub‐types, therefore, possess the enzymes for glycerate‐3‐P reduction in the mesophyll chloroplasts (Hatch and Osmond, 1976). However, even in those C₄ sub‐types that have photosystem 2 in the bundle‐sheath, glycerate‐3‐P is reduced in the mesophyll. It can thus be inferred that the glycerate‐3‐P/triose‐P shuttle between the bundle‐sheath and mesophyll serves an important function in C₄ metabolism. This could be (i) a means of ensuring H⁺ transport, and hence charge balance, between the two cell types; (ii) to decrease the amount of reductant (NADPH) required for reduction of glycerate‐3‐P in the bundle‐sheath and hence a decrease in photosynthetic O₂ evolution in the bundle sheath, thus improving the CO₂/O₂ ratio, favouring Rubisco carboxylation over oxygenation; (iii) co‐ordination of the Benson–Calvin cycle and C₄ cycle turnover. Co‐ordination of activities in the mesophyll and bundle‐sheath requires metabolic cross‐talk to occur. For example, triose phosphates and hexose phosphates signal between the C₃ and C₄ cycles by acting as positive effectors (‘metabolite messages’) of PEP carboxylase activity (Chollet et al., 1996; Walker and Leegood, 1999).

Just as mitochondria in leaves of C₃ plants have developed a high capacity to decarboxylate the glycine generated by photorespiration (Douce et al., 2001), mitochondria have a specialized role in photosynthesis in both NAD‐malic enzyme and PEPCK‐type C₄ plants, which use NAD‐malic enzyme to decarboxylate malate in the bundle sheath mitochondria (Carnal et al., 1993). Since photosynthetic fluxes vastly exceed respiratory fluxes, this requires some uncoupling of C₄ acid decarboxylation from normal mitochondrial metabolism by engagement of the alternative, cyanide‐insensitive pathway of respiration (Agostino et al., 1996).

Aquatic C₄ photosynthesis

The availability of inorganic carbon for photosynthesis in water is limited by diffusion (10⁴ lower than in air) and pH. In addition, ambient O₂ concentrations may rise to twice those in air‐saturated water. These are conditions which potentially lead to high Rubisco oxygenase activity and hence high rates of photorespiration. As a result, many aquatic photosynthesizers have developed a variety of CO₂‐concentrating mechanisms. In the majority of cases, in algae, cyanobacteria and many aquatic angiosperms, these involve biophysical solutions (the uptake of CO₂ or bicarbonate or proton extrusion to convert bicarbonate to CO₂). However, at least three members of the Hydrocharitaceae, Hydrilla verticillata, Egeria densa (Casati et al., 2000) and Elodea canadensis (de Groote and Kennedy, 1977) show evidence of C₄ metabolism. Elodea species and Egeria densa additionally have a mechanism of apoplastic acidification that enhances the conversion of

\({\mathrm{HCO}_{3}^{{-}}}\)

to CO₂ (van Ginkel et al., 2001). There is also evidence for C₄ metabolism in a marine macroalga, Udotea flabellum (Reiskind and Bowes, 1991) and in a marine diatom, Thalassiosira weissflogii (Reinfelder et al., 2000; though see Johnston et al., 2001). The best studied of these is Hydrilla verticillata, which is a freshwater, submerged angiosperm, a common introduction to freshwater lakes in the USA. In winter, H. verticillata plants are scattered in open water, but in summer they form dense mats of vegetation just below the surface. Plants in these mats encounter low dissolved inorganic carbon, especially low CO₂, alkaline pH values, high temperatures, elevated O₂ concentrations, and high irradiance (Reiskind et al., 1997). When growing under these conditions, C₄‐like characteristics are induced, accompanied by low CO₂ compensation points, low rates of photorespiratory CO₂ release, minimal inhibition of photosynthesis by O₂, and increased net photosynthesis (Salvucci and Bowes, 1981). This is accompanied by a suite of metabolic changes characteristic of C₄ photosynthesis, including the induction of PEP carboxylase (Salvucci and Bowes, 1981), pyruvate, Pi dikinase, NADP‐malic enzyme, and aminotransferases (Magnin et al., 1997), and transfer of label from C4 acids into intermediates of the Benson–Calvin cycle (Salvucci and Bowes, 1983). PEP carboxylase is cytosolic, whereas PPDK, NADP‐malic enzyme and NADP‐malate dehydrogenase are predominantly chloroplastic (Magnin et al., 1997) (Fig. 3). However, the structural features of H. verticillata show that it lacks the Kranz anatomy characteristic of terrestrial C₄ species (Reiskind et al., 1997). The lamina of the H. verticillata leaf is only two cells thick and no compartmentation of PEP carboxylase and Rubisco into separate cells has been detected. Direct measurements of the internal carbon pool indicated about a 5‐fold concentration over that in the surrounding water (equivalent to about 400 μM CO₂ in the chloroplast), but no such concentration in uninduced ‘C₃’ plants (Reiskind et al., 1997). This rules out the notion that, rather than having an active CO₂‐concentrating mechanism, H. verticillata is simply refixing photorespired CO₂, as in some C₃–C₄ intermediates, such as Moricandia arvensis (Monson and Rawsthorne, 2000).

There is the evidence that the C₄ mechanism is also inducible in the amphibious leafless sedge, Eleocharis vivipara, which has C₃ biochemical traits under submerged conditions, but develops C₄ biochemical traits, as well as Kranz anatomy, under aerial conditions, a process regulated by abscisic acid (Ueno, 1998). A C₄ system lacking Kranz anatomy also appears to operate in aquatic leaves of Orcuttia spp., which is a grass that germinates and establishes in seasonal pools, followed by a short aerial phase of growth after pools dry out (Keeley, 1998).

Could a single‐celled C₄ system work in air?

A number of attempts are being made to introduce the enzymic components of C₄ photosynthesis into C₃ plants, such as rice, tobacco and potato. These are based on the H. verticillata system (i.e. PEP carboxylation in the mesophyll and decarboxylation of C₄ acids in the chloroplast) (Fig. 3) and the premise that a single‐celled CO₂‐concentrating mechanism could operate in the mesophyll cells of a C₃ leaf in air.

Single cell CO₂‐concentrating mechanisms are effective in algae and cyanobacteria. However, it must be emphasized that both have internal compartments within the chloroplast (the pyrenoid in the algae and some bryophytes and the carboxysome in cyanobacteria) with a low CO₂ permeability (Table 1) that prevent CO₂ leakage from the site of CO₂ release (Badger and Spalding, 2000; Smith and Griffiths, 2000) and both function in the aquatic environment. How does a single‐cell C₄‐like system operate to concentrate CO₂ in the vicinity of Rubisco in H. verticillata? There is, as yet, no evidence for reduced membrane permeability or for any internal structures that could concentrate CO₂ within the chloroplast of H. verticillata. Although biological cells and membranes show considerable variability in their permeability to CO₂ (Table 1), perhaps partly influenced by the presence of aquaporins (Nakhoul et al., 1998; Prasad et al., 1998), at present it seems likely that the major factor tending to prevent CO₂ loss from the cells is the diffusive resistance provided by the unstirred layer at the boundary between the leaf and the environment (Keeley, 1998). This situation would also obtain in Orcuttia in its submerged state but, additionally, in Orcuttia the centripetal arrangement of mesophyll chloroplasts means that chloroplasts are placed at the greatest possible distance from the environment, which may facilitate the operation of this single‐celled C₄ system both in water and in air (Keeley, 1998).

It has only recently emerged that a fully terrestrial plant, the hygrohalophytic chenopod Borszczowia aralocaspica, has a single‐cell CO₂‐concentrating mechanism. One means of identifying the presence of a C₄‐like CO₂ concentrating mechanism is to measure carbon isotope composition. Differences in carbon isotope composition arise because Rubisco discriminates against naturally occurring ¹³CO₂ much more than PEP carboxylase. Hence C₃ plants are depleted in ¹³C, with carbon isotope signatures ranging from about −20 to −35‰, compared with C₄ plants with carbon isotope signatures ranging from about −10 to −14‰ (Cerling, 1999). The δ¹³C values of B. aralocaspica were −13.03‰ from young stems, and −13.78‰ from leaves (Freitag and Stichler, 2000), consistent with the occurrence of C₄ photosynthesis. In addition, photosynthesis in B. aralocaspica was not inhibited by O₂, indicating the absence of photorespiration (Voznesenskaya et al., 2001). B. aralocaspica has a novel type of leaf anatomy with single‐layered chlorenchyma in which elongated cells surrounding the central water storage tissue are radially arranged (Freitag and Stichler, 2000). At the inner, centripetal, end there are large chloroplasts that are granal, have starch and high amounts of Rubisco, and there are no intercellular air spaces. However, the centrifugal end is exposed to the intercellular space. There are some chloroplasts along the radial walls of the cells towards the centrifugal end, but these have few grana, little starch and low Rubisco. Immunolocalization has shown that NAD‐malic enzyme is specifically located in the mitochondria at the centripetal end of the cell, while PEP carboxylase is distributed throughout the cytosol (Voznesenskaya et al., 2001). The single‐layered photosynthetic tissue is thus dimorphic and combines the essential anatomical and functional characteristics of a two‐layered chlorenchyma of regular C₄ plants. It is inferred that the ‘bundle‐sheath’ reactions occur at the inner end of the cell, and that the ‘mesophyll’ reactions occur at the outer, centrifugal end. Clearly, there are a lot more questions to answer as to mechanism, but presumably it is able to function as a CO₂‐concentrating mechanism in air because of the long diffusion path from one end of the cell to the other. So the answer to the question ‘Could a single‐celled C₄ system work in air?’ is apparently ‘yes’, but only with leaf anatomy substantially different from that of a conventional C₃ leaf.

Biotechnological approaches to C₄ photosynthesis

The simplest theoretical single‐celled system that could be engineered into a C₃ leaf would be PEP carboxylation in the cytosol, oxaloacetate transport into the chloroplast, decarboxylation of oxaloacetate to PEP by PEP carboxykinase in the chloroplast, followed by transport of PEP back to the cytosol. This could be achieved at the expense of only 1ATP per CO₂ transferred, although no C₄ plant is known to possess such a system. Alternatively, the Hydrilla system utilizes two more steps with the decarboxylation of malate by NADP‐malic enzyme and the regeneration of PEP from pyruvate by pyruvate, Pi dikinase (Fig. 3). This would involve expressing carbonic anhydrase in the cytosol to ensure a supply of bicarbonate to PEP carboxylase, PEP carboxylase in the cytosol, NADP‐malic enzyme and pyruvate, Pi dikinase in the chloroplast, and transporters for PEP and oxaloacetate to ensure that the operation of the cycle was not limited by intracellular transport. It might be that the activity of NADP‐malate dehydrogenase would also be insufficient. Attempts have been made to introduce these various steps, mainly singly, into the leaves of C₃ plants. In all these attempts, it must be borne in mind that three of these enzymes (PEP carboxylase, PEP carboxykinase and pyruvate, Pi dikinase) are regulated by phosphorylation, so that the introduced enzyme may be down‐regulated in vivo by post‐translational regulation, thus defeating attempts at over‐expression. This can be avoided by introducing unregulated forms of the enzymes, either by site‐directed mutagenesis or from other organisms.

Introducing a single enzyme, or even an incomplete portion of the C₄‐cycle, is, of course, unlikely in itself to have a large impact on photosynthesis. However, there is some evidence that manipulations have led to the desired redirection of fluxes. Maize PEP carboxylase was over‐expressed in tobacco plants leading to 2‐fold higher activities. Transgenic tobacco plants had significantly elevated levels of titratable acidity and malic acid, consistent with this increased expression, but no physiological changes with respect to photosynthetic rate or CO₂ compensation point (Hudspeth et al., 1992). The authors concluded that PEP carboxylase was probably working maximally at only a few per cent of the rate of photosynthesis. Ku et al. introduced the intact gene of maize PEP carboxylase into rice giving remarkably high levels of expression (Ku et al., 1999). The activities of PEP carboxylase in leaves of some of these transgenic rice plants were 2–3‐fold higher even than those in maize, and the enzyme accounted for up to 12% of the total leaf‐soluble protein. It is clear from this work that increasing the amounts of PEP carboxylase in isolation does not have dramatic effects on photosynthesis, although it may alter stomatal conductance (Ku et al., 2000). The transgenic rice plants exhibited reduced O₂ inhibition of photosynthesis, but this is probably due to effects of phosphate recycling than effects on photorespiration (Matsuoka et al., 2000; Leegood and Furbank, 1986). It is likely that the PEP carboxylase introduced into rice is largely inactive in vivo because of down‐regulation by dephosphorylation (Chollet et al., 1996). Interestingly, in their early work with Atriplex crosses, Björkman et al. already concluded that elevated PEP carboxylase activity by itself (up to a 10‐fold increase) had no effect on the CO₂ compensation point, the photosynthetic rate, or the inhibitory effects of O₂ on photosynthesis (Björkman et al., 1971).

Attempts to engineer the Hydrilla‐type system into potato suggest that modest expression of PEP carboxylase and NADP‐malic enzyme, either singly or in combination, can influence photorespiratory characteristics. For example, substantial over‐expression of PEP carboxylase from Corynebacterium glutamicum in potato led to an increase in the rate of dark respiration and a small decrease in Γ*, the CO₂‐compensation point measured in the absence of dark respiration (Häusler et al., 1999). Double transformants with an additional 3–5‐fold over‐expression of Flaveria pringlei NADP‐malic enzyme in the chloroplast showed a temperature‐dependent decrease in the electron requirement for CO₂ assimilation, again suggesting a slight suppression of photorespiration (Lipka et al., 1999). Double transformants of potato (PEP carboxylase and NADP‐malic enzyme) exhibited the most consistent attenuating effect on photorespiration, as shown by reductions in Γ* as well as temperature and oxygen effects on photosynthesis. However, these are most likely a result of local changes in CO₂ concentration induced by increasing respiration (Häusler et al., 2002).

Suzuki et al. over‐expressed an unregulated phosphoenolpyruvate carboxykinase (PCK) from the C₄ plant, Urochloa panicoides, in the chloroplasts of rice plants (Suzuki et al., 2000). In ¹⁴CO₂ labelling experiments, up to 20% of the radioactivity was incorporated into C4 acids (malate, oxaloacetate, and aspartate) in leaves of transgenic plants, as compared with about 1% in excised leaves of control plants. When ¹⁴C‐malate was fed to excised leaves the extent of incorporation of radioactivity into sucrose was 3‐fold greater in transgenic plants than in control plants and the level of radiolabelled aspartate was significantly lower in transgenic plants. Thus, expression of PCK in rice chloroplasts led to a partial change in carbon flow in mesophyll cells into a C₄‐like photosynthetic pathway. In tobacco, introduction of the PEP carboxykinase gene from the bacterium Sinorhizobium meliloti, either singly or in combination with PEP carboxylase, had little effect on photosynthetic parameters (Häusler et al., 2001).

Potato plants over‐expressing maize pyruvate, Pi dikinase showed a decrease in pyruvate and an increase in malate and a small, but significant decrease in δ¹³C, suggesting increased PEP carboxylation (Ishimaru et al., 1998). There was, however, no change in CO₂‐compensation point. In rice, over‐expression of maize pyruvate, Pi dikinase (f*ckayama et al., 2001) has been claimed to lead to a higher photosynthetic rate, associated with higher stomatal conductance (Ku et al., 2000).

Potentially useful pleiotropic effects have also occurred in these manipulations. A 20–70‐fold increase in maize NADP‐malic enzyme in rice leaves (located mainly in the chloroplasts) led to aberrant chloroplast structure with agranal thylakoid membranes, and an inverse correlation between NADP‐malic enzyme activity and chlorophyll and photosystem II activity (Takeuchi et al., 2000). This is particularly interesting in relation to the presence of agranal chloroplasts in the bundle‐sheath of NADP‐malic species. Hoewever, other studies of rice over‐expressing maize NADP‐malic enzyme have also indicated a reduction in chlorophyll content, reduced growth and enhanced photoinhibition (Tsuchida et al., 2001), probably resulting from over‐reduction of the NADP pool as a result of a high activity of the over‐expressed enzyme in vivo (Takeuchi et al., 2000; Tsuchida et al., 2001).

Conclusions

Leaving aside all the arguments about whether or not it is desirable to engineer C₄ photosynthesis into C₃ plants (Sheehy et al., 2000), it seems unlikely that attempts to introduce single cell CO₂‐concentrating mechanisms will be successful without also introducing some of the structural characteristics of C₄ photosynthesis, that is, a compartment in which CO₂ could be concentrated. If the anatomical characteristics of C₄ plants were to be incorporated, this means understanding the complex factors that regulate the development of the different cell types in C₄ plants and, although this may only be controlled by a few genes, these have not yet been identified. Even after making the bundle‐sheath of an existing C₃ plant green by enhancing chloroplast number, and expressing the appropriate enzymes in the mesophyll and bundle‐sheath, the leaf structure would still be inappropriate for the operation of C₄ photosynthesis because of poor cell‐to‐cell communication. Equally, engineering a single‐celled C₄ system, such as that found in Borszczowia, would require a reworking of C₃ leaf structure. The alternative is to take a different approach to reducing photorespiration. One approach would be to introduce pyrenoids or carboxysomes (intracellular compartments for the concentration of CO₂ that are found in algae and cyanobacteria) into the chloroplasts of C₃ plants. Another would be to move the location of photorespiratory glycine metabolism to the bundle‐sheath, taking as a working example the suppression of photorespiration that occurs in C₃–C₄ intermediates (Winzer et al., 2001; Monson and Rawsthorne, 2000). The last alternative would be to express, in C₃ plants, improved forms of Rubisco, notably those from rhodophyte algae in which the relative specificity for CO₂ compared to O₂ is higher than that of higher plant Rubisco (Whitney et al., 2001).

Research in Sheffield into C₄ photosynthesis and photorespiratory metabolism was supported by the Biotechnology and Biological Sciences Research Council, UK.

References

© Society for Experimental Biology

Comments

0 Comments

I agree to the terms and conditions. You must accept the terms and conditions.

Email alerts

Citing articles via

More from Oxford Academic