C4-Ization of C3 Plants

Water equilibrated at normal atmospheric pressure dissolves 11-µM CO2, which forms 110-µM HCO3- at pH 7.2 and 25 °C (Yokota and Kitaoka, 1985). While RuBisCO fixes CO2, phosphoenolpyruvate carboxylase (PEPC) uses HCO3- as the substrate. This characteristic confers a tremendous advantage to C4 plants. Since the Km for HCO3- of maize PEPC is as low as 20 mM (Uedan and Sugiyama, 1976), this enzyme can exhibit submaximal activity in the mesophyll cytosol. In the case of the C4 plant maize, oxalacetate formed by PEPC in mesophyll cells is reduced to malate and then decarboxylated by NADP+ -dependent malic enzyme in the mitochondria of bundle sheath cells to give rise to CO2 and pyruvate (Heldt, 1997; Kanai and Edwards, 1999). Pyruvate returns to mesophyll chloroplasts to be salvaged to phosphoenolpyruvate (PEP) by pyruvate Pi dikinase (PPDK). The active operation of this pathway can convert HCO3- in mesophyll cytosol to CO2 concentrated in bundle sheath cells. The CO2 concentration around RuBisCO in chloroplasts of bundle sheath cells reaches 500 µM (von Caemmerer and Furbank, 1999), causing net CO2 fixation to be saturated at 10–15 Pa CO2 without any detectable photorespiration (Edwards and Walker, 1983). Thus, this auxiliary metabolic CO2-pumping system confers significantly better nitrogen investment and water-use efficiencies to C4 plants compared with C3 plants. If this CO2-pumping system could be introduced into C3 plants, the transgenic plants would be expected to show highly improved photosynthetic performance and productivity (Ku et al., 1996).

The maize PEPC gene has been introduced into rice chloroplasts (Ku et al., 1999). Although the severalfold higher PEPC activity in chloroplasts did not influence carbon metabolism (Häusler et al., 2002), transgenic plants expressing over 50 times more PEPC activity than wild type exhibited slightly higher CO2-fixation rates that were relatively insensitive to O2 (Ku et al., 1999). The primary CO2-fixation product in these transgenic plants was PGA, not C4 acid (Fukayama et al., 2000). However, the introduction of single C4 genes will not establish a metabolic CO2-pumping system since this transgenic rice depends on glycolysis for the supply of PEP (Matsuoka et al., 2001). Maize malic enzyme and PPDK have been individually introduced into rice plants, but positive effects on photosynthesis have not been observed (Fukayama et al., 2001; Tsuchida et al., 2001). One unexplained consequence of the ectopic expression of the maize NADP+ -malic enzyme in C3 chloroplasts has been either the lack or disturbance of grana, possibly indicating altered protein–protein interactions (Takeuchi et al., 2000). The incorporation of both PEPC and PPDK into rice, generated by crossing of single-gene transformants, has been achieved and the plants appeared to behave in a more C4-like fashion (Ku et al., 2001). Introduction of more than two C4 genes into C3 plants has not yet been attempted.

Unlike C4 plants, C3 plants transgenic for all three genes may not fix CO2 efficiently since the diffusion of CO2 in cytosol and through membranes is rapid. An observation that seems to support this prediction is that cyanobacteria concentrate HCO3- within cells to a level up to 103 times higher than the ambient CO2 concentration (Kaplan and Reinhold, 1999). The genes for the CO2-pumping systems have been identified (Shibata et al., 2002). Endogenous carbonic anhydrase is localized in carboxysomes where the HCO3- is dehydrated to CO2 to be fixed by RuBisCO (Kaplan and Reinhold, 1999). Induction of a high level of carbonic anhydrase activity in the cytosolic space caused conversion of HCO3- into CO2, which was released from the cells at a rate sufficient to nullify the pumping activity (Price and Badger, 1989). It will be important to learn more and understand how such high local concentrations of CO2 around RuBisCO can be maintained and possibly engineered into higher plant chloroplasts. In this context, the C4-type performance of Borszczowia aralocaspica (Chenopodiaceae) from the Gobi desert (Voznesenskaya et al., 2001) provides another interesting example. In this plant, RuBisCO and NAD+ -malic enzyme are localized in chloroplasts and mitochondria, respectively, and are located at the proximal end of cells. Chloroplasts reside in the distal part of the cells and contain PPDK, but not RuBisCO, while PEPC is located throughout the cell. Understanding how such a spatial arrangement of enzymes is accomplished and maintained will be important for the recreation of a functional C4 pathway in C3 plants.