Single and double overexpression of C₄‐cycle genes had differential effects on the pattern of endogenous enzymes, attenuation of photorespiration and on contents of UV protectants in transgenic potato and tobacco plants

Abstract

To improve the efficiency of CO₂ fixation in C₃ photosynthesis, C₄‐cycle genes were overexpressed in potato and tobacco plants either individually or in combination. Overexpression of the phosphoenolpyruvate carboxylase (PEPC) gene (ppc) from Corynebacterium glutamicum (cppc) or from potato (stppc, deprived of the phosphorylation site) in potato resulted in a 3–6‐fold induction of endogenous cytosolic NADP malic enzyme (ME) and an increase in the activities of NAD‐ME (3‐fold), NADP isocitrate dehydrogenase (ICDH), pyruvate kinase (PK), NADP glycerate‐3‐P dehydrogenase (NADP‐GAPDH), and PEP phosphatase (PEPP). In double transformants overexpressing cppc and chloroplastic NADP‐ME from Flaveria pringlei (fpMe1), cytosolic NADP‐ME was less induced and pleiotropic effects were diminished. There were no changes in enzyme pattern in single fpMe1 overexpressors. In cppc overexpressors of tobacco, the increase in endogenous cytosolic NADP‐ME activity was small and changes in other enzymes were less pronounced. Determinations of the CO₂ compensation point (Γ*) as well as temperature and oxygen effects on photosynthesis produced variational data suggesting that the desired decline in photorespiration occurred only under certain experimental conditions. Double transformants of potato (cppc/fpMe1) exhibited the most consistent attenuating effect on photorespiration. In contrast, photorespiration in tobacco plants appeared to be diminished most in single cppc overexpressors rather than in double transformants (cppc/fpMe1). In tobacco, introduction of the PEP carboxykinase (PEPCK) gene from the bacterium Sinorhizobium meliloti (pck) had little effect on photosynthetic parameters in single (pck) and double transformants (cppc/pck). In transgenic potato plants, increased PEPC activities resulted in a decline in UV protectants (flavonoids) in single cppc or stppc transformants, but not in double transformants (cppc/fpMe1). PEP provision to the shikimate pathway inside the plastids, from which flavonoids derive, might be restricted only in single PEPC overexpressors.

C₄‐cycle, malic enzyme, phosphoenolpyruvate carboxylase, phosphoenolpyruvate carboxykinase, transgenic potato/tobacco.

A, CO₂ assimilation, C_i, intercellular CO₂ concentration, e/A‐ratio, electron requirement for apparent CO₂ assimilation; Γ*, CO₂ compensation point in the absence of dark respiration in the light, NAD(P)‐GAPDH, NAD(P)‐dependent glycerate‐3‐phosphate dehydrogenase, NAD(P)‐ME, NAD(P)‐dependent malic enzyme, NADP‐ICDH, NADP‐dependent isocitrate dehydrogenase, PEPCK, phosphoenolpyruvate carboxykinase, PEPC, phosphoenolpyruvate carboxylase, PEPP, phosphoenolpyruvate phosphatase, PK, pyruvate kinase, PPDK, pyruvate orthophosphate:dikinase, R_d, dark respiration in the light, Rubisco, ribulose‐1,5‐bisphosphate carboxylase/oxygenase

Introduction

Photorespiration commences with the oxygenase reaction of Rubisco and results in the loss of at least 25% of CO₂ assimilated in air (Leegood et al., 1995). Environmental conditions such as high temperature and water stress lead to increases in the rate of photorespiration and thus decrease the efficiency of CO₂ assimilation in C₃ plants. In contrast, C₄ plants have developed a strategy to prevent losses of CO₂ by photorespiration in that the CO₂ concentration in the vicinity of Rubisco is increased and the oxygenation of RuP₂ is thus suppressed (Leegood, 1997). A prerequisite for this CO₂ concentrating mechanism is an anatomical and biochemical division into mesophyll cells and a layer of gas‐tight bundle sheath cells. The first step in the C₄‐cycle is the carboxylation of PEP by PEPC in the cytosol of the mesophyll cells using HCO₃⁻ as the inorganic carbon substrate. This yields the C₄ dicarboxylic acid oxaloacetate (OAA), which is either reduced to malate or transaminated to aspartate and is then transported into the bundle sheath cells, where CO₂ is released at high rates depending on the type of C₄ metabolism, either by NADP‐ME, NAD‐ME, or PEPCK. This strategy permits high rates of CO₂ assimilation at a relatively small stomatal aperture and enhances water use efficiency in C₄ plants compared to C₃ plants. Besides C₄ plants, there are a variety of C₃–C₄ intermediates (Monson et al., 1984), which exhibit different strategies to increase the CO₂ concentration at the site of Rubisco. Interestingly, there are also submersed aquatic plants, which display an inducible C₄ like metabolism, but lack the Kranz‐anatomy typical for C₄ plants (Magnin et al., 1997; Spencer et al., 1996). In Hydrilla verticillata, for instance, several enzymes typical for the C₄‐cycle are induced (i.e. PEPC, NAD/NADP‐ME, and PPDK) when the CO₂ concentration in solution drops (i.e. at increased water temperatures). This induction of C₄‐cycle enzymes is accompanied by a decrease in the CO₂ compensation point from about 40 μl l⁻¹ to below 20 μl l⁻¹ (Reiskind et al., 1997). The question arises whether similar features can be introduced into terrestrial C₃ plants by a molecular approach. First attempts into this direction have been encouraging (Gehlen et al., 1996; Hudspeth et al., 1992; Ishimaru et al., 1997; Kogami et al., 1994; Ku et al., 1999; Suzuki et al., 2000; Takeuchi et al., 2000). In transgenic potato, the introduction of PEPC from C. glutamicum resulted in a decrease in the CO₂ compensation point Γ* (Häusler et al., 1999). However, this was accompanied by an increase in CO₂ release by respiration in the light and dark (Gehlen et al., 1996; Häusler et al., 1999). Most likely, an increased PEPC activity acts as an HCO₃⁻/CO₂ pump in the mesophyll at the expense of photoassimilates. Thus, more CO₂ is released than is actually fixed by PEPC. This results in a decrease in Γ*, due to an increase in the CO₂/O₂ ratio, but also in a higher rate of CO₂ loss from the plants. Furthermore, in transgenic potato plants with a combined overexpression of PEPC and chloroplastic NADP‐ME, the electron requirement for CO₂ assimilation was lower than in the wild‐type or single PEPC or NADP‐ME overexpressors when the leaf temperature was increased above 30 °C. This suggests an appreciable attenuation of photorespiration (Lipka et al., 1999).

In this study, the physiological and biochemical responses of wild‐type potato and tobacco plants, two closely related solanaceous species, were compared with transformants overexpressing C₄‐cycle genes individually and in combination. There has been a focus here on changes in the pattern of enzymes that are directly or indirectly involved in the metabolism of PEP or products of PEP carboxylation and the effects on CO₂ assimilation and photosynthetic electron transport have been assessed. Changes in enzyme pattern and photosynthetic parameters observed in the transgenic plants are interpreted as compensational processes or adaptations to a redirection of metabolic fluxes.

Materials and methods

Plant material and gene constructs

Wild‐type potato (Solanum tuberosum cv. Désirée) and tobacco (Nicotiana tabacum L. cv. Petit Havanna SR1) plants were transformed with the leaf disc method (Horsch et al., 1985). As controls, wild‐type plants were transformed with the empty pS vector. Transgenic potato lines containing the genes cppc from C. glutamicum (designated 3LD and 9×9–1) or fpMe1 from F. pringlei (designated ME) have already been described in detail previously (Gehlen et al., 1996; Lipka et al., 1999). For tobacco, the single cppc overexpressing line cppc197‐4 was used.

Construction and overexpression of a malate‐insensitive potato PEPC in S. tuberosum

In potato, a modified endogenous PEPC gene (stppc) from potato (GenBank: X90982) was overexpressed under the control of the double 35S promoter. In order to reduce the malate sensitivity of ppc1 the phosphorylation site was modified: Glu₇, Lys₈, and Leu₉ were replaced by a Ser and Ser₁₁ was replaced by Asp. A further modification concerned 37 amino acids from positions 384 to 420. The DNA sequence coding for these amino acids was exchanged by the corresponding sequence from Flaveria trinervia. Modifications were accomplished by standard cloning techniques. The resulting gene was designated Stppc_S9D‐C4.

The bacterial expression vector pTrc99A was used to express modified proteins for further characterization. PEPC coded by Stppc_S9D‐C4 had a lower K_m(PEP) value (23 μM) than the endogenous PEPC (57 μM) and a strongly reduced malate sensitivity and was therefore used in further experiments. The modified gene was cloned into the plant vector pPAM (GeneBank Accession No. AY027531) together with an expression cassette containing the double 35S‐Promotor (35SS) (Kay et al., 1987), the 5′‐untranslated region of ‘tobacco etch virus’ (TL) (Carrington and Freed, 1990) and the polyadenylation site pA35S. The DNA sequence between the distal part of the 35SS promoter and the proximal part of the modified stppc gene was as follows, tttggagaggacc (35SS), tcgagaattct (transition sequence), caacacaaca tatacaaaac aaacgaatct caagcaatca agcattctac ttctattgca gcaatttaaa tcatttcttt taaagcaaaa gcaattttct gaaaattttc accatttacg aacgat (TL), agcc (transition sequence), atg gct acg agg aat ctg agc gca gat att gat (stppc_S9D‐C4). Potato transfomants carrying the modified version of stppc were designated SPsStSD. Tobacco double transformants carrying cppc and fpMe1 were essentially generated as described previously (Lipka et al., 1999) for potato plants and were designated (cppc+ME×cppc)(158–4×273–2 : 11) and (cppc+ME)(158–4×273–1 : 3).

Cloning of the PEPCK gene and functional expression in E. coli

Transgenic tobacco plants were generated containing the PEPCK gene (pckA) from Sinorhizobium meliloti 2011 (Aguilar et al., 1985) targeted to the chloroplast. PckA was amplified from plasmid pTH195 (Osteras et al., 1995) with the 5′‐oligonucleotide Lpck5′ (gg aat tca ggA CGC GTt agg tgc ATG gat gag ctc ggc agc cgc aat cc) and a 3′‐oligonucleotide complementary to the vector sequence including the EcoRI recognition sequence of the multiple cloning site. The 5′‐oligonucleotide Lpck5′ contained the start codon of pckA. Four triplets coding for the last amino acids from the carboxy terminus of Rubisco (rbcS1) preceded the ATG. The MluI restriction site was used to ligate the PCR product to the 5′‐part of rbcS1 resulting in an rbcS1‐pckA fusion. This oligonucleotide additionally contained an EcoRI‐ and a SstI‐site. The PCR‐amplified sequence was cloned into the EcoRI site of pUC19 resulting in pRm1, which was the basis for further cloning into the IPTG inducible bacterial expression vectors pKK223‐3, pTrc99A and pGEX‐5X‐3 (Pharmacia) as well as into the plant expression vector pS. The EcoRI fragment from pRm1 was inserted into the EcoRI sites of pKK223–3 and pGEX‐5X‐3 resulting in pRm5 and pRm7; a SstI fragment with pck was cloned into pTrc99A at the SstI‐site to yield pRm6. The three constructs were tested for complementation of an E. coli mutant (HG4) unable to grow on succinate because it lacks pck and the PEPsynthetase gene ( pps) (Goldie and Sanwal, 1980). HG4 mutants transformed with the individual three gene constructs were capable of growing on M9 minimal agar plates at 30 °C and 37 °C with succinate (0.5%) as sole carbon source. However, strong induction of the pck gene by addition of IPTG (5 mM) was fatal for growth indicating a perturbance of metabolism at very high PEPCK activities as was observed earlier (Chao and Liao, 1993). The protein pattern of the two strains on SDS‐PAGE revealed a strong band at the expected molecular weight positions of 58 kDa for HG4(pRm6) and 87 kDa for a GST‐PEPCK fusion protein in HG4(pRm7). Concomitantly, PEPCK activity increased 140‐ and 660‐fold, respectively, in induced overnight cultures. The E. coli strain BL21 (Grodberg and Dunn, 1988) was used for production of PEPCK‐GST fusion protein, required for raising specific antibodies against PEPCK from S. meliloti.

Transformation of tobacco plants with pS‐Lpck

The construction of the plant vector pS‐Lpck containing the pckA gene from S. meliloti fused to the Rubisco (rbcS1) transit sequence of potato was analogous to that of pSrLpps described earlier (Panstruga et al., 1997). The expression cassette from pBinHygTX (35S‐TripleX) (Gatz et al., 1992) had been inserted into a pUC19 vector via EcoRI and HindIII (pRP8). In addition, the pck gene from yeast in fusion with the rbcs1 transit sequence was ligated into the SmaI site of the cassette resulting in plasmids pRP19. Most part of the chimeric rbc‐pdk(yeast) gene (except for the proximal part of rbcS1 leader) was replaced by a MluI/BamHI fragment from pRm1, thus substituting the yeast pck completely by pck from S. meliloti fused to the distal part of rbc1. The complete rbcs1 transit sequence was regenerated by this procedure. To eliminate possible PCR errors, a 1759 bp SstI fragment was exchanged by the corresponding fragment of pTH195. The resulting plasmid was designated pRm3. As it was intended to express the gene under the original 35S promoter, the chimeric gene was excised with HpaI and BamHI and the ends (filled in by Kleenow polymerase) were ligated into the SmaI‐site of vector pS (Panstruga et al., 1997) to yield pS‐Lpck. Tobacco lines containing pS‐Lpck were designated pck.

For the generation of tobacco double transformants (cppc/pck) from cppc single transformants a hygromycin resistance‐mediating plasmid was used. Therefore, the whole expression cassette (including 35S promoter, Lpck and polyadenylation site pA_CaMV) from pS‐Lpck was excised with HindIII and ligated into the HindIII‐site of pRP10. The latter plasmid was derived from pBinHygTX by replacing the 35S TripleX/pAOcs expression cassette with a 16 bp EcoRI/HindIII adaptor. The resulting plasmid was designated pBIN‐Lpck.

Screening for PEPCK positive tobacco plants was performed immunologically, essentially as described for the detection of NADP‐ME (Lipka et al., 1999). Single transformants overexpressing pck were designated pck(270‐10) and pck(276‐10). Double transformant lines of tobacco containing cppc and pck were designated (cppc×pck)(100–4×276–10 : 1) and (cppc×pck)(100–4×609–2 : 6).

Isolation of potato ME cDNAs by RT‐PCR

cDNAs for the detection of transcript amounts of cytosolic and chloroplastic NADP‐ME as well as of NAD‐ME from potato were generated by RT‐PCR (Chelly et al., 1988) with total RNA isolated from wild‐type potato using gene specific primer pairs. For NAD‐ME the sequence data of the potato enzyme (62 kDa subunit) was used (Winning et al., 1994; Accession No. Z23023, forward: gttcttaatgctaggagg, reverse: tcctggatatcgaagggtata). This resulted in a PCR product with a length of 630 bp. For cytosolic and chloroplastic NADP‐ME, the sequence information published for tomato plants (Lycopersicon esculentum; leMe2; Accession No. AF001270, leMe1, Accession No. AF001269) was used to design degenerated primer pairs using stretches of the coding region, which were least homologous to published sequences of NADP‐MEs from C₃ and C₄ plants. Potato‐specific cDNAs were amplified with these primers. For the amplification of stMe2 the forward and reverse primers were, at(act)gtxgtxacxga(ct)ggxga and (ct)tc(ct)xgt xac(ct)tgxgcxgc, respectively. For stMe1 the following primer combination was used, tt(ct)aaraarccxtgggcxca (forward); xcc(ct)tt(ct)tcraartt(ct)tc(ct)tg (reverse). This resulted in PCR fragment sizes of 422 bp and 1037 bp for stMe2 and stMe1, respectively. The PCR products were quantitatively separated on agarose gels and electro‐eluted onto Nylon membranes. A portion of the cDNA was ligated into pGEM^®‐T Easy (Promega) and sequenced (Perkin Elmer, ABI Prism, 310 genetic analyser). The amino acid sequences derived from the individual PCR products were identical to those published.

Extraction of RNA and Northern blot analysis

For RNA isolation, leaf material of greenhouse‐grown potato plants was harvested at midday and frozen in liquid nitrogen. RNA was extracted using the RNase ALL medium (internet information; ‘bionet.molbio.methds‐regnts’). Total RNA (10 μg per lane) was separated on 1% agarose gels in Mops‐formaldehyde (Sambrook et al., 1989), subsequently transferred on Porablot NY amp membranes (downward transfer) and hybridized with ³²P‐labelled cDNA probes (according to Church and Gilbert, 1984). The signal strength was quantified using a phosphorimager (Storm 860, Molecular Dynamics) equipped with the Image Quant program (Molecular Dynamics, Sunnyvale, California, USA). The intensity of the signal was normalized for slight differences in the amount of 28S rRNA on the membrane.

Plant growth and propagation

Potato plants were grown from tubers in a growth chamber (14/10 h light/dark; 22 °C) and propagated by cuttings 3–4 weeks before an experiment. Tobacco plants were grown in a greenhouse (25 °C) with supplemental illumination (4000 lx, Osram HQI‐T400/D lamps) for 16 h and propagated by cuttings until experiments were performed.

Extraction of leaf material for enzyme assays

Samples for enzyme assays were taken from upper source leaves of 6‐week‐old potato plants or 5‐weeks‐old tobacco plants immediately after photosynthesis measurements (see below). Leaf material was frozen in liquid nitrogen and stored at −80 °C until use. Frozen leaf material was extracted in a medium containing 50 mM Hepes‐NaOH (pH 7.5), 5 mM MgCl₂, 2 mM EDTA, 10% (v/v) glycerol, 0.1% Triton‐X 100, and optional 10 mM DTT. The extracts were desalted by gel filtration on NAP‐5 columns (Pharmacia). Enzyme activities were either measured directly after extraction or in extracts, which were brought to 50% (v/v) glycerol and stored at −20 °C. Storage of extracts had no effect on enzyme activities for at least up to 4 weeks.

Enzyme assays

PEPC:

 PEPC activities were determined using two different reaction mixtures modified after Ashton et al. (Ashton et al., 1990). Assay 1 contained, 100 mM Tricine‐KOH (pH 8.0), 5 mM MgCl₂, 2 mM DTT, 1 mM NaHCO₃, 0.2 mM NADH, 2 U MDH, and 5 mM PEP; assay 2 contained, 100 mM Mes‐NaOH (pH 6.6), 50 mM NaHCO₃ (pH of the reaction mixture 7.2), 5 mM MnCl₂, 0.2 mM NADH, 5 mM DTT, 2 U MDH, and 5 mM PEP.

PEPCK:

PEPCK was assayed in, 100 mM Mes‐NaOH (pH 6.6), 50 mM NaHCO₃ (pH of the reaction mixture 7.2), 5 mM MnCl₂, 0.2 mM NADH, 5 mM DTT, 5 mM PEP, 4 mM ATP, and 2 U MDH.

NADP(NAD)‐ME:

The activity of NADP‐ME was determined as described previously (Häusler et al., 1987) in 100 mM Tricine‐NaOH (pH 8.4), 25 mM malate, 25 mM MgCl₂, and 0.6 mM NADP, and NAD‐ME was determined according to Ashton et al. (Ashton et al., 1990) in 50 mM Hepes‐NaOH (pH 7.2), 5 mM malate, 4 mM MnCl₂, 2 mM EDTA, 0.2 mM NAD, 5 mM DTT, and 0.1 mM Coenzyme A.

NADP(NAD)‐MDH:

In order to activate NADP‐MDH fully the extracts were mixed with DTT to give a final concentration of 50 mM and were incubated for 30 min at room temperature (Holaday et al., 1992). The assay mixture contained, 100 mM Tricine‐NaOH (pH 8.0), 5 mM EDTA, 2 mM OAA (prepared freshly prior to start of the reaction), 0.2 mM NADPH, and 20 mM DTT. NAD‐MDH was assayed in 100 mM Hepes‐NaOH (pH 7.5); 5 mM EDTA, and 2 mM NADH.

NADP(NAD)‐GAPDH:

The assay for NADP‐GAPDH (Leegood, 1990) contained, 100 mM Tricine‐NaOH (pH 7.8), 5 mM 3‐PGA, 0.2 mM NADPH, 2 mM ATP, 8 mM MgSO₄, 2 mM DTT, 1.0 mM EDTA, and 2 U ml⁻¹ PGK. For NAD‐GAPDH the same assay medium was used with the exception that NADPH was replaced by NADH and DTT was omitted.

PK, PEPP:

PK was assayed according to Plaxton (Plaxton, 1990) either in 100 mM Hepes‐NaOH (pH 6.9) or (pH 7.9), 2 mM PEP, 2 mM ADP, 50 mM KCl, 10 mM MgCl₂, 0.2 mM NADH, 0.2 mg ml⁻¹ BSA, 2 mM DTT, and 2 U LDH. PEPP activity was determined in the same assay mixtures as for PK in the absence of ADP.

NADP‐ICDH:

NADP‐ICDH was assayed according to Chen (Chen, 1998) in 100 mM K‐phosphate (pH 7.5), 5 mM MgCl₂, 2 mM isocitrate, and 0.1 mM NADP.

Isolation of chloroplasts from potato and tobacco leaves

Chloroplasts were isolated according to Barlett et al. at 0–4 °C on a reduced scale (Bartlett et al., 1982); leaf material (approximately 10 cm² leaf area) was disintegrated between two small interlocking ribbed rollers (MEKU, Wennigen, Germany) and 2 ml extraction buffer was added immediately. The homogenate was passed through one layer of Miracloth, collected in a 2 ml tube and centrifuged in a Microfuge for 2 s. The supernatant was removed with a syringe and the pellet resuspended. Chloroplast isolation was completed within 2 min.

Native polyacrylamide gel electrophoresis and NADP‐ME activity staining

Separation of native proteins and staining for NADP‐ME activity on polyacrylamide gels was done according to Häusler et al. (Häusler et al., 1987) with the system of Davis (Davis, 1964) using 7.5% separation gels. Activity of NADP‐ME was visualized using the standard reaction supplemented with 0.5 mM EDTA, PMS (10 μg ml⁻¹) and NBT (200 μg ml⁻¹). As a control, MgCl₂ was omitted.

Photosynthesis measurements

CO₂ assimilation and modulated Chl a fluorescence were measured as described elsewhere (Häusler et al., 1999). O₂ inhibition of CO₂ assimilation was measured either in a gas mixture containing 1490 μl l⁻¹ CO₂, 21% O₂ balanced with N₂ or 1510 μl l⁻¹ CO₂, 2% O₂ balanced with N₂ provided by gas cylinders. The CO₂ concentration of the incoming gas stream were lowered to values expected for ambient air with a CO₂ diluter (Analytical Development Company, UK). The intracellular CO₂ concentration (C_i) was adjusted by varying the concentrations of CO₂ in the incoming gas stream and was kept constant in all sets of experiments. For measurements of photosynthesis, fully developed upper source leaves of 6‐week‐old potato plants or 5‐week‐old tobacco plants were used. In order to consider diurnal changes in photosynthetic performance, measurements were done with the same plants after 4–7 h light on the first day and 9–13 h light on the second day. After the measurements, leaves were excised, plunged into liquid N₂ and stored at −80 °C. These leaf samples were used for the determination of enzyme pattern.

Determination of UV protectants (flavonoids)

For flavonoid measurements, leaf discs were extracted as described previously (Pinto et al., 1999) in MeOH : H₂O : HCl (79 : 20 : 1 by vol.) at 60 °C for 1 h (with 99.8% MeOH and 35.4% HCl) and the absorption at 310 nm was determined after cooling down the extracts to room temperature.

Statistical evaluation of experimental data

All data shown in the tables and figures are mean values ± standard deviation (SD) of 2–7 independent data acquisitions using leaves from different plants of individual transformant or control lines. Significant differences between wild‐type/control plants and transformants were assayed using the Welch‐test (Lorenz, 1984), which allows for unequal relative errors between two groups of measurements assuming that a Gauss distribution is applicable. The authors are aware that this is not necessarily the case for all sets of data. Note that a high variability within a set of data and/or a limited data number makes significant differences between wild‐type/control plants and transformants less probable, despite similar or identical mean values.

Results

PEPC from plants and C. glutamicum can be distinguished by differential assay compositions

The kinetic properties of PEPC from different sources may vary. For instance, the K_m(PEP) of PEPC from C. glutamicum ranges between 0.8–5.0 mM depending on the presence of acetyl CoA (100 μM), an activator of the bacterial enzyme, whereas the K_m(PEP) of PEPC from potato is lower (190 μM) (Gehlen et al., 1996). The activities of endogenous PEPC from potato and tobacco leaves and of the bacterial enzyme can be distinguished in vitro by modifications of the assay composition. As is shown in Table 1A and B, assay 2 is more specific for PEPC from C. glutamicum, which displays high activities only in those transgenic plants overexpressing the bacterial enzyme, either alone or in combination with other C₄‐cycle enzymes such as NADP‐ME (potato and tobacco) or PEPCK targeted to the chloroplast (tobacco only). By contrast, assay 1 yields high activities with endogenous PEPCs from potato (particularly in transgenic potato plants overexpressing the engineered version of endogenous PEPC; stppc, Table 1A) and tobacco plants (Table 1B) and from C₃ and C₄Flaveria species (Table 1C). On the basis of activity determinations with assay 2, overexpression of cppc increased PEPC activities in single and double transformants of potato and tobacco by a factor of 10–27 and 6–14, respectively (Table 1A, B). There was no such increase in PEPC activity in single transformants containing only fpMe1 (potato) or pck (tobacco). Moreover, in stppc overexpressors, PEPC activity measured with assay 1 increased 6–9‐fold. Compared to earlier reports (Gehlen et al., 1996; Häusler et al., 1999), PEPC activities in single cppc overexpressors of potato were slightly different. This is mainly due to differences in the assay composition, which was intermediate to assays 1 and 2 with respect to Mg²⁺ and HCO₃⁻ concentrations (compare Gehlen et al., 1996). High PEPC activities observed earlier in the cppc overexpressing line 9×9‐1 were not apparent in this study. In conclusion, as substrate and ion concentrations as well as pH in planta are less well defined, the use of different assay compositions helped to distinguish maximum activities of PEPC from bacterial and plant sources in vitro (similar to immunodetection of the individual proteins on Western blots), but would not allow an unequivocal estimate of the in vivo activities of either enzyme.

Table 1.

Activities of PEPC, NADP‐ME and NAD‐ME in (A) wild‐type potato (Désirée), a range of transgenic potato plants overexpressing bacterial PEPC (cppc, lines 3LD9, 3LD15 or 9×9‐1), a modified potato PEPC (stppc, lines SPsStSD41, 91, 21, and 72), chloroplast NADP‐ME (fpMe1, line ME18) individually or a combination of bacterial PEPC and chloroplast NADP‐Me (cppc/fpMe1, lines 3LD9/ME1, 3, 2, and 5), as well as (B) wild‐type tobacco (SR1), a range of transgenic tobacco plants with an individual overexpression of bacterial PEPC (cppc) and PEPCK (pck) or a combined overexpression of PEPC and PEPCK (cppc/pck) or chloroplast NADP‐ME (cppc/fpMe1) and (C) C₃and a C₄species of Flaveria

(A) Potato

Biotype	Enzyme activity (mU mg–1)
	PEPC	PEPC	NADP‐ME	NADP‐ME+	NAD‐ME
	(assay 1)	(assay 2)	(cytosol)	(chloroplast)
Désirée (Wt)	23.1±5.2	13.1±7.8	63.6±25.4	4.24±3.07	7.6±2.3
	(n=7)	(n=7)	(n=7)	(n=3)	(n=7)
PEPC overexpressors
(cppc)
3LD9	52.1±9.3 e	335.5±26.8 d	226.9±20.1 c	6.91±3.20	20.2±0.5 a
	(n=2)	(n=2)	(n=2)	(n=3)	(n=2)
3LD15	111.3	120.4	173	4.85±2.90	8.0
				(n=3)
9×9‐1	57.0	78.5	77.8	4.02±1.10	7.5
				(n=3)
(stppc)
SPsStSD41	123.0±20.9 a	22.6	160.4±20 a	0.80±0.48	9.5
	(n=7)		(n=3)	(n=3)
SPsStSD91	116.3±21.0 a	27.9	170.6±7.7 a	n.d.	14.8
	(n=7)		(n=3)
SPsStSD21	212.7±38.6 a	31.1	155.9±32 c	2.12±1.88	14.8
	(n=7)		(n=3)	(n=3)
SPsStSD72	159.6±26.9 a	36.0	205.1±47.7 c	n.d.	15.5
	(n=7)		(n=3)
NADP‐ME overexpressors
(fpMe1)
ME 18	28.1±13	16.0±3.2	65.5±4.6	30.2±8.4 b	8.1±1.4
	(n=3)	(n=3)	(n=3)	(n=3)	(n=3)
PEPC/NADP‐ME double transformants
(cppc/fpMe1)
3LD9/ME1	50±17.6 e	365±165.6 c	97.5±16.1 e	14.1±1.9 d	9.4±0.1 e
	(n=2)	(n=2)	(n=2)	(n=3)	(n=2)
3LD9/ME2	26.2±9.7	38.3±4.4 a	67.1±7.7	n.d.	9.3±1.4
	(n=3)	(n=3)	(n=3)		(n=3)
3LD9/ME3	36.5±13.6	161.7±29.9 b	82.4±12.2	n.d.	9.4±2.0
	(n=3)	(n=3)	(n=3)		(n=3)
3LD9/ME5	49.4±15.9 c	314.9±59.1 a	109.5±30.6 d	20.6±2.3 b	11.2±2.8 e
	(n=5)	(n=5)	(n=5)	(n=3)	(n=5)

(B) Tobacco
Biotype	Enzyme activity (mU mg−1)
	PEPC	PEPC	NADP‐ME	NADP‐ME+	PEPCK+	NAD‐ME
	(assay 1)	(assay 2)	(cytosol)	(chloroplast)	(chloroplast)
SR1 (Wt)	92.3±20.3	20.3±3.2	35.9±15.2	0.00	0.2±0.3	3.5±0.7
	(n=4)	(n=4)	(n=4)	(n=3)	(n=3)	(n=4)
PEPC overexpressors	113.9±16	286.1±81.5 c	48.6±15.6	n.d.	n.d.	5.5±0.2 b
(cppc)	(n=2)	(n=2)	(n=2)			(n=2)
PEPCK overexpressors	90.8±23.7	13.7±7.9	45.8±1.2	n.d	7.0±2.0 c	4.3±1.8
(pck)	(n=3)	(n=3)	(n=3)		(n=3)	(n=3)
PEPC/NADP‐ME	82.3±23.0	197.7±25.5 a	67.6±13.8 c	47.7±23.0	n.d.	5.2±0.9 c
double transformants	(n=5)	(n=5)	(n=5)	(n=3)		(n=5)
(cppc/fpMe1)
PEPC/PEPCK	83.3±6.5	135.7±25.4 b	47.9±11.6	n.d.	5.45±0.7 b	4.6±0.8 e
double transformants	(n=4)	(n=4)	(n=4)		(n=3)	(n=4)
(cppc/pck)

(C) Flaveria species
Flaveria species	Enzyme activity (mU mg−1)
	PEPC	PEPC	Total NADP‐ME	NAD‐ME
	(assay 1)	(assay 2)
F. angustifolium (C3)	82.3	42.8	14.3	7.8
F. pringlei (C3)	25.7	14.6	13.1	7.7
F. trinervia (C4)	887.4	146.4	737.9	36.4

Where indicated (+)NADP‐ME and/or PEPCK activities were also determined in isolated chloroplasts of potato and/or tobacco plants and referred to chloroplast protein. These data were corrected for cytosolic contaminations. PEPC was either measured at a pH of 8.0, with 5 mM PEP, 1 mM HCO₃⁻ (assay 1) or at pH 7.2, in 50 mM Mes‐NaOH (pH 6.6); 50 mM HCO₃⁻ and 5 mM PEP (assay 2). Where indicated, the data represent the mean±SD, where n represents the number of leaf samples taken from different plants of the individual lines. The data sets were tested for significant differences between transformants and the corresponding wild‐type plants using the Welch method (see Materials and methods). The corresponding P‐values were labelled by letters: a, P<0.001; b, P<0.01; c, P<0.02; d, P<0.05; e, P<0.1 (n.d., not determined). Data without SD represent single measurements.

Overexpression of pck targeted to chloroplasts in tobacco plants

Plant PEPCK is a cytosolic enzyme, which is involved in the CO₂ concentrating mechanism in C₄ and CAM plants as well as in gluconeogenesis in a variety of tissues and specialized cells of C₃ plants (Leegood et al., 1999). High activities of PEPC and PEPCK in one compartment would result in a futile cycle and the waste of ATP. As it was desired to release CO₂ in the vicinity of Rubisco and to prevent futile cycling, the bacterial pck gene from S. meliloti was targeted to the chloroplasts of transgenic tobacco plants by fusing it to the Rubisco (rbcS1) transit sequence of potato. The functionality of the PEPCK protein was tested by complementation of an E. coli strain unable to grow on succinate as the sole carbon source (see Materials and methods).

PEPCK activity was barely detectable in leaf extracts of wild‐type tobacco or transformants overexpressing the pck gene targeted to the chloroplast. Moreover, in plants containing pck in combination with cppc, the high background of bacterial PEPC activity prevented an accurate determination of PEPCK activity. However, PEPCK activity determined in isolated chloroplasts increased from 0.2 mU mg⁻¹ in the wild type by a factor of 35 and 27 in single and double transformants carrying pck alone and in combination with cppc (Table 1B). The poor activity of PEPCK in leaf extracts could be due to a high sensitivity of the PEPCK protein towards proteolytic degradation (Leegood et al., 1999). However, limited proteolysis does not affect V_max.

The activity of endogenous NADP‐ME increases as a result of PEPC overexpression

In transgenic potato plants, total activities of NADP‐ME were raised up to 4‐fold in leaf extracts of PEPC (cppc, lines 3LD9, 3LD15 9×9‐1; stppc lines SPsStSD41, 91, 21, and 72) overexpressors as well as in double transformants (lines 3LD/ME1, 2, 3, 5) containing cppc in combination with fpMe1 (Table 1A). There was also an increase in mitochondrial NAD‐ME, which was significant in the single cppc overexpressing line 3LD9 and prominent in most of the stppc overexpressing lines tested (single measurements, Table 1A). The increase in total NADP‐ME activity in the double transformants (cppc/fpMe1) was not due to overexpression of fpMe1 per se as stromal NADP‐ME activity in single fpMe1 transformants (line ME18) could only be unequivocally detected in chloroplast, but not in leaf extracts (Lipka et al., 1999). NADP‐ME activities measured in leaf extracts of individual fpMe1 overexpressors were therefore similar to wild‐type plants, but they were considerably increased in chloroplast extracts of single and double transformants of potato as well as of tobacco plants carrying the fpMe1 gene (Table 1A). Moreover, the large increase in NADP‐ME activity in leaf extracts of PEPC overexpressors was not observed in chloroplast extracts suggesting that this activity was localized outside the chloroplasts (Table 1A).

On the basis of chloroplast protein contents, endogenous chloroplast NADP‐ME activity was low and ranged between 0 and 4.24 mU mg⁻¹ in wild‐type tobacco and potato plants or single PEPC overexpressors, but increased significantly (between 14 and 47 mU mg⁻¹) in chloroplasts extracted from plants overexpressing fpMe1 (i.e. lines ME18, 3LD9/ME1 and 5 for potato and cppc/fpMe1 for tobacco). From the data shown in Table 1A, it appears likely that the increase in extraplastidial NADP‐ME responds to the increase in PEPC in a dose‐dependent manner both in cppc and stppc overexpressors of potato. Interestingly, in double transformants of potato carrying the fpMe1 gene in combination with cppc, total NADP‐ME activity was lower compared to single cppc overexpressors despite similar magnitudes of foreign PEPC activity suggesting that the introduction of the second gene (i.e. fpMe1) triggers the decline of endogenous cytosolic NADP‐ME activity (Table 1A).

For tobacco, there was also the trend of an increase in total NADP‐ME in cppc overexpressors, which was more apparent after NADP‐ME activity staining on native gels (see below). An increase in NADP‐ME activity (2‐fold) was significant only in double transformants carrying cppc in combination with fpMe1 (Table 1B). Hence, an overexpression of PEPC has differential effects on changes in cytosolic NADP‐ME activity between different species, even between closely related species such as potato and tobacco.

Staining of NADP‐ME activity isolated from leaf and chloroplast extracts on native gels

Staining for NADP‐ME activity on native polyacrylamide gels reveals that, despite distinct migration distances, there was only one major NADP‐ME activity band extracted from potato and tobacco leaves that was considerably increased in single cppc (line 3LD9 for potato and line cppc197‐4 for tobacco) or double cppc/fpMe1 (line 3LD9/ME5, potato only) overexpressors (Fig. 1A). It should be noted that the activity of cytosolic NADP‐ME in the respective wild‐type plants shown in Fig. 1 was extremely low compared to other experiments (not shown). The faint second activity band, which was observed only in wild‐type potato, was most probably due to endogenous chloroplastic NADP‐ME (Fig. 1A). The stained bands in the absence of MgCl₂ (control) represent most likely chloroplastic NADP‐MDH. NADP‐ME activity in chloroplast extracts of single and double transformants of potato could only be detected when the plants express chloroplastic NADP‐ME (fpMe1) (Fig. 1B). This is consistent with the large increase of NADP‐ME activities in the respective chloroplast extracts (Table 1A). Furthermore, a comparison of the R_f values indicated that the relative migration distance of the activity band of introduced NADP‐ME was different from the major activity band found in whole leaf extracts (not shown).

Fig. 1.

Activity staining of NADP‐ME after separation of leaf or chloroplast extracts on native polyacrylamide gels. (A) Leaf extracts were prepared from wild‐type potato (Désirée) or tobacco (SR1) or transgenic plants overexpressing PEPC (cppc) individually or in combination with chloroplast NADP‐ME (fpMe1; 3LD9/ME5; for potato only). The solid arrows indicate the location of cytosolic NADP‐ME activity bands, whereas the dotted arrow indicates a faint activity band, which is presumably due to chloroplastic NADP‐ME. As a control MgCl₂ was omitted. The activity bands in the absence of MgCl₂ represents most likely NADP‐MDH. In (B) chloroplast extracts prepared from wild‐type/control potato plants as well as from single PEPC (cppc) and NADP‐ME (fpMe1) overexpressors and double transformants carrying PEPC in combination with chloroplast NADP‐ME (cppc/fpMe1) were separated on native gels and stained for NADP‐ME activity in the presence of MgCl₂.

Transcript levels of endogenous cytosolic NADP‐ME are increased in transgenic potato plants overexpressing cppc or stppc

In order to explore whether the observed increase in cytosolic NADP‐ME activity in potato plants overexpressing cppc is regulated on the transcriptional or post‐transcriptional level, Northern blot analysis was performed. For this purpose the respective cDNA used as a probe was isolated by RT‐PCR with degenerated primer pairs designed from gene‐specific sequences published for the leMe2 gene from tomato (see Materials and methods). As shown in Fig. 2 (left panel), transcript amounts of cytosolic NADP‐ME from potato (stMe2) were low in control plants, but increased by a factor of 2–6 in individual cppc overexpressors of potato and were slightly elevated in stppc overexpressors (Fig. 2; left panel). The transcript amount of stMe2 in transgenic potato plants carrying the fpMe1 gene alone was even lower compared to the control plants, whereas the abundance of stMe2 transcripts was again slightly increased in individual double transformants carrying cppc in combination with fpMe1 compared to the wild type. Thus, the rise in NADP‐ME activity (Table 1A) is correlated with the increase in transcript amounts of stMe2 in the respective transgenic lines (compare Table 1A). The transcript amounts of stMe2 in leaves of wild‐type potato are also subjected to changes in the nutritional status of the plants (cf. Pinto et al., 1999). Nutrient‐deficient wild‐type potato plants exhibited higher transcript amounts of stMe2 compared to nutrient‐sufficient plants (Fig. 2; right panel [L]). Moreover, there were no large differences in the transcript amounts between leaf blade and midrib (Fig. 2; right panel [LB, MR]).

The mRNAs of endogenous chloroplastic NADP‐ME or mitochondrial NAD‐ME were not detectable on Northern blots suggesting low transcript amounts. However, both transcripts must have been present in leaves, as the corresponding cDNAs were isolated by RT‐PCR (see Materials and methods).

Fig. 2.

Northern blot analysis of cytosolic NADP‐ME (stMe2) transcripts in wild‐type and transgenic potato plants overexpressing either PEPC (cppc or stppc) or chloroplast NADP‐ME (fpMe1) individually or in combination (cppc/fpMe1) (left panel). Each lane contained 10 μg total RNA. The right panel shows transcript amounts of Désirée wild‐type plants, which were nutrient‐sufficient or nutrient‐deficient, and where L, LB and MR denote whole leaf, leaf blade and midrib, respectively.

Pattern of enzymes involved in ‘PEP metabolism’ are affected in transgenic potato and tobacco plants overexpressing C₄‐cycle genes

Effects of overexpressing C₄‐cycle genes on the activity pattern of other enzymes involved in the metabolism of PEP or in the products of PEP carboxylation or ‘proliferation’, were investigated in transgenic potato plants (Table 2A). Besides NAD(P)‐ME (Table 1A), there was an increase in the activity of PK (measured either at the pH optimum of the putative chloroplastic or cytosolic enzymes; i.e. pH 7.9 or pH 6.9) only in single cppc overexpressors, which would be consistent with an increase in glycolytic fluxes. PEPP, an enzyme which catalyses a similar reaction as PK, but without ATP generation, was elevated in single transformants of potato carrying the cppc gene as well as in two lines of the cppc/fpMe1 double transformants when measured at pH 7.9 (Table 2A; PEPP [2]). An increase in PEPP has been discussed as an indicator for phosphate depletion (Duff et al., 1989). Activities of NADP‐MDH and NAD‐MDH, which catalyse the conversion of OAA to malate in the stroma or extra‐chloroplastic compartments, were not affected in single cppc or fpMe1 overexpressors. For the double cppc/fpMe1 overexpressors, only the line 3LD9/ME2 showed an increase in both NAD‐MDH and NADP‐MDH activities. NADP‐ and NAD‐GAPDH catalysing 3PGA/GAP conversions in the stroma or the cytosol were increased only in the double transformants (cppc/fpMe1), but not in the respective single transformants (cppc or fpMe1). NADP‐ICDH activity, which is involved in anaplerotic generation of carbon skeletons for amino acid biosynthesis, showed the trend of a slight increase only in cppc overexpressors.

As for potato, there was an appreciable increase in the activity of PK in individual cppc overexpressors of tobacco, which was prominent for the putative cytosolic enzyme (measured at pH 6.9). In contrast to potato, PK activity was also increased in double transformants of tobacco carrying cppc and fpMe1 (Table 2B). Interestingly, the trend of an increase in PK activity observed in cppc and cppc/fpMe1 overexpressors was less in transformants carrying pck alone or in combination with cppc (Table 2B). In transgenic tobacco plants, there was no large effect on the activities of NADP‐GAPDH, NAD‐GAPDH and NADP‐ICDH. However, NADP‐MDH activity was lowered in PEPCK ( pck) overexpressors, particularly in transgenic tobacco plants overexpressing both cppc and pck (Table 2B). NAD‐MDH activities were not affected in transgenic tobacco plants compared to the wild type. There was also a trend of a decline in PEPP (1) in cppc/fpMe1 double transformants and in cppc overexpressors of tobacco.

Table 2.

Enzyme activities in (A) wild‐type potato and a range of transgenic potato lines overexpressing bacterial PEPC (cppc, line3LD9) and chloroplast NADP‐ME (fpMe1, line ME18) individually or in combination (cppc/fpMe1, lines 3LD9/ME1, 2, 3, 5) and (B) wild‐type tobacco and a range of transgenic lines overexpressing bacterial PEPC (cppc) or PEPCK (pck) individually as well as a combination of PEPC and NADP‐ME (cppc/fpMe1) or PEPC and PEPCK (cppc/pck); PEPP was assayed either at pH 6.9 (1) or pH 7.9 (2)

Biotype

Enzyme activity (mU mg−1)

PEPC (%)

NADP‐MDH

NAD‐MDH

PK (cyt)

PK (p)

PEPP (1)

PEPP (2)

NADP‐GAPDH

NAD‐GAPDH

NADP‐ICDH

(A) Potato

Désirée (Wt) (n=7)

100

16.7±3.4

2520±703

45.9±12.4

36.8±17.1

20.7±5.5

28.9±5.3

175.0±52.2

77.7±30.3

6.0±0.39

PEPC overexpressors

(cppc)

3LD9 (n=2)

2561

17.9±0.1

2584±10

96.3±15.6 e

79.0±21 e

28.2±2.7 e

41.2±2.7 c

171.7±6.5

84.2±0.6

10.2±2.9

NADP‐ME overexpressors

(fpMe1)

ME18 (n=3)

122

16.7±0.8

2570±140

40.6±6.6

31.0±5.5

18.0±3.5

32.5±3.5

164.3±24.1

73.0±11.1

6.6±0.39

PEPC/NADP‐ME double transformants

(cppc/fpMe1)

3LD9/ME1 (n=2)

2786

17.1±1.4

3154±21 e

51.8±3

34.1±1.57

19.2±0.7

41.6±0.7 a

273.6±3.6 b

143.5±.3 b

6.9±3.9

3LD9/ME2 (n=3)

292

24.2±6.4 e

4517±1155 e

62±15.8

56.4±17.6

31.2±8.1

39.8±8.1

270.1±66.6

161.5±69.8

4.3±2.7

3LD9/ME3 (n=3)

1229

16.7±0.8

2829±114

51.4±5.9

40.7±3.1

23.7±10.

34.8±10.8

220.7±25.7

97.2±14.0

5.4±1.9

3LD9/ME5 (n=5)

2404

16.5±3.5

3735±1776

42.3±7.7

40.3±7.5

17.2±4.8

40.2±4.8 b

240.5±64.2 e

141.9±50.5 d

4.7±3.2

(B) Tobacco

SR1 (Wt) (n=4)

100

24.7±7.7

3269±508

12.4±5.2

23.4±6.2

43.0±8.0

34.9±3.4

120.8±60.3

53.1±30.6

10.9±4. 0

PEPC overexpressors

(cppc) (n=2)

1409

25.0±0.8

3299±273

52.1±9.3 e

30.5±8.1

33.2±5.7

39.2±19.2

138.5±56.7

60.2±12.8

12.1±4.4

PEPCK overexpressors

(pck) (n=3)

67

16.4±0.4

2737±409

28.1±13

25.6±6.6

39.2±0.1

30.7±5.8

84.3±60.2

49.0±35.8

9.1±2.3

PEPC/NADP‐ME double transformants

(cppc/fpMe1) (n=5)

974

19.1±6.2

2978±430

50±17.6 b

19.6±7.9

33.2±3.2 e

33.3±6.4

82.1±60.9

43.9±33.9

9.5±3.1

PEPC/PEPCK double transformants

(cppc/pck) (n=4)

668

10.9±4.5 d

3496±687

26.2±9.7 e

21.6±4.6

38.1±5.9

33.1±4.0

145.2±16.3

68.3±8.5

11.9±1.1

The data represent the mean±SD where n represents the number of samples taken from different plants of the individual lines. The data sets were tested for significant differences between transformants and the corresponding wild‐type plants using the Welch method (see Materials and methods). The calculated P‐values were labelled by letters: a, P<0.001; b, P<0.01; c, P<0.02; d, P<0.05; e, P<0.1.The percentage PEPC activity was calculated from the data shown in Table 1A (for potato) and Table 1B (for tobacco). Before sampling photosynthesis parameters were determined with the same leaves (compare Table 3A and B and Figs 4 and 5).

The CO₂ compensation point (Γ*) and dark respiration in the light (R_d) are differentially affected in transgenic potato plants

The effect of single and double overexpression of cppc and fpMe1 on the CO₂ compensation point in the absence of dark respiration in the light (Γ*) and on dark respiration in the light (R_d) was studied with transgenic potato plants using the method of Brooks and Farquhar (Brooks and Farquhar, 1985) described in detail by Häusler et al. (Häusler et al., 1999). Dependencies of CO₂ assimilation (A) on the intercellular CO₂ concentration (C_i) measured for control potato (line pS13) and a cppc overexpressor (line 3LD13) at a range of different non‐saturating photon flux densities (PFD) and at a saturating PFD had a single intersection, which allows the determination of Γ* on the C_i‐axis and −R_d on the A‐axis (Fig. 3A, B). These were, Γ*=38 μl l⁻¹ and 32 μl l⁻¹, and R_d=0.77 μmol m⁻² s⁻¹ and 1.2 μmol m⁻² s⁻¹ for the control plant (line pS13) and the PEPC (cppc) overexpressor (line 3LD13), respectively. However, A/C_i curves of fpMe1 overexpressors (line ME19) and double cppc/fpMe1 transformants (lines 3LD9/ME5 and 2) look more complex because a clear single intersection of the A/C_i curves is missing (Fig. 3C–E). On average, Γ* appeared to be slightly increased in ME19 (Fig. 3C) compared to the control (Fig. 3A). A closer inspection of A/C_i curves obtained for the double transformants 3LD9/ME5 and 3LD9/ME2 suggests a decline in Γ* from 38 μl l⁻¹ in the control plant to 23 μl l⁻¹ and 18 μl l⁻¹ in the lines 3LD9/ME5 and 3LD9/ME2, respectively, only when the PFDs ranged between 150–300 μmol m⁻² s⁻¹. This appeared to be accompanied by an increase in R_d from 0.9 μmol m⁻² s⁻¹ in the control plant to about 1.7 μmol m⁻² s⁻¹ in the double transformants. However, the A/C_i curves at low light (15 μmol m⁻² s⁻¹) and at high light (600 μmol m⁻² s⁻¹) dramatically diverged indicating that this approach for Γ* and R_d determination is not feasible. It is likely that stromal redox potentials are perturbed with increasing PFDs by the introduction of chloroplast NADP‐ME, particularly when it operates in concert with overexpressed PEPC.

Fig. 3.

Typical determinations of CO₂ compensation points (Γ*) and rates of dark respiration in the light (R_d) in the potato control plant pS13 (A), a transgenic line overexpressing PEPC (cppc; line 3LD13) (B) and chloroplast NADP‐ME (fpMe1, line ME18) (C) individually or in combination (cppc/fpMe1, lines 3LD9/ME5 [D] and 3LD9/ME2 [E]). For the control (A) and the single PEPC overexpressor (B), the PFDs were 11, 74, 100, and 320 μmol m⁻² s⁻¹. For single NADP‐ME (fpMe1) overexpressors (C) and double transformants carrying cppc in combination with fpMe1 (D, E), the PFDs were 15(□), 150(•), 300(○), and 600(▵) μmol m⁻² s⁻¹. Arrows point at possible intersections of the curves. ((A) and (B) were taken from Häusler et al., 1999).

Fig. 4.

Temperature dependency of A (A) and e/A‐ratios (B) of potato wild type as well as transgenic lines overexpressing PEPC (cppc) or chloroplast NADP‐ME (fpMe1) either individually (lines 3LD9 or ME18) or in combination (cppc/fpMe1 lines 3LD9/ME1, 3, 2, and 5,) measured in 21% O₂ (black bars) or 2% O₂ (grey bars). Care was taken that the C_i during the measurements remained constant at 200±10 μl l⁻¹, the PFD ranged between 490 and 510 μmol m⁻² s⁻¹. The differences in A and e/A ratios between 35 °C and 25 °C were calculated as negative or positive changes per degree (°C). Negative changes indicate a decline in the measured parameters, whereas positive changes indicate an increase in the parameter with an increase in temperature. The data represent the mean±SD where n represents the number of leaf samples taken from different plants of the individual lines. The data sets were tested for significant differences between transformants and the corresponding wild‐type plants using the Welch method (see Materials and methods). The calculated P‐values are shown in parenthesis with P<0.001 (1), P<0.01 (2), P<0.02 (3), P<0.05 (4), and P<0.1 (5). Note that enzyme activities were measured in extracts prepared from the same leaves used for the determination of photosynthesis parameters (compare Tables 2A and 3A for the corresponding n‐values).

Fig. 5.

Temperature dependency of A(A) and e/A‐ratios (B) in tobacco wild type as well as transgenic tobacco lines overexpressing PEPC (cppc) or PEPCK (pck) individually as well as a combination of PEPC and NADP‐ME (cppc/fpMe1) or PEPC and PEPCK (cppc/pck) measured in measured in 21% O₂ (black bars) or 2% O₂ (grey bars). Care was taken that the C_i during the measurements remained constant at 200±10 μl l⁻¹ and the PFD ranged between 490 and 510 μmol m⁻² s⁻¹. The differences in A and e/A‐ ratios between 35 °C and 25 °C were calculated as negative or positive changes per degree (°C). Negative changes indicate a decline in the measured parameters whereas positive changes indicate an increase in the parameter with an increase in temperature. The data represent the mean±SD where n represents the number of samples taken from different plants of the individual lines. The data sets were tested for significant differences between transformants and the corresponding wild‐type plants using the Welch method (see Materials and methods). The calculated P‐values are shown in parenthesis with P<0.001 (1), P<0.01 (2), P<0.02 (3), P<0.05 (4), and P<0.1 (5). Note that enzyme activities were measured in extracts prepared from the same leaves used for the determination of photosynthesis parameters (compare Tables 2B and 3B for the corresponding n‐values).

Oxygen‐dependent inhibition of CO₂ assimilation and enhancement of electron requirements for CO₂ assimilation are diminished in transgenic potato plants with a combined overexpression of PEPC and NADP‐ME

An increase in atmospheric O₂ from 2% to 21% and/or in temperature from 25 °C to 35 °C (at 21% O₂) increases the rate of oxygenation over the rate of carboxylation of RuBP catalysed by Rubisco (Chen and Spreitzer, 1992) and results in a decline in A and in an increase in the electron requirement for A (e/A‐ratio) as a consequence of higher photorespiration rates. This increase in the e/A‐ratio is due to higher losses of CO₂ per electron transported and to the additional requirement of electrons for the recycling of carbon back into the Calvin cycle.

Photosynthetic parameters of wild‐type and transgenic potato plants were compared on the basis of O₂ and temperature effects on A and e/A‐ratios (for the latter compare Lipka et al., 1999). The percentage O₂ inhibition of A can be expressed as (A_npr−A_pr)/A_npr×100, where npr and pr denote non‐photorespiratory (i.e. 2% O₂) and photorespiratory (i.e. 21% O₂) conditions, respectively (Ku et al., 1999). Likewise, the O₂ effect on the e/A‐ratio can be expressed as (e/A_pr−e/A_npr)/e/A_pr×100 reflecting the relative increase in the e/A‐ratio as photorespiration increases. In the wild type, O₂ inhibition of A and O₂ dependent e/A‐enhancement were increased from about 42.5% and 43.5% at 25 °C to 55.6% and 56.9% at 35 °C, respectively (Table 3A). This shows that both parameters, the fractional O₂ inhibition of A and enhancement of e/A‐ratios are closely linked in the wild type and, as a closer inspection of Table 3A reveals, in all transgenic potato lines investigated. In cppc overexpressors (line 3LD9), both parameters were indistinguishable from the wild type (Table 3A), whereas they were lowered appreciably in fpMe1 overexpressors (line ME18) at 35 °C, but not at 25 °C. The O₂ effect on both parameters was most pronounced for the double transformant lines (3LD9/ME3, 2 and 5). However, there were differential effects on both parameters between different double transformant lines with regard to the temperature. The relative attenuation of both parameters was similar in the line 3LD9/ME5 at 25 °C and 35 °C, whereas it was much more pronounced in the line 3LD9/ME2 at 35 °C than at 25 °C (Table 3A). Thus, both parameters decreased by about 10–12% at 35 °C in the line 3LD9/ME2.

Table 3.

Effect of varying O₂ concentrations (2% and 21%) on CO₂ assimilation (A) and on the electron requirement for CO₂ assimilation (e/A) in (A) wild‐type potato and a range of transgenic potato lines overexpressing bacterial PEPC (cppc, line 3LD9) or chloroplast NADP‐ME (fpMe1, line ME18) individually or in combination (cppc/fpMe1, lines 3LD9/ME1, 2, 3, and 5) as well as (B) wild‐type tobacco and a range of transgenic tobacco lines overexpressing bacterial PEPC (cppc) or PEPCK (pck) individually as well as a combination of PEPC and NADP‐ME (cppc/fpMe1) or PEPC and PEPCK (cppc/pck)

A and photosynthetic electron transport were determined at a PFD of between 490 and 510 μmol m⁻² s⁻¹, in an atmosphere containing either 21% O₂ (photorespiratory [pr] conditions) or 2.0% O₂ (non‐photorespiratory [npr] conditions). The C_i was adjusted in a range between 190 and 210 μl l⁻¹ by diminishing the external CO₂ concentration (1450 μl l⁻¹) using a CO₂ diluter. The percentage O₂ inhibition of A and the O₂‐dependent enhancement of e/A‐ratio was calculated from (A_npr–A_pr)/A_npr×100 and (e/A_pr–e/A_npr)/e/A_pr×100. The data sets were tested for significant differences between transformants and the corresponding wild‐type plants using the Welch method (see Materials and method).

Biotype		O2 effect on photosynthesis
		O2 inhibition of A (%)			O2 enhancement of e/A‐ ratios (%)
		25 °C	35 °C	25 °C	35 °C
(A) Potato
Désirée (Wt)	(n=7)	42.5±3.6	55.6±2.1	43.5±3.6	56.9±5.3
PEPC overexpressors
(cppc)
3LD9	(n=2)	40.8±1.0	52.0±3.5	40.0±1.0 e	54.6±1.3
NADP‐ME overexpressors
(fpMe1)
ME18	(n=3)	40.1±1.9	44.9±5.8 e	39.5±2.6	47.6±6.7 e
PEPC/NADP‐ME double transformants
(cppc/fpMe1			)
3LD9/ME1	(n=2)	37.7±2.3	50.7±3.5	39.5±0.6 d	53.0±1.3
3LD9/ME3	(n=3)	39.4±1.6 e	46.9±4.7 e	39.4±5.3	47.1±5.3 e
3LD9/ME2	(n=3)	38.4±1.5 d	43.5±5.9 e	39.2±6.0	47.0±6.0 e
3LD9/ME5	(n=5)	36.9±2.8 c	47.5±4.3 c	38.0±4.6 e	50.9±4.6 e
(B) Tobacco
SR1 (Wt)	(n=4)	39.9±1.0	51.2±5.1	40.5±0.5	53.1±4.2
PEPC overexpressors
(cppc)	(n=2)	23.9±8.9 e	37.4±14 e	26.9±9.6 e	38.1±18
PEPCK overexpressors
(pck)	(n=3)	41.3±2.9	49.1±1.0	42.0±0.2	53.3±6.5
PEPC/NADP‐ME double transformants
(cppc/fpMe1)	(n=5)	35.0±4.3 e	46.6±4.4	40.3±1.8	51.0±3.9
PEPC/PEPCK double transformants
(cppc/pck)	(n=4)	28.0±9.5 e	45.3±10.5	38.2±7.1	50.7±7.1

The calculated P‐values were labelled by letters: a, P<0.001; b, P<0.01; c, P<0.02; d, P<0.05; e, P<0.1. Note that the same leaves were used for the determination of enzyme activities shown in Table 2A (for potato) and Table 2B (for tobacco).

Temperature dependencies of A and e/A‐ratios in transgenic potato plants

Changes in A and in e/A‐ratios are linearly correlated with the temperature between 25 °C and 35 °C in wild‐type potato and transgenic lines overexpressing fpme1 alone or in combination with cppc (Lipka et al., 1999). Different transgenic lines were compared on the basis of the slopes of A and e/A‐ratios measured between 25 °C and 35 °C (A₃₅−A₂₅)/10 in a gas mixture containing either 21% or 2% O₂. If the rate of photorespiration were decreased in 21% O₂ (i.e. under photorespiratory conditions), a less steep decline in A or a less steep increase in e/A‐ratios would be expected. The slopes (A₃₅−A₂₅)/10 were negative for wild‐type and transgenic potato plants with an individual overexpression of either cppc or fpMe1 (Fig. 4A) indicating that A declines with increasing temperature (see above). There was a decrease in the slopes (A₃₅−A₂₅)/10 to less negative values in double transformants overexpressing cppc/fpMe1 with the most pronounced effect for the line 3LD9/ME2. In 2% O₂, conditions under which the oxygenase reaction of Rubisco is strongly suppressed, A was less affected in the wild‐type and single cppc and fpMe1 overexpressors. However, there was an marked increase in the slopes (A₃₅−A₂₅)/10 for the double transformants, in which A increased with increasing temperature.

The fractional changes of the e/A‐ratio were similar, but not identical to those of A (Fig. 4B). There was a slight decline in the slope of e/A‐ratios in single cppc overexpressors, a more pronounced decline in the fpMe1 overexpressors and a clear decrease in three of the four double transformants (Fig. 4B). In the line 3LD9/ME2, there was even a negative slope suggesting a decrease rather than an increase in the e/A‐ratio as the temperature is raised. In 2% O₂, slopes of e/A‐ratios were negative and not appreciably affected in transgenic plants compared to the wild type. For the cppc/fpMe1 double transformants, temperature effects on A and e/A ratios were consistent with the observed O₂ effects on both parameters shown in Table 3A.

Oxygen‐dependent inhibition of CO₂ assimilation and enhancement of electron requirements for CO₂ assimilation are altered in transgenic tobacco plants with an individual or a combined overexpression of PEPC, NADP‐ME, and PEPCK

The impact of O₂ on photosynthesis was measured in transgenic tobacco plants overexpressing cppc individually or in combination with fpMe1 or pck (Table 3B). In contrast to potato, an appreciable effect on the attenuation of O₂ inhibition of A or the enhancement of e/A‐ratios was observed for PEPC overexpressors. There was apparently no effect on both parameters in plants overexpressing pck alone and a trend of a decline when plants overexpress cppc and pck in combination. Beside cppc overexpressors, double transformants overexpressing cppc and fpMe1 showed a decline in the O₂ inhibition of A at 25 °C, but not at 35 °C. However, this was less marked than in individual cppc overexpressors. Moreover, for the double PEPC/NADP‐ME transformants (cppc/fpMe1) there was a pronounced discrepancy between the attenuation of O₂ inhibition of A (28.0±9.5%) and the O₂ enhancement of e/A‐ratios at 25 °C (38.2±7.1%).

Temperature dependencies of A and e/A‐ratios in transgenic tobacco plants

The slopes (A₃₅−A₂₅)/10 measured in 21% O₂ (Fig. 5A) were negative for wild‐type and transgenic tobacco plants overexpressing cppc alone or in combination with fpMe1. However, an overexpression of pck individually or in combination with cppc resulted in a lack of temperature response of A, i.e. the slopes (A₃₅−A₂₅)/10 were close to zero (Fig. 5A). Under non‐photorespiratory conditions (2% O₂), the increase in A with increasing temperature was similar for wild‐type plants and the individual transgenic lines. A trend of a decline in the e/A‐ratio was observed only in transgenic tobacco plants overexpressing cppc and pck in combination, but not in the respective single transformants (Fig. 5B). In 2% O₂, there was also a decline in the slope of the e/A‐ ratio in transgenic tobacco lines overexpressing pck, cppc/pck and cppc/ME. In conclusion, temperature effects on A and e/A‐ratios (Fig. 5) were not consistent with the observed O₂ effects on both parameters for the individual transformant lines (Table 3B).

Effect of elevated PEPC activity on the content of UV protectants (flavonoids) in potato leaves

In a recent report, Pinto et al. observed a direct linear correlation between the activity of cytosolic NADP‐ME and flavonoid contents in bean leaves after treatment of N‐deficient plant with UV‐radiation (Pinto et al., 1999). The question was asked whether an increase in cytosolic NADP‐ME in PEPC overexpressors of potato causes a similar increase in UV‐protectants. Surprisingly, the combined increase in the activity of PEPC and endogenous cytosolic NADP‐ME in cppc or stppc overexpressors diminished rather than increased flavonoid contents in transgenic potato. As is shown in Fig. 6, flavonoid contents declined by 15% (cppc) to 44% (stppc) in single PEPC overexpressors, but remained unchanged compared to the wild type in fpMe1 overexpressors or in double transformants overexpressing cppc in combination with fpMe1. These data suggest that initial steps of the shikimate pathway in the chloroplasts, which provide the phenolic ring for flavonoid biosynthesis, are impaired when PEP consumption is increased in the cytosol by overexpressing either cppc or stppc.

Fig. 6.

Relative contents of UV protectants (flavonoids) isolated from leaves of control potato and transgenic lines overexpressing bacterial PEPC (cppc, lines 3LD13, 15 and 9), modified potato PEPC (stppc, lines SPsStSD21 and 72), chloroplast NADP‐ME (fpMe1, lines ME18 and 19) as well a combination of bacterial PEPC and NADP‐ME (cppc/fpMe1, lines 3LD9/ME1, 8 and 5). The data represent the mean±SD where n represents the number of samples taken from different plants of the individual lines (n=5 each). The data sets were tested for significant differences between transformants and the corresponding wild‐type plants using the Welch method (see Materials and methods). The calculated P‐values are shown in parenthesis with P<0.001 (1), P<0.01 (2), P<0.02 (3), P<0.05 (4), and P<0.1 (5).

Discussion

An attempt was made to lower the rate of photorespiration and to increase thereby the efficiency of CO₂ assimilation in C₃ plants by introducing a combination of genes involved in the CO₂ concentrating mechanism of C₄ plants into the two closely related solanaceous species, potato and tobacco. An increase in the activity of individual C₄‐cycle enzymes might cause perturbations of metabolism in C₃ plants, but it can also result in substantial effects on certain aspects of photosynthetic performance (Häusler et al., 1999; Ku et al., 1999; Lipka et al., 1999). The question arises whether a fully operational C₄‐cycle, as it is realized in the submerged aquatic plant H. verticillata (see Introduction) would circumvent metabolic perturbations and increases the efficiency of CO₂ assimilation in terrestrial C₃ plants.

Compensational changes in endogenous enzyme activities in transgenic potato and tobacco plants

Enzymes involved in the metabolism of PEP or in the products of PEP carboxylation were altered in transgenic potato and tobacco plants. The most striking effect was observed for cytosolic NADP‐ME, which increased up to 4‐fold in transgenic potato plants as a response to increased PEPC activities. Staining for NADP‐ME activity on native gels suggested a much higher increase in some experiments (Fig. 1). Introduction of the bacterial PEPC (cppc) as well as of a modified version of the endogenous gene (stppc) yielding a protein, which is less sensitive to malate inhibition (see Materials and methods), induced cytosolic NADP‐ME at a transcriptional level. Even wild‐type potato and tobacco plants exhibit relatively high activities of cytosolic NADP‐ME compared to other C₃ plants as, for instance, F. pringlei (Table 1C). This is consistent with the observation that leaves of solanaceous species contain substantial activities of this enzyme (Knee et al., 1996). In leaves of C₃ plants the role of cytosolic NADP‐ME is unclear. There are indications that the cytosolic enzyme is associated with vascular bundles, particularly with developing xylem and internal phloem (Schaaf et al., 1995) suggesting that NADPH produced by this reaction is required for lignin biosynthesis. Moreover, antisense repression of cytosolic NADP‐ME in tobacco resulted in a decline in lignin deposition (Schuch et al., 1990). The highest expression was found in stems and roots, and the lowest in the leaves. There are also a number of indications for the involvement of NADP‐ME in stress responses (Casati et al., 1999), such as in bean plants, where the expression of NADP‐ME increased as a result of N‐limitation (for potato, Fig. 2; right panel) and correlates with the abundance of UV‐absorbing compounds (see below). This points to an involvement of NADP‐ME in the biosynthesis (probably via NADPH proliferation) of these compounds (Pinto et al., 1999).

For the transgenic potato plants, the observed increase in the activities of endogenous cytosolic NADP‐ME is most likely the consequence of metabolic perturbation caused by the introduction of PEPC. The increase in cytosolic NADP‐ME may counteract interfering effects on cell pH as a consequence of excessive accumulation of malic acid in the cytosol. In a broad range of transgenic potato plants overexpressing either cppc or stppc, the increase in endogenous cytosolic NADP‐ME activity appeared to be directly correlated with the increase in PEPC activities. Interestingly, the induction of NADP‐ME was attenuated in transgenic potato plants overexpressing both cppc and fpMe1 despite the identical range of PEPC activities in single and double transformants (Table 1A). This suggests that pleiotropic changes in enzyme activities caused by the introduction of the first gene (i.e. cppc) are diminished or counteracted when metabolism is redirected by the introduction of the second gene (i.e. fpMe1).

In transgenic tobacco overexpressing cppc, there was only a moderate increase in cytosolic NADP‐ME activity as compared to potato transformants (compare Table 1B with Fig. 1). This is consistent with the notion that PEPC overexpressors of tobacco are capable of accumulating substantial amounts of malate (Hudspeth et al., 1992). In contrast to potato, the increase in cytosolic NADP‐ME activity was even further amplified in double transformants of tobacco with a combined overexpression of cppc and fpMe1. This indicates that both closely related plant species have the capacity to respond differentially to an increase in C₄‐cycle enzymes.

In transgenic potato plants, there was also an appreciable increase in the activity of mitochondrial NAD‐ME, PK and PEPP, and a higher NADP‐ICDH activity in cppc overexpressors compared to the wild type. The increase in these enzymes together with the induction of cytosolic NADP‐ME is consistent with the notion that an overexpression of PEPC in potato plants results in higher glycolytic fluxes and in increased rates of carbon losses from photoassimilates by respiration (Häusler et al., 1999). PK activities were also raised in transgenic tobacco lines overexpressing cppc individually or in combination with fpMe1. Interestingly, in contrast to transgenic potato lines, activities of PEPP were not increased in transgenic tobacco lines. An increase in PEPP is an indicator of phosphate limitation (Duff et al., 1989). It is conceivable that inorganic phosphate pools respond differentially towards the introduction of C₄‐cycle genes.

In double transformants of potato (cppc/fpMe1), there was an increase in NADP and NAD‐GAPDH activities, which was not apparent in the respective double transformants of tobacco. For potato, changes in NAD(P)‐GAPDH activities are difficult to interpret, as single overexpression of cppc or fpMe1 had no effect on the activities of these enzymes. It is conceivable that stromal as well as cytosolic (mitochondrial) redox potentials (i.e. NAD(P)H/NAD(P)‐ratios) are severely impaired when cytosolic PEPC and chloroplastic NADP‐ME work in concert, which might trigger an increase in both enzymes. However, as discussed previously (Lipka et al., 1999), there are several open questions concerning the substrates transported across the chloroplast envelope for the operation of chloroplastic NADP‐ME. It is, for instance, uncertain whether malate or OAA, the principal products of PEP carboxylation, are imported into the chloroplasts or whether pyruvate produced by chloroplastic NADP‐ME is further metabolized inside the chloroplasts or exported.

The introduction of PEPCK ( pck) into tobacco chloroplasts had only minor effects on the enzyme pattern (Table 2B). However, overexpression of PEPCK in combination with PEPC (cppc/pck) resulted in an appreciable decline in the activity of chloroplastic NADP‐MDH. NADP‐MDH is proposed to be a part of the malate valve, which shuttles excessive redox equivalents produced in the stroma into the cytosol via reduction of OAA to malate (Scheibe et al., 1987). At this stage it is not clear, which signal is responsible for the decline in NADP‐MDH in PEPC/PEPCK (cppc/pck) double transformants of tobacco. As for PEPC/NADP‐ME (cppc/fpMe1) double transformants of potato, it is again conceivable that redox potentials (or ATP/ADP‐ratios) are perturbed in the stroma as a consequence of a higher rate of ATP consumption by PEPCK in the chloroplasts, when production of cytosolic OAA by PEPC increases in transgenic tobacco. Furthermore, there is no information available about the fate of PEP generated by PEPCK in the chloroplast; is it further metabolized in the stroma or exported from the chloroplasts? In the latter case, PEP could be recycled as a substrate for PEPC in the cytosol.

Photosynthesis and photorespiration are affected by the introduction of C₄‐cycle genes into potato and tobacco plants

It has been shown previously that overexpression of PEPC (cppc) in transgenic potato plants resulted in a small, but significant decline in Γ*, which was accompanied by higher rates of carbon losses by respiration in the light and dark at the expense of carbohydrates (Gehlen et al., 1996; Häusler et al., 1999). Temperature effects on e/A‐ratios determined with double transformants of potato overexpressing NADP‐ME targeted to the chloroplast in the background of the PEPC overexpressors suggested that photorespiration can be attenuated if CO₂ is released directly in the vicinity of Rubisco (Lipka et al., 1999). Moreover, a substantial decrease in the oxygen inhibition of A in transgenic rice plants with an up to 110‐fold increase in PEPC activities has been proposed (Ku et al., 1999). In order to dissect individual components of photosynthesis and/or photorespiration, which are likely to be affected in transgenic potato and tobacco plants with single and double overexpression of C₄‐cycle genes, several approaches have been combined. These were (1) determinations of Γ* and R_d, (2) evaluations of the effect of increasing temperature on the relative O₂ inhibition of A and the enhancement of e/A‐ratios and (3) assessments of direct temperature responses of A and the e/A‐ratio. These data will be discussed on the basis of the observed changes in endogenous enzymes (measured in the same leaves used for photosynthesis measurements) caused by the introduction of C₄‐cycle genes.

Effects on Γ* and R_d:

The method of Brooks and Farquhar was applied to investigate whether a combined overexpression of cppc and fpMe1 in transgenic potato plants affects Γ* and R_d (Brooks and Farquar, 1985). A/C_i curves measured at increasing non‐saturating PFDs and one saturating PFD yielded single intersections (from which Γ* and R_d could be determined) with wild‐type potato and PEPC overexpressors only (Fig. 4A, B). However, in single (fpMe1) and double (cppc/fpMe1) overexpressors the intersection was equivocal. An accurate determination of Γ* and R_d was therefore not feasible. If it is assumed that overexpressed NADP‐ME produces additional NADPH in the stroma, the observed discrepancy in A/C_i curves between potato wild type (control) or cppc overexpressors and single fpMe1 or double (cppc/fpMe1) overexpressors might be explained on the basis of perturbations of stromal NADP/NADPH ratios combined with light‐dependent changes in both the release of CO₂ from malate and the activation states of enzymes regulated by the thioredoxin system (i.e. NADP‐GAP, NADP‐MDH, etc.). (i) In low light, stromal NADP is more abundant, which would allow an appreciable rate of oxidative decarboxylation of malate by NADP‐ME in fpMe1 and cppc/ fpMe1 overexpressors provided that malate is transported into the chloroplasts (see above). However, as Calvin cycle enzymes are less activated under low light conditions, CO₂ assimilation might not match the rate of malate decarboxylation by NADP‐ME and CO₂ is likely to be released in the stroma without successive refixation by Rubisco. This is reflected in negative values for A in the double transformants at low light (Fig. 3D, E). (ii) In intermediate light, NADP‐ME and Calvin cycle enzymes as well as NADP‐MDH may work in concert and the CO₂ released from malate in the stroma increases the CO₂/O₂ ratio in the vicinity of Rubisco, which would result in an attenuation of photorespiration (the estimated Γ* in intermediate light dropped substantially to 18–20 μl l⁻¹ in the line 3LD9/ME2, Fig. 3E). (iii) At high light, the stromal redox state increases due to high electron transport rates and might limit NADP availability for NADP‐ME. Interestingly, the double transgenic line 3LD9/ME2, which displayed increased NADP‐MDH and NAD‐MDH activities (Table 2A), exhibited the most pronounced effect with respect to an attenuation of photorespiration in intermediate light (Table 3A; Figs 3, 4). Most likely this reflects an additional relief in the limitation of oxidative malate decarboxylation by the capacity of OAA reduction within or outside of the chloroplast catalysed by NADP‐ or NAD‐MDH. Moreover, the activities of NADP‐GAPDH are increased in all double transformants suggesting an up‐regulation of this enzyme only when PEPC and chloroplastic NADP‐ME work in concert.

In contrast to single fpMe1 and double cppc/fpMe1 overexpressors, oxidative decarboxylation of malate by induced endogenous cytosolic NADP‐ME, such that may occur at increased rates in single PEPC overexpressors, would neither perturb stromal redox states nor the shapes of A/C_i‐curves with increasing PFDs.

O₂ inhibition of A and enhancement of e/A‐ratios:

 In a recent report (Ku et al., 1999), it was proposed that an up to 110‐fold increase of PEPC by overexpressing the maize gene in transgenic rice plants had a substantial attenuating effect on the O₂ inhibition of CO₂ assimilation, which would be consistent with a decrease in the rate of photorespiration relative to CO₂ assimilation. However, more recent data revealed that this effect was caused by the phosphate limitation of photosynthesis rather than by a decrease in photorespiration (Fukayama et al., 2000). Lipka et al. observed a decline in the e/A‐ratio in double transformants of potato containing PEPC and chloroplastic NADP‐ME relative to the wild type when the leaf temperature was raised above 30 °C (Lipka et al., 1999), which is strong support for a suppression in the rate of photorespiration.

In this study, both these approaches have been combined by dissecting the O₂ effects on A and on the e/A ratios at different leaf temperatures in a variety of single and double transformants of potato and tobacco (Table 3A, B). As photosynthetic electron transport was determined by an independent method compared to gas exchange measurements, an appreciable correlation between the percentage of O₂‐dependent inhibition of A and the enhancement of e/A‐ratios at 25 °C and 35 °C observed for transgenic potato plants underlines the validity of this approach (Table 3A).

A comparison of the data in Tables 3A and 3B reveals that A and e/A‐ratios were differentially affected by O₂ and temperature in transgenic potato and tobacco plants. Single overexpression of PEPC had no effect on both parameters in transgenic potato transformants (Table 3A), whereas they were diminished in single cppc transformants of tobacco (Table 3B). Differences in the O₂ effects on photosynthesis reflect most likely differential capacities to accumulate malate between both species. A substantial accumulation of malate has been reported for transgenic tobacco plants overexpressing PEPC from maize (Hudspeth et al., 1992), whereas transgenic potato plants overexpressing cppc exhibit only a moderate increase in malate at the end of the light period (Häusler et al., 1999). The lack of malate accumulation in PEPC overexpressors of potato is most likely a consequence of the 4‐fold induction of endogenous cytosolic NADP‐ME combined with the increase in NAD‐ME, PK and NADP‐ICDH. Thus, malate produced following PEP carboxylation is readily decarboxylated and the CO₂ released. In contrast to PEPC overexpressors of potato, the above enzymes were increased only moderately in transgenic tobacco which infers that, in tobacco, higher fluxes are directed into malate formation and subsequent accumulation. The latter scenario would also bring about an attenuation of O₂ inhibition of A combined with a lowered enhancement of e/A‐ratios.

Transgenic potato plants overexpressing chloroplast NADP‐ME (fpMe1) individually as well as in combination with PEPC (cppc) exhibited a decline in O₂ inhibition of A and enhancement of e/A‐ratios only at 35 °C. At 25 °C, there were no appreciable changes in both parameters compared to the wild type except for the line 3LD9/ME5 (Table 3A), which showed high PEPC activity combined with high chloroplastic NADP‐ME activity (Table 1A). In contrast, O₂ inhibition of A and attenuation of e/A‐ratios declined appreciably in the line 3LD9/ME2 at both temperatures despite only a small increase in PEPC activity (Table 1A). In this particular line, NAD‐MDH and NADP‐MDH are increased suggesting that one or both of these enzymes may control the flux into oxidative decarboxylation of malate inside the chloroplast (see above). Moreover, activity of cytosolic NADP‐ME in the line 3LD9/ME2 was similar to the wild type, but was high in the line 3LD9/ME5 suggesting that competitive effects between chloroplastic and cytosolic NADP‐ME for the common substrate should be considered as well. In tobacco plants overexpressing cppc/fpMe1, there was a significant attenuation of O₂ inhibition of A only at 25 °C (Table 3B). However, this was not reflected in an appreciable attenuation of the e/A ratios suggesting that the relief of O₂ inhibition of A was not entirely caused by a decline in photorespiration.

Overexpression of PEPCK alone or in combination with PEPC in transgenic tobacco plants had neither an effect on O₂ inhibition of A nor on O₂ enhancement of e/A‐ratios (Table 3B), whereas a strong attenuating response was observed in single cppc overexpressors of tobacco (see above). It should be noted that PEPC activity in single cppc transformants (286 mU mg⁻¹) was appreciably higher compared to the double cppc/pck transformants (136 mU mg⁻¹), which might, at least in part, explain the observed differences.

Temperature dependencies of A and e/A‐ratios:

 It has been shown previously that changes in A and in e/A‐ratios are directly correlated with the temperature between 25 °C and 35 °C (Lipka et al., 1999). Thus, transgenic potato and tobacco plants can be compared on the basis of the slopes of A and e/A‐ratios measured between 25 °C and 35 °C under photorespiratory (21% O₂) and non‐photorespiratory (2% O₂) conditions. For transgenic potato, the temperature‐dependent decline in A and the enhancement of e/A ratios was diminished most noticeably in PEPC/NADP‐ME (cppc/fpMe1) double transformants under photorespiratory conditions, particularly in the line 3LD9/ME2. This is consistent with the attenuation of O₂ effects on A and e/A‐ratios in the same transgenic lines (see above). In tobacco, O₂ effects on A and e/A ratios were attenuated most in single cppc overexpressors. However, this was neither supported by a diminished temperature response of A nor by the relative decline in the e/A ratio (compare Table 3B and Fig. 5A, B). Temperature effects on A were attenuated most obviously in pck, cppc/pck and cppc/ME overexpressors of tobacco. Again this was not supported by the data on temperature effects on e/A‐ratios (Fig. 5B) or on O₂ effects on A and e/A‐ratios (Table 3B).

Effects on UV protectants

It has recently been shown that the abundance of UV‐protectants directly correlates with the activity of cytosolic NADP‐ME in bean plants exposed to UV‐B radiation (Pinto et al., 1999). It was proposed that a higher requirement of redox equivalents is responsible for this effect. Surprisingly, in leaves of transgenic potato plants overexpressing either cppc or stppc, the contents of UV protectants were diminished despite a significant increase in the activity of cytosolic NADP‐ME (Fig. 6). Moreover, this decline in UV protectants was not apparent in double (ccpc/fpMe1) or single (fpMe1) transformants of potato. In single PEPC overexpressors higher rates of PEP consumption in the cytosol caused by increased PEPC activities might deplete cytosolic PEP pools. PEP is one of the precursors for the shikimate pathway, localized inside the plastids (Schmid and Amrhein, 1995). PEP provision from the cytosol is required because chloroplasts lack the ability to produce PEP from glycolysis as the activities of phosphoglycerate mutase and enolase are either absent or very low (Stitt and ap Rees, 1979). This view was reinforced recently by investigations on mutants of Arabidopsis thaliana lacking the phosphoenolpyruvate translocator (PPT) (Streatfield et al., 1999), which has recently been cloned (Fischer et al., 1997). If it is considered that flavonoids derive from products of the shikimate pathway with cytosolic PEP as one of the precursors, a restriction in PEP supply to the chloroplasts by high activities of PEPC in the cytosol, might cause a decline in these components and shows that overexpression of C₄‐cycle enzymes not only affects primary metabolism and photosynthesis, but also affects secondary metabolism. However, at this stage it is not clear why a combined overexpression of PEPC and chloroplastic NADP‐ME replenishes the contents of UV‐protectants.

Conclusions

The data presented in this paper demonstrate (i) that potato and tobacco plants are responding differently towards single or combined overexpression of C₄‐cycle genes with respect to compensational changes in enzyme pattern and photosynthetic properties. (ii) A combined overexpression of PEPC and chloroplastic NADP‐ME diminished pleiotropic effects caused by single overexpression of PEPC. In particular, the attenuated induction of cytosolic NADP‐ME in double cppc/fpMe1 overexpressors suggests that the introduction of a fully operational, self‐sufficient C₄‐like cycle in C₃ plants (including PEPC, chloroplastic NADP‐ME combined with PPDK or PEPCK as well as the PPT) might minimize pleiotropic effects as observed after the introduction of the individual transgenes. The ‘futile cycle’ established thereby would result in an efficient CO₂ pump at the expense of ATP and would be decoupled from the residual metabolism, which is not the case when single C₄‐cycle genes are overexpressed. (iii) In order to achieve an operational C₄‐cycle, all of its essential components have to be overexpressed in a balanced fashion including chloroplast metabolite transporters.

Acknowledgement

This work was supported in part by a grant from the Deutsche Forschungsgemeinschaft (DFG).