•Understanding of the influences of root-zone CO2 concentration on nitrogen (N) metabolism is limited.
•The influences of root-zone CO2 concentration on growth, N uptake, N metabolism and the partitioning of root assimilated 14C were determined in tomato (Lycopersicon esculentum).
•Root, but not leaf, nitrate reductase activity was increased in plants supplied with increased root-zone CO2. Root phosphoenolpyruvate carboxylase activity was lower with NO3−- than with NH4+-nutrition, and in the latter, was also suppressed by increased root-zone CO2. Increased growth rate in NO3−-fed plants with elevated root-zone CO2 concentrations was associated with transfer of root-derived organic acids to the shoot and conversion to carbohydrates. With NH4+-fed plants, growth and total N were not altered by elevated root-zone CO2 concentrations, although 14C partitioning to amino acid synthesis was increased.
•Effects of root-zone CO2 concentration on N uptake and metabolism over longer periods (> 1 d) were probably limited by feedback inhibition. Root-derived organic acids contributed to the carbon budget of the leaves through decarboxylation of the organic acids and photosynthetic refixation of released CO2.
The anaplerotic role of phosphoenolpyruvate carboxylase (PEPc) in root tissue is widely recognized. The dependence of this enzyme on the concentration of dissolved inorganic carbon (DIC) in the root tissue is less often considered, since it is assumed that this occurs at high concentrations as a result of respiratory activity. Unlike aerated hydroponic solutions, soils have high CO2 concentrations (c. 2000 to 200 000 p.p.m., Norstadt & Porter, 1984) mainly because of limited diffusion and the biological components of the soil. Although the overwhelming flux of CO2 is out of the root, the root tissue DIC concentrations are determined by both the rate of root respiration and the gradient of CO2 between the root tissue and the external root environment. Root-zone DIC concentration influences photosynthesis (Cramer & Richards, 1999), respiration (van der Westhuizen & Cramer, 1998), NO3− uptake (Cramer et al., 1996), partitioning of C and N to organic and amino acid synthesis (Cramer & Lewis, 1993) and growth (Cramer & Richards, 1999). Positive effects of root-zone DIC on plant growth have been reported previously (Vapaavuori & Pelkonen, 1985), particularly in plants growing under high irradiances, salinity stress or high shoot temperatures (Cramer & Richards, 1999).
In short-term (6–12 h) uptake experiments, increased root-zone DIC promoted NO3− uptake compared with ambient root-zone DIC, whereas NH4+ uptake was decreased or unchanged (Cramer et al., 1996). Although increased NO3− uptake could be caused by increased availability of organic acid for root assimilation of NO3−, NH4+ assimilation also requires C skeletons from the TCA cycle for amino acid synthesis (Schweizer & Erismann, 1985). This and the fact that NO3− uptake by nitrate reductase (NR)-deficient mutants of barley was increased by elevated root-zone DIC indicates that there may be a direct stimulatory effect of DIC on NO3− uptake (Cramer et al., 1996). However, the question arises as to whether the influence of DIC on N uptake is sustained for longer periods of time (days).
The biological demand for CO2 or HCO3− in nongreen tissues frequently exceeds the uncatalysed equilibrium between CO2 and HCO3− (Raven & Newman, 1994). Carbonic anhydrase (CA) catalyses the reversible hydration of CO2 (Rengel, 1995). Carbonic anhydrase activity has found in both the stroma of chloroplasts (87% of total cellular activity) and the cytosol (13%) of Solanum tuberosum leaves (Rumeau et al., 1996). It is also known to occur in root nodules at high activities where it is responsible for equilibrating CO2 and HCO3−, possibly supplying PEPc with HCO3− for anaplerotic replenishment of TCA cycle intermediates or for facilitating release of CO2 from the nodules which have low gas permeability (Atkins et al., 2001). Non-nodular root tissue CA activities are lower than those of nodules. However, CA activity in Zea mays root tips was found to be 200 times higher than that of PEPc in vivo (Chang & Roberts, 1992), indicating that its activity was high enough to play a role. However, the possible variation in CA activity in roots exposed to varying CO2 concentrations is unknown.
The disparity between in vitro and in situ root PEPc activities may result from regulation in vivo (Cramer et al., 1999). Regulation is through protein phosphorylation which results in a decrease in the sensitivity of PEPc to allosteric inhibitors such as malate (Jiao & Chollet, 1991) and an increase in catalytic activity. Koga & Ikeda (1997) found that root PEPc activity was 2- to 2.5-fold higher with NH4+ nutrition than with NO3− nutrition and reached higher values in NH4+-fed wheat, barley and tomato plants compared with NO3−-fed plants. Concentration of PEPc protein, measured by Western blot analysis, was greater in the roots of plants supplied with NH4+ (Koga & Ikeda, 1997) and methionine sulfoximine suppressed the increased PEPc activity elicited by NH4+ nutrition (Koga & Ikeda, 2000). It was concluded that NH4+ assimilation is required for induction of de novo synthesis of PEPc by NH4+ nutrition. Since root-zone DIC concentrations influence NO3− and NH4+ uptake, it is possible that it would influence the production of PEPc protein or the regulation of PEPc activity.
Elevated root-zone DIC was found to stimulate NR activity in vitro and in situ in barley plants (Cramer et al., 1996). Nitrate reductase is affected by several factors including NO3− availability, pH (Kaiser & Brendle-Behnisch, 1995), light/dark, inhibitor proteins (Glaab & Kaiser, 1995) and photosynthesis (Kaiser & Brendle-Behnisch, 1991). It is active in the dephosphorylated form and partly inactivated in the phosphorylated form, although, complete inactivation requires binding of an inhibitor protein (Glaab & Kaiser, 1995). It is activated by cytosolic acidification through dephosphorylation and is inactivated by phosphorylation in response to cytosolic alkalization (Kaiser & Brendle-Behnisch, 1995). Since dissolved CO2 is a weak acid, exposure to CO2 may change the cytosolic pH, with consequences for the regulation of plant metabolism (Bown, 1985).
Although the influence of soil CO2 concentrations on soil chemistry is well known, the influence of high concentrations of soil CO2 on plant metabolism is not widely recognized. There are several reports of the influence of root-zone CO2 concentration on N acquisition/assimilation and root C fixation (Cramer, 2002). These reports are largely based on short-term experiments (1–8 h). In this investigation we extended this previous work to longer periods of time (up to 15 d). The influence of root-zone CO2 concentration on the allocation and partitioning of root-assimilated C were explored by means of a root-supplied 14CO2 pulse-chase experiment in which the changes were followed over a 24-h chase-period. The activities of key enzymes in N assimilation (NR, PEPc and CA) were also measured to determine whether the activities and regulation of these could account for changes in N and C acquisition and assimilation.
Materials and Methods
Seedlings (14 d old) of Lycopersicon esculentum (L.) cv. F144 grown on a 1 : 1 mixture of vermiculite and compost were transferred to hydroponic culture after rinsing the roots in water. The hypocotyls of the plants were wrapped in black, closed-cell foam rubber and inserted through collars in the lids of 22-l hydroponic tanks with eight plants per tank. The tanks were completely opaque and contained 20 l Long Ashton nutrient medium (Hewitt, 1966) modified to contain 2 mm of either NaNO3 or NH4Cl as a N source and 0.09 mm Fe-ethylene diamine tetra-acetic acid (Fe-EDTA) as an iron source. The nutrient medium was changed weekly and the pH of the medium was maintained at 5.8 by adjusting the pH with HCl or NaOH daily. Plants were grown in a temperature controlled (minimum 15°C, maximum 25°C) glasshouse at the University of Stellenbosch (South Africa) during spring (September and October). Nutrient solutions were strongly aerated with ambient air (380 p.p.m. CO2) or with air containing elevated root-zone CO2 (5000 p.p.m. CO2) produced by enriching ambient air with CO2 from a cylinder of industrial grade CO2 (Afrox, Cape Town, South Africa). Gas chromatograph analysis was performed on an undiluted sample of CO2 gas (Mosgas, Mossel Bay, South Africa) and no ethylene was detected. Plants grown for the in vitro NR, PEPc and CA assays and for the 14C-feeding experiments were aerated with 0 p.p.m. instead of 380 p.p.m. root-zone CO2 to accentuate the differences between low and elevated root-zone CO2 treatments. Carbon dioxide was removed from the air by passing ambient air through 2 m NaOH and a column (4 cm diameter and 30 cm long) containing 4–8 mesh soda lime (Saarchem, Krugersdorp, South Africa). The CO2 concentration was monitored continuously using an ADC Mk3 (Analytical Development Corporation, Hoddeston, UK) infrared gas analyser (IRGA). To prevent diffusion of CO2 from the root-zone and the consequent enrichment of atmosphere around the shoots, the lids of hydroponic tanks were sealed with closed-cell foam rubber around the rim and clamped onto the tanks. The air-space between the surface of the nutrient solution and the lid was maintained under a partial vacuum to ensure that net air flow was inwards. Plants were used for experiments when the biomass was c. 6 g.
Relative growth rates and nitrogen use
The plants were grown in 1 l bottles in a controlled environment chamber (Controlled Environments Ltd, Winnipeg, Canada) with an irradiance of 450 µmol m−2 s−1, 14-h photoperiod, relative humidity of 60% and a light/dark temperature of 25°C/18°C. The nutrient solutions were aerated either with ambient air (380 p.p.m. CO2) or air containing elevated root-zone CO2 (5000 p.p.m) and the pH was maintained at 5.8 by adjusting the pH with HCl or NaOH daily. There were four treatments comprising of either 2 mm NaNO3 or NH4Cl combined with either 380 p.p.m. or 5000 p.p.m. CO2. The air-space between the surface of the nutrient solution and the lid was maintained under partial vacuum.
The fresh weights of the seedlings were determined regularly over the course of 15 d by carefully blotting the roots of the seedlings and weighing. In a preliminary trial it was shown that this procedure did not significantly reduce the biomass accumulation of the plants. The relative growth rates (RGRs) were calculated from linear regression of the logarithms of the fresh weights vs time. Fresh nutrient solution was supplied after each weighing and samples of the nutrient solution retained for analysis of the N content. The NO3− and the NH4+ concentrations of the samples were determined according to the salicylic acid method of Cataldo et al. (1975) and phenol-hypochlorite method of Solorzano (1969), respectively. After 15 d the plants were harvested and fresh weights of the shoots and roots determined after which the plants were dried in an oven at 80°C for 48 h and reweighed.
For total N determination the oven-dried plant components were milled in a Wiley mill using a 0.5 mm mesh (Arthur H. Thomas, Philadelphia, PA, USA). The digestion was carried out with 0.05 g of milled plant material in a digestion block (Gerhardt, Königswinter, Germany) with 3 ml 3.4% (w : v) salicylic acid in concentrated sulphuric acid, 1 ml of distilled H2O and a selenium pellet (Saarchem). The samples, including titriplex V standards, were digested at room temperature for 2 h, at 200°C for 1 h, 270°C for 1 h and at 370°C until they were clear. The concentration of NH4+ was determined on the diluted digest according to the method of Solorzano (1969).
Nitrate uptake was measured using eight replicate plants per treatment. The plants were transferred to bottles containing 300 ml Long Ashton nutrient solution (pH 5.8) with 0.2 mm NaNO3 and were preincubated for 12 h at an irradiance of c. 1200 µmol m−2 s−1. The hypocotyls of the plants were wrapped in closed-cell foam rubber and inserted through the lids of the bottles. The roots were aerated with either 380 p.p.m. or 5000 p.p.m. root-zone CO2. The roots receiving 5000 p.p.m. root-zone CO2 had columns containing soda lime attached to the lids of the bottles to trap the CO2 released from the solution and thus prevent enrichment of the atmosphere surrounding the shoots. The bottles were placed in a water-bath and the temperature of both the water and the surrounding air maintained at 20°C.
After preincubation three plants of each treatment were harvested and divided into roots, stems and leaves, weighed, quenched in liquid N2 and stored at −80°C. These samples were later used to determine the initial tissue NO3− concentrations. The remaining plants were supplied with 300 ml fresh Long Ashton nutrient solution (pH 5.8) with 1 mm NaNO3 and incubated for a further 6 h. Subsamples of 1 ml were taken at the start of the experiment, and after 1, 3 and 6 h for determination of NO3− uptake rates measured by NO3− depletion. Thereafter the plants were harvested and divided into roots, stems and leaves, weighed, quenched in liquid N2 and stored at −80°C.
The in situ nitrate reductase activity (NRA) was calculated from the difference between NO3− taken up from the nutrient medium and the concentrations of NO3− in the plant tissue after the uptake period, less the initial concentrations in the plant tissue. The NO3− concentration in the nutrient solutions and the plant tissue were measured using the method of Cataldo et al. (1975). Tissue NO3− was extracted by vacuum infiltrating a homogenous sample of tissue (c. 0.3 g) in 10 ml distilled water and extracting in a waterbath at 80°C for 2 h. Each extract was mixed and subsamples of 1 ml centrifuged at 1300 g for 5 min, after which the NO3− concentration was determined.
Nitrate reductase activity (in vitro)
Eight replicate plants from each treatment were harvested, divided into root and leaf material, quenched in liquid N2 and stored at −80°C until assaying for in vitro NRA. The enzyme was assayed according to a modification of the method of Kaiser and Huber (1997). The frozen tissue was homogenized with acid-washed sand in a precooled mortar and pestle in 4 ml g−1 f. wt extraction buffer containing 100 mmN-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES)-KOH (pH 7.6), 3 mm dithiothreitol (DTT), 10 µm flavin adenine dinucleotide (FAD) (Sigma Chemical Co., St Louis, MO, USA), 2 mm EDTA, 10% (v/v) glycerol, 2% (w/v) casein, 2.5% (w/v) polyvinylpolypyrrolidone (PVPP) and 1 µm sodium molybdate. The homogenate was centrifuged at 16 000 g and 4°C for 5 min and then 100 µl of the supernatant incubated with 5 µl of either (1) 200 mm MgCl2, (2) 300 mm EDTA or (3) a mixture of 100 mm AMP, 200 mm KH2PO4 and 300 mm EDTA for 10 min at 30°C. Assaying with MgCl2 provided an estimate of NR activity in vivo while assaying with EDTA allowed estimation of the activity of the phosphorylated NR without the inhibitor protein (IP) while the maximum activity (equivalent to total NR protein) of the enzyme was assayed with the AMP, KH2PO4 and EDTA mixture. The incubation period was reduced from the 30 min used by Kaiser & Huber (1997) to 10 min as a result of degradation of NR activity over the longer time period (data not shown). Reaction medium (900 µl) consisting of 100 mm HEPES-KOH (pH 7.6), 1 mm DTT, 10 µm FAD, 20 mm KNO3, and 0.2 mm NADH was added to the supernatant and incubated for 5 min at 30°C in a waterbath after which the reaction was stopped by addition of 125 µl of 0.5 m zinc acetate. The samples were centrifuged for 1 min in a microfuge and 1 ml of a 1 : 1 mixture of 1% (w : v) sulphanilamide in 1.5 m HCl and 0.01% (w : v) N-(1-napthyl) ethylenediamine hydrochloride (NED) added to 300 µl of reaction medium and the absorbance determined at 540 nm after 15 min. The NRA was calculated from the amount of NO2− formed.
Phosphoenolpyruvate carboxylase activity (in vitro)
Eight replicate plants from each treatment were harvested, divided into roots and leaves, quenched in liquid N2 and stored at −80°C until assayed for PEPc activity according to a modification of the method of Coombs (1987). The frozen tissue was homogenized with acid-washed sand in a precooled mortar and pestle in 4 ml g−1 f. wt extraction buffer containing 100 mm Tris-HCl (pH 8.0), 10 mm MgCl2, 5 mm DTT, 20% (v : v) glycerol, 5 mm NaF, 20 µm leupeptin (Sigma), 2% (w : v) casein and 2% (w : v) PVPP. The homogenate was centrifuged at 25 000 g for 15 min at 4°C and 75 µl of the supernatant added to 500 µl of reaction medium consisting of 50 mm HEPES-KOH (pH 8.0), 5 mm MgCl2, 5 mm NaF, and 3 mm phosphoenolpyruvate (PEP) (Sigma). Malate sensitivity, which gives an indication of phosphorylation status of the enzyme, was determined by the addition of malate to a final concentration of 0.8 mm to the reaction medium (Foyer et al., 1998). The reaction was initiated by addition of 50 µl of 11.7 mm NaH14CO3 (specific activity 0.98 µCi µmol−1) and incubated at 30°C. From this reaction mixture 50 µl samples were added to 250 µl of 20% (v : v) of saturated dinitrophenylhydrazine in 2 m HCl. The samples were left to stand overnight in a fumehood after which 2 m NaOH was added to neutralize the pH of the samples. The samples were counted on a LS 1801 liquid scintillation counter (Beckman Instruments Inc., Fullerton, CA, USA) with 2 ml of Readygel (Beckman).
Carbonic anhydrase activity (in vitro)
Eight replicate plants from each treatment were harvested, divided into roots and leaves, quenched in liquid N2 and stored at −80°C until assayed for CA activity according to a modification of the method of Makino et al., 1992). The frozen tissue was homogenized with acid-washed sand in a precooled mortar and pestle in 4 ml g−1 f. wt extraction buffer containing 50 mm Tris-HCl (pH 7.5), 10 mm DTT, 0.5 mm EDTA, 10% (v : v) glycerol, 0.1% (v : v) Triton-X100 (Sigma), 2% (w : v) casein and 2% (w : v) PVPP. The homogenate was centrifuged at 15 000 g for 5 min at 4°C and the supernatant kept on ice until addition to the reaction cuvette. Reaction buffer (4 ml) consisting of 20 mm Tris (pH 8.3) was added to the temperature controlled (3 ± 0.5°C) reaction cuvette together with 500 µl of supernatant. After stabilization of the pH, 2 ml of the reaction buffer saturated with CO2 was added to the reaction cuvette and the rate of pH change was measured between pH 8.5 and 7.9 using a Corning 445 pH meter (Corning, New York, NY, USA). The observed rate of change of pH was converted to equivalent µmol H+ generated by comparison with a calibration established by titrating the reaction buffer between pH 8.5 and 7.9 with HCl.
The seedlings were pretreated overnight in 300 ml containers in a Long Ashton nutrient solution with 2 mm of either NaNO3 or NH4Cl and aerated with 0 or 5000 p.p.m. CO2, the latter corresponding to 0.21 mm total dissolved inorganic C at equilibrium (25°C and pH 5.8). The seedlings were then transferred to fresh 300 ml Long Ashton solution containing 2 mm of either NaNO3 or NH4Cl. NaHCO3 containing 0.093 MBq NaH14CO3 was then added to the nutrient solutions, a 1 ml subsample taken from each container and the aeration discontinued. Solutions were aerated for 15 s every 15 min over a period of 1 h to prevent anaerobic conditions developing and to agitate the solutions. Over the 1 h labelling period the release of respiratory CO2 was estimated to result in an increase in the CO2 concentration in the solutions of c. 0.08 mm (based on a nominal respiration rate of 100 nmol g−1 root d. wt s−1; van der Westhuizen & Cramer, 1998). The plants were removed from the nutrient solutions, the roots rinsed in two volumes of 2 mm CaSO4 and transferred to fresh nutrient solutions containing 2 mm of either NaNO3 or NH4Cl. The roots of three plants from each treatment were blotted dry and the plants were divided into leaf, stem and root components, weighed, quenched in liquid N2 and stored at −18°C. This procedure was repeated after 6, 12 and 24 h. Subsamples of the nutrient solutions were taken at each harvest interval and stored at −18°C.
The plant components were homogenized in liquid N and the suspended in 50 ml of cold 80% (v : v) ethanol and stored at −18°C for 48 h prior to extraction for 60 min at 45°C. The samples were filtered through Whatman No. 1 filter paper (Whatman International Ltd, Maidstone, UK) and the filtrate was made up to volume (75 ml). A subsample of 500 µl from each soluble fraction was acidified with 50 µl of 0.3 m HCl and stood in a fume hood for 12 h on a shaker. Then 2 ml of Readygel (Beckman) was added, the samples mixed and counted on a LS 1801 scintillation counter (Beckman). The residue was dried in an oven for 48 h at 80°C and the weights of the insoluble fraction determined. The biscuits of the insoluble fraction were finely ground in a mortar and pestle and a subsample of 100 mg weighed out and acidified with 500 µl of 0.3 m HCl, stood on a shaker in a fume hood for 12 h. Thereafter 2 ml of Soluene-350 (Packard Instrument BV Chemical Operations, Groningen, The Netherlands) was added and samples shaken for a further 12 h. Ethanol (4 ml, 96% v : v) was added and the samples were mixed. A subsample of 100 µl was taken and counted with 2 ml of Readygel (Beckman) on a LS 1801 liquid scintillation counter (Beckman). Samples of 75 ml of the soluble fractions were evaporated and made up to 20 ml. Subsamples of 1 ml were loaded onto 1 ml Dowex 50W-X8 and Dowex 1W-X8 (Sigma) columns in 1 ml disposable syringes and further eluted, collected and counted in the same way as the nutrient solutions.
Subsamples of 20 ml of the nutrient solutions were evaporated and made up to 5 ml. The samples were separated into basic, acidic and neutral fractions using ion exchange resins prepared according to Atkins & Canvin (1971). Samples were loaded onto 1 ml Dowex 50 W-X8 and Dowex 1 W-X8 (Sigma) columns in 1 ml disposable syringes and eluted with 25 ml 50% (v : v) ethanol. The eluate was collected as the neutral fraction. The basic fraction was eluted from the Dowex 50 W-X8 column with 10 ml 2 m HCl and the acidic fraction was eluted from the Dowex 1 W-X8 column with 10 ml of 6 m formic acid. Subsamples of 1 ml were taken of each fraction and counted with 2 ml Readygel (Beckman) on a scintillation counter. The basic fraction consists mainly of amino acids, the acidic fraction mainly of organic acids and monophosphate esters and the neutral fraction mainly of sugars (Atkins & Canvin, 1971). This was verified by recovery experiments with 14C-labelled succinate, glucose and leucine.
Results were subjected to either analysis of variance or to Student's t-tests to determine the significance of differences between the responses to the treatments. Where percentage data were used these were arcsine transformed (Zar, 1984) prior to statistical analysis. Where analysis of variance was performed, post hoc Fisher's protected least significant difference (LSD) tests (95%) were conducted to determine the differences between the individual treatments using statgraphics version 7.0 (Statistical Graphics Corporation, 1993).
Relative growth rates and nitrogen uptake
The RGR values, measured as the increase in fresh weight of intact plants, increased over time and then declined as the plants became larger (Fig. 1). The final weights of the plants supplied with NO3− nutrition were (mean ± SE) 6.7 ± 0.3 g and 10.1 ± 0.5 g for the plants supplied with 380 p.p.m. and 5000 p.p.m. root-zone CO2, respectively. The corresponding final weights of the plants supplied with NH4+ nutrition were 3.7 ± 0.2 g and 4.4 ± 0.1 g for the plants supplied with 380 p.p.m. and 5000 p.p.m. root-zone CO2, respectively. The RGRs of NO3− fed plants grown with both 380 p.p.m and 5000 p.p.m. root-zone CO2 were initially similar, but after 8 d the plants grown with 5000 p.p.m. had higher RGR values (P < 0.05, anova followed by LSD multiple ranges tests) than those of plants grown with 380 p.p.m. root-zone CO2 (Fig. 1). The RGRs of NH4+-fed plants grown with 5000 p.p.m. root-zone CO2 were higher than those of NH4 ± fed plants grown with 380 p.p.m. root-zone CO2 for the first 5 d of the experiment (P < 0.05, anova followed by LSD multiple ranges tests), after which no differences between RGRs could be discerned. The mean RGR over 15 d was c. 1.1-fold higher for NO3−-fed plants grown with 5000 p.p.m. root-zone CO2 compared with those grown with 380 p.p.m. root-zone CO2 (Fig. 1). In plants supplied with NH4+ nutrition the root-zone CO2 concentration did not influence the mean RGR.
The NH4+ uptake rate was higher than that of NO3−, but the uptake rates were not significantly altered in these long-term experiments by exposure to altered rhizosphere CO2 concentrations (Fig. 2). No significant difference was found between the NO3− uptake rates (expressed per gram root dry weight over the last day of the long-term growth experiment) of plants grown at 380 p.p.m. root-zone CO2 compared with 5000 p.p.m. root-zone CO2 (Fig. 2). However, there was a 1.4-fold increase in the concentration of total N of the shoots of NO3−-fed plants supplied with 5000 p.p.m. compared with 380 p.p.m. root-zone CO2, but no effect on the root total N (Fig. 3). There were also no significant differences in the total N concentration of shoots or roots in NH4+-fed plants as a result of variation in root-zone CO2 concentration. The total N shoot : root ratios for NO3−-fed plants were higher at elevated root-zone CO2 (Fig. 3), but root-zone CO2 did not significantly change the distribution of total N within NH4+-fed plants. Thus, elevated root-zone CO2 concentration could favour the translocation of NO3−, or its reduction products, from the root to the shoot.
The in situ NRA (µmol g−1 f. wt h−1), calculated from the uptake rates and the change in the concentration of NO3− in the tissue during the uptake period, was 11.9 ± 1.1 for plants grown with 380 p.p.m. root-zone CO2 compared with 17.0 ± 0.6 for plants grown with 5000 p.p.m. root-zone CO2 (Students’t-test P = 0.004). In vitro root NRA determined after preincubation with MgCl2, EDTA or AMP was also higher for plants grown with 5000 p.p.m. root-zone CO2 compared with 0 p.p.m. root-zone CO2 (Table 1). By contrast, the equivalent in vitro leaf NRA was significantly lower for plants grown with 5000 p.p.m. root-zone CO2 compared with 0 p.p.m. root-zone CO2. In the roots, NRA assayed after preincubation with MgCl2 had the lowest activity, while assaying after preincubation with AMP had the highest activity for both root-zone CO2 concentrations. Assuming that the assay with preincubation in AMP, KH2PO4 and EDTA represents the maximum NRA, and that preincubation in EDTA alone represents the NRA with the inhibitory effect of phosphorylation, the extent of inhibition by phosphorylation was estimated to be c. 2% in leaves and 19% in roots. Incubation with only MgCl2 represents the phosphorylated, inhibitor protein-bound NRA as an estimate of in situ activity. This activity was c. 30% and 35% of the maximum activity in leaves and roots, respectively. The root-zone CO2 concentration had no influence on the proportion of NR that was either phosphorylated or phosphorylated with the inhibitor protein bound.
Table 1. Nitrate reductase activity (NRA) for tomato (Lycopersicon esculentum) plants grown with 2 mm NO3− combined with 0 or 5000 p.p.m. root-zone CO2, phosphoenolpyruvate carboxylase (PEPc) and carbonic anhydrase activity (CA) of plants grown on either 2 mm NO3− or NH4+ combined with 0 or 5000 p.p.m. root-zone CO2
Activities Leaf 0 p.p.m. CO2
5000 p.p.m. CO2
Root 0 p.p.m. CO2
5000 p.p.m. CO2
In vitro NRA is shown with the different activation states of NR. Assaying NRA with MgCl2 provided an estimate of in vivo NRA; assaying with ethylenediaminetetraacetic acid (EDTA) provided an estimate of phosphorylated NRA without the inhibitor protein (IP) bound and ‘maximum’ activity of the enzyme was assayed with AMP, KH2PO4 and EDTA. Values in parentheses after NRA values indicate percentage activity of ‘maximum’ activity. Values in parentheses after PEPc values indicate the percentage PEPc activity remaining when 0.8 mm malate was included in assay medium. Means ± SE are shown (n = 8). Different letters after values indicate significant differences between treatments tested using analysis of variance (anova) with post hoc LSD tests. Different tissues were tested separately.
NRA (µmol g−1 f. wt h−1)
13.1 ± 0.6ab (69)
10.8 ± 0.4a (70)
6.2 ± 0.4a (63)
8.9 ± 0.7b (67)
18.7 ± 1.0c (97)
15.4 ± 1.0b (98)
7.8 ± 0.4ab (81)
10.8 ± 0.8d (82)
AMP + KH2PO4 + EDTA
19.4 ± 1.4c
15.6 ± 0.8b
9.7 ± 0.4cd
13.2 ± 0.9e
PEPc (µmol g−1 f. wt h−1)
9.7 ± 1.8b (97)
9.2 ± 1.1b (96)
3.7 ± 2.6a (72)
4.9 ± 1.4a (70)
3.3 ± 0.7a (99)
3.7 ± 1.5a (94)
14.2 ± 0.9b (51)
9.4 ± 0.9a (52)
CA (µmol g−1 f. wt s−1)
440 ± 25b
299 ± 40a
9.3 ± 3.1a
9.8 ± 2.7a
413 ± 34b
400 ± 59b
12.4 ± 3.2a
16.0 ± 5.2a
Total PEPc activity in the leaves was c. threefold higher in NO3−-fed plants compared with NH4+-fed plants (Table 1). In roots the total PEPc activity was also higher with NH4+ nutrition than with NO3− nutrition when supplied with 0 p.p.m. CO2, although this difference was smaller at 5000 p.p.m. Nitrate-fed plants had a c. 1.3-fold higher root PEPc activity at 5000 p.p.m. root-zone CO2 compared with 0 p.p.m. root-zone CO2. By contrast, in plants supplied with NH4+ nutrition, the root PEPc activity at 5000 p.p.m. was lower (34%) than that of plants supplied with 0 p.p.m. root-zone CO2. Assay of leaf PEPc activity in the presence of malate only modified the PEPc activity slightly (< 6%) indicating that most PEPc was phosphorylated (active form). Malate inhibited root PEPc activity by c. 30% in plants supplied with NO3− and by c. 50% in plants supplied with NH4+. There was, however, no effect of root-zone CO2 concentration on PEPc sensitivity to malate.
Leaf CA activity was c. 30% lower in NO3−-fed plants grown with 5000 p.p.m. root-zone CO2 than in those grown with 0 p.p.m. root-zone CO2 (Table 1). The concentration of root-zone CO2 had no significant effect on leaf CA activity of NH4+-grown plants. Root CA activity was only c. 3% of that found in the leaves and no significant difference in root CA activity could be discerned for either N source or root-zone CO2 concentration. The CA activity was far in excess of PEPc activity in both leaves and roots.
The incorporation of 14CO2 was expressed as a rate of CO2 incorporation by using the scintillation counts (Bq g−1 f. wt) combined with the initial specific activities of the DI14C supplied. This accounts for the fact that the amount of DI14C supplied was a smaller proportion of the total DIC at higher root-zone CO2 concentrations and estimates the amount of inorganic C incorporated by the roots into acid-stable products (i.e. excluding inorganic C). As a result of the release of respiratory CO2, this specific activity would have declined over the 1-h labelling period and therefore the calculated rate of incorporation was probably under-estimated. Only a small proportion of the 14CO2 taken up by roots was allocated to the acid-stable, 80% ethanol-insoluble fraction over the 24-h chase period (Fig. 4). In general, the amount of assimilated C allocated to the insoluble fraction increased over the chase period, while that of the acid-stable 80% ethanol soluble fraction decreased.
Plants supplied with NH4+ nutrition incorporated c. 10-fold more CO2 than did NO3−-fed plants. Plants supplied with 5000 p.p.m. incorporated c. 10-fold more CO2 than did plants supplied low amounts of carrier (designated 0 p.p.m. CO2). There were only small amounts of root-derived 80% ethanol soluble C in the leaves. Despite this, significant amounts of 80% ethanol insoluble C accumulated in the leaves. The patterns of change in the soluble and insoluble fractions over the chase period were comparable between N and CO2 treatments, although the extents of incorporation were different. In the root, soluble labelled C decreased over time while that in the stem and leaves lagged the root changes with an initial increase followed by a decrease (Fig. 4).
The proportions of 80% ethanol-soluble labelled C allocated to the neutral (mostly carbohydrates), acidic (mostly organic acids) and basic (mostly amino acids) fractions in leaves, stems and roots were not strongly influenced by the root-zone CO2 concentrations (Fig. 5). Allocation to the acidic fraction in the leaves and stems was also not influenced by the N treatments and accounted for c. 25 and 30% of the label in the soluble fraction, respectively. However, plants supplied with NO3− nutrition allocated c. twofold more labelled C to the acidic fraction in roots. The proportion of labelled C in this fraction decreased strongly over time. Plants supplied with NH4+ nutrition allocated more than c. twofold more label to the basic fraction than did plants supplied with NO3− nutrition. The proportion of label in the neutral fraction was greater in plants supplied with NO3− nutrition than in those supplied with NH4+ nutrition and was greatest in leaves, increasing over time.
On transfer from the DI14C-containing labelling solution, a large proportion of acid-stable labelled C was washed off the roots (Fig. 6). Exudation was more rapid with NH4+ than with NO3− nutrition. The plants supplied with NH4+-nutrition exuded a c. twofold larger proportion of label in the basic fraction than did the plants supplied with NO3−-nutrition. This was compensated for by a smaller proportion of label in the neutral fraction of these plants. The proportion of the exudates associated with these fractions was not strongly influenced by the concentration of CO2 supplied to the root zone (Fig. 7).
The increase in RGR of NO3−-fed plants caused by elevated root-zone CO2 (Fig. 1) could be due to changes in one or more of several processes including photosynthesis, respiration, NO3− uptake or partitioning of C and N (Cramer & Richards, 1999). However, there was no evidence for a sustained increase in NO3− uptake with elevated root-zone CO2 concentration (Fig. 2), although more rapid initial uptake may have contributed to greater shoot total N and biomass accumulation. The higher RGR of NO3−-compared with NH4+-fed plants (Fig. 1) was not associated with more rapid sustained uptake of N by the plants supplied with NO3− (Fig. 2), but rather may have been due to the toxicity of accumulated NH4+. Root-based synthesis of amino acids from nutrient NH4+ uptake competes with root growth resulting in increased shoot : root ratios and wilting (Cramer & Lewis, 1993). Furthermore, NH4+ can competitively exclude other cations, result in intense ATP demand for NH4+ efflux and cause rhizosphere acidification (Britto & Kronzucker, 2002). The fact that increased root-zone CO2 concentration did not increase the growth of plants supplied with NH4+ nutrition indicates that anaplerotic C alone could not ameliorate the negative effects of NH4+. Anaplerotic C fixation may not influence NH4+-induced growth depression since both anaplerotic C fixation and growth depend on shoot-derived C.
Nitrate uptake was previously found to be strongly stimulated by elevated root-zone CO2 in NR-deficient mutants of barley (Cramer et al., 1996) and in tomato (van der Merwe & Cramer, 2000) over 6–8 h uptake periods. A similar increase was observed in this study when plants were preincubated for 12 h in solutions containing low NO3− concentrations (0.2 mm) prior to the uptake experiment (Fig. 2). However, root-zone CO2 concentration had no influence on the uptake of NO3− over the longer term (15 d, Fig. 2). In previous work, the plants were either grown on NH4+ or deprived of N for c. 12 h prior to determination of NO3− uptake rates. Thus, the stimulation of NO3− uptake by elevated root-zone CO2 concentration requires depletion of NO3− within the tissue. Since stimulation of NO3− uptake has been found to be independent of NRA (Cramer et al., 1996), it has been proposed to involve an electrochemical modification facilitating NO3− uptake (M. D. Cramer & A. J. Miller, unpubl. data). Thus, stimulation of NO3− uptake by elevated root-zone CO2 could be a transient effect on the uptake kinetics, which is damped by subsequent feedback control. Inhibition of NH4+ uptake in roots aerated with 5000 p.p.m. root-zone CO2 (Fig. 2) could be due depletion of C skeletons for glutamate synthesis (van der Westhuizen & Cramer, 1998). However, increased synthesis of amino acids (Fig. 5) could also downregulate NH4+ uptake, as reported previously (Causin & Barneix, 1993; Feng et al., 1994; Glass et al., 1997).
Shoot total N concentrations of NH4+-fed plants grown with 380 p.p.m. root-zone CO2 were higher than those of NO3−- fed plants, possibly owing to a greater uptake of NH4+ and translocation to the shoot or to slower growth with NH4+-than with NO3−-nutrition (Fig. 3). The higher total N concentrations for shoots and higher total N shoot : root ratios found for NO3−-fed plants grown with 5000 p.p.m. root-zone CO2 compared with 380 p.p.m. root-zone CO2 (Fig. 3) might result from greater initial uptake, followed by translocation of reduced N to the shoot. This increased transfer to the shoot could be facilitated by the availability of anaplerotic C for amino acid synthesis. Although the proportion of 14C localized in the soluble basic fraction was not strongly influenced by the concentration of root-zone CO2 (Fig. 5), there was greater total incorporation of inorganic C (Fig. 4) and thus a greater synthesis of amino compounds.
In situ NRA was greater in plants supplied with elevated root-zone CO2. Since this was measured on plants preincubated with low concentration of NO3− (0.2 mm) for 12 h, it does not reflect the steady-state condition. In vitro NRA in leaves was decreased by elevated CO2 (Table 1) and this may represent downregulation of NRA in response to increased root assimilation of NO3− and the accompanying synthesis of amino acids (Figs 4 and 5). The increased assimilation within the root might have limited the flux of NO3− to the shoot and thus limited NRA there. Although the extent of phosphorylation seemed to have little influence on the NRA (particularly in the leaves), the removal of the inhibitor protein did significantly increase the rates of NRA. However, there were only minor differences in the phosphorylation and inhibition status of NRA between the root-zone CO2 treatments. This indicates that NRA probably responded to induction of NR expression by NO3−, rather than to post-translational regulation, and that changes in NRA activity induced by varying root-zone CO2 concentrations (Cramer et al., 1996) probably depend on NO3− uptake.
Phosphoenolpyruvate carboxylation may serve as a source of anaplerotic C during NH4+ assimilation to compensate for the loss of TCA cycle intermediates to amino acid synthesis or as a source of C for organic acid synthesis in NO3−-fed plants to maintain ionic balance in the xylem sap (Schweizer & Erismann, 1985; Cramer et al., 1993; Vuorinen & Kaiser, 1997). The reduction and assimilation of NO3− in the shoots of tomato plants could account for the higher total PEPc activity in leaves of NO3−-fed plants compared with NH4+-fed plants grown with 5000 p.p.m. root-zone CO2 (Table 1). Similar results were reported by Schweizer & Erismann (1985). Root-zone CO2 did not influence leaf or root PEPc activity of NO3−-fed plants. This reflects the steady-state situation in which NO3− uptake was not increased and 14C incorporation was not strongly increased. Decreased total root PEPc activity of NH4+-fed plants grown with 5000 p.p.m. root-zone CO2 compared with 0 p.p.m. root-zone CO2 was probably due to inhibition of NH4+ uptake by 5000 p.p.m. root-zone CO2 (Fig. 2), possibly as a consequence of feedback inhibition. According to Koga & Ikeda (1997), the increased anaplerotic PEPc activity in roots of wheat, barley and tomato transferred to NH4+ from NO3− nutrition was dependent on de novo protein synthesis. Interestingly, although PEPc activity was greater in the roots of plants supplied with NH4+ nutrition, PEPc from the NH4+-grown plants was more sensitive to inhibition by malate, indicating that the enzyme was less phosphorylated (Jiao & Chollet, 1991). Since NH4+ nutrition decreases 14C associated with organic acids (Fig. 5) and probably depletes organic acids, especially at low root-zone CO2 concentrations, it may be that PEPc in vivo was highly active in the relative absence of malate. This high root PEPc activity was also associated with greater incorporation of 14C (Fig. 4).
It is well known that CA occurs in root nodules at high activities where it is responsible for equilibrating CO2 and HCO3− (Atkins et al., 2001). Nodular CA activities range between 334 and 1863 µmol g−1 f. wt nodule s−1 (Atkins et al., 2001). This is much higher than the CA rates reported in the present investigation for root tissue (c. 10 µmol g−1 f. wt s−1). The uncatalysed conversion of CO2 to HCO3− probably occurs at rates sufficient to support the activities of root PEPc (Raven & Newman, 1994). However, inhibitors of CA activity supplied to tomato roots were previously found to inhibit respiratory O2 consumption and DI14C incorporation by tomato roots (van der Westhuizen & Cramer, 1998). Furthermore, the in vitro rates of CA activity found in this study were far higher than the in vitro rates of PEPc activity. This concurs with the in vivo results of Chang & Roberts (1992), indicating that CA could facilitate the in vivo equilibrium between HCO3− and CO2, if required.
In C3 leaves most of the CA activity is in the chloroplast stroma (Atkins et al., 1972; Fett & Coleman, 1994; Rumeau et al., 1996) and this enzyme may facilitate inorganic C diffusion into and across the chloroplast for photosynthesis by accelerating the dehydration of HCO3− to CO2 (Badger & Price, 1994; Majeau & Coleman, 1994). Elevated atmospheric CO2 concentrations have been found to result in inhibition of Rubisco and CA in Pisum sativum (Majeau & Coleman, 1996). Increased root-zone CO2 concentration results in a decrease in photosynthetic carboxylation rates (Cramer & Richards, 1999). This was suggested to result from the transport of organic C to the shoot and the decarboxylation of this in the shoot which could increase intercellular CO2 concentrations (Arteca & Poovaiah, 1982) and result in the downregulation of Rubisco (Cramer & Richards, 1999). Hibberd & Quick (2002) found evidence for a C4-like reassimilation of decarboxylation-derived CO2 in the stems of C3 plants. Possibly the decrease in leaf CA activity of NO3−-fed plants grown with 5000 p.p.m. root-zone CO2 compared with 0 p.p.m. root-zone CO2 could be due to increased availability of CO2 in the shoot. This could particularly be the case with plants supplied with NO3−, which initially accumulated a large proportion of label in the acidic fraction (organic acids) in the root tissue (Fig. 5). These organic acids when translocated to the shoot may provide a ready source of CO2 by being metabolized in the TCA cycle to liberate CO2 which is then refixed through photosynthesis. It is noteworthy that in plants supplied with NO3−, a large proportion of the soluble fraction in the leaves was associated with the neutral fraction (carbohydrates), possibly derived from refixation of CO2 derived from organic acid decarboxylation. In NH4+ plants, which are likely to transfer a greater proportion of C through the xylem as amino acids, there is perhaps less potential for such a refixation mechanism to be important.
The products of root PEPc fixation were retained to some extent in the plant tissue or exuded into the root solution. Regardless of N source, plants grown with 0 p.p.m. root-zone CO2 concentration retained (tissue + exudates) c. 66% of incorporated 14C after 24 h. By contrast, with 5000 p.p.m. root-zone CO2, NO3−- and NH4+-fed plants retained c. 86% and 77% of their incorporated 14C after 24 h, respectively. Thus, plants supplied with higher concentrations of CO2 to the root-zone lost less of the 14C supplied through respiration. This could be a result of the suppression of root CO2 loss by increased incorporation of DIC at elevated root-zone CO2 concentrations, as has been reported previously (van der Westhuizen & Cramer, 1998). High concentrations of CO2 are also known to repress the rate of respiration in roots (Nobel & Palta, 1989; Palta & Nobel, 1989) and harvested vegetables (Wills et al., 1979; Herner, 1987).
Ammonium nutrition resulted in greater DI14C incorporation than did NO3− nutrition. This was probably due to the more rapid uptake and assimilation of NH4+ than of NO3− (Murphy & Lewis, 1987), and the associated requirement for C skeletons for amino acid synthesis (Cramer et al., 1993). The initially high proportion of label associated with the basic fraction in the leaves of plants supplied with NH4+ nutrition resulted from rapid transfer of labelled amino acids from the roots to the shoots. The subsequent decrease in the amino acid fractions in the leaves was probably caused by diversion to insoluble components, metabolism of the amino acids, respiratory losses or cycling of the amino compounds back from the shoot to the root. Extensive cycling of N between the shoot and root has been documented previously (Lambers et al., 1982).
The diversion of incorporated 14C into organic acids in NO3−-fed plants and subsequent export to the shoots (Fig. 6) was consistent with previous results (Cramer et al., 1993; Cramer & Richards, 1999). Organic acids translocated from the roots to the shoots could be decarboxylated in the shoots and the CO2 reassimilated to form carbohydrates (Cramer & Richards, 1999), accounting for the greater proportion of the neutral fraction in the shoots of plants supplied with NO3− (Fig. 5). The decrease in the proportion of label associated with the acidic fraction in the roots over time could also reflect exudation from the root (Fig. 7), increased synthesis of amino acids, and subsequent allocation of amino acids and carbohydrates to structural material, as indicated by increasing allocation to the insoluble fraction over time (Fig. 4).
Greater net exudation with elevated root-zone CO2 (Fig. 6) was probably due to greater PEPc CO2 refixation, which, especially when combined with NH4+ nutrition, could result in an increased amino acid synthesis and subsequent exudation of these compounds (Cramer & van der Westhuizen, 2000). The reason for the exudation of amino acids is not clear, although it could simply be a consequence of the accumulation of amino compounds in the root or a regulatory mechanism to reduce tissue amino acid levels. A small proportion of the label exuded was associated with the organic acid fraction. The fact that the organic acids accounted for a relatively small proportion of the exuded label probably reflects the fact that the plants were grown with adequate P nutrition (2 mm). Phloem-derived sugars translocated through the root apoplast may ‘leak’ into the external solution (Jones & Darrah, 1993; Marschner, 1995), although this notion has been challenged by work on Leptochloa fusca, which may sustain diazotrophic bacteria with sugar exudates (Mahmood et al., 2002).
Changes in RGR induced by varying root-zone CO2 concentrations are generally small in the absence of specific stresses, such as salinity and high temperatures (Cramer & Richards, 1999), and aluminium (Cramer & Titus, 2001). In plants supplied with NO3−, large-scale changes in aspects of physiology (e.g. NO3− uptake, NRA and PEPc activity) induced by transition to elevated root-zone CO2 concentrations or upon initial exposure to NO3− were not sustained in the longer term. Although the RGR of plants grown with NH4+ nutrition was not altered by root-zone CO2, there were sustained changes in PEPc activity and 14C incorporation. This indicated that plants supplied with NH4+ nutrition utilized root-zone CO2 to meet the demand for anaplerotic carbon for amino acid synthesis on a sustained basis, although feedback regulation due to amino acid accumulation could have limited the capacity for RGR to respond to this. The results presented on CA activity and 14C partitioning are consistent with organic acid produced in roots, especially with NO3− nutrition, being partly decarboxylated in the shoot and the released CO2 refixed through photosynthesis. This refixation of root-derived CO2 may contribute to photosynthetic carbon acquisition and to enhanced RGR, especially under stress conditions (Cramer & Richards, 1999).
We thank the National Research Foundation for financial support.