Nitrate assimilation in plant shoots depends on photorespiration

Abstract

Photorespiration, a process that diminishes net photosynthesis by ≈25% in most plants, has been viewed as the unfavorable consequence of plants having evolved when the atmosphere contained much higher levels of carbon dioxide than it does today. Here we used two independent methods to show that exposure of Arabidopsis and wheat shoots to conditions that inhibited photorespiration also strongly inhibited nitrate assimilation. Thus, nitrate assimilation in both dicotyledonous and monocotyledonous species depends on photorespiration. This previously undescribed role for photorespiration (i) explains several responses of plants to rising carbon dioxide concentrations, including the inability of many plants to sustain rapid growth under elevated levels of carbon dioxide; and (ii) raises concerns about genetic manipulations to diminish photorespiration in crops.

Rubisco, the most prevalent protein in plants, indeed in the biosphere, catalyzes the reaction of ribulose-1,5-bisphosphate with either CO₂ or O₂ and thereby initiates, respectively, the CO₂ assimilatory (C₃ reductive) or photorespiratory (C₂ oxidative) pathways. The balance between the two reactions depends on the relative concentrations of CO₂ and O₂ at the site of catalysis. At current atmospheric levels of CO₂ (≈360 μmol·mol^-1) and O₂ (≈209,700 μmol·mol^-1), photorespiration in C₃ plants dissipates >25% of the carbon fixed during CO₂ assimilation (1). Thus, photorespiration has been viewed as a wasteful process, a vestige of the high CO₂ atmospheres under which plants evolved (2). At best, according to current thought, photorespiration may mitigate photoinhibition under high light and drought stress (2, 3) or may generate amino acids such as glycine for other metabolic pathways (4). Genetic modification of Rubisco to minimize photorespiration in crop plants has been the goal of many investigations (5).

Atmospheric CO₂ concentrations will rise to somewhere between 600 and 1,000 μmol·mol^-1 by the end of the 21st century (6). Transferring C₃ plants from ambient (≈360 μmol·mol^-1) to elevated (≈720 μmol·mol^-1) CO₂ concentrations decreases photorespiration and initially stimulates net CO₂ assimilation and growth by ≈30% (7). With longer exposures to elevated CO₂ concentrations (days to weeks), however, net CO₂ assimilation and plant growth slow down until they stabilize at rates that average 12% (8) and 8% (9), respectively, above those of plants kept at ambient CO₂ concentrations. This phenomenon, known as CO₂ acclimation, is often associated with diminished activities of Rubisco and other enzymes in the C₃ reductive photosynthetic carbon cycle (10, 11), but the influence of elevated CO₂ may not be specific to these enzymes (12). Rather, CO₂ acclimation follows a 14% decline in overall shoot nitrogen concentrations (13), a change nearly double what would be expected if a given amount of nitrogen were diluted by the additional biomass that accumulates under elevated CO₂ concentrations (9, 12).

We proposed a relatively simple explanation for these responses: elevated CO₂ concentrations inhibit the assimilation of nitrate () in shoots of C₃ plants (14-16). Because is the prominent form of inorganic nitrogen available to plants from temperate well aerated soils (17), diminished assimilation dramatically alters the nitrogen balance in C₃ plants (15). Much of our evidence was based on estimates of shoot assimilation derived from calculations of the difference in the assimilatory quotient (ΔAQ, ratio of net CO₂ consumption to net O₂ evolution) between plants that received as their sole nitrogen source and those that received ammonium () as their sole source. Here, we establish ΔAQ as a measure of assimilation using genotypes of Arabidopsis in which reductase activities are enhanced or deficient. We then use both ΔAQ and an independent measure to demonstrate that assimilation depends on photorespiration in a dicotyledon (Arabidopsis) and a monocotyledon (wheat). These results offer a different perspective on the importance of photorespiration and on attempts to minimize it.

Materials and Methods

Materials and Growth Conditions. We used three genotypes of Arabidopsis thaliana cv. Columbia: (i) the wild type, (ii) a transgenic line harboring the chimeric gene Lhch1^*3::Nia1^*2 that overexpresses one form of reductase (18), and (iii) a genotype with mutations in both structural genes for reductase, nia1 nia2 (19). Seeds were germinated on plates filled with a dilute Murashige-Skoog medium (2.3 g·liter^-1) in 0.75% Phytagar (GIBCO/BRL). The plates were placed in controlled environment chambers (Conviron, Winnipeg, MB, Canada) at ambient CO₂ levels and received 9 h of 350 μmol·m^-2·s^-1 photosynthetically active radiation and 24°C. After 10 d, seedlings were transferred one at a time to 5 × 40-mm pieces of rock wool (Grodania, Hovedgaden, Denmark). Twenty seedlings were transplanted to an opaque 4-liter polyethylene container, the end of the rock wool opposite the seedling being immersed in an aerated nutrient solution containing 200 μM NH₄Cl and 200 μM KNO₃ as nitrogen sources (20). This solution was changed every 3 d. The container was placed in the same controlled environment chamber as the plates.

We surface-sterilized wheat (Triticum aestivum cv. Veery 10) seeds for 1 min in 2.6% NaClO, washed them thoroughly with water, and germinated them for several days on thick paper toweling saturated with 10 mM CaSO₄. Twenty seedlings were transplanted to a 19-liter opaque polyethylene tub filled with an aerated nutrient solution containing 200 μM NH₄NO₃ (21). The solution was replenished every 3 d. The tubs were placed in a controlled environment chamber (Conviron), providing a photosynthetic photon flux density (PFD) of 650 μmol of quanta m^-2·s^-1 at plant height and a 16 h/25°C day and 8 h/15°C night. After ≈14 d, we transferred a seedling that had three true leaves into a gas-exchange measurement system.

Nitrate Reductase Activity. To assess reductase activity in Arabidopsis, 1 g of leaf material was ground with fine glass beads in a cold mortar that contained 4 ml of 0.1 M K-phosphate (pH 7.5), 1 mM EDTA, 3 mM cysteine, and 3% (wt/vol) casein (22). The homogenate was centrifuged at 30,000 × g for 10 min and the supernatant assayed for in vivo and fully activated reductase activity according to the procedure of Kaiser et al. (23).

Gas-Exchange Measurements. A plant was sealed by a rubber stopper around its stem into a shoot and root cuvette (24, 25). Leaves in the shoot cuvette were at their normal orientation; thus the angle of incidence was between 0° and 45° for Arabidopsis and 70° and 80° for wheat. Net gas fluxes from the shoot were monitored with the instrumentation described previously (15, 24). In brief, an infrared gas analyzer (Horiba VIA-500R, Kyoto) measured CO₂ fluxes, a custom O₂ analyzer based on heated zirconium oxide ceramic cells measured O₂ fluxes, and relative humidity sensors (Vaisala, Helsinki) measured water vapor fluxes. Mass flow controllers (Tylan, Torrance, CA) prepared the various gas mixtures, and a pressure transducer (Validyne, North Ridge, CA) monitored the gas flows through the shoot cuvette. We also placed wheat leaves in a leaf cuvette (LI-6400-40, Li-Cor, Lincoln, NE) and estimated the gross O₂ exchange from chlorophyll fluorescence, but this measure did not respond to nitrogen source or CO₂ level (26).

Nitrate Absorption and Accumulation. Wild-type Arabidopsis and wheat were grown as described above, except that 3 d before measurement for Arabidopsis and 2 d for wheat, the plants were shifted from a medium containing 200 μM NH₄Cl and 200 μM KNO₃ to one devoid of nitrogen. This protocol induced absorption and reductase but then depleted the plant tissue of free . The night before measurements, five to eight plants were transferred to a multiplant measurement system (27). The next morning, Arabidopsis or wheat plants received, respectively, 500 or 1,000 μmol·m^-2·s^-1 photosynthetically active radiation at plant height. The plants were exposed to an atmosphere of (i) 360 μmol·mol^-1 CO₂ and 21% O₂, (ii) 720 μmol·mol^-1 CO₂ and 21% O₂, or (iii) 360 μmol·mol^-1 CO₂ and 2% O₂. Then during a measurement period of 1 h for the Arabidopsis and 2 h for wheat, the plants were shifted to an aerated medium containing 0 or 5.5 μmol . Absorption was assessed by the amount of remaining in the medium after the measurement period. After the measurement period, the plants were divided into shoots and roots, oven-dried, and ground to a powder in a ball mill. Water extracts of the powder were analyzed for via HPLC (28), and accumulation in the shoots and roots were calculated from the difference in content between the plants that had received during the measurement period and those that had not. Nitrate assimilation was calculated as the difference in the rates of absorption and plant accumulation. The rate of shoot accumulation was the amount of accumulated in the shoots during the measurement period divided by the time.

Statistics. A repeated-measures analysis of variance was performed by using the mixed procedure in sas (PROC MIXED, SAS Institute, Cary, NC). The PFD was considered to be a repeated factor, because each canopy was measured at all five levels of PFD. Effects of the treatments and their interactions were considered significant when P < 0.05.

Results

Nitrate Reductase Activities. In Arabidopsis, reductase in the shoot was nearly fully activated (Fig. 1). In 36-d-old wild-type plants, the fully activated rates of reduction in μmol of per g of fresh mass per min (mean ± SE, n = 10) were 0.13 ± 0.02 in the shoots (Fig. 1) and 0.030 ± 0.001 in the roots at ambient CO₂ concentrations. The short-day regime under which the Arabidopsis plants were grown prevented them from flowering, but as the wild-type plants aged from 36 to 48 d, reductase activity in the shoots diminished markedly (Fig. 1). A transgenic line that harbored the chimeric gene Lhch1^*3::Nia1^*2 (29) had twice the reductase activity of the wild type, whereas a genotype with mutations in both structural genes for reductase, nia1 nia2 (19), had no significant activity (Fig. 1). In wheat, the fully activated rates of reductase activity in μmol of per g of fresh mass per min (mean ± SE, n = 6) were 0.58 ± 0.03 and 0.021 ± 0.003 in the shoots and roots, respectively, at ambient CO₂ concentrations and 0.46 ± 0.06 and 0.023 ± 0.002 in the shoots and roots, respectively, at elevated CO₂ concentrations (15).

Fig. 1.

reductase activity (μmol of generated per g of fresh mass per min) as a function of plant age (d) in leaves of a wild-type A. thaliana cv. Columbia (WT), a transgenic line harboring the chimeric gene Lhch1^*3::Nia1^*2 (OE), and a genotype (nia1 nia2) with mutations in both structural genes for reductase (Mut). Because reductase is regulated through phosphorylation, leaf tissue was assayed under conditions that either dephosphorylated the enzyme (fully activated) or did not change its phosphorylation (in vivo). Shown are the mean ± SE (n = 5-8 plants).

Shoot Gas Fluxes. We simultaneously monitored net CO₂ and O₂ fluxes from shoots of intact Arabidopsis and wheat plants as a function of light level. There were six treatments: plants received either or as a nitrogen source and an atmospheric gas composition of either (i) 360 μmol·mol^-1 CO₂ and 21% O₂ (ambient CO₂ and O₂), (ii) 700 or 720 μmol·mol^-1 CO₂ and 21% O₂ (elevated CO₂), or (iii) 360 μmol·mol^-1 CO₂ and 2% O₂ (low O₂). Net CO₂ consumption was stimulated under elevated CO₂ or low O₂ concentrations but was similar for both nitrogen treatments (Figs. 5 and 6, which are published as supporting information on the PNAS web site), a response typical for C₃ plants that have received ample amounts of nitrogen (30). Net O₂ evolution differed most between and nutrition under ambient CO₂ and O₂ atmospheres (Figs. 5 and 6).

The ΔAQ, the change in the AQ (the ratio of net CO₂ consumption to net O₂ evolution) with a shift from to nutrition, highlights these differences (Figs. 2 and 3). Under ambient CO₂ and O₂ atmospheres, ΔAQ was positive in plants having significant activities (36-d-old wild-type Arabidopsis, Fig. 2A; transgenic Arabidopsis overexpressing reductase, Fig. 2D; and wheat, Fig. 3), but did not deviate from zero in plants with diminished reductase activities (48-d-old wild-type Arabidopsis, Fig. 2B; and the Arabidopsis knockout mutants, Fig. 2C). In Arabidopsis and wheat plants having significant activities, ΔAQ decreased at low O₂ concentrations and became negligible at elevated CO₂ concentrations (Figs. 2 A and D and 3).

Fig. 2.

Changes in assimilatory quotient with the shift from to (ΔAQ) as a function of photosynthetic PFD in shoots of A. thaliana cv. Columbia. Thirty-six-day-old wild-type plants (A), 48-d-old wild-type plants (B), a genotype with mutations in the two structural genes for reductase (nia1 nia2) (C), and a transgenic line harboring the chimeric gene Lhch1^*3::Nia1^*2 (D). The plants were grown under ambient CO₂ (360 μmol·mol^-1) and measured under ambient CO₂ and O₂ (360 μmol·mol^-1 CO₂ and 21% O₂; circles), elevated CO₂ (720 μmol·mol^-1 CO₂ and 21% O₂; triangles), or low O₂ (360 μmol·mol^-1 CO₂ and 2% O₂; squares). Shown are the mean ± SE, n = 5-8 plants.

Fig. 3.

Changes in assimilatory quotient with the shift from to (ΔAQ) as a function of photosynthetic PFD in shoots of wheat (T. aestivum cv. Veery 10). The plants were grown under ambient CO₂ (360 μmol·mol^-1) and measured under ambient CO₂ and O₂ (360 μmol·mol^-1 CO₂ and 21% O₂; circles), elevated CO₂ (700 μmol·mol^-1 CO₂ and 21% O₂; triangles), or low O₂ (360 μmol·mol^-1 CO₂ and 2% O₂; squares). Shown are the mean ± SE, n = 5-8 plants. The data for ambient CO₂ and O₂ and elevated CO₂ and ambient O₂ have been published (15).

Nitrate Accumulation. Another measure of assimilation is the difference between the amount of that a plant absorbs and that it accumulates in its tissues. According to this measure, both elevated CO₂ and low O₂ concentrations inhibited plant assimilation in Arabidopsis and wheat (Fig. 4), although the influence of low O₂ concentrations was significant only at P < 0.2 in Arabidopsis. Absorption of also declined at elevated CO₂ and low O₂ concentrations but to a lesser extent than assimilation (Fig. 4). Moreover, the rates at which accumulated in the shoots of either species did not differ significantly among treatments (data not shown).

Fig. 4.

In wild-type Arabidopsis and wheat, uptake as the amount of depleted from a medium and assimilation as the difference between the rates of net uptake and net accumulation of free in plant tissues. Thirty-six-d-old Arabidopsis plants (A) or 10-d-old wheat (B) were exposed to either 360 μmol·mol^-1 CO₂ and 21% O₂ (gray), 720 μmol·mol^-1 CO₂ and 21% O₂ (black), or 360 μmol·mol^-1 CO₂ and 2% O₂ (white). Shown are the mean ± SE (n = 13-16). Treatments labeled with different letters differ significantly (P ≤ 0.05). The light levels were 500 and 1,000 μmol·m^-2·s^-1 PAR for Arabidopsis and wheat, respectively.

Discussion

Two independent methods indicated that assimilation in Arabidopsis and wheat decreased under both elevated CO₂ and low O₂ atmospheres.

The first method was a real-time continuous measure involving AQ, the ratio of net CO₂ consumption to net O₂ evolution. The AQ decreases as assimilation increases: additional electrons generated from the light-dependent reactions of photosynthesis are transferred to and hence to , stimulating net O₂ evolution while having little effect on CO₂ consumption (15, 24, 31, 32). We present ΔAQ, the change in AQ under versus nutrition rather than AQ, because several other biochemical processes such as lipid metabolism can influence AQ, but these processes do not change rapidly with nitrogen source, so ΔAQ should predominantly reflect assimilation (32). The ΔAQ also has appropriate scaling, because it should be zero when assimilation is negligible and should increase as nitrate assimilation increases. Here (Figs. 2 and 3), ΔAQ differed from zero only in plants with relatively high reductase activities, affirming its relationship with assimilation.

The second method for assessing assimilation was a traditional one based on the difference between the total amount of absorbed and that which accumulated in plant tissues (e.g., refs. 33-38). This method has several difficulties.

1. It estimates assimilation in the whole plant, not just in the shoots. Nonetheless, the observed changes in total assimilation with CO₂ levels (Fig. 4) probably reflected mostly the responses of the shoots, because assimilation in the roots usually comprises only a minor percentage of the total during the day (39) and is relatively insensitive to CO₂ levels (15). For example, reductase activity was 27 times greater in wheat shoots than roots and 4.3 times greater in 36-d-old wild-type Arabidopsis shoots than roots.

2. This method requires destructive tissue analysis after the uptake measurement and thus cannot be conducted in real time.

3. Although the plants were deprived of nitrogen for 3 d, free in the tissues of the controls (those that did not receive during the uptake measurements) spanned a broad range, causing variation in the estimates of accumulation.

4. Uptake measurements were conducted during the transition from nitrogen deprivation to nitrogen sufficiency. The rates at which accumulated in the shoots, however, were similar in all treatments (data not shown), indicating that availability in the shoots did not limit assimilation at elevated CO₂ concentrations.

Despite these difficulties, the decline in assimilation rates under elevated CO₂ or low O₂ concentrations determined by this method (Fig. 4) paralleled the results based on the ΔAQ (Figs. 2 and 3).

A physiological response common to elevated CO₂ and low O₂ is diminished photorespiration (40). The observed shifts in ΔAQ under elevated CO₂ or low O₂ concentrations did not result directly from photorespiration. Photorespiration releases CO₂ and consumes O₂ in equal amounts (41); therefore, if only the photorespiratory pathway were involved, ΔAQ would shift in the opposite direction to the one we observed. For example, the 36-d-old wild-type Arabidopsis under ambient CO₂ and O₂ had an AQ of 0.94 ± 0.01 under and 1.04 ± 0.01 under (mean ± SE for the five light levels); equal fluxes of CO₂ and O₂ from photorespiration would bring the AQ values for these treatments closer together as photorespiration increases and further apart as it decreases. A straightforward interpretation for the decline in ΔAQ at elevated CO₂ or low O₂ is that assimilation depends on photorespiration. Our results with the second method for assessing assimilation (Fig. 4) affirm this interpretation.

Possible Mechanisms. One part of the photorespiratory pathway is the export of malate from the chloroplast through the cytoplasm and into the peroxisome, where it generates NADH, which reduces hydroxypyruvate. This malate “valve” or “shuttle” increases the NADH/NAD ratio in the cytoplasm (42) and thereby may provide NADH instrumental in the reduction of to . Malate also serves as a counterion that prevents alkalinization when , an anion, becomes incorporated into a neutral amino acid (43). Such processes could explain the observations that assimilation was fastest in Arabidopsis and wheat under ambient CO₂ and O₂ concentrations (Figs. 2, 3, 4), the treatment under which photorespiration was highest.

The influence of elevated CO₂ concentrations on assimilation was more pronounced than that of low concentrations of O₂ (Figs. 2 A and D, 3, and 4). Two additional mechanisms contribute to the inhibitory effect of elevated CO₂ concentrations on assimilation. (i) Transport of from the cytosol into the chloroplast involves the net diffusion of HNO₂ or cotransport of protons and across the chloroplast membrane. This requires the stroma to be more alkaline than the cytosol (44, 45). Elevated concentrations of CO₂ can dissipate some of this pH gradient, because additional CO₂ movement into the chloroplast acidifies the stroma. As a result, elevated CO₂ concentrations inhibited transport into the chloroplast (15). (ii) Several competing processes, the C₃ reductive photosynthetic carbon cycle, the reduction of to , and the incorporation of into amino acids, occur in the chloroplast stroma (46) and require reduced ferredoxin generated by photosynthetic electron transport (47). Key enzymes in these processes have different affinities for reduced ferredoxin: ferredoxin-NADP reductase has a K_m of 0.1 μM, nitrite reductase has a K_m of 0.6 μM, and glutamate synthase has a K_m of 60 μM (48). As a result, assimilation proceeds only if the availability of reduced ferredoxin exceeds that needed for NADPH formation (49, 50). For wheat (Fig. 3) and tomato (16), this occurred only at high light intensities under ambient CO₂ and O₂ concentrations, conditions under which CO₂ availability limited C₃ photosynthesis.

The responses of CO₂ and O₂ fluxes to the various treatments were similar in the wild-type Arabidopsis and the transgenic that overexpresses reductase (Fig. 2 A and D). This similarity supports the contention that reductase activity by itself limits neither assimilation (23) nor plant performance (51).

Implications. Our finding that CO₂ inhibits assimilation in shoots of Arabidopsis and wheat is consistent with previous studies on barley (24), tomato (16), and wheat (14, 15). If CO₂ inhibition of shoot assimilation were common among C₃ species, it could account for several responses of plants to elevated CO₂, including the decline in shoot protein and the diminished activities of photosynthetic enzymes. Nitrogen availability determines plant responses to elevated CO₂ concentrations more than any other environmental factor (52, 53), but ecosystems show a broad range of responses to elevated CO₂ concentrations, possibly as a result of the seasonal and spatial fluctuations in the relative availabilities of and . For instance, ecosystems in which is the dominant nitrogen form, such as pine forests (54) or wetlands (55), show a relatively large increase (≈25%) in net primary productivity under CO₂ enrichment, whereas ecosystems in which is dominant, such as grasslands (56) or wheat fields, at standard fertilizer levels (low fertilizer treatment at Maricopa, AZ; ref. 57) show declines in net primary productivity under CO₂ enrichment.

Plants vary in their relative dependence on and as nitrogen sources and in their balance between shoot and root assimilation (17). Our results suggest that rising atmospheric CO₂ levels will favor taxa that prefer as a nitrogen source or assimilate primarily in their roots.

Extensive efforts to increase the specificity of Rubisco for CO₂ relative to O₂ and thereby increase the productivity of C₃ crops have proved unsuccessful (5). Our results indicate that such efforts might have hitherto unforeseen consequences: in agricultural systems where is the dominant form of inorganic nitrogen, minimizing photorespiration may be associated with nitrogen deprivation.