In Vivo Respiratory Metabolism of Illuminated Leaves

Abstract

Day respiration of illuminated C₃ leaves is not well understood and particularly, the metabolic origin of the day respiratory CO₂ production is poorly known. This issue was addressed in leaves of French bean (Phaseolus vulgaris) using ¹²C/¹³C stable isotope techniques on illuminated leaves fed with ¹³C-enriched glucose or pyruvate. The ¹³CO₂ production in light was measured using the deviation of the photosynthetic carbon isotope discrimination induced by the decarboxylation of the ¹³C-enriched compounds. Using different positional ¹³C-enrichments, it is shown that the Krebs cycle is reduced by 95% in the light and that the pyruvate dehydrogenase reaction is much less reduced, by 27% or less. Glucose molecules are scarcely metabolized to liberate CO₂ in the light, simply suggesting that they can rarely enter glycolysis. Nuclear magnetic resonance analysis confirmed this view; when leaves are fed with ¹³C-glucose, leaf sucrose and glucose represent nearly 90% of the leaf ¹³C content, demonstrating that glucose is mainly directed to sucrose synthesis. Taken together, these data indicate that several metabolic down-regulations (glycolysis, Krebs cycle) accompany the light/dark transition and emphasize the decrease of the Krebs cycle decarboxylations as a metabolic basis of the light-dependent inhibition of mitochondrial respiration.

Illuminated leaves simultaneously assimilate CO₂ through the photosynthetic carbon reduction cycle and lose CO₂ through photorespiration and day respiration. In darkness, leaves no longer assimilate CO₂ via the photosynthetic carbon reduction cycle but produce CO₂ through dark respiration. Although dark respiration is known to involve glycolysis and CO₂ production through pyruvate dehydrogenation and the degradative Krebs cycle (Trethewey and ap Rees, 1994; Plaxton, 1996), the carbon metabolism that is responsible for the CO₂ respiratory release in the light is almost unknown. This is so because the day respiratory CO₂ flux is very low and masked by the photosynthetic carbon fixation and the photorespiratory CO₂ production in the light, and is thus difficult to study.

Nevertheless, it has been repeatedly shown, using either the Laisk's (Laisk, 1977) or Kok's method (Kok, 1948), that the rate of day respiration (R_d) is less than that of dark respiration (R_n; for review, see Atkin et al., 2000) so that light is known to inhibit respiration, with a R_d/R_n value (usually denoted as μ) ranging from 30% to 100% (for a recent study, see Peisker and Apel, 2001). Pioneering gas exchange measurements on mustard suggested that some enzymatic activities are inhibited in the light so that substrates accumulate (Cornic, 1973), explaining the respiratory burst when leaves are darkened: the light enhanced dark respiration. More recently, it has been shown in the unicellular alga Selenastrum minutum that pyruvate kinase (Lin et al., 1989) is inhibited by light. It is also the case of the pyruvate dehydrogenase complex that is partly inactivated by (reversible) phosphorylation in extracts from illuminated leaves (Budde and Randall, 1990; Tovar-Mendez et al., 2003). Photorespiration is also probably involved in the inhibition of pyruvate dehydrogenase as it has been shown that this enzyme is down-regulated by NH₃, which is a byproduct of the photorespiratory Gly decarboxylation (Krömer, 1995). Enzymes of the Krebs cycle are also assumed to be inhibited in the light because of a high mitochondrial NADH level due to photorespiratory Gly decarboxylation (Atkin et al., 2000). Additionally, it has been shown that the mitochondrial isocitrate dehydrogenase is inhibited by the high NADPH/NADP ratios that occur in the light (Igamberdiev and Gardeström, 2003).

Although all these enzymatic data suggest that the respiratory pathway is down-regulated in the light regarding both glycolysis and the Krebs cycle, respiratory metabolic fluxes in vivo in leaves are not well known. Some labeling experiments with carbon isotopes (¹³C or ¹⁴C) have already been done to disentangle respiratory metabolic fluxes in vivo in the light and in the dark, but surprisingly, studies that have focused on labeling of the resulting respired CO₂ are scarce. Using ¹⁴CO₂ labeling techniques, Pärnik et al. (2002) suggested that CO₂ production in the light is composed of (1) decarboxylation of primary products like triose phosphates and malate (between 10% and 50% from one species to another), and (2) decarboxylation of end-products like Suc and starch to a greater extent (50%–90%). ¹²CO₂ (respiratory) production in a ¹³C atmosphere has been used to show that day respiration is less than night respiration, and its rate as well as the ratio μ decreased at high CO₂ concentration (Pinelli and Loreto, 2003). It has thus been proposed that inhibition of CO₂ production is determined by the CO₂ fixation flux (Atkin et al., 1998; Pinelli and Loreto, 2003). The ability of leaves to oxidize some metabolites through respiration has also been investigated with feeding experiments using labeled organic compounds. Supplying ¹³C-enriched Glc to myrtle leaves in the light with CO₂-free air did enrich the CO₂ subsequently respired in the dark, although the CO₂ was not completely labeled (Affek and Yakir, 2003). Nevertheless, the ¹³C amount in the light-respired CO₂ was not measured in that study.

So already published data do not show what the respiratory metabolic pathway is actually in illuminated leaves; that is, what the metabolic fluxes associated with the day respiratory CO₂ production are. We address this question here by feeding illuminated French bean (Phaseolus vulgaris) leaves with positional ¹³C-enriched Glc and pyruvate and measuring the resulting ¹³C-enrichment in both day and dark-respired CO₂ and in intermediary respiratory compounds, using isotope ratio mass spectrometry and nuclear magnetic resonance (NMR), respectively. The isotopic analysis of CO₂ respired in the dark and the measurement of the carbon isotope discrimination during photosynthesis allows us to calculate the ¹³C-content in dark- and light-respired CO₂, respectively. Respiration of supplied ¹³C₁- and ¹³C₃-enriched Glc is found to be almost completely inhibited by light. By contrast, decarboxylation of ¹³C₁-enriched pyruvate through the pyruvate decarboxylation is much less inhibited by light, unlike ¹³C-2-pyruvate. These data are supported by NMR spectra obtained from illuminated leaves and taken as a whole, they suggest that (1) the Krebs cycle and glycolysis are strongly inhibited by light, with little interconversion between triose phosphates and hexose phosphates through the triose phosphate isomerase and aldolase reactions, and (2) the pyruvate dehydrogenation is only partly inhibited by light, with the acetyl-CoA molecules being directed toward purposes other than respiration.

THEORY

This “Theory” section describes the mathematical background used to calculate the R_d decarboxylations when detached leaves were supplied with ¹³C-enriched molecules (Fig. 1).

Figure 1.

Metabolic model used in this paper for decarboxylations of ¹³C-enriched substrates and respiratory variables taken into account (r_PDH and r_K and the sum, r_day or r_night). r_PDH and r_K are the rates of decarboxylation of ¹³C-enriched substrates through the PDH reaction and the Krebs cycle, respectively. r_day and r_night are the sum of r_PDH and r_K in the light and in the dark, respectively. Different symbols are used for different carbon atom positions in order to see the pathways to CO₂ production.

In the following, the isotope composition (δ¹³C) and the isotope ratio (¹³C/¹²C) are denoted as δ and R, respectively, and the percentage of ¹³C is denoted as λ. λ is simply deduced from δ through the following relationship:

As where R_st is the ¹³C/¹²C ratio in the Pee Dee reference material (R_st = 0.0112372), we have:

(1)

Rate of ¹³C-Enriched Substrate Decarboxylation in the Light (r_day)

The on-line discrimination value was obtained using the method of Evans et al. (1986):

where the isotope compositions of air entering and leaving the cuvette are δ_e and δ_o, respectively. ξ is equal to c_e/(c_e − c_o) where c_e and c_o are the CO₂ molar fractions in the entering and leaving air, respectively.

The Δ_obs value obtained before the substrate addition is denoted as Δ. It is assumed that after substrate addition, the air leaving the leaf cuvette is the sum of the CO₂ left by photosynthetic discrimination and additional CO₂ released from substrate respiration. The latter flux is denoted as r_day. If the leaf area is denoted as S and c_fixed is the CO₂ amount (in μL/L) fixed by photosynthesis, the mass balance equation for CO₂ is as follows:

where d (L/s) is the air flow through the leaf cuvette and V_M the molar volume. Thus, we have:

(2)

Case of Feeding with a Compound That Has a Homogeneous Isotope Composition

Homogeneous isotopic compounds are not involved in this study, but the calculation is explained here as it gives a basis to understand the next step, that is, the use of compounds with a heterogeneous isotope composition.

The isotope composition of the substrate fed to the leaf is denoted as δ_s. Isotopic mass balance is so that the ¹³C amount entering the cuvette (c_eλ_e) plus that from the additional CO₂ release (r_daySV_Mλ_s/d) is equal to the sum of the ¹³C amount fixed by the leaf (c_fixed λ_fixed) and the ¹³C amount leaving the cuvette (c_oλ_o). That is,

(3)

Note that the similar relationship with deltas is not correct as the strong ¹³C-enrichment does not allow one to neglect R compared to 1 so that λ ≠ R. λ_fixed is obtained with the usual relationship and Equation 1, where Δ is obtained before feeding by on-line isotopic measurements (Evans et al., 1986). Substituting Equation 2 in Equation 3 gives:

(4)

It should be noted that possible ¹²C/¹³C fractionations that can occur during the absorption of labeled compounds are neglected here. The effect of such a fractionation is clearly negligible, that is, in the per mil order of magnitude, compared to the labeling level, which is in the percent order of magnitude or more.

Case of Feeding with a Compound That Has a Nonhomogeneous Positional Isotope Composition

The underlying assumption of the previous paragraph is that the substrate is isotopically homogeneous. However, the isotope composition of the feeding substrates used in these experiments is nonhomogeneous so that its different carbon atom positions do not have the same Δ values. In such a case, the metabolic reactions that are responsible for the decarboxylation of the carbon atoms and their rates should be taken into account. For example, this occurs when Glc or pyruvate is added: the C-1 atom of pyruvate is decarboxylated by pyruvate dehydrogenase, while the C-2 and C-3 positions are decarboxylated by the Krebs cycle (Fig. 1). Similarly, the C-3 and C-4 atom positions of Glc are decarboxylated by the pyruvate dehydrogenase reaction, the other being decarboxylated by the Krebs cycle. Advantage can then be taken from this with positionally ¹³C-enriched substrates; ¹³C-1-pyruvate would specifically enrich the CO₂ produced by pyruvate dehydrogenase, while ¹³C-2-pyruvate would specifically enrich the CO₂ that comes from the Krebs cycle. The same applies to positional ¹³C-enrichment in Glc.

The additional decarboxylations through the pyruvate dehydrogenase (PDH) reaction and the Krebs cycle are denoted as r_PDH and r_K, respectively (Fig. 1). With the relationship r_day = r_PDH + r_K, Equation 2 still works. For ¹³C-1-pyruvate, Equation 3 becomes:

(5)

where λ₁ is the ¹³C percentage in the C-1 position of the labeled pyruvate. λ_c is the ¹³C percentage of the other (unlabeled) positions. For ¹³C-2-pyruvate, we have:

(6)

where λ₂ is the ¹³C percentage in the C-2 (labeled) position of pyruvate. λ_c (unlabeled positions) is the same as λ_c of Equation 5 (¹³C-1 enrichment). It can be seen that the two conditions of ¹³C-enrichment give two Equations (5 and 6) so that r_PDH and r_K can be deduced by a substitution procedure. A similar procedure is used for the positional ¹³C-enrichment of Glc.

Rate of ¹³C-Enriched Substrate Decarboxylation in Darkness

The CO₂ that is produced in darkness after a light period with ¹³C-enriched substrate feeding comes from respiratory oxidation of new photosynthates (¹³C percentage in the fixed carbon λ_fixed), photosynthates from the previous light period in the greenhouse (¹³C percentage λ_previous), and additional C coming from the ¹³C-enriched substrate fed to the leaf (¹³C percentage λ_s). It has been previously shown that the contribution of new photosynthates to dark respiration after 3-h light in French bean is 40% (Nogués et al., 2004). So the ¹³C percentage in photosynthates feeding respiration is given by λ_p = 0.4 λ_fixed + 0.6 λ_previous. It should be noted that possible variations in the coefficients due to some physiological reasons do only introduce a minor error in the estimate of the ¹³C-enriched substrate decarboxylation r_night because of the strong ¹³C-enrichment in the substrate.

The total CO₂ production in the dark is denoted as R_n. The ¹³C percentage in dark respired CO₂ (denoted as λ_global) is calculated with the δ¹³C value and Equation 1. The ¹³C mass balance gives the following relationship:

(7)

Rearranging, it gives:

(8)

When substrates do not have a homogeneous isotopic distribution (positional enrichment), Equation 7 is completed to:

(9)

for the ¹³C-1-pyruvate feeding. And similarly, for the ¹³C-2-enrichment, it gives:

(10)

Equations 9 and 10 allow one to extract r_PDH and r_K with a substitution procedure. The method is similar for ¹³C-Glc.

RESULTS

On-Line Carbon Isotope Discrimination of Leaves Fed with ¹³C-Enriched Substrates

The carbon isotope discrimination during photosynthesis of detached French bean leaves before and after addition of ¹³C-enriched carbohydrates is shown in Figure 2. Before adding substrates, the carbon isotope discrimination was around 20‰ in all cases. This is in accordance with the p_i/p_a value around 0.7 (data not shown). At t = 60 min, ¹³C-enriched substrates were added. The overall signature of Glc was 5,500‰ and that of pyruvate was 2,500‰. When ¹³C₁-pyruvate was supplied, the carbon isotope discrimination then increased to Δ_obs = 180‰ as a consequence of ¹³C₁-enriched pyruvate decarboxylation. The carbon isotope discrimination then slightly decreased and reached approximately 140‰. The on-line carbon isotope discrimination value increased much less with ¹³C₂-enriched pyruvate, with a maximum value of Δ_obs around 50‰, indicating that decarboxylations following Pyr dehydrogenation (i.e. Krebs cycle) had very small rates. Glc was hardly decarboxylated in the light, thus with a very small increase in Δ_obs (up to 35‰–40‰ only).

Figure 2.

Left, Development of the on-line carbon isotope discrimination (Δ_obs) of detached bean leaves before and after feeding with Glc that is ¹³C-enriched in C-1 (⋄) or C-3 (□), or pyruvate that is ¹³C-enriched in C-1 (▾) or C-2 (▴). In both positional enrichment cases, the overall δ¹³C values of Glc and pyruvate are 5,500‰ and 2,500‰, respectively. The detached leaves are first put in distilled water and the light is turned on. The vertical dotted line indicates the moment at which the substrate (Glc or pyruvate) is supplied (t = 60 min). The horizontal dotted line represents the mean photosynthetic fractionation before substrate feeding (Δ_obs approximately equal to 20‰). The gas exchange conditions were 350 μL L⁻¹ CO₂ in 21% O₂, 22°C, and 450 μmol m⁻² s⁻¹ light. Note that the Δ_obs measurements only begin after photosynthesis stabilizes that is, after approximately 30 min in the light. At t = 180 min, the light is switched off for dark-respired CO₂ measurements. The trends of the on-line isotope discrimination with Pyr feeding are indicated with a solid line. Right, δ¹³C value (in per mil) of the CO₂ respired in the dark just after having switched off the light. Same symbols as for the left section.

Rates of Decarboxylation in the Light

The rates of decarboxylation of the ¹³C-enriched substrates (see Fig. 1 that summarizes the parameters considered) in the light were calculated using the carbon isotope discrimination Δ_obs and are given in Table I. These calculations used a two variable model (see “Theory” section) that is, r_PDH, the rate of decarboxylation through the PDH, and r_K, the decarboxylation through the Krebs cycle. These rates may be compared to the overall R_d. As expected from conclusions of Figure 2, the decarboxylation rate of pyruvate through the PDH reaction was around 0.05 to 0.06 μmol m⁻² s⁻¹, the other decarboxylation rates being very low, under 0.01 μmol m⁻² s⁻¹. Although the effect of decarboxylation of ¹³C₁-enriched Pyr was striking in Figure 2, the decarboxylation rate was low compared to day respiration (R_d, around 0.6 μmol m⁻² s⁻¹) simply because the labeling was strong in Pyr. It is noteworthy that the decarboxylation rates were negligible compared to the overall R_d, so that we argue that the respiratory pathway was not artefactually enhanced in our feeding experiment. Accordingly, the overall R_d of fed leaves was similar to that of the control. Nevertheless, the decarboxylation CO₂ flux could have been somewhat underestimated because of refixation (see below).

Table I.

Respiration rate and calculated decarboxylations (in μmol m⁻² s⁻¹) of Glc and pyruvate supplied to detached French bean leaves through the PDH and the Krebs cycle in the light and in the dark (see “Theory” section for calculation details)

Day respiration rates were measured as the slope of Γ/r_iR_n relationships (see “Gas Exchange Measurements” section) and are thus total day respiratory fluxes in the light. Decarboxylation data are mean and se of three independent measurements. The mean inhibition of decarboxylation by light was calculated as the ratio mean light decarboxylation/mean dark decarboxylation.

		Control	Pyruvate	Glc
Light
Rd	Day respiration rate	0.632 ± 0.125	0.625 ± 0.150	0.539 ± 0.143
rPDH	Decarboxylation from the PDH reaction		0.058 ± 0.009	0.004 ± 0.001
rK	Decarboxylation from the Krebs cycle		0.005 ± 0.004	0.009 ± 0.004
Darkness
Rn	Dark respiration rate	1.178 ± 0.039	1.266 ± 0.071	1.331 ± 0.115
rPDH	Decarboxylation from the PDH reaction		0.079 ± 0.023	0.070 ± 0.015
rK	Decarboxylation from the Krebs cycle		0.116 ± 0.011	0.236 ± 0.050
iPDH	Mean inhibition of the PDH reaction by light		27%	94%
iK	Mean inhibition of the Krebs cycle by light		95%	96%

Rates of Decarboxylation in Darkness

After 180 min (see Fig. 2), light was switched off and the carbon isotope composition of the CO₂ evolved in the first 30 min of darkness was measured. The δ¹³C value of respired CO₂ with ¹³C₁- and ¹³C₃-Glc feeding was approximately 740‰ and 640‰, respectively, and the δ¹³C value of respired CO₂ with ¹³C₁- and ¹³C₂-pyruvate feeding was approximately 520‰ and 290‰, respectively (Fig. 2, right). The decarboxylation rates calculated using these values are shown in Table I. r_PDH was around 0.070 μmol m⁻² s⁻¹ with either Glc or pyruvate and r_K with Glc was around the double (0.2 μmol m⁻² s⁻¹) of that with pyruvate. Clearly, these decarboxylations witness that leaves oxidized molecules via glycolysis in the dark. A similar result was already obtained by Stitt and ap Rees (1978) with ¹⁴CO₂. When compared to the day decarboxylation values, the inhibition by light was around 95% for both r_PDH and r_K with Glc and around 30% and 95% for r_PDH and r_K with pyruvate, respectively. In other words, light inhibited only partially the pyruvate dehydrogenase reaction and almost stopped the Krebs cycle. In the light, the respiratory breakdown of Glc into CO₂ did not occur through the Krebs cycle or the pyruvate dehydrogenase reaction, simply demonstrating that the Glc molecules could not reach these metabolic steps in illuminated leaves.

Metabolic Pathways That Consumed ¹³C-Enriched Glc

The weakness of Glc or pyruvate decarboxylation in the light raises the question of the fate of these molecules in the leaf. That is why starch purification on sample of the experiment of Figure 2 was made and NMR analysis of illuminated leaves fed with positional, fully ¹³C-labeled substrates (99% ¹³C in a given position) was done. The results are shown in Table II and Figure 3, respectively.

Table II.

Amount and carbon isotope composition (δ¹³C) of starch from detached leaves supplied with ¹³C-enriched Glc (δ¹³C 5,500‰) or ¹³C-enriched Pyr (δ¹³C 2,500‰), 15 mmol L⁻¹ for 2 h in the light with 350 μL L⁻¹ CO₂ at −51‰ (see Fig. 1) after 1 h in the light without feeding

Data are mean and sd of three independent measurements.

Conditions	Amount	δ13C
	μg mg−1 fresh weight	‰
Control (greenhouse)	13.1 ± 5.3	−28.5 ± 0.5
No feeding	15.1 ± 5.0	−31.4 ± 1.3
13C1-Glc	15.0 ± 3.3	−1.9 ± 9.0
13C3-Glc	13.6 ± 2.9	−9.0 ± 9.5
13C1-Pyr	11.7 ± 5.4	−16.5 ± 3.4
13C2-Pyr	12.2 ± 3.2	−29.9 ± 1.6

Figure 3.

Distribution of ¹³C from NMR analysis of leaves fed with ¹³C₁- or ¹³C₃-Glc (A and B), ¹³C₁- or ¹³C₂-pyruvate (C and D) for 2 h in the light at ambient CO₂. In each case, the compound supplied to leaves is indicated in the top left of each section. Simplified metabolic pathways are represented in order to emphasize the relationships between metabolites. Cit, Citrate; TP, triose phosphates; Org Ac, organic acids. The dashed line stands for the production of organic acids from pyruvate via the phosphoenolpyruvate carboxylation. The bottom continuous line in C and D stands for the refixation of decarboxylated CO₂. Carbon atoms are colored in accordance with their labeling level: black, no labeling, gray, low labeling, white, strong labeling. For each labeled carbon, numbers are in percent of ¹³C that is in the sample. Carbon atoms that are considered to be labeled are those which proportion of ¹³C (¹³C/¹²C + ¹³C) is equal or more than 2% (the natural abundance of ¹³C is approximately 1.1%).

¹³C-enriched Glc supplied to leaves was directed to Suc synthesis; the ¹³C content of Suc as a whole approached 60% in both (C-1 and C-3) enrichment conditions (Fig. 3, A and B). When ¹³C₃-Glc was used, the scrambling of carbon atoms appeared, with a small labeling of all the carbon atom positions in the Glc moiety of Suc (¹³C content around 4%). This presumably came from the pentose phosphates cycle. Accordingly, this effect was not seen with ¹³C₁-Glc because the first steps of the cycle are the dehydrogenation and decarboxylation of the C-1 carbon atom of Glc.

Surprisingly, starch was also labeled by Glc (Table II); with ¹³C₁-Glc and ¹³C₃-Glc, the δ¹³C value of starch was strongly higher (−1.9‰ and −9‰, respectively) than in unfed leaves (−31.4‰). In other words, Glc could feed starch synthesis. Its contribution should have nevertheless been low as the starch amount was not significantly different between fed and unfed leaves (Table II). The contribution of feeding Glc (in percent of total starch amount), denoted here as σ, can be calculated with the following mass-balance relationship:

that gives σ = 0.5% ± 0.2% and 0.4% ± 0.2% for ¹³C₁-Glc and ¹³C₃-Glc, respectively. This contribution was very low, but was clearly seen in starch (Table II) simply because the δ¹³C value of the labeling Glc was very high. This result indicates that Glc molecules from the cytoplasm (or, more generally, C from the fed Glc) could reach the chloroplastic compartment and enter the starch synthetic pathway.

The enriched carbon atoms of Glc were almost not redistributed to other positions by metabolic pathways; when supplied with ¹³C₁-Glc, the C-6 carbon atoms in leaf Glc, Fru, or Suc were not labeled (Fig. 3A). Similarly, when supplied with ¹³C₃-Glc, the C-4 carbon atoms in leaf Glc and Fru were not labeled and the corresponding positions in Suc were only weakly labeled (4% of the ¹³C content; Fig. 3B). In other words, the scrambling of carbon atoms through the triose phosphates and hexose phosphates interconversion (with triose phosphate isomerase and aldolase) was very low. This result is consistent with the gas exchange measurements of Figure 2 in which glycolysis appeared to be stopped so that Glc was not (or very weakly) decarboxylated.

Metabolic Pathways That Consumed ¹³C-Enriched Pyruvate

When leaves were fed with ¹³C₂-Pyr (99% of ¹³C in C-2), Ala was (weakly) labeled in C-2, indicating that Pyr had been aminated and citrate was labeled, strongly suggesting that some of the Pyr molecules entered the Krebs cycle (Fig. 3D). Surprisingly, when fed with ¹³C₁-Pyr, leaves had only a few labeled carbon atoms (Fig. 3C). Ala appeared as only weakly labeled in C-1, but this originated mainly from the low detectability of carboxyl (−COOH) carbon atoms with NMR. Moreover, the CO₂ produced in the light accounted for an important ¹³C loss; using a decarboxylation rate of 0.05 μmol m⁻² s⁻¹ (Table I), there was a ¹³C loss of 350 μmol ¹³C m⁻² during the light treatment (2 h),while approximately 1,950 μmol Pyr m⁻² entered the leaf (=Pyr concentration × transpiration rate = 15 mol L⁻¹ × 18 μL m⁻² s⁻¹ × 2 h light), that is, approximately 18% of the ¹³C supplied in C-1. A similar calculation gives, with ¹³C₂-Pyr, a ¹³C loss of only approximately 1%.

Refixation of Decarboxylated CO₂

The rate of the decarboxylation of ¹³C-enriched substrates in the light may have been underestimated by possible refixation of CO₂. If so, the absolute rate of refixation is expected to be important when ¹³C₁-Pyr was supplied to leaves as decarboxylation was high in that case (Table I). Nevertheless, Glc and Fru were not labeled and Suc was only weakly labeled in the C-3 position of the Fru moiety (Fig. 3C). Although it is difficult to quantify such small quantities, this labeling in Suc accounted, as a maximum, for 0.5 μmol g⁻¹ of ¹³C, that is, a ¹³C (re)fixation rate of 0.009 μmol m⁻² s⁻¹ during 2 h of the treatment in the light. When ¹³C₁-Pyr was supplied to leaves, starch was also labeled (the isotope signature is around −20‰, compared to −31‰ expected; Table II) and this was also a consequence of refixation. The amount of ¹³C-enriched carbon in starch was calculated to be 0.2% (similar calculation as for section “Metabolic Pathways That Consumed ¹³C-Enriched Glc”). The starch amount and the percentage of carbon in starch were 0.015 mg mg⁻¹ fresh weight and approximately 40%, respectively. This gives a refixation rate (of labeled carbon) of 0.008 μmol m⁻² s⁻¹. The overall refixation rate would thus have been, as a maximum, 0.017 μmol m⁻² s⁻¹. This means that the total decarboxylation rate of ¹³C₁-Pyr would have been 0.058 (see Table I) + 0.017 = 0.075 μmol m⁻² s⁻¹ and so, the inhibition of the PDH reaction by light would have been approximately 5% only. The effect of refixation in the other case (¹³C₂-Pyr) was clearly negligible: starch was not labeled (Table II) and no labeling could be seen in Suc, Glc, or Fru (Fig. 3D). As the rates of decarboxylation were similar with C-1 and C-3 ¹³C-enriched Glc, refixation was likely to be negligible in both cases.

DISCUSSION

The respiratory metabolism in illuminated leaves is inhibited compared to darkness, as repeatedly shown by gas-exchange measurements (for review, see Atkin et al., 2000). However, the metabolic basis of such an inhibition is not well known. We addressed this question by feeding experiments using ¹³C-enriched substrates and followed the ¹³C atoms with isotope ratio mass spectrometry and NMR to determine which metabolic pathways are inhibited in the light.

The Pyruvate Dehydrogenase Activity in the Light

The two main steps responsible for the respiratory CO₂ production are the dehydrogenation of Pyr (PDH reaction) and the Krebs cycle (Fig. 1). The carbon atoms that are decarboxylated are not the same in both: the PDH reaction decarboxylates the C-1 of Pyr, while the Krebs cycle decarboxylates the two others. Feeding illuminated leaves with ¹³C-enriched Pyr in C-1 significantly enriched the respiratory CO₂ produced in the light, as revealed by the strong modification of the photosynthetic carbon isotope discrimination measured on-line, from the steady value of 20‰ to approximately 160‰ (Fig. 2). Clearly, the PDH reaction consumes the Pyr molecules, and so is not totally inhibited in the light. When compared to the dark decarboxylation rate, the calculated inhibition is 27% (Table I). Refixation of decarboxylated CO₂ can nevertheless occur and lead to an overestimation of the inhibition level. Indeed, there was a small ¹³C enrichment in Suc (Fig. 3) as well as in starch (Table II) and a calculation (see “Results”) gives a refixation rate of 0.017 μmol m⁻² s⁻¹; that is, a total decarboxylation rate of Pyr of 0.075 μmol m⁻² s⁻¹. The refixation rate obtained here stands for 22% of the (total) decarboxylation of ¹³C-enriched Pyr, a value that is in accordance with calculations that use ¹⁴C data by Gerbaud and André (1987) on sunflower leaves (between 15% and 22%), but a little low compared to Pinelli and Loreto (2003) who found a refixation rate of 40% in mint leaves maintained at 350 μL L⁻¹. In this study, we looked at only ¹³C enrichment in the two main compounds of CO₂ fixation; that is, Suc and starch. Nevertheless, we recognize that some ¹³C atoms in other compounds were missed, leading to a small underestimation of the refixation rate. Taking our value as correct, this would give a light inhibition value of the PDH reaction of only 5%.

However, it is not possible in our experiment to distinguish the chloroplastic PDH activity from the mitochondrial one so that our values do represent the total cellular PDH activity. It has been shown that the mitochondrial PDH is partly inhibited in illuminated leaves by phosphorylation (Budde and Randall, 1990). In addition, photorespiratory produced NH₃ is also assumed to inhibit this enzyme (Krömer, 1995). By contrast, the chloroplastic PDH is not regulated by phosphorylation and is assumed to be active in the light (Plaxton, 1996). The decarboxylation rate measured in this study (Table I) is likely to be the sum of both PDH activity in which the chloroplastic enzyme plays the major role.

Furthermore, the PDH reaction may be artefactually activated by the amount of Pyr (thermodynamic mass action law) introduced in the leaf. We nevertheless think that this effect is minor; the pyruvate concentration in the feeding solution is only 15 mmol L⁻¹ and although it is difficult to detect carboxyl carbon atoms by NMR, it was not possible to see the corresponding C-1 atom of Pyr on the NMR spectrum, indicating that the Pyr concentration in the leaf is certainly less than 2 mmol L⁻¹ in the cell.

Glycolysis and the Krebs Cycle Are Inhibited in the Light

In contrast with ¹³C₁-Pyr, when leaves were fed with ¹³C₂-Pyr, the photosynthetic carbon isotope discrimination was hardly modified (Fig. 2), indicating that there was nearly no decarboxylation of the C-2 carbon atom of Pyr. In other words, the Krebs cycle activity was very low and so there is a very small ¹³CO₂ production. The calculated decarboxylation rate is 0.005 μmol m⁻² s⁻¹ only in the light; that is, the Krebs cycle is inhibited by 95% compared to the dark decarboxylation rate (Table I). This is consistent with the NMR results: (1) the intermediary products of the Krebs cycle such as succinate and fumarate are not detected, Glu (that is the amino acid that corresponds to α-ketoglutarate) was hardly labeled, (2) the two carboxyl atoms of citrate were ¹³C-labeled so that citrate stood for 15% of the sample ¹³C content, i.e. each carboxyl carbon atom represented 7.5% only of the ¹³C content. These data are in accordance with previous observations of Hanning and Heldt (1993) that mitochondria extracted from illuminated leaves had a low metabolic flux throughout the Krebs cycle. In addition, the mitochondrial matrix in the light is supposed to be reduced because of the photorespiratory Gly decarboxylation that lead to a high NAD(P)H/NAD(P)⁺ ratio. Some enzymes of the Krebs cycle are inhibited by the high NADH/NAD⁺ ratios (for review, see Siedow and Day, 2000), and it has been recently found that isocitrate dehydrogenase from pea is inhibited by high NADPH/NADP⁺ ratios, a feature that occurs in illuminated mitochondria (Igamberdiev and Gardeström, 2003).

The fact that the Krebs cycle is slowed down in the light raises the question of the fate of acetyl-CoA molecules produced by the PDH reaction. In our study, acetyl-CoA probably accumulated a little: the decrease in the photosynthetic on-line discrimination after 115 min (Fig. 2) suggests that the PDH reaction was less active. This was likely because of the retroinhibition of Pyr dehydrogenase by its product acetyl-CoA (Harding et al., 1970; Rapp et al., 1987), as the Krebs cycle activity that consumes acetyl-CoA was severely diminished in the light. A significant part of the acetyl-CoA molecules was directed to fatty acid production in the chloroplast (Ohlrogge and Jaworski, 1997). Accordingly, the mutant line of Arabidopsis (Arabidopsis thaliana) that produces the antisense RNA of the PDH kinase (thus enhancing the mitochondrial PDH reaction), accumulated ¹⁴C-labeled fatty acids when ¹⁴C-Pyr was fed to (photosynthetic) stems (Marillia et al., 2003), strongly suggesting that fatty acid synthesis can act as outfall for acetyl-CoA molecules.

¹³C-enriched Glc was only weakly decarboxylated in the light whatever the position of the ¹³C-enrichment was (Fig. 2). Thus, Glc molecules could hardly reach the PDH step, very likely because they could not enter glycolysis. Instead, the Glc molecules were directed to Suc synthesis (Fig. 3). Further, when fed with ¹³C₁-Glc, the Fru moiety in Suc was only very weakly labeled in C-6, and similarly, when fed with ¹³C₃-Glc, the Fru moiety in Suc was hardly labeled in C-4 (Fig. 3). Clearly, this shows that the C-1/C-6 and C-3/C-4 interconversion through the triose phosphates isomerase reaction was minor. In other words, only a small fraction (less than approximately 5% of the ¹³C-Glc fed to the leaf) of the Glc molecules reached this step, strongly suggesting a high metabolic resistance to the glycolytic breakdown of hexoses. Noteworthy, these results are in accordance with the regulation of the enzymes responsible for the phosphorylation/dephosphorylation of Fru-6-P to Fru-2,6-bisphosphate: (1) in the light, the high triose phosphates/inorganic phosphate decreases the Fru-2,6-bisphosphate concentration, promoting the dephosphorylation of Fru-1,6-bisphosphate in the cytosol (Stitt, 1990), and (2) in the chloroplast, phosphofructokinase is thought to be inhibited in the light (Plaxton, 1996). The rationale of such enzymatic regulations is that hexose molecules are prevented from entering the glycolytic breakdown, and so respiration does not consume Suc as soon as it is synthesized in the cytoplasm.

Cytoplasmic Glc Contributes to Starch Synthesis in the Light

When fed with ¹³C-enriched Glc (at 5,500 per mil), leaves produced ¹³C-labeled starch (Table II), with a δ¹³C value around 30‰ higher than without Glc feeding, indicating that Glc contributed to starch synthesis. The contribution was nevertheless low, around only 0.5% of the starch amount (that is, approximately 0.075 μg mg⁻¹ fresh weight) was labeled (see the “Result” section for calculations). The flux of starch synthesis from ¹³C-enriched Glc during the 2 h of feeding treatment in the light was then 0.075 μg mg⁻¹/2 h ∼ 0.038 μg mg⁻¹ h⁻¹; that is, 0.045 μmol C m⁻² s⁻¹, while the total starch synthetic flux was around 2 μg mg⁻¹/3 h ∼ 0.7 μg mg⁻¹ h⁻¹ (Table II); that is, 0.84 μmol C m⁻² s⁻¹. So, Glc fed 0.045/0.84 ∼ 5% of starch synthesis in the leaf during the feeding experiment. The rate of starch synthesis from ¹³C-enriched Glc is consistent with that found in potato by Quick et al. (1995): chloroplasts extracted from intact leaves and illuminated with ¹⁴C-Glc (in a medium containing HCO₃⁻) synthesized starch from ¹⁴C-Glc with a rate of approximately 0.04 μmol C m⁻² s⁻¹ (recalculated assuming a realistic chlorophyll amount of 0.4 g Chl m⁻²).

It has also been found that other carbon sources than photosynthetic CO₂ can feed starch synthesis in intact illuminated leaves: Nogués et al. (2004) showed that it was not possible to completely label starch with CO₂ and while the starch amount increased, the δ¹³C value of starch reached a plateau.

These observations might be paralleled with the presence of a diffusion-driven Glc phosphate translocator on the inner chloroplast membrane (Schäfer et al., 1977; Quick et al., 1995), which can feed starch synthesis with cytoplasmic hexose molecules, although the associated flux appears to be quantitatively minor in this study.

Concluding Remarks

To our knowledge, this paper describes the first in vivo study on the metabolic basis of inhibition by light of leaf respiration. Clearly, the main inhibited steps are the entrance of hexose molecules into the glycolytic pathway and the Krebs cycle. We nevertheless recognize that our experiments were made in typical conditions (21% O₂, 350 μL L⁻¹ CO₂, 22°C) and the results may be influenced by environmental parameters. Indeed, it has been shown that there lies no inhibition of respiration by light in some physiological conditions (Sharp et al., 1984, and refs. therein). Two parameters are of particular interest: temperature, which is known to enhance respiratory enzymatic activities, and the photorespiratory rate (oxygen partial pressure), which has an effect on the redox status of the mitochondria. Further experimental data are now needed to investigate the effect of such environmental conditions on day respiratory metabolism. Moreover, the fact that the Krebs cycle appears to be slowed down in the light raises the question of energy in leaf cells and more precisely, of NADH feeding of the respiratory chain. Although one may suggest that photorespiration or the cytoplasmic-mitochondrial malate shuttle have such a role, further metabolic studies are needed to determine their respective contribution in vivo.

MATERIALS AND METHODS

Plant Material

French bean (Phaseolus vulgaris) L. cv Contender plants were grown from seed in 1-L pots of potting mix in a greenhouse, as described by Tcherkez et al. (2003). Minimum photosynthetic photon flux density during a 16-h photoperiod was maintained at approximately 500 μmol m⁻² s⁻¹ by supplemental lighting. Temperature and vapor pressure deficit were maintained at approximately 25.5°C/18.5°C and 1.4/1.2 kPa day/night, respectively. Carbon isotope composition (δ¹³C) of CO₂ in the greenhouse air was −9.5‰ ± 0.3‰. The first trifoliar fully expanded leaves were used for all measurements.

Gas Exchange Measurements

Closed System (Dark Respiration)

The respiration chamber was placed in a closed system, which was directly coupled to an elemental analyzer (EA) NA-1500 (Carlo-Erba, Milan) through a 15-mL loop, as described by Tcherkez et al. (2003). Briefly, molar fractions of respiratory CO₂ were measured with an infrared gas analyzer (IRGA; Finor, Maihak, Germany) placed in the closed system that was first flushed with CO₂-free air. The loop was shunted when CO₂ reached around 300 μL L⁻¹ and the gas inside the loop was introduced into the EA with helium for gas chromatography. The connection valve between the elemental analyzer and the isotope ratio mass spectrometer (VG Optima, Micromass, Villeurbanne, France) was opened when the CO₂ peak emerged from the EA.

Open System (Photosynthesis and On-Line Carbon Isotope Discrimination)

The assimilation chamber was connected in parallel to the sample air hose of the LI-6400 (LI-COR, Lincoln, NE). This aluminum chamber ([20 × 12 × 6] 10⁻⁶ m³) had a clear Plexiglas lid that allowed us to accommodate the middle leaflet (typical leaf surface approximately 0.01 m²). Two fans placed in the chamber gave a boundary layer conductance to water of approximately 6.7 mol m⁻² s⁻¹. Leaf temperature was controlled at 20°C with circulating water from a cooling water bath to the jacket of the leaf chamber and was measured with a copper-constantan thermocouple plugged to the thermocouple sensor connector of the LI-6400 chamber/IRGA. Ingoing air was dried (at approximately 1 mmol H₂O mol⁻¹) and passed through the chamber at a rate of 1 L min⁻¹, monitored by the LI-6400. Molar fractions of CO₂ were measured with the IRGA of the LI-6400. Light was supplied by a 500-W halogen lamp (Massive N. V., Kontich, Belgium). The lamp was placed about 30 cm above the chamber and 5 cm of deionized water and 1 cm of glass in the container filtered the radiation. The photosynthetic photon flux density at leaf level inside the chamber was maintained at 450 μmol m⁻² s⁻¹ during the labeling period. Inlet CO₂ was obtained from a gas cylinder (Air Liquide, Grigny, France) with a δ¹³C of −51.2‰ ± 0.2‰. The outlet air of the chamber was regularly shunted and was sent to the loop to measure its isotope composition and thus the on-line carbon isotopic discrimination (Δ_obs). The gas inside the loop was introduced into the EA for gas chromatography as described above. Δ_obs during photosynthesis was measured following the method described by Evans et al. (1986; see “Theory”).

Day Respiration Measurements

Day respiration was measured with a LICOR-6400 open system, according to the method described in Peisker and Apel (2001). Briefly, A/c_i curves are done at different light levels and the linear fit gives the CO₂ compensation point Γ and the internal leaf resistance r_i. Γ is plotted as a function of the product r_iR_n and the slope of the linear regression is μ = R_d/R_n and so gives R_d. In bean, the effect of light on the Γ/r_i relationship was slight and did not modify significantly the estimate of μ.

Starch Extraction Procedure

The extraction procedures for starch were similar to that described by Tcherkez et al. (2003). Leaf powder was suspended with 1 mL of distilled water in an Eppendorf tube (Eppendorf Scientific, Hamburg, Germany). After centrifugation, the pellet was washed four times with 95% ethanol at room temperature and starch was extracted with HCl solubilization and precipitated with cold methanol. After lyophilization, starch is transferred to tin capsules (Courtage Analyze Service, Mont Saint-Aignan, France) for isotope analysis.

NMR

Perchloric acid (PCA) extracts were prepared from 5 g of frozen leaf material as described by Aubert et al. (1996) for phloem cells. Spectra were obtained on a spectrometer (AMX 400) equipped with a 10-mm multinuclear probe tuned at 161.9 and 100.6 MHz for ³¹P- and ¹³C-NMR, respectively. The deuterium resonance of ²H₂O (100 μL added per mL of extract) was used as a lock signal.

Conditions for ¹³C-NMR acquisition utilized 19-μs pulses (90°) at 6-s intervals and a sweep width of 20 kHz. Broad-band decoupling at 2.5 W during acquisition and 0.5 W during the delay was applied using the Waltz sequence; the signal was digitized using 32,000 data points zero-filled to 64,000 and processed with a 0.2-Hz line broadening. ¹³C-NMR spectra are referenced to hexamethyldisiloxane at 2.7 ppm. Mn²⁺ ions were chelated by the addition of 1 mmol L⁻¹ 1,2-cyclohexylenedinitrilotetraacetic acid. The assignments of resonance of ¹³C peaks were carried out according to Gout et al. (1993). Identified compounds were quantified from the height of their resonance peaks using fully relaxed conditions for spectra acquisition (pulses at 20-s intervals). Peak intensities were normalized to a known amount of the reference compound (maleate) that is added to the sample (internal standard). A carbon atom is here considered to be labeled when its estimated positional ¹³C proportion ¹³C/(¹³C + ¹²C) is more than 2% (the natural abundance is nearly 1.1%).

¹³C-Enriched Molecules

The positional ¹³C-labeled molecules (99% ¹³C in the considered position) were purchased to Eurisotop (Saclay, France). Pyruvate was dissolved in distilled water and pH was corrected to 6.8 with NaOH. To obtain non-fully labeled solutions, the labeled compounds were mixed with industrial Glc (δ¹³C = −9‰) or pyruvate (δ¹³C = −21‰) from Sigma. The resulting overall composition of the Glc and pyruvate solutions was checked to be 5,500‰ and 2,500‰, respectively. In other words, the ¹³C-enriched carbon atom position had a composition of 47,750‰ and 8,000‰, respectively (the other positions being at −9‰ and −21‰ for Glc and Pyr, respectively). The final concentration was 0.015 mol L⁻¹ in all cases. The solutions were poured in an Eppendorf tube and fed to the leaves through the transpiration stream.