Metabolic Fate of the Carboxyl Groups of Malate and Pyruvate and their Influence on δ¹³C of Leaf-Respired CO₂ during Light Enhanced Dark Respiration

Abstract

The enhanced CO₂ release of illuminated leaves transferred into darkness, termed “light enhanced dark respiration (LEDR)”, is often associated with an increase in the carbon isotope ratio of the respired CO₂ (δ¹³C_LEDR). The latter has been hypothesized to result from different respiratory substrates and decarboxylation reactions in various metabolic pathways, which are poorly understood so far. To provide a better insight into the underlying metabolic processes of δ¹³C_LEDR, we fed position-specific ¹³C-labeled malate and pyruvate via the xylem stream to leaves of species with high and low δ¹³C_LEDR values (Halimium halimifolium and Oxalis triangularis, respectively). During respective label application, we determined label-derived leaf ¹³CO₂ respiration using laser spectroscopy and the ¹³C allocation to metabolic fractions during light–dark transitions. Our results clearly show that both carboxyl groups (C-1 and C-4 position) of malate similarly influence respiration and metabolic fractions in both species, indicating possible isotope randomization of the carboxyl groups of malate by the fumarase reaction. While C-2 position of pyruvate was only weakly respired, the species-specific difference in natural δ¹³C_LEDR patterns were best reflected by the ¹³CO₂ respiration patterns of the C-1 position of pyruvate. Furthermore, ¹³C label from malate and pyruvate were mainly allocated to amino and organic acid fractions in both species and only little to sugar and lipid fractions. In summary, our results suggest that respiration of both carboxyl groups of malate (via fumarase) by tricarboxylic acid cycle reactions or by NAD-malic enzyme influences δ¹³C_LEDR. The latter supplies the pyruvate dehydrogenase reaction, which in turn determines natural δ¹³C_LEDR pattern by releasing the C-1 position of pyruvate.

Introduction

Transferring light-acclimated leaves into darkness lead to rapid changes in various metabolic processes, causing halt of any photosynthetic activity and increased respiration rates (Florez-Sarasa et al., 2012; Griffin and Turnbull, 2012). During such light–dark transitions, leaves show enhanced O₂ consumption (Stone and Ganf, 1981; Azcon-Bieto et al., 1983) concurrently with enhanced CO₂ release (Hill and Bryce, 1992; Atkin et al., 1998). The extent of this increasing CO₂ emissions after darkening, defined as “light enhanced dark respiration (LEDR)”, depends on light intensity of the previous photoperiod and lasts about 20–30 min, with a maximum CO₂ release within the first 5 min (Atkin et al., 1998; Griffin and Turnbull, 2012). LEDR does not interfere with the post-illumination burst, which occurs within <20 s after darkening and is related to photorespiratory glycine oxidation (Atkin et al., 1998).

During LEDR the carbon isotope ratio of leaf-respired CO₂ (δ¹³C_LEDR) is also changing, showing a strong initial increase which is followed by a progressive decrease (Werner et al., 2009; Wegener et al., 2010; Lehmann et al., 2016). Such a δ¹³C_LEDR pattern was also observed under natural conditions in a grassland ecosystem (Barbour et al., 2011), demonstrating that this phenomenon has also potential relevance at higher scales. δ¹³C_LEDR is also relevant for cases when leaf dark-respired CO₂ is sampled from light-acclimated leaves. Many field (Prater et al., 2006; Sun et al., 2009; Dubbert et al., 2012) and laboratory experiments (Wegener et al., 2010; Lehmann et al., 2015, 2016) have shown that δ¹³C_LEDR is often less negative compared to respiratory substrates. The ¹³C enrichment in leaf-respired CO₂ during LEDR progressively increases during daytime, which has been shown to be strongly related to the cumulative assimilation of CO₂ (Hymus et al., 2005). Moreover, δ¹³C_LEDR may retain information of plant internal carbon allocation and was observed to differ among plant functional groups (Wegener et al., 2010, 2015): evergreen, slow-growing or aromatic species showed a very high increase in δ¹³C_LEDR during the day (up to 14.8‰), while fast-growing, non-aromatic herbaceous species showed no or a small increase in δ¹³C_LEDR (Lehmann et al., 2016). However, although leaf dark-respired CO₂ is a potential and easy accessible indicator for leaf internal metabolic processes, high-resolution measurements of δ¹³C_LEDR are scarce (Werner et al., 2009; Jardine et al., 2014). Also the interpretation of this data is still difficult, since its respiratory substrates and associated enzymatic reactions are not fully understood so far (Werner and Gessler, 2011; Ghashghaie and Badeck, 2014).

δ¹³C_LEDR may depend on the heterogeneous intramolecular isotope distribution within respiratory substrates (e.g., sugars). For instance, the C-3 and C-4 positions of glucose are enriched in ¹³C compared to other positions of the same molecule (Rossmann et al., 1991; Gilbert et al., 2012), probably due to an equilibrium isotope effect of the aldolase reaction (Gleixner and Schmidt, 1997). On the one hand, ¹³C enriched positions of glucose molecules broken down during glycolysis are relocated to the C-1 position of pyruvate molecules and can be released as CO₂ by the mitochondrial pyruvate dehydrogenase (mtPDH) reaction in the dark, influencing δ¹³C_LEDR. Additionally, the chloroplast pyruvate dehydrogenase (cpPDH) reaction can consume photosynthetically produced pyruvate in the light (Tovar-Mendez et al., 2003). On the other hand, ¹³C depleted positions of glucose molecules relocated in the acetyl-CoA residue (C-2 and C-3 position of pyruvate) can be used in dark for respiration or for biosynthesis of various substrates such as amino acids, lipids, or organic acids (Tcherkez et al., 2009). Leaf feeding studies showed that the extent of ¹³CO₂ respiration from ¹³C-1 pyruvate was higher than from pyruvate molecules labeled in other positions (Tcherkez et al., 2005, 2009). Other studies indicated that species-specific differences in ¹³CO₂ respiration from ¹³C-1 pyruvate might explain the species-specific differences in δ¹³C_LEDR (Priault et al., 2009; Wegener et al., 2010).

Furthermore, also other respiratory substrates are expected to have a significant influence on δ¹³C_LEDR. A study by Gessler et al. (2009) showed a clear malate breakdown shortly upon darkening in Ricinus communis, estimating that malate decarboxylation represents 22% of the respired CO₂ during LEDR. The role of malate during LEDR was supported by a more recent study from Lehmann et al. (2015) who identified malate as an important and relatively ¹³C enriched respiratory substrate in Solanum tuberosum. The biochemical link between malate and δ¹³C_LEDR was explained with an anaplerotic flux via the PEPC reaction that replenishes TCA cycle intermediates, which are withdrawn for biosynthetic use (Werner et al., 2011; Lehmann et al., 2015). This is in line with other studies, suggesting that organic acids such as malate derived from PEPC may accumulate in leaves under illumination due to inhibition of enzymatic reactions in the mitochondrion such as fumarase, NAD-dependent isocitrate dehydrogenase (mtIDH), and 2-oxoglutarate dehydrogenase (OGDH) (Werner et al., 2011; Eprintsev et al., 2016), causing an open and only partially active TCA cycle (Hanning and Heldt, 1993; Tcherkez et al., 2005; Araújo et al., 2012). Thus, a rapid breakdown of malate via TCA recycling in the dark may be in support of enhanced CO₂ release during LEDR and simultaneous increases in δ¹³C_LEDR, but also other individual organic acids might be involved (Lehmann et al., 2016). In addition, malate derived from PEPC via oxaloacetate and the malate dehydrogenase reaction (MDH) is expected to be enriched in ¹³C at the C-4 position compared to other molecule positions (Melzer and O’Leary, 1987; Savidge and Blair, 2004), since PEPC fixes HCO₃^- with a net isotope fractionation of -5.7‰ (with respect to dissolved CO₂; Farquhar, 1983). Thus, decarboxylation of this malate molecule position in the dark by the previously light-inhibited mitochondrial NAD-ME reaction, as observed in Hordeum vulgare protoplasts (Hill and Bryce, 1992; Igamberdiev et al., 2001), might also contribute to the observed increase in δ¹³C_LEDR (Barbour et al., 2007; Gessler et al., 2009; Werner et al., 2011; Tcherkez et al., 2012), although NAD-ME might fractionate against ¹³C by about 20‰ as observed in Crassula plants (Grissom et al., 1987). Direct experimental evidence for respiration of malate is still missing and thus the respiratory and metabolic fate of the carboxyl groups (C-1 and C-4 position) of malate during light–dark transitions and their influence on δ¹³C_LEDR remains to be proven.

Here, we hypothesize that δ¹³C_LEDR is determined by decarboxylation of malate and pyruvate, but we expect position-specific and species-specific differences in the intensity of these reactions. Our main objectives are (1) to determine if and to which extent malate and pyruvate influence respiration during light–dark transitions, (2) to identify position- and species-specific differences of both substrates in respiration, (3) and in ¹³C allocation to metabolic fractions. Therefore, we fed different position-specific ¹³C-labeled malate (¹³C-1, ¹³C-4) and pyruvate (¹³C-1, ¹³C-2) via the xylem stream to leaves of two different species, with known high (Halimium halimifolium) and low (Oxalis triangularis) increases in δ¹³C_LEDR. We measured the R_Label during light–dark transitions on-line with high time-resolved isotope laser spectroscopy and determined the ¹³C allocation to leaf metabolic fractions using isotope ratio mass spectrometry.

Materials and Methods

Plant Material

We chose species from two functional groups as described in Priault et al. (2009): the evergreen Mediterranean shrub Halimium halimifolium L. and the fast-growing herb Oxalis triangularis A. St.-Hil. While H. halimifolium plants were potted in 5 L pots filled with sand (plant height 40–60 cm), O. triangularis plants were potted in 1 L pots with potting soil (plant height 10–15 cm). Both species have a similar LEDR duration, with the main CO₂-peak declining within approximately 30 min after darkening followed by a slow decrease for another 30 min (Lehmann et al., 2016). All plants were grown under controlled conditions in walk-in climate chambers with a constant temperature of 23°C and relative humidity of about 60% during a 12 h light period (09:00–21:00 h) with a photosynthetic photon flux density (PPFD) of about 770 μmol m^-2 s^-1.

Acquiring of Position-Specific ¹³C-Labeled Malate and Pyruvate

¹³C-1 and ¹³C-4 labeled malate substrates were synthesized with coupled enzymatic reactions modified after Rosenberg and O’Leary (1985): 4.4 mmol NADH disodium salt (Roth, Arlesheim, Switzerland), 2.8 mmol 2-oxoglutarate (Sigma–Aldrich, Buchs, Switzerland), 2.9 mmol ¹³C-1 or ¹³C-4 aspartate (99%, both Sercon, Crewe, UK), and 100 U glutamate-oxaloacetate transaminase (Roche, Rotkreuz, Switzerland) were dissolved in 50 ml of 0.2 M KH₂PO₄ (pH 7.5) buffer solution. Subsequently, pH 7.5 was readjusted with 11 ml of 1 M KOH and the reaction solution continuously stirred for 10 min at 25°C. The enzymatic reaction was then started by adding 100 U of a malate dehydrogenase solution (Roche, Rotkreuz, Switzerland). Aliquots of the reaction solution were analyzed at 340 nm with a 96-well microplate reader (ELx800, BioTek, Luzern, Switzerland) to follow the NADH degradation, which ended after 2 h. Thereafter, all enzyme residues of the reaction solution were removed with pre-washed centrifugation filters with a molecular weight cut-off of 5000 da (Vivaspin 15R, 5000 MWCO HY, Sartorius, Göttingen, Germany). Subsequently, aliquots of the reaction solutions were separated by ion exchange chromatography using Dowex material (see below). Malate containing fractions of the reaction solution were eluted from Dowex 1X8 columns with 40 ml 1 M HCl, frozen (-20°C), and lyophilized. Finally, pellets were re-suspended in deionized water and aliquots merged to one labeling solution. Synthesis of ¹³C-1 and ¹³C-4 labeled malate was verified by NMR (Supplementary Figure S1), showing the appropriate position-specific ¹³C-labeling, with only marginal impurities. ¹³C-1 pyruvate (Cambridge Isotope Laboratories, Tewksbury, MA, USA) and ¹³C-2 pyruvate (Sigma-Aldrich, Buchs, Switzerland) were commercially acquired.

Experimental Setup for Laser-Based On-line Measurements

Leaf ¹³CO₂ respiration was determined on-line by a cavity ringdown laser spectrometer (CRDS, G2101-I, Picarro, Santa Clara, CA, USA). The CRDS system holds a wavelength monitor, which quantifies the spectral signature of CO₂ isotopologs with a time-based measurement technique in an optical cavity at 1603 nm. Temperature and pressure within the measurement cavity were controlled at 40°C and 140 Torr, respectively. Data were monitored continuously with a temporal resolution of 0.75 Hz, thus 45 single values were averaged per minute for further analysis.

For all laser-based on-line gas exchange measurements, a transparent 500 ml glass cylinder, with one inlet and one outlet, was used as a plant chamber that enclosed twigs or leaves. Inlets of all plant chambers were flushed continuously with fresh air at a molar mass flow rate of 640 μmol s^-1. Outlets were connected via Teflon tubing to the CRDS, thus determining ¹³CO₂ of sample gas (¹³CO_2SG). For ¹³CO₂ measurements of reference gas (¹³CO_2RG), an empty plant chamber of the same size was used. The open bottom side of all plant chambers was sealed with airtight non-fractionating plastic foil (FEP, 4PTFE, Stuhr, Germany). Before and after each ¹³CO_2SG measurement of about 40–60 min, ¹³CO_2RG was analyzed for about 10–20 min. ¹³CO_2RG concentrations (μmol ¹³CO₂ mol^-1) were interpolated with a generic regression equation (y = mx + b) for the period of ¹³CO_2SG measurements. Switching between ¹³CO_2SG and ¹³CO_2RG was done manually by re-plugging the CRDS Teflon tubing. ¹³CO₂ readings were discarded within the first minutes after switching between gases. Compressed air from gas cylinders (4.4 ppm ¹³CO₂, Riessner-Gase, Lichtenfels, Germany) was analyzed two times per day for about 10–20 min, showing a total variation of SD ≤ 0.008 ppm in ¹³CO₂, thus no significant drift during the experiments occurred.

During the experiments, one twig with leaves of H. halimifolium or three leaves of O. triangularis still attached to the potted plant were enclosed in the plant chambers. The system was sealed with airtight plastic foil around chamber and stalk of the twigs or leaf petioles. Respiration rates of twigs and leaves were measured in the first 10–20 min to determine the non-labeled leaf ¹³CO₂ respiration. Subsequently, twigs or leaves within the plant chamber were detached from the plant by cutting and immediately transferred into tap water and cut again under water to prevent air embolism in the xylem. The truncated ends of these twigs and leaves were then directly transferred in reaction vials, which contained the ¹³C-labeling solutions and were placed outside of the plant chamber. The ¹³C-labeling solutions were refilled during the experiments to provide continuously ¹³C-labeled substrates to the plant. Subsequently, ¹³CO₂ respiration was measured for about 20 min in the light and 20 min in the dark by covering the plant chamber with a light-impermeable cloth. Three to five replicates were measured with each position-specific ¹³C-labeled substrate for each species. Leaf area of all leaves was determined after each experiment.

Tests with dilution series of ¹³C-labeled malate (¹³C-1, ¹³C-4) in H. halimifolium and in O. triangularis showed that 2 mM malate solutions produced sufficient ¹³CO₂ emissions and thus this concentration was applied for all further experiments. For ¹³C-labeled pyruvate (¹³C-1, ¹³C-2), 6 mM solutions yielded the best signals. Differences in molar strength by a factor of three for the labeling solutions were accounted for when computing the R_Label and the Δ_Label (see Eqs 1 and 3). All ¹³C-labeled substrates were purely provided without isotopic dilution by unlabeled substrates. No significant changes in total CO₂ respiration (¹²CO₂ + ¹³CO₂) occurred after feeding plants of both species with any position-specific ¹³C-labeled substrate, indicating that the applied concentrations of malate and pyruvate solutions had no significant effect on respiratory CO₂ emissions. Transpiration rates were monitored by CRDS and checked to be in steady-state during substrate application. Mean transpiration rates during the light in H. halimifolium (1.06 ± 0.13 mmol m^-2 s^-1) and O. triangularis (0.89 ± 0.04 mmol m^-2 s^-1) showed no significant differences for all ¹³C-labeled substrates (both species, p > 0.05; ANOVA + Tukey-HSD), indicating that absorption rates were similar for all experiments in both species. Air temperature and PPFD within the plant chambers during the measurements was about 29°C and 720 μmol m^-2 s^-1, respectively.

Equation 1 was used to calculate R_Label, which was corrected for non-labeled leaf ¹³CO₂ respiration:

$$R_{Label}=\left(\frac{f{(}^{13}{CO}_{2_{SG}}{-}^{13}{CO}_{2_{RG}})}{a}\right)-R_{NL}$$ (1)

where, f is the molar mass flow rate (μmol s^-1), a the leaf area (m²), ¹³CO_2SG and ¹³CO_2RG the sample and the reference gas (μmol ¹³CO₂ mol^-1), and R_NL the non-labeled leaf ¹³CO₂ respiration (μmol ¹³CO₂ mol^-1).

Eq. 2 was used to calculate sums of ¹³CO₂ respiration (∫ _R) for a distinct period:

$${\displaystyle {\int }_R}={\displaystyle \sum _{i=1}^n{R}_{Label}}$$ (2)

where n is the number of seconds in the light or in the dark.

Isotopic Analysis of Leaf Metabolic Fractions

Similar to the above-described experiments regarding respiration, an additional experiment was carried out to determine ¹³C allocation of position-specific ¹³C-labeled malate and pyruvate to metabolic fractions. One twig with leaves of H. halimifolium or three leaves of O. triangularis were placed into reaction vials containing one of the four different position-specific ¹³C-labeled substrates or deionized water to correct for non-labeling conditions. To trace the ¹³C allocation in leaf metabolic fractions during light dark-transitions, leaves of both species were harvested after 20 min in the light, as well as after 40 min (20 min in the light and 20 min in the dark), using additional twigs/leaves of the same plants. Subsequently, leaves from both species were immediately frozen with liquid N₂, lyophilized, and milled to a fine powder with a ball mill. 100 mg of the plant powder was dissolved in 1.5 ml MCW (methanol, chloroform, water, 12:3:5, v:v:v) and boiled for 30 min at 70°C in a water bath to extract the water soluble fraction, as described in Richter et al. (2009; method I2). Samples were centrifuged for 2 min at 10000 × g, and 800 μl of the supernatant were transferred into a new reaction vial. After adding 250 μl chloroform, samples were shaken intensively and centrifuged for 2 min at 10000 × g. For the isotopic analysis of lipids, aliquots of the lower chloroform phase were carefully taken to avoid contamination with the upper phase and transferred into tin capsules. In a next step, 1.2 ml of the upper phase was transferred into a new reaction vial, mixed roughly with 500 μl chloroform, and centrifuged 2 min at 10000 × g. Finally, 1 ml of the upper phase, which contained the total hydrophilic fraction of the extract, was transferred into a new reaction vial and stored at -20°C.

The hydrophilic fraction was further separated to sugar, amino acid, and organic acid fractions with ion-exchange chromatography using Dowex material (100–200 mesh, Sigma–Aldrich, Buchs, Switzerland; see Richter et al., 2009). In brief, different Dowex materials were conditioned with 1M HCL (Dowex 50WX8) or with 1M sodium formate (Dowex 1X8) and transferred to columns, which were placed above each other on a rack. Columns were extensively prewashed with deionized water to fully remove potential carbon residues. Subsequently, the hydrophilic fraction was added and the neutral, sugar containing fraction was eluted with 35 ml deionized water, while the flow-through was collected in 50 ml falcon tubes. The amino acid fraction was absorbed by Dowex 50WX8 and eluted with 30 ml 3M NH₃, while the organic acid fraction was absorbed by Dowex 1X8 and eluted with 35 ml 1M HCL. Eluates of all three fractions were frozen at -20°C, lyophilized, and the pellets re-suspended in deionized water. Subsequently, aliquots of the fractions were transferred into tin capsules. All capsules (including lipids) were oven-dried at 60°C and δ¹³C values analyzed with an EA-IRMS (Thermo Fisher Scientific, Bremen, Germany). Measurements and referencing for the IRMS were done after Werner et al. (1999) and Werner and Brand (2001). The IRMS long-term precision of a quality control standard (-43.11‰, Caffeine, Fluka, Buchs, Switzerland) was ≤0.15‰ (SD).

Equation 3 was used to calculate the Δ_Label in leaf metabolic fractions, which was corrected for δ¹³C values under non-labeling conditions:

$${\Delta }_{Label}= {\delta }^{13}{C}_L- {\delta }^{13}{C}_{NL}$$ (3)

where δ¹³C_L are the δ¹³C values in leaf metabolic fractions (lipids, sugars, amino and organic acids) of plants fed with position-specific ¹³C-labeled substrates, and δ¹³C_NL is the mean δ¹³C value of the corresponding leaf metabolic fraction under non-labeling conditions.

Statistics

General linear models were performed to test for position-specific, species-specific, and temporal differences in ∫ _R and Δ_Label of different metabolic fractions. One-way analysis of variance (ANOVA) and Tukey-HSD post hoc tests were used to show differences in Δ_Label of different metabolic fractions for each ¹³C-labeled substrate. Unless otherwise specified, means and standard errors are given. All statistical analyses were performed in R (version 3.1.3; R Core Team, 2015).

Results

Leaf ¹³CO₂ Respiration of Different Position-Specific ¹³C-Labeled Substrates and Species

To identify the respiratory substrates and enzymatic reactions, determining the species-specific differences in δ¹³C_LEDR at natural isotope abundances in H. halimifolium (high increases in δ¹³C_LEDR) and in O. triangularis (low increases in δ¹³C_LEDR), we fed those plants with different ¹³C-labeled malate and pyruvate substrates, expecting position- and species-specific differences in ¹³CO₂ respiration during light–dark transitions.

¹³CO₂ respiration rates (R_Label) from ¹³C-1 and ¹³C-4 malate were low in the light, but steeply increased shortly after darkening during LEDR, with a peak of about 0.02 μmol ¹³CO₂ m^-2 s^-1 (Figures 1 and 2). Clear light–dark differences for sums of the ¹³CO₂ respiration rates (∫ _R) reflect the observed R_Label patterns (Tables 1 and 2). Furthermore, ∫ _R of malate substrates showed no significant position-specific and species-specific difference, showing that ¹³CO₂ respiration from ¹³C-1 malate did not differ significantly compared to ¹³CO₂ respiration from ¹³C-4 malate in both species during light–dark transitions (Tables 1 and 2).

FIGURE 1.

Label-derived leaf ¹³CO₂ respiration (R_Label, μmol ¹³CO₂ m^-2 s^-1) during light–dark transitions in Halimium halimifolium. Twigs with leaves were fed with position-specific ¹³C-labeled malate (¹³C-1, ¹³C-4) or pyruvate (¹³C-1, ¹³C-2) for about 40 min (20 min in the light and 20 min in the dark). Gray areas denote dark periods. Black bold lines denote mean values, while symbols indicate single replicates (n = 3–5 individuals).

FIGURE 2.

Label-derived leaf ¹³CO₂ respiration (R_Label, μmol ¹³CO₂ m^-2 s^-1) during light–dark transitions in Oxalis triangularis. Leaves were fed with position-specific ¹³C-labeled malate (¹³C-1, ¹³C-4) or pyruvate (¹³C-1, ¹³C-2) for about 40 min (20 min in the light and 20 min in the dark). Gray areas denote dark periods. Black bold lines denote mean values, while symbols indicate single replicates (n = 3–4 individuals).

Table 1.

Sums of ¹³CO₂ respiration (∫ _R) derived from different ¹³C-labeled substrates.

Species	Label	∫ R-light	∫ R-dark
Halimium halimifolium	13C-1 Malate	0.06 ± 0.02	0.22 ± 0.08
	13C-4 Malate	0.05 ± 0.02	0.33 ± 0.11
	13C-1 Pyruvate	0.37 ± 0.10	0.71 ± 0.24
	13C-2 Pyruvate	0.03 ± 0.01	0.08 ± 0.02
Oxalis triangularis	13C-1 Malate	0.05 ± 0.01	0.17 ± 0.05
	13C-4 Malate	0.08 ± 0.03	0.38 ± 0.19
	13C-1 Pyruvate	0.16 ± 0.06	0.23 ± 0.07
	13C-2 Pyruvate	0.03 ± 0.00	0.02 ± 0.00

∫ _R values from different ¹³C-labeled malate and pyruvate substrates were calculated for light (∫ _R-light, μmol ¹³CO₂ m^-2 s^-1) and dark periods (∫ _R-dark, μmol ¹³CO₂ m^-2 s^-1) in Halimium halimifolium and Oxalis triangularis plants. Means and SE are given (n = 3–5 individuals). Please refer to Table 2 for statistical analyses.

Table 2.

Statistical analyses of ∫ _R and Δ_Label values from different metabolic fractions.

Parameter	∫ R	Lipids	Sugars	Organic Acids	Amino Acids
Malate	0.191	0.163	0.764	0.860	0.094
Species	0.960	0.001	NA	0.007	0.032
Malate∗Species	0.614	0.117	NA	0.666	0.240
Time	0.001	0.138	0.902	0.142	0.537
Pyruvate	0.001	0.001	0.065	0.004	0.273
Species	0.015	0.001	NA	0.001	0.782
Pyruvate∗Species	0.039	0.544	NA	0.068	0.183
Time	0.129	0.028	0.925	0.101	0.494

Results of general linear models show light–dark (Time), position-specific (Malate, Pyruvate), and species-specific (Species) differences and their interactions (e.g., Malate^∗Species). Please note that we separately performed statistical tests on malate and pyruvate, since a comparison of the absolute ∫ _R and Δ_Label values from both substrates is potentially biased by differences in substrate transport from the assimilation site to the site of reaction, mixing with leaf internal substrate pools, and substrate compartmentalization. P-values are indicated. Significant values (P ≤ 0.05) are marked in bold.

In contrast to both ¹³C-labeled malate substrates, which can be released by different reactions (NAD-ME, TCA cycle), ¹³CO₂ respiration rates from ¹³C-1 pyruvate reflect the direct CO₂ release by a pyruvate dehydrogenase reaction (cpPDH or mtPDH). Highest R_Label values from ¹³C-1 pyruvate were observed in H. halimifolium, with about 0.025 μmol ¹³CO₂ m^-2 s^-1 in the light, and with a pronounced peak of 0.06 μmol ¹³CO₂ m^-2 s^-1 within the first minutes in the dark during LEDR, which was followed by a clear continuous decrease (Figure 1). A very different ¹³CO₂ respiration pattern from ¹³C-1 pyruvate was found in O. triangularis, with no clear peak during LEDR (Figure 2). R_Label from ¹³C-2 pyruvate, reflecting the activity of CO₂ releasing reactions within the TCA cycle, showed hardly any increase in both species (Figures 1 and 2). Statistical analyses on ∫ _R values are in agreement with the observed ¹³CO₂ respiration patterns from pyruvate substrates, showing a species-dependency for changes caused by the ¹³C-labeled substrates and no clear light–dark differences (Tables 1 and 2). A Tukey HSD post hoc test revealed significant position-specific differences for ∫ _R from pyruvate substrates during light–dark transitions in H. halimifolium (P < 0.001), but not in O. triangularis (P = 0.354). It also shows species-specific differences for ∫ _R from ¹³C-1 pyruvate (P ≤ 0.012), but not for ∫ _R from ¹³C-2 pyruvate (P = 0.987). In summary, clear position- and species-specific differences during LEDR were only observed in the ¹³CO₂ respiration pattern obtained from pyruvate.

¹³C Allocation to Leaf Metabolic Fractions

Carbon isotope ratios in leaf metabolic fractions (lipids, sugars, amino and organic acids) of H. halimifolium and O. triangularis plants did not show significant light–dark differences between the two sampling time points, except for lipids from plants fed with ¹³C-2 pyruvate (Table 2). Thus, mean values of both time points were used for further analysis, i.e., to determine ¹³C allocation of position-specific ¹³C-labeled substrates to metabolic fractions.

The Δ_Label derived from both ¹³C-labeled pyruvate substrates was highest in the amino acid fraction of both species, while Δ_Label derived from ¹³C-labeled malate substrates was found to be highest in both, the amino acid and the organic acid fractions (Figure 3). In contrast, Δ_Label derived from any ¹³C-labeled substrate was lowest in lipid and sugar fractions of in both species, or showed no distinct ¹³C allocation. Significant position-specific differences were only found in lipid and organic acid fractions derived from ¹³C-labeled pyruvate substrates, with higher Δ_Label values from ¹³C-2 pyruvate compared to those from ¹³C-1 pyruvate (Table 2). Significant species-specific differences were observed across different ¹³C-labeled substrates and fractions. In summary, ¹³C from malate and pyruvate was species-specifically allocated to leaf metabolic fractions, mainly to amino and organic acids under light conditions, while position-specific differences in ¹³C allocation were only observed for pyruvate.

FIGURE 3.

¹³C allocation of position-specific¹³C-labeled malate (¹³C-1, ¹³C-4) and pyruvate (¹³C-1, ¹³C-2) to metabolic fractions in Halimium halimifolium and Oxalis triangularis. The label-induced ¹³C increase corrected for non-labeling conditions (Δ_Label, ‰) indicates the amount of total ¹³C allocated to leaf metabolic fractions. Mean Δ_Label values for the two harvests (20 min light or 20 min light and 20 min dark) were calculated, since no significant light–dark differences were observed for Δ_Label, except for lipids in plants fed with ¹³C-2 pyruvate, which showed higher δ¹³C values in the dark than in the light (Table 2). Please refer to Table 2 for the statistical analyses. n.a. indicates no observable Δ_Label. Capital letters denote significant differences among metabolic fractions of plants fed with the same ¹³C-labeled substrate (ANOVA and Tukey-HSD). Means and SE are given (n = 3 individuals).

Discussion

Carboxyl Groups of Malate Similarly Fuel Respiration during Light–Dark Transitions

Here, we investigated the ¹³CO₂ respiration pattern from different position-specific ¹³C-labeled malate and pyruvate in species with marked differences in δ¹³C_LEDR to identify potential respiratory substrates and enzymatic reactions. Both, ¹³C-1 and ¹³C-4 malate were clearly respired during the light–dark transitions in H. halimifolium and O. triangularis (Figures 1 and 2; Table 1), demonstrating that both carboxyl groups of malate contribute to respiration, especially during LEDR. Interestingly, the ¹³CO₂ respiration rates from ¹³C-1 and ¹³C-4 malate were similar in both species, which is surprising since only the C-4 position of malate have been suggested to be ¹³C enriched compared to other molecule positions (Melzer and O’Leary, 1987; Savidge and Blair, 2004) and thus expected to be of high importance for the ¹³C enrichment in respired CO₂ during LEDR at natural isotope abundances (Barbour et al., 2007; Gessler et al., 2009; Werner et al., 2011; Tcherkez et al., 2012). The equal contribution of both carboxyl groups of malate to respiration in both species is most likely provoked by the fumarase activity, causing a partial isotope randomization of the carboxyl groups of malate (Bradbeer et al., 1958; Gout et al., 1993). For instance, ¹³C-label at the C-4 position of malate can be transferred to the C-1 position of malate, which can be decarboxylated by NAD-ME, producing ¹³C-1 labeled pyruvate that fuels mtPDH (Figure 4). This respiratory flux is in line with high NAD-ME activity during LEDR (Igamberdiev et al., 2001), which may be stimulated by fumarate in support of oxidation of photosynthetic substrates (Tronconi et al., 2015). Moreover, if randomization of malate by the fumarase reaction is assumed, the ¹³C-label from both carboxyl groups of malate could be released by the action of the NAD-dependent mtIDH and the OGDH in the TCA cycle. Otherwise, these reactions would only release one carboxyl group of malate; C-1 by mtIDH and C-4 by OGDH (Figure 4).

FIGURE 4.

Overview of ¹³C allocation from position-specific ¹³C-labeled malate (¹³C-1, ¹³C-4) and pyruvate (¹³C-1; ¹³C-2) during light–dark transitions. Sun (☼) and half-moon () symbols denote temporary ¹³C allocation during light or dark periods (LEDR), respectively. Solid arrows indicate ¹³C allocation of highest importance, while dashed arrows indicate reactions with intermediate steps. ¹³C-labeled malate substrates in brackets denote alternative respiratory pathways due to the influence of NAD-ME (decarboxylation of C-4 position of malate) and fumarase (randomization of C-1 and C-4 position of malate). The following abbreviations are used: cyIDH and mtIDH, NADP-cytosolic and NAD-mitochondrial isocitrate dehydrogenase; cyMDH and mtMDH, cytosolic and mitochondrial malate dehydrogenase; NAD-ME, NAD-malic enzyme; OAA, oxaloacetate; OGDH, 2-oxoglutarate dehydrogenase; cpPDH and mtPDH, chloroplast and mitochondrial pyruvate dehydrogenase; PEPC, phosphoenolpyruvate carboxylase; TCA cycle, tricarboxylic acid cycle.

Decarboxylation of ¹³C-labeled malate in the light was generally not very strong, but also visible in all species (Figures 1 and 2; Table 1). The theoretical background for malate respiration in light was previously described by Werner et al. (2011). At least for ¹³C-1 malate, this might be explained by conversion of malate via the mtMDH reaction to oxaloacetate, which is used together with glycolytic acetyl-CoA by the citrate synthase reaction to build up citrate. This citrate can then be transferred to the cytosol via the citrate shuttle, where it is converted to isocitrate and subsequently decarboxylated to α-ketoglutarate and CO₂ by the NADP-dependent cytosolic isocitrate dehydrogenase reaction (cyIDH, Figure 4) (Werner et al., 2011). On the other hand, less is known about the biochemical pathways for direct respiration of ¹³C-4 malate in the light, since NAD-ME (Hill and Bryce, 1992; Igamberdiev et al., 2001) and TCA cycle reactions (Hanning and Heldt, 1993; Tcherkez et al., 2005; Araújo et al., 2012) are known to be partially down-regulated. Cytosolic fumarase might be involved, which was found to be expressed in some species (Pracharoenwattana et al., 2010; Eprintsev et al., 2014). However, the role of the enzyme in malate randomization in the light and during LEDR has not been elucidated yet.

Position- and Species-Specific Differences in Pyruvate Respiration Determine Natural δ¹³C_LEDR Patterns

In contrast to ¹³C-labeled malate substrates, position-specific differences in R_Label were induced by the ¹³C-labeled pyruvate substrates (Figures 1 and 2; Table 1), which were, however, species-dependent (Table 2). Differences in ¹³CO₂ respiration from ¹³C-1 pyruvate and ¹³C-2 pyruvate were significant in H. halimifolium, the species with high natural increases in δ¹³C_LEDR (up to 14.8‰; Wegener et al., 2010), but not in O. triangularis, the species which exhibits only low natural increases in δ¹³C_LEDR (up to 3.4 ‰; Wegener et al., 2010). This demonstrates a biochemical link between respiration of the C-1 position of pyruvate and the known increases in δ¹³C_LEDR and suggests a higher mtPDH activity during LEDR in H. halimifolium compared to O. triangularis (Priault et al., 2009; Wegener et al., 2010). Pyruvate for the mtPDH reaction may be provided by different fluxes, e.g., by glycolysis, by anaplerotic PEPC fluxes (via PEPC, MDH, NAD-ME), and by fluxes from carbon storage pools such as citrate and malate (via decarboxylation reactions from TCA cycle and NAD-ME; Figure 4).

We also observed an increase in ¹³CO₂ respiration rates from ¹³C-1 pyruvate in the light, especially in H. halimifolium (Figures 1 and 2; Table 1). This was most likely mainly caused by the chloroplast PDH reaction (cpPDH, Figure 4; Tcherkez et al., 2005, 2012), since the mitochondrial PDH reaction is known to be strongly down regulated under illumination (mtPDH, Figure 4), (Budde and Randall, 1990). In contrast, ¹³C-2 pyruvate was only weakly respired during the light–dark transitions in both species (Figures 1 and 2; Table 1). This suggests that the C-2 position of pyruvate (reflecting the acetyl-CoA residue) is predominantly used for biosynthesis of compounds during light–dark transitions. Indeed, this was further supported by position-specific ¹³C allocation differences in leaf metabolic fractions originated from ¹³C-labeled pyruvate substrates as described in the next paragraph.

¹³C-Label from Malate and Pyruvate Is Mainly Allocated to Amino and Organic Acids

Our final objective was to determine the ¹³C allocation from position-specific ¹³C-labeled substrates to metabolic fractions. We typically observed that ¹³C allocation to metabolic fractions (lipids, sugars, amino and organic acids) was high under light conditions, but immediately decreased when plants were transferred into darkness. This observation can be explained by well-known changes in transpiration and respiration rates in the dark. On the one hand, darkening causes a decrease of transpiration rates (due to decrease in vapor pressure deficit) and thus lower substrate uptake and less ¹³C allocation to metabolic fractions. On the other hand, most of the ¹³C-labeled substrates are used for respiration rather than for biosynthesis in the dark, causing an increase in ¹³CO₂ respiration rates during LEDR in both species (Figures 1 and 2). With one exception, lipids in plants fed with ¹³C-2 pyruvate were significantly more enriched in ¹³C after darkening (Figure 3; Table 2), suggesting that the acetyl-CoA produced from pyruvate is especially used for lipogenesis during light–dark transitions. This also partially explains the low ¹³CO₂ respiration rates from ¹³C-2 pyruvate compared to those from ¹³C-1 pyruvate in both species (Figures 1 and 2).

Furthermore, our results show that the ¹³C-label from both malate and pyruvate were generally allocated to the amino and organic acid fractions in both species. As summarized in Figure 4, the substrates were mostly used as precursors for the synthesis of amino acids derived from pyruvate, oxaloacetate, 2-oxoglutarate, as well as for the synthesis of organic acids such as citrate and malate, which are TCA cycle intermediates (Werner et al., 2011) or used as carbon storage molecules (Zell et al., 2010). However, with respect to the ¹³C-labeled substrates, we cannot distinguish if the organic acids taken up via the xylem stream were fully used for biosynthesis or if they partly accumulated in the leaves. Δ_Label values in lipids for both species suggested that only small amounts of the position-specific ¹³C-labeled malate and pyruvate were used for lipogenesis (Figures 3 and 4); similar results was found in a recent ¹⁴C labeling study (Grimberg, 2014). Low Δ_Label values in sugars in H. halimifolium might be explained by photosynthetic fixation of respired CO₂ derived from the ¹³C-labeled substrates, as previously observed for pyruvate (Tcherkez et al., 2005) or malate substrates (Hibberd and Quick, 2002).

Conclusion

Based on our experiments, we provide first direct experimental evidence that both carboxyl groups of malate are used as respiratory substrates during LEDR in different species. Surprisingly, plants fed with ¹³C-1 or ¹³C-4 malate showed no position- and species-specific differences in respiration pattern and metabolic fractions, strongly indicating that the carboxyl groups of malate undergo leaf internal isotope randomization by fumarase. Thus, our findings introduce a new level of complexity to the interpretation of respiratory substrates and enzymatic reactions influencing δ¹³C of leaf-respired CO₂ at natural isotope abundances. We conclude that malate respiration via different enzymatic reactions and metabolic pathways (e.g., fumarase, NAD-ME, and TCA cycle) supports and supplies pyruvate respiration during LEDR, which determines the species-specific differences of natural δ¹³C_LEDR patterns via decarboxylation of the C-1 position of pyruvate by PDH reactions, as shown within this study.

Author Contributions

ML and FW equally contributed to the manuscript. ML, FW, CW, and RW designed the experiments. ML synthesized the position-specific ¹³C-labeled malate. FW assembled the experimental setup. ML and FW performed the experiments and acquired the raw data. ML, FW, and MB processed the data. RW, CW, NB, RS, and VM supervised data analyses and interpretation. All authors helped drafting the manuscript and gave essential input to the work.

Conflict of Interest Statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Abbreviations

δ¹³C_LEDR

carbon isotopic composition of leaf-respired CO₂ during LEDR

Δ_Label

label-induced ¹³C increase

cyIDH and mtIDH

cytosolic and mitochondrial isocitrate dehydrogenase

LEDR

light enhanced dark respiration

cyMDH and mtMDH

cytosolic and mitochondrial malate dehydrogenase

NAD-ME

NAD-malic enzyme

OGDH

2-oxoglutarate dehydrogenase

cpPDH and mtPDH

chloroplast and mitochondrial pyruvate dehydrogenase

PEPC

phosphoenolpyruvate carboxylase

R_Label

label-derived leaf ¹³CO₂ respiration

TCA cycle

tricarboxylic acid cycle

Supplementary Material

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2016.00739