Light-responsive metabolite and transcript levels are maintained following a dark-adaptation period in leaves of Arabidopsis thaliana

Authors

  • Igor Florez-Sarasa,

    1. Grup de Recerca en Biologia de les Plantes en Condicions Mediterranies, Departament de Biologia, Universitat de les Illes Balears, Carretera de Valldemossa Km 7.5, 07122 Palma de Mallorca, Spain
    Search for more papers by this author
  • Wagner L. Araújo,

    1. Max-Planck-Institut für Molekulare Pflanzenphysiologie, Am Mühlenberg 1, 14476 Potsdam-Golm, Germany
    2. Departamento de Biologia Vegetal, Universidade Federal de Viçosa, 36570-000 Viçosa, MG, Brazil
    Search for more papers by this author
  • Sabá V. Wallström,

    1. Department of Biology, Lund University, Biology Building, Sölvegatan 35, SE-22362 Lund, Sweden
    Search for more papers by this author
  • Allan G. Rasmusson,

    1. Department of Biology, Lund University, Biology Building, Sölvegatan 35, SE-22362 Lund, Sweden
    Search for more papers by this author
  • Alisdair R. Fernie,

    1. Max-Planck-Institut für Molekulare Pflanzenphysiologie, Am Mühlenberg 1, 14476 Potsdam-Golm, Germany
    Search for more papers by this author
  • Miquel Ribas-Carbo

    1. Grup de Recerca en Biologia de les Plantes en Condicions Mediterranies, Departament de Biologia, Universitat de les Illes Balears, Carretera de Valldemossa Km 7.5, 07122 Palma de Mallorca, Spain
    Search for more papers by this author

Author for correspondence:
Miquel Ribas-Carbo
Tel: +34 971172051
Email: mribas@uib.cat

Summary

  • The effect of previous light conditions on metabolite and transcript levels was investigated in leaves of Arabidopsis thaliana during illumination and after light-enhanced dark respiration (LEDR), when dark respiration was measured.
  • Primary carbon metabolites and the expression of light-responsive respiratory genes were determined in A. thaliana leaves before and after 30 min of darkness following different light conditions. In addition, metabolite levels were determined in the middle of the night and the in vivo activities of cytochrome and alternative respiratory pathways were determined by oxygen isotope fractionation.
  • A large number of metabolites were increased in leaves of plants growing in or transiently exposed to higher light intensities. Transcript levels of respiratory genes were also increased after high light treatment. For the majority of the light-induced metabolites and transcripts, the levels were maintained after 30 min of darkness, where higher and persistent respiratory activities were also observed. The levels of many metabolites were lower at night than after 30 min of darkness imposed in the day, but respiratory activities remained similar.
  • The results obtained suggest that ‘dark’ respiration measurements, as usually performed, are probably made under conditions in which the overall status of metabolites is strongly influenced by the previous light conditions.

Introduction

Respiration provides ATP, reducing equivalents and carbon intermediates needed for growth, maintenance and transport processes in plants. In photosynthetic tissues, plant respiratory metabolism is modulated under illumination as a result of interactions with photosynthetic metabolism (Hurry et al., 2005). Photosynthetic reactions produce triose phosphates (TPs) that are exported to the cytosol and metabolized via glycolysis to produce substrates for the tricarboxylic acid (TCA) cycle in mitochondria. TPs can also be used for the synthesis of sucrose that can be exported to other tissues. The ATP needed for sucrose synthesis and transport is thought to be supplied by mitochondrial oxidative phosphorylation (Krömer et al., 1993). In addition, photorespiratory 2-phosphoglycolate production, produced by the oxygenase reaction of Rubisco, leads to the production of glycine, which becomes a major substrate for oxygen consumption in the light (Dry et al., 1983; Bauwe et al., 2010; Maurino & Peterhänsel, 2010). Glycine decarboxylation in mitochondria produces NH4+ which can diffuse to the chloroplast, where it can be re-assimilated by the glutamine synthetase/glutamate synthase (GS/GOGAT) system, thereby interconnecting mitochondrial respiratory metabolism with photosynthetic nitrogen assimilation (Gardeström et al., 2002; Noguchi & Yoshida, 2008; Foyer et al., 2011). A wide range of evidence has been accumulated suggesting that the TCA cycle in the light can provide the carbon precursors and the reducing power needed for nitrogen assimilation (for a review, see Nunes-Nesi et al., 2007b). In addition, transport of malate and TPs from the chloroplasts to the cytosol and mitochondria may directly dissipate reductants from the chloroplast under conditions of excess light (Noguchi & Yoshida, 2008). Thus, mitochondria can exert a strong control over the redox balance of the cell (Noctor et al., 2007).

Light, via phytochrome and/or cryptochrome signaling, has been observed to up-regulate genes encoding respiratory chain components, including the type II NAD(P)H dehydrogenase gene NDA1 and the alternative oxidase gene AOX1a (Escobar et al., 2004; Rasmusson & Escobar, 2007), and to down-regulate the succinate dehydrogenase (Popov et al., 2010). Light-dependent changes in expression also include the induction of the mitochondrial malate dehydrogenases, that are possibly involved in the export of reductant derived from photorespiratory glycine decarboxylation, and the suppression of other TCA cycle genes (Rasmusson & Escobar, 2007). Likewise, changes in the metabolite profile which occur diurnally reflect the interactions between photosynthetic and respiratory metabolism (Urbanczyk-Wochniak et al., 2005). Moreover, transgenic plants with altered expression of enzymes related to the TCA cycle exhibit changes in photosynthetic activity and metabolism (for a review, see Nunes-Nesi et al., 2011). Similarly, alterations in the expression of components of the mitochondrial electron transport chain suggest that its correct functioning is required in order to achieve an optimal photosynthetic performance (Igamberdiev et al., 2001; Dutilleul et al., 2003, 2005; Raghavendra & Padmasree, 2003; Bartoli et al., 2005; Sweetlove et al., 2006; Juszczuk et al., 2007; Yoshida et al., 2011). However, in these instances, there are few common responses at the metabolite level.

All of these metabolic interactions between photosynthetic and respiratory metabolism affect the extent of respiratory activity during illumination. Measurements of CO2 and O2 exchange in the light are difficult to interpret because of the combination of processes of oxygenation and carboxylation that occur in illuminated leaves (Hurry et al., 2005). Several studies using different methodologies have concluded that CO2 evolution is inhibited in the light (Atkin et al., 2000; Bykova et al., 2005; Hurry et al., 2005; Tcherkez et al., 2005; Pärnik et al., 2007). Such an inhibition of CO2 evolution in the light could, at least partially, be explained by the inhibitory phosphorylation of the mitochondrial pyruvate dehydrogenase, which is known to occur in response to light, and by the action of NH4+, produced by photorespiration (Budde & Randall, 1990; Tovar-Mendez et al., 2003). Furthermore, under photorespiratory conditions, glycine decarboxylation could increase the NADH/NAD+ ratio in the mitochondrial matrix, inhibiting TCA cycle dehydrogenases on illumination (Igamberdiev & Gardeström, 2003; Bykova et al., 2005; Tcherkez et al., 2008). In sharp contrast with the results concerning CO2 evolution, the extent of O2 uptake in the light remains unclear. In experiments using O2 isotopes, there are contradictory reports showing inhibition (Canvin et al., 1980), no effect (Avelange et al., 1991) or even stimulation (Xue et al., 1996) of O2 uptake in the light. However, these reports did not take into consideration the different O2 fractionations by the two terminal oxidases of the mitochondrial electron transport chain (Guy et al., 1989). Therefore, a change in electron partitioning between the two main respiratory pathways in the light would affect the O2 isotope ratio and, consequently, interfere with the 18O : 16O ratios detected in the light. It seems highly likely that respiration is, at least to some extent, inhibited in the light, when measured as CO2 evolution. However, the genetic studies described above reveal that the remaining respiratory activity is of vital importance for a normal metabolic function within the illuminated leaf.

Regardless of the effects of light on respiration during illumination, several studies have reported that a previous light exposure affects respiration measured in the dark (reviewed in Noguchi, 2005). The light intensity affects the carbohydrate status, which, in turn, can regulate dark respiration via altered substrate availability (Noguchi, 2005; Florez-Sarasa et al., 2009). Nevertheless, light can regulate the electron partitioning between cytochrome and alternative pathways independent of carbohydrate-mediated processes (Ribas-Carbo et al., 2000). In all of these studies, measurements of dark respiration on previously illuminated leaves were performed after a period of darkness in order to avoid light-enhanced dark respiration (LEDR). LEDR has been described as a post-illumination burst of CO2 release and/or O2 uptake (for reviews, see Atkin et al., 2000 and Padmasree et al., 2002) and has been well characterized (Reddy et al., 1991; Xue et al., 1996; Atkin et al., 1998). It has been observed that LEDR lasts roughly 30 min in different plant species and under several different experimental conditions (Azcon-Bieto & Osmond, 1983; Azcon-Bieto et al., 1983; Reddy et al., 1991; Atkin et al., 1997, 1998). Therefore, it has been assumed that, after this time, leaf tissues resume ‘normal’ dark respiration (Atkin et al., 2000; Padmasree et al., 2002). However, the reports mentioned above also demonstrated changes in dark respiration and electron partitioning between cytochrome and alternative pathways (after the LEDR period) caused by changes in the previous light conditions. Bearing this in mind, it is likely that, if dark respiration is modulated by the previous light conditions, the metabolite state of the leaf tissue after LEDR, when dark respiration is measured, might be more similar to light than to dark conditions.

In order to investigate to what extent previous light conditions can affect the overall status of metabolites during dark respiration, two different experiments were performed. In the first experiment, changes in primary carbon metabolites and expression of light-responsive respiratory genes were determined before and after a 30-min dark period that followed a 2-h high light treatment (800 μmol m−2 s−1, HLT). In a second experiment, the levels of primary carbon metabolites were determined before and after a 30-min dark period in plants grown at low (80 μmol m−2 s−1, LGL) and moderate (300 μmol m−2 s−1, MGL) growth light intensities. In addition, metabolite levels were determined in the middle of the night in plants grown at LGL and MGL in order to obtain a comparison with ‘normal’ dark conditions. In parallel, the oxygen uptake rate and electron partitioning between cytochrome and alternative pathways were determined using the oxygen isotope fractionation technique (Ribas-Carbo et al., 2005). Our objective was to determine major changes in a broad range of metabolite levels and in transcript levels of some well-known light-inducible respiratory genes that occurred following 30 min of darkness, the time point at which respiration is traditionally measured.

Materials and Methods

Plant material and growth conditions

Arabidopsis thaliana (L.) Heynh ecotype Col-0 seeds were sown in prewetted pots with a mixture of peat, perlite and vermiculite (2 : 1 : 1, v/v). After sowing, pots were kept for 48 h in darkness at 4°C for seed stratification. Pots (volume, 0.4 l) were placed in plastic trays that were used for subirrigation with half-strength Hoagland’s solution applied once a week. Three plants per pot were grown in a growth chamber for 35 d (after sowing) with a 10 h : 14 h light : dark photoperiod, with day/ night temperatures of 26°C/24°C and relative humidity above 40%. Two different experiments were performed as specified below.

In the first experiment, plants were grown at a light intensity of 80 ± 10 μmol m−2 s−1, considered as LGL, and, 2 h into the light cycle, a set of plants was moved to a light intensity of 800 ± 20 μmol m−2 s−1 for 2 h, hereafter considered as HLT. The temperature during HLT was kept between 25 and 26°C. Rosettes were harvested for metabolite profiling and transcript level analyses (see section below) at 2 h into the light cycle under LGL, at the end of 2 h of HLT and after 30 min of dark treatment following HLT (HLT + Dark).

In the second experiment, plants were grown at LGL or at a light intensity of 300 ± 20 μmol m−2 s−1, considered as MGL. Plants grown either in LGL or MGL conditions started to bolt between days 35 and 37, indicating a similar developmental stage. Rosettes were harvested for metabolite profiling before bolting at 2 h into the light cycle under LGL, 2 h into the light photoperiod under MGL, after 30 min of dark treatment after 2 h under LGL (LGL + Dark), after 30 min of dark treatment after 2 h under MGL (MGL + Dark), 7 h into the dark period in plants grown at LGL (LGL-Night) and 7 h into the dark period in plants grown at MGL (MGL-Night).

Respiration and oxygen isotope fractionation measurements

Leaves from plants growing at LGL, MGL and after HLT were placed in the dark for 30 min to avoid LEDR. Respiration and oxygen isotope fractionation measurements were performed as described in Florez-Sarasa et al. (2007) and lasted 50–60 min. Respiratory partitioning between the two respiratory pathways was calculated from the oxygen isotope fractionation by the alternative oxidase (Δa) and the cytochrome oxidase (Δc). Δa, determined in the presence of 10 mM of KCN, was 30.4‰, whereas Δc of 20.9‰ was taken from Florez-Sarasa et al. (2007). The leaf area was determined with a portable leaf area meter, AM 300 (ADC Bioscientific Ltd., Great Amwell, Herts, UK), after respiration measurements. Thereafter, dry weights were determined after drying leaves for 2 d at 60°C. Four or five replicates of leaves from three different plants were analyzed.

Metabolic profiling

Metabolite extractions were performed as described previously (Liu et al., 2009). Derivatization and gas chromatography-time of flight-mass spectrometry (GC-TOF-MS) analyses were carried out as described previously (Lisec et al., 2006). The GC-TOF-MS system was composed of a CTC CombiPAL autosampler, an Agilent 6890N gas chromatograph and a LECO Pegasus III time-of-flight mass spectrometer running in EI+ mode. Metabolites were identified by comparison with database entries of authentic standards (Kopka et al., 2005; Schauer et al., 2005).

Data were normalized with respect to the mean response calculated for the plants growing at low light conditions (LGL, 80 μmol m−2 s−1) (to allow statistical assessment, individual plants from this set were normalized in the same way), except for data presented in Fig. 6 and Supporting Information Table S3. In these, data were normalized to the mean response for the plants at LGL + Dark conditions. Values presented are the mean of the relative values ± SE of six to seven measurements from a pool of three plants each. In Tables S1–S3, values in bold were determined by Student’s t-test to be significantly different (< 0.05) from the LGL-treated plants for Tables S1, S2, and from the LGL + Dark-treated plants for Table S3.

Transcript analyses

Total RNA from leaf samples was isolated using the RNeasy Plant Mini Kit (Qiagen, Hilden, Germany). Three RNA preparations each for the three light treatments in the first experiment (see the section on Plant material and growth conditions above) were employed for cDNA synthesis performed using the RevertAid H Minus First Strand cDNA Synthesis Kit (Fermentas, New England Biolabs) with Oligo (dT)18 primers. The cDNA obtained was then treated with RNase H (Fermentas, New England Biolabs). Real-time PCR primers used for NDA1 were as described in Michalecka et al. (2003) and, for AOX1a, as described in Escobar et al. (2004). Real-time PCRs were run in a Rotor Gene 2072 Real-Time Cycler (Corbett Research, Sydney, Australia) as described previously (Escobar et al., 2004). Annealing temperatures for the primers were 60°C for NDA1 and 59°C for AOX1a.

Chlorophyll fluorescence measurements

The maximum quantum efficiency of photosystem II (Fv/Fm) was obtained from chlorophyll fluorescence measurements on 6–12 rosette leaves per light treatment (LGL, MGL and HLT) with a portable pulse amplitude modulation fluorometer (PAM-2000, Walz, Effeltrich, Germany), as described previously (Galle et al., 2007).

Statistical analyses

Student’s t-tests were used for statistical analyses of metabolite levels and respiratory activities from Table 1, comparing LGL or LGL + Dark with the rest of the light and dark treatments. ANOVA and Duncan’s test were used in SPSS v. 17 (SPSS Inc., Chicago, IL, USA) for comparisons among transcript levels from Fig. 3 and respiratory parameters from Table 2 at different light and dark conditions. Significant differences relate to < 0.05. In addition, Pearson correlation coefficients (r) were obtained from the levels of metabolites at all dark conditions and their corresponding respiratory activities on a per dry weight basis.

Table 1.   Respiration and electron partitioning after high light treatment
 LGL + DarkHLT + Dark
  1. Total respiration (Vt), electron partitioning to the alternative pathway (τa), cytochrome pathway activity (νcyt) and alternative pathway activity (νalt) are shown for dark-adapted leaves of Arabidopsis thaliana plants at low growth light (LGL, 80 μmol quanta m−2 s−1) and after a high light treatment (HLT, 800 μmol m−2 s−1) (for details, see the Materials and Methods section). Values are means ± SE of five biological replicates. Significant differences (< 0.05) from the LGL + dark treatment are denoted in bold.

Vtμmol O2 m−2 s−10.32 ± 0.010.43 ± 0.02
nmol O2 gDW−1 s−129.7 ± 0.837.0 ± 1.0
τa 0.34 ± 0.030.37 ± 0.03
νcytμmol O2 m−2 s−10.21 ± 0.010.27 ± 0.03
nmol O2 gDW−1 s−119.5 ± 0.723.2 ± 0.9
νaltμmol O2 m−2 s−10.11 ± 0.010.16 ± 0.01
nmol O2 gDW−1 s−110.1 ± 1.313.8 ± 1.3
Table 2.   Respiration and electron partitioning under different growth light intensities
 LGL + DarkLGL-NightMGL + DarkMGL-Night
  1. Total respiration (Vt), electron partitioning to the alternative pathway (τa), cytochrome pathway activity (νcyt) and alternative pathway activity (νalt) in leaves of Arabidopsis thaliana plants taken 2 h into the light cycle and dark adapted ( + Dark) and 7 h into the dark photoperiod (-Night) and grown at either low (LGL, 80 μmol quanta m−2 s−1) or moderate (MGL, 300 μmol quanta m−2 s−1) growth light intensities. Values are means ± SE of four biological replicates. Different letters denote significant differences (< 0.05) between light and dark treatments.

Vtμmol O2 m−2 s−10.28 ± 0.01a0.29 ± 0.02a0.66 ± 0.03b0.64 ± 0.03b
nmol O2 gDW−1 s−130.8 ± 0.7a30.6 ± 1.1a38.9 ± 1.7b40.6 ± 1.8b
τa 0.33 ± 0.04a0.37 ± 0.02a0.39 ± 0.04a0.36 ± 0.01a
νcytμmol O2 m−2 s−10.19 ± 0.02a0.18 ± 0.02a0.40 ± 0.05b0.41 ± 0.02b
nmol O2 gDW−1 s−120.5 ± 0.8a19.5 ± 1.4a23.9 ± 2.7ab26.0 ± 1.3b
νaltμmol O2 m−2 s−10.09 ± 0.01a0.10 ± 0.00a0.25 ± 0.02b0.23 ± 0.01b
nmol O2 gDW−1 s−110.3 ± 1.2a11.1 ± 0.3a15.0 ± 1.2b14.6 ± 0.9b

Results

Light-induced changes in respiratory activities measured after dark treatment

In the first experiment, LGL-grown plants were subjected to HLT. Leaf oxygen uptake (Vt), measured after 30 min of dark adaptation (+ Dark), increased significantly by 37% and 25% on an area and dry weight basis, respectively, compared with LGL grown + dark-treated leaves (Table 1). However, no significant difference in the electron partitioning to the alternative pathway (τa) was observed. Therefore, cytochrome and alternative pathways were similarly increased after HLT (Table 1).

In the second experiment, plants were grown at LGL and MGL conditions. In this case, Vt was 133% and 27% higher in plants at MGL + Dark than in those at LGL + Dark on an area and dry weight basis, respectively (Table 2). With regard to respiration, both cytochrome and alternative pathways were higher in MGL + Dark than in LGL + Dark plants, but τa was not changed significantly (Table 2). Finally, no significant differences were observed between the respiration rates of plants taken 2 h into the light cycle and dark adapted for 30 min and those measured in the middle of the night (Table 2).

Maintenance of respiration and oxygen isotope fractionation after dark treatment

Leaf respiratory parameters, except for measurements performed at night, were determined after a dark treatment of 30 min in previously light-exposed plants. In these plants, the oxygen isotope fractionation and oxygen consumption were constant over the whole measurement period (i.e. 50–60 min after the 30-min dark adaptation) as denoted by the linear regressions between – loge f and loge(R/R0) (Fig. 1a,c) and between the amount of O2 consumed and time (Fig. 1b,d).

Figure 1.

Linearity of respiration and electron partitioning in darkness. Arabidopsis thaliana plants grown at moderate growth light (MGL) intensity and after a 2-h high light treatment (HLT) were placed in complete darkness for 30 min before the start of the respiration and oxygen isotope fractionation measurements. After the 30-min dark period, respiration measurements were performed for 50–60 min. Examples of linear regressions between loge f and loge(R/R0) obtained from a measurement of the oxygen isotope fractionation in 30-min dark-treated leaves of MGL-grown plants (a) and HLT-exposed plants (c). R and R0 are the oxygen isotope ratios at a given time and at the initial time of the measurement, respectively, and f is the fraction of oxygen remaining at a given time ([O2]/[O2]0). In the equations presented, the slopes correspond to the oxygen isotope fractionation values. The R2 values of the linear regressions between loge f and loge(R/R0) were higher (R2 ≥ 0.997) than the minimum acceptable level (Ribas-Carbo et al., 1997). Examples of linear regressions between the amount of O2 consumed and time (min) are shown in (b) and (d) during the same measurements as performed in (a) and (c), respectively. The slopes from the equations correspond to the oxygen consumption rates per leaf area. All the figures show data from one representative experiment of five showing similar results.

Maintenance of high light-induced changes after dark treatment

In the first experiment, HLT caused significant changes in 31 metabolite levels of the 44 analyzed when compared with the LGL values (Fig. 2 and Table S1), from which only the level of aspartic acid was decreased significantly by 26%. The levels of fructose, glucose and mannose were 2.0, 2.8 and 1.7 times higher after HLT relative to LGL (Fig. 2a and Table S1). TCA cycle intermediates (cis-aconitic, citric, fumaric, isocitric and malic acids) were also increased by between 1.2 and 3.1 times after HLT. Moreover, a large number of amino acids showed increased levels after HLT, with glycine (22.1-fold), phenylalanine (5.8-fold), alanine (4.1-fold) and isoleucine (3.4-fold) being the most light-induced metabolites (Fig. 2a and Table S1).

Figure 2.

Metabolite profiling of high light- and dark-treated Arabidopsis thaliana plants. Plants were harvested for metabolite profiling at low growth light (LGL) conditions, just before the high light treatment (HLT), and at the end of the high light and dark treatments (HLT + Dark) (for details, see the Materials and Methods section). (a) Heat map showing the values expressed as fold changes, with the levels in LGL individually set to unity. (b) The log2 values for the levels of 44 metabolites after the 30-min dark treatment are plotted against the levels of the same metabolites after HLT. Zero represents the value for each metabolite at LGL. The continuous line denotes a linear regression and is complemented with the R2 value. The dashed line marks unity. In (c), data are shown for the four metabolites that exhibited a significant difference of at least 25% in HLT + dark-treated plants relative to HLT-exposed plants. Values are means ± SE for six biological replicates.

When plants subjected to HLT were moved to darkness for 30 min, the light-induced changes in the levels of most metabolites were maintained, and the profile after dark treatment was thus strongly different from the LGL values (Fig. 2a and Table S1). For increased clarity, the values of the levels of the 44 metabolites analyzed were set on a log2 scale relative to the LGL values. The log2 values after HLT and 30 min of darkness were plotted against the values for the same metabolites immediately after HLT (Fig. 2b). The high correlation (R2 = 0.93) between metabolite levels after HLT and those after 30 min of darkness, as well as the closeness of the slope to unity, indicates that the levels of most metabolites were characterized by relatively few changes before and after dark treatment (Fig. 2b). The minority of metabolites that showed a significant difference of at least 25% in HLT + dark-treated relative to HLT plants are presented in Fig. 2(c). After 30 min of dark treatment, lysine and pyruvic acid returned to the levels of LGL conditions (Fig. 2c), whereas alanine and phenylalanine were decreased after dark treatment but still exhibited higher levels when compared with those observed at LGL intensity (Fig. 2c).

In parallel, we investigated changes in the expression of two light-responsive respiratory genes on HLT and subsequent dark treatment. After 2 h of HLT, the transcript levels of NDA1 and AOX1a were increased significantly (1.3 and 3.5 times, respectively) in comparison with LGL intensity levels (Fig. 3). After HLT, a 30-min dark period did not result in any noticeable changes in the HL-induced levels of either NDA1 or AOX1a transcripts.

Figure 3.

Transcript levels after high light treatment (HLT), and after a subsequent 30-min dark period (HLT + Dark). Arabidopsis thaliana plants grown, treated and harvested as in Fig. 2 were analyzed for the expression levels of the light-responsive genes NAD(P)H dehydrogenase (NDA1) (black bars) and alternative oxidase (AOX1a) (gray bars) by real-time PCR. Values are means ± SE for three biological replicates. The values are expressed relative to the mean expression levels under low growth light (LGL) conditions. Significant differences (< 0.05) in comparison with LGL conditions are denoted by different letters.

Maintenance of metabolite profiles in LGL and MGL plants after dark treatment

In the second experiment, the metabolite profile of plants grown at LGL was generally similar to that after an additional 30 min of darkness (LGL + Dark) (Fig. 4a and Table S2). Among the 44 metabolites analyzed, only seven showed a significant difference of at least 25% between LGL + Dark and LGL levels (Fig. 4b). Among them, glucose (39%), fructose (48%), sorbitol (53%), glycine (63%) and sucrose (65%) decreased, whereas cis-acotinic acid (199%) and arginine (117%) increased after the 30-min dark period (Fig. 4b).

Figure 4.

Metabolite profiling of Arabidopsis thaliana plants grown at low growth light (LGL) intensity and after 30 min of dark treatment (LGL + Dark). Plants were harvested for metabolite profiling at LGL conditions, just before a time point 2 h into the normal light cycle and at the end of a subsequent 30-min dark period (for details, see the Materials and Methods section). (a) Heat map showing the values expressed as fold changes, with the levels in LGL individually set to unity. In (b), data are shown for the seven metabolites that exhibited a significant difference of at least 25% in dark-treated plants (LGL + Dark) relative to LGL plants. Values are means ± SE for six biological replicates.

Growing plants at MGL caused significant changes in the levels of 31 metabolites when compared with LGL-grown plants (Fig. 5 and Table S2). Levels of glucose, mannose and sucrose increased by 1.4, 1.4 and 1.6 times, respectively, in plants growing at MGL. Similar to plants after the onset of HLT, TCA cycle intermediates, such as cis-aconitic, citric, fumaric, isocitric and malic acids, were 2.7, 1.8, 2.5, 3.1 and 1.6 times higher, respectively, in MGL than in LGL. Moreover, several amino acids increased significantly in MGL plants (Fig. 5a and Table S2). It was observed that phenylalanine (5.3-fold), isoleucine (4.8-fold), alanine (4.7-fold), tryptophan (2.9-fold) and arginine (2.5-fold) were the amino acids showing the highest relative levels at MGL (Fig. 5a and Table S2).

Figure 5.

Metabolite profiling of Arabidopsis thaliana plants grown at moderate growth light (MGL) intensity and after 30 min of dark treatment (MGL + Dark). Plants were harvested for metabolite profiling at MGL conditions, just before 2 h into the normal light cycle and at the end of a subsequent 30-min dark period (for details, see the Materials and Methods section). (a) Heat map showing the values expressed as fold changes, with the levels in LGL individually set to unity. (b) The log2 values for the levels of 44 metabolites after the 30-min dark treatment (MGL + Dark) are plotted against the levels of the same metabolites at MGL. Zero represents the value for each metabolite at LGL. The continuous line denotes a linear regression and is complemented with the R2 value. The dashed line marks unity. In (c), data are shown for the eight metabolites that exhibited a significant difference of at least 25% in dark-treated plants (MGL + Dark) relative to plants at MGL just before the dark treatment. Values are means ± SE for six biological replicates.

The levels of most metabolites observed in plants grown under MGL were maintained after 30 min of darkness (MGL + Dark), that is, both profiles strongly deviated from the LGL levels (Fig. 5a and Table S2). The log2 values (relative to LGL) for MGL + Dark were plotted against the values for the same metabolites at MGL (Fig. 5b). The high correlation (R2 = 0.72) between metabolites at MGL and MGL + Dark, as well as the closeness of the slope to unity, indicates that the levels of most metabolites were characterized by relatively few changes before and after the dark treatment (Fig. 5b). Only eight of the 44 metabolites showed a significant difference of at least 25% in response to the dark treatment (Fig. 5c). The metabolites that decreased the most during the 30-min dark treatment were sucrose (55%), glycine (40%), alanine (34%), pyruvic acid (29%), isoleucine (28%) and asparagine (26%), whereas isocitric acid and glutamine were increased by 84% and 89%, respectively, after the dark treatment (Fig. 3c). Thus, of the 31 metabolites displaying significantly higher levels in MGL relative to LGL, only five showed a value after MGL + Dark that was similar or lower than the LGL value (Table S2).

Changes in metabolite profiles under dark conditions and correlation with respiration

In addition to metabolite profiles after 30 min of darkness (LGL + Dark and MGL + Dark), the levels of metabolites under ‘normal’ dark conditions were determined in the middle of the night in plants grown at both low (LGL-Night) and moderate (MGL-Night) growth light intensities. Metabolite levels determined under the four dark conditions were normalized to the levels at LGL + Dark and presented in Fig. 6(a) and Table S3.

Figure 6.

Metabolite profiling of plants under dark conditions and its relationship with respiratory activities. Arabidopsis thaliana plants were harvested for metabolite profiling at 7 h into the ‘normal’ dark period in plants grown in both low (LGL-Night) and moderate (MGL-Night) light intensities (for details, see the Material and Methods section), and compared with the levels of the plants harvested as in Figs 4, 5 after 30 min of darkness (LGL + Dark and MGL + Dark) (for details, see the Material and Methods section). (a) Heat map showing the values expressed as fold changes, with the levels in LGL + Dark individually set to unity. (b) Pearson correlation coefficients (r) between metabolite levels and total (Vt), cytochrome (νcyt) and alternative (νalt) respiration on a dry weight basis. Values in (a) are means ± SE for six biological replicates. The r values in (b) were obtained from four-point linear correlations and ordered from the highest to the lowest correlation with Vt.

Significant changes were observed when metabolite profiles after 30 min of darkness were compared with those in the middle of the night. The metabolite profile at LGL-Night was markedly different from that at LGL + Dark (Fig. 6a and Table S3). The levels of 17 metabolites were significantly different between LGL + Dark and LGL-Night conditions, most showing decreased levels at LGL-Night. The levels of fructose and glucose were strongly decreased by 95% and 83%, respectively, whereas the level of sucrose was increased by 179% at LGL-Night. The levels of several organic acids were also decreased at LGL-Night, such as pyruvic acid (33%), glutaric acid (33%), isocitric acid (61%) and malic acid (71%), whereas only fumaric acid was increased (30%). In addition, the levels of several amino acids were decreased, including arginine (22%), threonine (23%), phenylalanine (31%), proline (39%), serine (52%) and glycine (84%), whereas tryptophan was increased (33%). Moreover, the levels of metabolites at MGL-Night were also strongly different from those at MGL + Dark, similar to the comparison between LGL + Dark and LGL-Night described above (Fig. 6a and Table S3).

Moreover, the comparison of metabolite profiles between LGL- and MGL-grown plants in each dark condition can also be performed. The levels of several metabolites at MGL + Dark were higher than those at LGL + Dark conditions (Fig. 6a and Table S3), showing similar differences to those observed between MGL and LGL conditions (Fig 5a). Levels of mannose, glucose, fructose and sucrose were 1.4, 1.7, 1.9 and 2.0 times higher, respectively, in plants at MGL + Dark than in those at LGL + Dark. In addition, TCA cycle intermediates, such as malic, 2-oxoglutaric, citric, fumaric and isocitric acids, were 1.2, 1.4, 1.4, 2.2 and 5.6 times higher, respectively, in MGL + Dark relative to LGL + Dark. Moreover, almost all of the amino acids observed to be light induced in Fig. 5 remained at higher levels in MGL + Dark than in LGL + Dark (Fig. 6a and Table S3). In addition, the levels of most amino acids and organic acids were higher at MGL-Night than at LGL-Night, whereas some metabolite levels were similar, such as the levels of glucose, fructose, arginine, glycine, phenylalanine, malic acid and pyruvic acid among others.

Finally, metabolite levels in all dark conditions (LGL + Dark, LGL-Night, MGL + Dark and MGL-Night) were compared with the respiratory activities (on a per dry weight basis) measured under identical conditions. Pearson correlation coefficients were obtained and metabolites are presented from higher to lower levels of correlation with total respiration in Fig. 6(b). Metabolites showing higher correlations with respiration were amino acids, such as valine, glutamine, β-alanine, tryptophan and ornithine, and organic acids, such as lactic, glutamic, fumaric, 2-oxoglutaric and glutaric acids. Of the metabolites showing the highest correlations, lactic acid presented a negative correlation, whereas the rest were positively correlated with respiratory activities. However, sugars (i.e. glucose, fructose, sucrose), other amino acids (i.e. glycine, phenylalanine, arginine, homoserine) and organic acids (i.e. pyruvic and malic acids) presented a very low correlation with respiratory activities. Similar profiles were observed for the cytochrome and alternative pathway activities.

Discussion

Dark respiration measurements in leaves have generally been performed after several minutes (generally 30 min) in darkness to avoid LEDR. Respiration rates measured after this time have been proposed to represent ‘normal’ dark respiration (Atkin et al., 2000; Padmasree et al., 2002), and it should thus be assumed that all light-responsive metabolites that affect respiratory flux or partitioning would have reverted to dark levels. However, light-induced changes in dark respiration, when measured after the LEDR period, and variations in the in vivo activities of cytochrome and alternative pathways, depending on growth light intensity, have been reported previously (Noguchi et al., 2001; Noguchi, 2005; Florez-Sarasa et al., 2009, 2011). Moreover, in vivo respiratory activities are enhanced in plants exposed to high light conditions for a short period of time (i.e. hours; Yoshida et al., 2007, 2008; Florez-Sarasa et al., 2011), raising the question as to whether the overall metabolite state of the leaf tissue is closer to dark or light conditions. In the present study, A. thaliana Col-0 plants were grown at two different light intensities: 80 μmol quanta m−2 s−1, considered as LGL, and 300 μmol quanta m−2 s−1, considered as MGL, and plants grown at LGL were subjected to HLT of 800 μmol quanta m−2 s−1 for 2 h. The total respiration rate (Vt) in plants grown at MGL, after 30 min in darkness, was higher than in plants grown at LGL, both during the day (after LEDR) and at night (Table 2). Furthermore, the exposure of LGL-grown plants to HLT caused an increase in Vt of almost 40% (Table 1). Together with the dark respiration rate, the oxygen isotope fractionation was persistent during the subsequent 50–60 min after darkness, thus allowing the determination of the activities of the cytochrome and alternative pathways under these conditions (Fig. 1, Tables 1, 2). These results indicated that there was no substrate limitation for respiratory activities during the monitored time interval.

In the light, interactions between photosynthesis and respiration induce changes in metabolic pathways of primary carbon metabolism (Hurry et al., 2005). Respiration and photosynthesis have been formerly considered as separate pathways; however, an increasing amount of evidence suggests that their activities are closely coordinated through intracellular metabolite pools (Carrari et al., 2003; Raghavendra & Padmasree, 2003; Nunes-Nesi et al., 2005, 2007a, 2011; Noguchi & Yoshida, 2008). Given the fact that the increased respiratory activities observed in plants after HLT and when grown at MGL (Tables 1, 2) were sustained in the dark (Fig. 1), metabolite levels were, for both cases, determined before and after a 30-min dark period. It was observed that several sugars and TCA cycle intermediate levels were higher in HLT-exposed and MGL-grown plants than in LGL-grown plants (Figs 2a, 5a). Moreover, the levels of several amino acids showed a more pronounced increase (Figs 2a, 5a), suggesting a higher supply of TCA cycle carbon metabolites for amino acid synthesis in the light, in agreement with previous observations (Nunes-Nesi et al., 2007b; Tcherkez et al., 2008). The levels of phenylalanine, alanine and isoleucine were highly light induced after HLT and were also higher in plants grown under MGL. Nevertheless, glycine was dramatically induced after HLT, but was much less increased after MGL (Figs 2a, 5a). These results suggest a large increase in photorespiratory activity and a limitation in the conversion of glycine to serine caused by HLT. Accordingly, the maximum quantum efficiency of photosystem II (Fv/Fm) was lower in HLT-exposed (0.66 ± 0.01) than in MGL- (0.80 ± 0.01) and LGL (0.81 ± 0.01)-grown plants, indicating photoinhibition caused specifically by HLT. In addition to changes in metabolite profiles, a more than three-fold up-regulation of the expression of AtAOX1a was observed after HLT (Fig. 3). This is in good agreement with the previously reported up-regulation of genes encoding alternative oxidases after HLT (Yoshida & Noguchi, 2009).

After 30 min of darkness, the HLT-induced levels of most metabolites remained very different from the LGL levels (Fig. 2). In addition, the metabolite profile in dark-treated plants grown at MGL was similar to that before dark treatment, but very different from the profiles of LGL plants in the light (Fig. 5) or after 30 min of darkness (Fig. 6). The effect of the suddenly imposed dark treatments observed here is consistent with previously reported changes over the diurnal light period. For example, in rosettes of Arabidopsis, a large range of metabolites did not show changes in the short light–dark diurnal transitions, but, instead, changes were observed after several hours of light or dark periods (Gibon et al., 2006). In this context, diurnal changes in amino acid levels have also been observed in potato leaves, in which the majority of amino acids increased during the light and decreased during the dark period (Urbanczyk-Wochniak et al., 2005). However, tomato plants showed no major changes in the total amino acid content or in the levels of individual amino acids between light and dark periods (Carrari et al., 2003). In spite of the general maintenance of the light-induced levels of the majority of amino acids after the dark treatment (Figs 2, 5), a manifest decrease in the levels of a few amino acids was observed after the dark treatment (Figs 2c, 5c). This could be a result of the absence of photosynthetic nitrogen assimilation, which would decrease amino acid synthesis, as observed previously to be one of the main processes driving respiratory carbon metabolism under light conditions (Nunes-Nesi et al., 2007b; Tcherkez et al., 2008). In addition, the levels of glycine and some sugars were decreased after 30 min of darkness in both MGL- and LGL-grown plants (Figs 4b, 5c). The levels of sugars have been proposed to correlate with respiration rates in leaves of some species, whereas, in others, the respiration rate is likely to be controlled by energy demand (Noguchi, 2005). In our study, it seems unlikely that the levels of sugars and glycine become limiting for respiration after 30 min of darkness. In particular, the levels of glycine and some sugars were much lower at night (LGL-Night and MGL-Night) than in the day after 30 min of darkness (LGL + Dark and MGL + Dark) (Fig. 6a), whereas respiration remained similar (Table 2). Moreover, night respiratory activities remained higher in MGL- than in LGL-grown plants (Table 2), despite having similar glycine and sugar levels (Fig. 6a). Indeed, changes in sugar and glycine levels did not correlate with respiration rates, whereas variations in the levels of several amino acids and organic acids correlated better with respiration (Fig. 6b). Organic acids, such as malate and fumarate, have been demonstrated recently to be important carbon substrates for respiration in A. thaliana (Zell et al., 2010). However, further analyses of diurnal respiration time courses, in parallel with metabolite profiling (Urbanczyk-Wochniak et al., 2005; Gibon et al., 2006), in a wide range of species remain to be examined in future experiments to investigate the limiting substrates for respiration, as well as the implications of natural light intensities. For improved resolution, such studies will probably require an additional expansion of the metabolite coverage to include important compounds, such as the pyridine and adenine nucleotides, and may even require modeling of the subcellular compartmentation of these pools. Nevertheless, the maintenance of higher levels of TCA cycle intermediates is likely to be responsible for the higher respiratory activity observed in the current study after HLT and in plants growing at MGL. In close agreement with the latter hypothesis, reduced respiration rates were coupled to changes in the levels of TCA cycle intermediates caused by chemical inhibition of the 2-oxoglutarate dehydrogenase complex (Araújo et al., 2008). Furthermore, lower respiration rates, determined by following CO2 evolution, were also observed in tomato aconitase mutants (Aco-1) exhibiting reduced TCA cycle intermediates (Carrari et al., 2003). With regard to respiratory gene expression, the significantly elevated levels of AOX1a and NDA1 transcripts induced after HLT were maintained after 30 min of darkness. The stability of AOX1a and NDA1 transcript levels therefore indicates that the signals governing their expression are also stable over the 30-min dark treatment.

In conclusion, a higher growth light intensity, as well as a short-term HLT, induced changes in the levels of many metabolites in leaves. In addition, respiratory gene expression was induced after HLT. The great majority of these changes remain in effect even after 30 min of darkness, when dark respiration is normally measured. Therefore, these changes can favor the higher and persistent respiratory activity that is observed in plants grown at a higher light intensity and after HLT. These results suggest that ‘dark’ respiration measurements, as currently performed, are likely to be made under a metabolite state similar to that in the light. In other words, there is no general dark condition, but, in each case, it is preconditioned by the previous light regime.

Acknowledgements

This study was financed by the Spanish Ministry of Science and Innovation (MICINN) – project BFU2008-1072/BFI, the Max-Planck-Society (W.L.A. and A.R.F.) and the Swedish Research Council (A.G.R.). I.F-S. is supported by a FPI Fellowship of the Spanish Ministry of Education and Science. We would like to thank Dr Biel Martorell for his technical help with the isotope ratio mass spectrometer (IRMS), and all the staff at the Serveis Cientifico-Tecnics of the Universitat de les Illes Balears for their help whilst running these experiments. We are very grateful to Professor Jaume Flexas and Hans Lambers for many helpful discussions, and to Monica Higuera for her help in manuscript editing.

Ancillary