Distinct responses of the mitochondrial respiratory chain to long- and short-term high-light environments in Arabidopsis thaliana

Authors

  • KEISUKE YOSHIDA,

    Corresponding author
    1. Department of Biological Sciences, Graduate School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
    2. Chemical Resources Laboratory, Tokyo Institute of Technology, Nagatsuta 4259-R1-8, Midori-ku, Yokohama 226-8503, Japan
      K. Yoshida. e-mail: yoshida.k.ao@m.titech.ac.jp
    Search for more papers by this author
  • CHIHIRO K. WATANABE,

    1. Department of Biological Sciences, Graduate School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
    Search for more papers by this author
  • TAKUSHI HACHIYA,

    1. Department of Biological Sciences, Graduate School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
    Search for more papers by this author
  • DANNY THOLEN,

    1. Plant Systems Biology Group, Partner Institute of Computational Biology, 320 Yue Yang Road, Shanghai 200031, China
    Search for more papers by this author
  • MASARU SHIBATA,

    1. Department of Materials Engineering, Nagaoka National College of Technology, 888 Nishikatakai, Nagaoka, Niigata 940-8532, Japan
    Search for more papers by this author
  • ICHIRO TERASHIMA,

    1. Department of Biological Sciences, Graduate School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
    Search for more papers by this author
  • KO NOGUCHI

    1. Department of Biological Sciences, Graduate School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
    Search for more papers by this author

K. Yoshida. e-mail: yoshida.k.ao@m.titech.ac.jp

ABSTRACT

In order to ensure the cooperative function with the photosynthetic system, the mitochondrial respiratory chain needs to flexibly acclimate to a fluctuating light environment. The non-phosphorylating alternative oxidase (AOX) is a notable respiratory component that may support a cellular redox homeostasis under high-light (HL) conditions. Here we report the distinct acclimatory manner of the respiratory chain to long- and short-term HL conditions and the crucial function of AOX in Arabidopsis thaliana leaves. Plants grown under HL conditions (HL plants) possessed a larger ubiquinone (UQ) pool and a higher amount of cytochrome c oxidase than plants grown under low light conditions (LL plants). These responses in HL plants may be functional for efficient ATP production and sustain the fast plant growth. When LL plants were exposed to short-term HL stress (sHL), the UQ reduction level was transiently elevated. In the wild-type plant, the UQ pool was re-oxidized concomitantly with an up-regulation of AOX. On the other hand, the UQ reduction level of the AOX-deficient aox1a mutant remained high. Furthermore, the plastoquinone pool was also more reduced in the aox1a mutant under such conditions. These results suggest that AOX plays an important role in rapid acclimation of the respiratory chain to sHL, which may support efficient photosynthetic performance.

INTRODUCTION

The mitochondrial respiratory system plays a central role in supplying ATP necessary for a variety of cellular events. A series of electron transport reactions in the mitochondrial respiratory chain generate a proton gradient across the mitochondrial inner membrane, which is used for ATP production. This process is common in all aerobic organisms, but higher plant mitochondria additionally possess several respiratory pathways that are uncoupled from the proton translocation and thereby from the ATP production. In particular, the alternative oxidase (AOX) is the best known example of such energy-wasteful components. Contrary to the cytochrome pathway (CP), AOX transfers electrons directly from reduced ubiquinone (UQ) to oxygen without proton translocation (Finnegan, Soole & Umbach 2004). Several roles of AOX have been suggested so far, including heat generation for attracting pollinators in thermogenic species (Seymour 2001) and suppression of reactive oxygen species (ROS) generation especially under stressful situations (Maxwell, Wang & McIntosh 1999; Umbach, Fiorani & Siedow 2005). A recent study indicated that AOX has a role in determining normal development and function in soybean gametophyte (Chai et al. 2010). Nevertheless, the physiological importance of AOX is still not completely understood.

Another characteristic of plant mitochondria in photosynthetic tissues is the tight metabolic interaction with the chloroplasts. To date, multiple modes of this organelle crosstalk have been proposed (Gardeström, Igamberdiev & Raghavendra 2002; Raghavendra & Padmasree 2003; Nunes-Nesi, Sweetlove & Fernie 2007; Noguchi & Yoshida 2008). In the research area concerned with the organelle crosstalk, the possible role of AOX in illuminated leaves has been discussed. Chemical inhibition of AOX suppressed photosynthetic performance in mesophyll protoplasts (Padmasree & Raghavendra 1999) and intact leaves (Bartoli et al. 2005; Yoshida, Terashima & Noguchi 2006). Furthermore, when exposed to combined light and drought stress, the Arabidopsis aox1a mutant showed a stress-related phenotype, a lowered photosynthetic efficiency and an elevated ROS generation (Giraud et al. 2008). It has been suggested that AOX serves to keep the cellular redox balance by efficiently dissipating excess reducing equivalents produced by the photochemical reactions and the photorespiratory glycine oxidation (Yoshida, Terashima & Noguchi 2007; Strodtkötter et al. 2009). AOX may also be involved in the control of ascorbate synthesis in the light (Bartoli et al. 2006).

To ensure optimal functionality in the light, the mitochondrial respiratory system needs to acclimate to a variable light environment. It has often been observed that the leaf respiratory rate is influenced by growth light conditions (e.g. Noguchi et al. 2005). Such changes in respiratory rates may be accompanied by the modifications of the respiratory components ranging from transcriptional to post-translational control. In fact, expression of several respiratory genes is subject to light and diurnal regulations (Rasmusson & Escobar 2007). AOX is one of the components showing clear light-dependent up-regulation at the transcript and protein levels (Svensson & Rasmusson 2001; Yoshida et al. 2008; Yoshida & Noguchi 2009). It was also demonstrated that the capacity or in vivo activity of AOX are enhanced in a light-dependent manner (Azcón-Bieto, Lambers & Day 1983; Atkin, Cummins & Collier 1993; Ribas-Carbo et al. 2000; Noguchi et al. 2001). These results suggest that the adjustment of the mitochondrial respiratory system, notably AOX, is one of the key strategies for plants to acclimate to a fluctuating light environment. However, the regulatory mechanisms of the respiratory system in the light remain to be fully clarified. For example, it is largely unknown whether the AOX up-regulation is specific in plants suddenly exposed to strong light conditions or is also observed in plants placed under such conditions for a long time. Considering that the light intensity in plant habitat drastically varies in a wide range of timescale, this is an important issue in an eco-physiological context. Furthermore, the impact of a lack of AOX on the light acclimation of the respiratory system is unclear.

In the present study, we assessed the responses of the respiratory chain to long- and short-term high light (HL) environments. To examine the effect of the AOX deficiency, we used aox1a T-DNA insertion mutant of Arabidopsis thaliana. The obtained results strongly suggest that the light-acclimatory manner of the respiratory chain is strikingly different between long- and short-term HL environments. The physiological significance of these distinct responses is discussed.

MATERIALS AND METHODS

Plant materials and growth conditions

Arabidopsis (Arabidopsis thaliana L. Heynh.) wild-type (Columbia-0) and aox1a T-DNA insertion line (SALK_084897) were grown in soil in a controlled growth chamber. Soil was composed of Metro Mix 350 (Sun Gro Horticulture, Bellevue, WA, USA), red clay (Green Tec, Tochigi, Japan) and vermiculite (Vern-piece; Hakugen, Tokyo, Japan) at the ratio of 3:1:1 in volume. Plants were grown under two different light conditions, low light (LL; 80 µmol photons m−2 s−1) and HL (350 µmol photons m−2 s−1). Day/night cycle was 12 h/12 h. Temperature was 22–24 °C. Plants just before the bolting stage [plants grown under LL (LL plants): 5–6 weeks, plants grown under HL (HL plants): 3–4 weeks] were used for the experiments.

For short-term HL stress (sHL) treatment, LL plants were transferred to a new environment with a light intensity of 650 µmol photons m−2 s−1. This transfer took place after a dark incubation for 12 h (i.e. at the end of night period). During sHL treatment, temperature was kept at the same level as during the growth conditions.

Estimation of leaf day mass per area

For the calculation of leaf dry mass per area (LMA), the leaf area was measured using a scanner and software, ImageJ (National Institute of Mental Health, Bethesda, MD, USA). Then, the leaves were dried at 80 °C for 48 h, allowing measurement of dry weight.

Chlorophyll (Chl) measurements

Chl content and a/b ratio were determined spectrophotometrically after extraction with 80% (v/v) acetone according to Porra, Thompson & Kriedemann (1989).

Protein analysis

Protein extraction from whole leaf tissue was performed as previously described (Yoshida et al. 2007). Total protein content was determined according to Peterson (1977). The extracted proteins were separated by 12.5% sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE). For immunoreaction experiments, proteins were transferred to a polyvinylidene fluoride membrane (Hybond-P, GE Healthcare, Piscataway, NJ, USA). Immunodetection of AOX was performed with monoclonal antibody (Elthon, Nickels & McIntosh 1989) at a dilution of 1:50. Immunodetection of cytochrome c oxidase subunit II (COXII) and light-harvesting complex II protein (Lhcb1) was performed with polyclonal antibodies commercially available (Agrisera, Vännäs, Sweden) at a dilution of 1:1000 and 1:5000, respectively. For immunodetection of other mitochondrial proteins, F1-ATP synthase α and β subunits, pyruvate dehydrogenase complex E2 subunit, chaperonin 60 and voltage-dependent anion channel (VDAC), monoclonal antibodies designed by GT Monoclonal Antibodies (Lincoln, NE, USA) were used. These antibodies were used at a dilution of 1:100, except for VDAC (1:1000). Immunodetection of ribulose 1·5-bisphosphate carboxylase/oxygenase (Rubisco) large and small subunits was performed using polyclonal antibody provided Dr. M. Nakazono at a dilution of 1:10 000. Chemiluminescence was used for detection of horseradish peroxidase-conjugated secondary antibodies and visualized using ECL Advance Western Blotting Detection Kit (GE Healthcare) and LAS 4000 (Fuji, Tokyo, Japan). For quantification of visualized protein, Multi Gauge software (Fuji) was used. Experiments were repeated at least twice using independent sample preparations, and representative results are shown.

Determination of quinone contents and redox states

Detailed methods for the extraction of quinones from leaf tissues and the high-performance liquid chromatography (HPLC)-based analysis of quinone contents were described in our previous study (Yoshida et al. 2010). In brief, the extraction of quinones was conducted using freezer-cold acetone after the homogenization of leaves in liquid nitrogen. To prevent undesired changes in the quinone redox state, this process was conducted as quickly as possible. The extraction using acetone was repeated twice, and finally hexane was used to retrieve the residual quinones.

The HPLC system was composed as follows. As detectors, a photodiode array detector (SPD-M20A, Shimadzu, Kyoto, Japan) and a fluorescence detector (RF-10A, Shimadzu) were used. Separation of quinones was carried out by a reverse-phase column (TSKgel ODS-100V, ϕ4.6 × 150 mm, particles 3 µm, Tosoh, Tokyo, Japan) at 40 °C. Flow rate was 1.5 mL min−1[acetonitrile/ethanol = 3/1 (v/v)].

Although plant UQ is classified into two forms, UQ-9 and UQ-10, by the length of a prenyl side chain, only UQ-9 is involved in the respiratory electron transport in the Arabidopsis leaf (Yoshida et al. 2010). Thus, UQ-9 was referred to as UQ in the present study. We previously confirmed that the change in the redox status of quinones during the extraction and analysis was only marginal (Yoshida et al. 2010).

Extraction of total RNA and determination of transcript levels by real-time PCR

Total RNA was extracted using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. cDNA was obtained from the extracted RNA by reverse transcription using a ThermoScript RT-PCR System (Invitrogen) with oligo(dT)20 as the primer.

Transcript levels were measured using a 7000 Sequence Detection System (Applied Biosystems, Foster City, CA, USA). cDNA (1 µL) was amplified in the presence of 12.5 µL 2 × Power SYBR Green PCR Master Mix (Applied Biosystems), 0.5 µL specific primers (final concentration, 0.2 µm) and 10.5 µL sterilized water. PCR conditions were as follows: 50 °C for 2 min, 95 °C for 10 min and 40 cycles of 95 °C for 15 s followed by 60 °C for 1 min. The transcript levels were calculated by the comparative cycle threshold method. For internal control gene, 18 s rRNA was used. The primer sequences and AGI numbers of each gene are shown in Supporting Information Table S1.

Statistical analysis

Statistical analyses were conducted with Microsoft Excel software for the Student's t-test, and SPSS 12.0J software (SPSS, Chicago, IL, USA) for Tukey–Kramer's multiple comparison test.

RESULTS

Responses of the respiratory chain to different growth light conditions

Arabidopsis wild type and aox1a mutant were grown under two different light intensities, LL (80 µmol photons m−2 s−1) and HL (350 µmol photons m−2 s−1). HL plants grew faster and had a shorter life cycle than LL plants (Fig. 1a). HL plants attained the bolting stage at 3–4 weeks after sowing, while LL plants needed 5–6 weeks. HL plants developed typical sun-type leaves, characterized by a greater LMA, a lower Chl content per fresh weight (FW), and a higher Chl a/b ratio, than LL plants (Fig. 1b). Development of sun-type leaves in HL plants was also evidenced by a lower content of light-harvesting complex II protein and a higher content of Rubisco on a leaf protein basis (Fig. 1c). Higher content of plastoquinone (PQ) on a Chl basis in HL plants supports the idea that HL plants had higher photosynthetic capacity than LL plants (Fig. 1b). There was no phenotypic difference between the wild type and aox1a irrespective of growth light conditions.

Figure 1.

Phenotypic characteristics of the Arabidopsis wild type (WT) and aox1a mutant grown under high light (HL) and low light (LL). A, Images of plants grown under LL (LL plants) and HL (HL plants). The pictures were obtained 40 d (LL plants) and 25 d (HL plants) after sowing. B, Leaf characteristics of LL and HL plants. Leaf dry mass per area (LMA), chlorophyll (Chl) content, Chl a/b ratio and plastoquinone (PQ) content were examined. Each value represents the mean ± SE (n = 5–12). Different letters denote a statistically significant difference among the plants (P < 0.001, Tukey–Kramer's multiple comparison test). C, Immunodetection of light harvesting complex II protein (Lhcb1), Rubisco large subunit (RbcL) and small subunit (RbcS). The same amount of leaf total protein was loaded into each lane.

We examined the amount of some proteins localized in the respiratory chain by immunoblotting (Fig. 2a). The AOX amount in the wild type was comparable between LL and HL plants. In contrast, the amount of COXII was greater in HL plants than in LL plants. We quantified the COXII amount and estimated that the COXII amount in HL plants was 1.6- to 1.8-fold higher than that in LL plants. Other mitochondrial proteins addressed in the present study, F1-ATP synthase subunits and VDAC, did not show significant difference in their amounts between LL and HL plants. In addition, a deficiency of AOX had little effect on the amount of mitochondrial proteins under both light conditions.

Figure 2.

Leaf mitochondrial properties of the Arabidopsis wild type (WT) and aox1a mutant grown under high light (HL) and low light (LL). A, Immunodetection of mitochondrial proteins. Proteins examined were alternative oxidase (AOX), cytochrome c oxidase subunit II (COXII), F1-ATP synthase α and β subunits (F1-ATPase α and F1-ATPase β, respectively), and voltage-dependent anion channel (VDAC). Leaf total protein was extracted during the day period (5–7 h after the onset of day period) and the same amount of protein was loaded into each lane. B, Ubiquinone (UQ) content [(upper panel: oxidized form (UQox), lower panel: reduced form (UQred)]. C, UQ reduction level expressed as the ratio of UQred content to the total. UQ was extracted during the night period (at the end of night period) and day period (5–7 h after the onset of day period). Each value represents the mean ± SE (n = 6–8). Different letters denote a statistically significant difference among the plants (P < 0.05, Tukey–Kramer's multiple comparison test).

We then assessed the pool size and redox state of UQ, an electron carrier in the respiratory chain. Quinones were extracted during the day and night periods. To avoid undesired changes in the UQ redox state, we immediately extracted quinones from the leaves without measuring the leaf FW. As quinones were extracted with acetone, the Chl content could be measured. Therefore, the UQ content was expressed on a Chl basis (Fig. 2b). HL plants possessed a higher content of both oxidized and reduced forms of UQ (UQox and UQred) compared with LL plants. Even when normalized by FW using the data shown in Fig. 1b, the larger size of UQ pool in HL plants was evident (Supporting Information Fig. S1). UQ reduction level, expressed as the ratio of UQred content to the total, was maintained at a stable level irrespective of growth light conditions or the day/night period (Fig. 2c). Although the UQox content in HL plants was 25% lower in the day than in the night (Fig. 2b), the UQ reduction level was not significantly different between day and night (Fig. 2c). A deficiency of AOX had only a marginal effect on the UQ pool size and redox state. Results shown in Fig. 2 suggest that AOX is not involved in the acclimation of the respiratory chain to long-term HL in Arabidopsis leaves.

Responses of the respiratory chain to short-term HL stress

Next, we examined the response of the respiratory chain when LL plants were exposed to sHL (650 µmol photons m−2 s−1). Using the HPLC system for the simultaneous determination of PQ and UQ redox states (Yoshida et al. 2010), we examined the transient pattern of PQ and UQ redox states after the exposure to sHL (Fig. 3). As mentioned earlier, quinones were extracted from the leaves as quickly as possible. The data shown in Fig. 3 are represented on a Chl basis. We confirmed that the Chl content per FW did not change during sHL (Supporting Information Fig. S2).

Figure 3.

Changes in the content and reduction level of quinones [plastoquinone (PQ) and ubiquinone (UQ)] after short-term high light stress (sHL) in the Arabidopsis wild type (WT) and aox1a grown under low light. Oxidized (ox) and reduced (red) forms of quinones were quantified. The reduction level was expressed as the ratio of the content of the reduced form to the total. Data for WT (0–30 min sHL) were re-plotted from Yoshida et al. (2010). Each value represents the mean ± SE (n = 5–8). Asterisks denote a statistically significant difference between WT and aox1a (P < 0.05, Student's t-test).

As shown in Fig. 3, exposure of plants to sHL resulted in a decrease in oxidized form of PQ (PQox) and an increase in reduced form of PQ (PQred). These responses were observed immediately after the transfer to sHL (within 2 min). Consequently, the PQ reduction level was rapidly elevated by sHL. This transient reduction of the PQ pool was followed by a gradual re-oxidation. The UQ reduction level was also transiently elevated by sHL, but it showed a trend distinct from that for PQ. The sHL-dependent decrease in UQox and increase in UQred occurred more slowly than the corresponding changes in PQ. A saturating level for the reduction of the UQ pool was attained 10–20 min after the transfer to sHL. We argued in a previous study that this time-lagged elevation of the UQ reduction level may be caused by the metabolic interaction with the chloroplasts (Yoshida et al. 2010). The changes in the PQ and UQ reduction levels showed similar transient patterns in the wild type and aox1a until 2 h after the transfer to sHL. Further exposure to sHL differently affected the UQ reduction level between the wild type and aox1a. The UQ reduction level in the wild type was maintained at a moderate level at 4 and 8 h after the transfer to sHL, while that in aox1a was kept at significantly higher level compared with the wild type. Interestingly, the PQ reduction level was also significantly higher in aox1a than in the wild type at 4 h after the transfer to sHL.

To assess the underlying mechanisms responsible for the sHL-dependent variations of quinone redox states and the differences observed between the wild type and aox1a (Fig. 3), we examined the amount of several mitochondrial proteins and the transcript abundance of genes encoding respiratory chain components. Figure 4 shows the effect of sHL on the amount of several mitochondrial proteins. The AOX amount in the wild type clearly increased in a sHL-dependent manner. Quantification indicated that the AOX amount in the wild type exhibited a more than twofold increase at 4–8 h after the transfer to sHL. COXII, which accumulated at higher levels in HL plants than in LL plants (Fig. 2a), did not change in the amount by sHL. Similarly, the amount of other mitochondrial proteins showed no significant change. As expected, sHL-dependent AOX up-regulation at a protein level was not observed in aox1a. The amount of other mitochondrial proteins in aox1a was comparable with that in the wild type and did not show significant changes as a result of the sHL treatment.

Figure 4.

Changes in the amount of mitochondrial proteins after short-term high light stress (sHL) in the Arabidopsis wild type (WT) and aox1a grown under low light. Proteins examined were alternative oxidase (AOX), cytochrome c oxidase subunit II (COXII), F1-ATP synthase α and β subunits (F1-ATPase α and F1-ATPase β, respectively), pyruvate dehydrogenase complex E2 subunit (PDC E2), chaperonin 60 (CPN60) and voltage-dependent anion channel (VDAC). The same amount of leaf total protein was loaded into each lane.

We previously reported the Arabidopsis respiratory genes whose expression is regulated by light (Yoshida & Noguchi 2009). With a focus on such light-regulated genes, we investigated the effect of sHL on the transcript abundance of respiratory genes and on possible differences between the wild type and aox1a (Fig. 5). The AOX1a transcript level in the wild type sharply increased from 10 to 60 min after the transfer to sHL and attained a maximum that was 20–35-fold the level before the onset of sHL. Among other AOX genes, only AOX1c showed a clear sHL-dependent increase in the transcript level. The expression of genes encoding type II NAD(P)H dehydrogenases, NDA1, NDB2 and NDC1, was also enhanced by sHL, although it was also likely that the fluctuations of NDA1 and NDB2 transcript levels were a result of diurnal regulation (Michalecka et al. 2003; Elhafez et al. 2006). The expression of other genes addressed here, UCP1[encoding uncoupling protein (UCP)], COX6b (encoding COX subunit 6b) and CI76 (encoding complex I 76 kD subunit), was only weakly influenced by sHL, and the fluctuations of these transcripts were limited to 0.5–2.5-fold the levels before the treatment. One remarkable feature of the data shown in Fig. 5 is that aox1a did not show any significant modification of the expression profile in respiratory genes assessed in the present study (except for AOX1a).

Figure 5.

Changes in the transcript levels of respiratory genes after short-term high light stress (sHL) in the Arabidopsis wild type (WT) and aox1a grown under low light. The transcript levels were calculated from the difference in the threshold cycle (Ct) between 18 s rRNA and each gene. AGI numbers and primer sequences are shown in Supporting Information Table S1. Each value represents the mean ± SE (n = 4).

DISCUSSION

Among five AOX genes in Arabidopsis, AOX1a is the major isoform in the leaf (Clifton, Millar & Whelan 2006; see also the absolute transcript level shown in Fig. 5). It was previously demonstrated that the Arabidopsis aox1a mutant has no detectable AOX protein (Giraud et al. 2008) and shows little cyanide-resistant respiratory activity (Watanabe et al. 2008). In the present study, we also confirmed that the accumulation of AOX protein was fully impaired in aox1a (Figs. 2 & 4). Using this mutant, we evaluated the acclimatory manner of the respiratory chain to long- and short-term HL conditions and the physiological role of AOX under HL environments.

CP capacity is enhanced under long-term HL conditions and supports high ATP production

Our previous studies demonstrated that the AOX protein amount increased in response to substantially strong light (300–400 µmol photons m−2 s−1) in Arabidopsis (e.g. Yoshida & Noguchi 2009). Thus, we anticipated that HL plants would possess a higher AOX amount than LL plants, but our results showed that the AOX amount was not different between LL and HL plants (Fig. 2a). Instead, COXII was up-regulated in HL plants (Fig. 2a). Furthermore, as a notable characteristic of the respiratory chain in HL plants, a larger size of UQ pool was observed (Fig. 2b).

COX consists of multiple subunits, with each subunit showing a coordinated accumulation at the protein level (Giegéet al. 2005). Therefore, a high amount of COXII observed in HL plants most likely indicates a high accumulation of COX at the level of protein-complex assembly, resulting in high COX capacity. Not only COX, but also the phosphorylating respiratory components including complex I and III might be entirely up-regulated in HL plants. A co-expression analysis using ATTED-II (http://atted.jp/) indicated that several phosphorylating respiratory components show a coordinated gene expression pattern (data not shown). Given that HL plants also possessed a larger UQ pool than LL plants, it is conceivable that HL plants had higher respiratory capacity ensured by the preferable up-regulation of the phosphorylating pathway, leading to higher ATP production (Fig. 6). This idea is supported by some studies showing the relationship between the light intensity of growth conditions and the leaf respiratory rate (e.g. Sims & Pearcy 1991; Noguchi, Nakajima & Terashima 2001). For example, spinach grown under HL condition (490 µmol photons m−2 s−1) showed a higher respiratory rate than that grown under LL condition (200 µmol photons m−2 s−1) (Noguchi et al. 2005). In this case, the COX protein amount and capacity were enhanced in the HL-grown plant, as observed in the present study.

Figure 6.

Simplified model of the distinct responses of the mitochondrial respiratory chain to long- and short-term high light environments in Arabidopsis. AOX; alternative oxidase, COX; cytochrome c oxidase, MRR; mitochondrial retrograde regulation, UQ; ubiquinone.

It has been demonstrated that the electron flow to CP and the theoretical ATP yield via oxidative phosphorylation were positively correlated with the relative growth rate of plants (Millar et al. 1998; Florez-Sarasa et al. 2007). Although we did not estimate the in vivo electron partitioning to CP, the faster growth of HL plants (Fig. 1) may thus be facilitated by the higher ATP production. For example, mitochondrial ATP production supports the sucrose synthesis in the cytosol, which may be important for sustaining an increased photosynthetic capacity in sun-type leaves of HL plants (Krömer, Stitt & Heldt 1988).

Intriguingly, a deficiency of AOX did not affect the UQ redox state under long-term HL conditions (Fig. 2b). Therefore, AOX does not appear necessary as a stabilizer of the UQ redox state in this case.

AOX is induced in response to sHL and prevents over-reduction of the respiratory chain

In accordance with our previous studies, sHL-dependent AOX induction both at transcript and protein levels was observed in the wild type (Figs. 4 & 5). Concomitant with AOX protein accumulation, the UQ reduction level was maintained at a modest level (Fig. 3). On the other hand, aox1a failed to lower the UQ reduction level under sHL. Several studies have demonstrated that AOX deficiency results in compensatory expression of other anti-oxidant genes, which may engage in re-adjustment of cellular redox status (Fiorani, Umbach & Siedow 2005; Amirsadeghi et al. 2006; Giraud et al. 2008; Watanabe et al. 2008). When Arabidopsis AOX1a gene expression is impaired, expression of other AOX isoforms may be induced under certain situations (Strodtkötter et al. 2009). However, we were not able to observe such compensatory induction in aox1a (Figs. 4 & 5). Furthermore, as the expression pattern of other respiratory genes was not modified in aox1a (Fig. 5), it is not likely that non-phosphorylating bypass pathways including the type II NAD(P)H dehydrogenases and UCP could mitigate the disturbance of the UQ redox state in aox1a. These results suggest that AOX1a plays a critical role in determining the UQ reduction level in Arabidopsis leaves, at least under conditions employed in the present study.

Some studies have reported that, apart from the reduction during thermogenesis (Wagner, Wagner & Moore 1998), the redox state of UQ pool is usually kept at a stable level even under environmental stresses (Wagner & Wagner 1995; Millenaar et al. 2000). This homeostatic property is possibly a result of the systems finely and flexibly regulating the UQ reduction level; such systems would be essential for preventing mitochondrial ROS generation and subsequent deleterious effects on mitochondrial metabolism (Sweetlove & Foyer 2004). It is obvious that, when suddenly exposed to stress-inducing HL conditions, leaf AOX is up-regulated and regulates the UQ redox state (Fig. 6). This function of AOX, possibly in cooperation with other non-phosphorylating respiratory components, enables a more flexible metabolism in the mitochondria (e.g. photorespiratory glycine oxidation) and even in the whole cell. Notably, the PQ reduction level was slightly, but significantly, higher in aox1a than in the wild type under sHL (Fig. 3), implying that the efficient photosynthetic electron transport was disturbed in aox1a. This may be caused by an impairment of the photorespiratory metabolism and an accumulation of cellular reducing equivalents (Noguchi & Yoshida 2008). However, leaf PQ is partly electron transport-independent (Kruk & Karpinski 2006). Therefore, a more detailed analysis is required to draw conclusions about the alteration of photosynthetic characteristics in aox1a.

How is AOX induced under sHL?

Several studies using cultured cells have demonstrated that AOX expression is strongly induced in response to the inhibition of respiratory electron transport and resulting reduction of UQ pool (e.g. Clifton et al. 2005). This regulatory manner is now well known as a major pathway of mitochondrial retrograde regulation (MRR; Rhoads & Subbaiah 2007). We previously suggested that, also in the Arabidopsis leaf, light-dependent induction of AOX1a expression is achieved in this way (Yoshida & Noguchi 2009). In the present study, AOX1a expression was induced within 10–60 min after the transfer to sHL (Fig. 5). During this period, the UQ pool was markedly reduced (Fig. 3). This linkage supports the idea that light-dependent induction of AOX1a expression is driven via MRR (Fig. 6). Nevertheless, further studies addressing the association between the UQ redox state and the activity of AOX1a promoter region are needed to prove this hypothesis.

Concluding remarks

Our work revealed that the acclimatory manner of the respiratory chain to long- and short-term HL conditions is vastly different (Fig. 6). Under long-term HL, CP, but not AOX, is preferentially up-regulated. This may serve to supply more ATP and sustain the faster plant growth rate. By contrast, when plants are suddenly exposed to excess light, the respiratory chain becomes over-reduced state and AOX expression is enhanced. AOX then stabilizes the redox balance of the respiratory chain. This role of AOX may be essential for an optimal photosynthetic performance. In other words, AOX acts as a rescue system, enabling a rapid and flexible re-adjustment of intra-cellular environment under stressful light conditions.

ACKNOWLEDGMENTS

We are grateful to Drs. T.E. Elthon and M. Nakazono for their generous donation of antibodies. We also acknowledge Drs M. Kawaguchi, H. Hirano and M. Taira for the usages of instruments. This study was financially supported by the Ministry of Education, Science, Sports and Culture (Nos. 16207002, 19039009, 30300560) and Japan Society for the Promotion of Science (to K.Y.).

Ancillary