Molecular Plant • Volume 7 • Number 1 • Pages 58–70 • January 2014
RESEARCH ARTICLE
Linking Chloroplast Antioxidant Defense to Carbohydrate Availability: The Transcript Abundance of Stromal Ascorbate Peroxidase Is Sugar-Controlled via Ascorbate Biosynthesis a Former address: Plant Physiology and Biochemistry, Bielefeld University, Universitätsstraße 25, 33615 Bielefeld, Germany b Plant Physiology, Free University of Berlin, Dahlem Center of Plant Sciences, Königin-Luise-Str. 12–16, 14195 Berlin, Germany
ABSTRACT All genes encoding chloroplast antioxidant enzymes are nuclear-encoded and posttranscriptionally targeted to chloroplasts. The transcript levels of most of them decreased upon sucrose feeding like the transcript levels of many genes encoding components of the photosynthetic electron transport chain. However, the transcript abundance of stromal ascorbate peroxidase (s-APX; At4g08390) increased. Due to mild sugar application conditions, the plants kept the phosphorylation status of the ADP+ATP pool and the redox states of the NADPH+NADP+ and the ascorbate pools under control, which excludes them as signals in s-APX regulation. Correlation with ascorbate pool size regulation and comparison of transcript abundance regulation in the starch-biosynthetic mutant adg1, the ascorbate biosynthesis mutant vtc1, and the abscisic acid (ABA) biosynthetic mutant aba2 showed a link between sugar induction of s-APX and ascorbate biosynthesis. Key words: ascorbate; antioxidant; carbohydrate; chloroplast; gene expression; stromal ascorbate peroxidase.
Introduction Sugar availability regulates the expression of many genes for chloroplast proteins (Rook and Bevan, 2003; Rolland et al., 2006). While low sugar concentrations support seedling growth, excess carbohydrates antagonize greening and seedling development (Koch, 1999). Carbohydrate surplus inhibits the Calvin-Cycle (Macdonald and Buchanan, 1992) and promotes generation of reactive oxygen species (ROS) via feedback inhibition of photosynthetic electron transport (Rook et al., 2006a). Antioxidant enzymes and low-molecular-weight antioxidants, such as ascorbate (Asc) and glutathione (GSH), form a ROS-protective network (Asada, 1999; Dat et al., 2001; Dietz et al., 2002; Chang et al., 2009). Stromal and thylakoidbound ascorbate peroxidase (s-APX (At4g08390) and t-APX (At1g77490)) detoxify H2O2 on the expense of ascorbate (Asada, 1999). The co-substrate is regenerated by monodehydroascorbate and dehydroascorbate reductases (MDHAR and DHAR). In parallel, glutathione peroxidases (GPX) and peroxiredoxins (PRX) reduce peroxides via ascorbate-independent thiol-mediated pathways (Dietz et al., 2002). These enzymes are nuclear-encoded and posttranslationally targeted to the organelles by N-terminal transit peptides (Pitsch et al., 2010). Most proteins, like the four peroxiredoxins and t-APX, for
example, are exclusively targeted to chloroplasts. s-APX and MDHAR (At1g63940) can be alternatively targeted to chloroplasts and (the intermembrane space of) mitochondria (Chew et al., 2003), yet they display strong preference towards chloroplasts. To acclimate the chloroplast antioxidant protection upon (photo-)oxidative stress, the genes for chloroplast antioxidant enzymes respond to organellar signals. Organelle-to-nucleus signaling has been best studied for photosynthesis associated genes, like cab (encoding chlorophyll-a/b-binding proteins/light-harvesting complex proteins) and rbc-S (encoding the small subunits of ribulose-1,5-bisphosphate carboxylase/oxygenase). They are suppressed by carbohydrates (Arenas-Huertero et al., 2000), while apl3 (encoding a large subunit of ADP-glucose pyrophosphorylase; At4g39210) is induced by sugars (Rook et al., 2006b).
1 To whom correspondence should be addressed. E-mail margarete.baier@ fu-berlin.de, fax +49-30-838-51688, tel. +49-(0)30-838-53183
© The Author 2013. Published by the Molecular Plant Shanghai Editorial Office in association with Oxford University Press on behalf of CSPB and IPPE, SIBS, CAS. doi:10.1093/mp/sst154, Advance Access publication 7 November 2013 Received 25 June 2013; accepted 27 October 2013
Downloaded from http://mplant.oxfordjournals.org/ at University of California, San Diego on November 17, 2014
Isabelle Heibera, Wenguo Caib, and Margarete Baiera,1
Heiber et al. • s-APX Transcript Abundance Is Regulated by Ascorbate
RESULTS Optimization of the Experimental Set-Up In nature, the mucilage of the outer seed coat provides the first carbohydrate source upon germination. In experiments, most of the mucilage is washed off during seed sterilization. Here, Arabidopsis was screened for the optimal background sucrose (Suc) concentration prior to gene-expression analyses. In continuous moderate light (50 μmol quanta m–2 s–1), the seedling biomass increased on Suc-supplemented MS-media up to 0.8% (w/v) Suc (data not shown). With Suc concentrations higher than 1.6% (w/v), anthocyanins accumulated in hypocotyls and cotyledons indicating excess carbohydrates (Solfanelli et al., 2006). Suc concentrations higher than 2.5% (w/v) resulted in growth inhibition. Based on these
observations, 1% (w/v) Suc was chosen as optimal sugar concentration and 0.5% (w/v) Suc was used to establish moderate sugar starvation. Sorbitol (Sor) was used as an osmotic control. Compared to Suc, it is taken up and metabolized much less efficiently (Shabala and Lew, 2002; Gibson, 2005). Chemically, the osmolarities of Suc and Sor are similar. In planta, hydrolysis of Suc into glucose and fructose increases the osmolarity and the heterotrophic metabolism decreases it. The Sor concentration best suited as an osmotic control for Suc treatments was determined empirically based on biomass comparison. At 10 d, the fresh weights of seedlings grown on 1% (w/v) Suc supplemented with 0.8–1.3% (w/v) Sor were in the range of seedlings grown on 2% (w/v) Suc (data not shown). Finally, 1% Sor was chosen as control for 2% Suc application.
Long-Term Carbohydrate Effect in Arabidopsis Seedlings Long-Term Carbohydrate Effect on the Transcript Abundance of Nuclear-Encoded Chloroplast Proteins Long-term effects of Suc and Sor on the expression of genes encoding chloroplast antioxidant enzymes were studied in ACTIN-2-standardized cDNA samples by RT–PCR (Figure 1). As secondary control, the transcript level of ubiquitin-11 (UBQ11; At4g05050) was quantified. Relative to ACTIN-2 transcript levels, the UBQ11 transcript levels were not significantly changed by any of the treatments (Figure 1). APL3 (encoding a large subunit of ADP-glucose-pyrophosphorylase) served as a control for induction by Suc (Rook et al., 2001). RBC-S (encoding the small subunit of Ribulose-1,6-bisphosphatcarboylase/oxygenase; At5g38430) and STP1 (encoding a high-affinity monosaccharide/proton symporter; At1g11260) were included to monitor suppression by surplus amounts of carbohydrates (Figure 1; Koch, 1999; Sherson et al., 2000). This study focuses on regulation of chloroplast antioxidant enzymes. On 2% (w/v) Suc, the transcript levels of the four chloroplast peroxiredoxins, 2CPA (At3g11630), 2CPB (At5g06290), Prx-Q (At3g26060), and Prx-IIE (At3g52960), of CuZn-superoxide dismutase 2 (CSD2; At2g28190), MDHAR (At1g63940), t-APX (At1g77490), the Rieske protein PET-C (At4g03280), the photosystem-II antenna protein LHCB2.2 (At2g05070), and of the cytochrome b6f proteins PET-M (At2g26500) were decreased compared to 1% (w/v) Suc and the osmotic control containing 1% (w/v) Sor demonstrating a broad inactivation of genes encoding chloroplast proteins. On 3% (w/v) Suc, the transcript levels were below those observed on 1% (w/v) and 2% (w/v) Suc. On 1% (w/v) Suc supplemented with 2% (w/v) Sor, they were barely detectable (Figure 1). PET-E1 (At1g76100) and PET-E2 (At1g20340) levels, encoding plastocyanins, were unchanged irrespective of the treatment. The s-APX mRNA level was increased on 2% (w/v) Suc, but not on 1% (w/v) Suc supplemented with 1% (w/v) Sor demonstrating Suc-dependent induction.
Downloaded from http://mplant.oxfordjournals.org/ at University of California, San Diego on November 17, 2014
Screens for sugar signaling mutants demonstrated a strong, but diverse, crosstalk of ABA, sugar, and ROS/redox signaling (Rook et al., 2001; Gonzalez-Guzman et al., 2002; Staneloni et al., 2008). So far, little is known about the transcriptional regulation of genes encoding chloroplast antioxidant enzymes. Array studies and RT–PCR demonstrate that transcript abundances often increase only by mild stress, but are decreased upon severe stress due to general inactivation of chloroplast function (summarized in Baier et al. (2010)). Regulation of the 2-Cys peroxiredoxin-A gene (At3g11630) has been investigated in detail (Baier et al., 2004b): cis-acting motifs located within 200 bp upstream of the transcription initiation site mediate developmental responses. Photosynthetic redox signals and ABA-signals integrate antagonistically at a more distally located promoter region, designated as redoxbox. There, the APETALA-2-type transcription factor RAP2.4a induces 2cpa transcription upon moderate oxidative stress (Shaikhali et al., 2008). Upon severe stress, RAP2.4a oligomerizes. As a consequence, 2cpa transcription activity decreases (Shaikhali et al., 2008). Transcriptional regulation of other genes for chloroplast antioxidant enzymes has not been studied so far. Mutants impaired in redox regulation of the 2cpa promoter (Heiber et al., 2007) and knockout lines of the 2cpa redox-regulating transcription factor RAP2.4a (Shaikhali et al., 2008) show links between s-APX, t-APX, and 2CPA in redox regulation. To investigate the impact of the photosynthetically controlled parameters, carbohydrate availability, light, and ABA on the expression of nuclear genes for chloroplast antioxidant enzymes, here the transcript abundances were analyzed in Arabidopsis following light and sugar treatments. The responses were compared to those in the ascorbate-biosynthetic mutant vtc1 (Conklin et al., 1997), the starch-biosynthetic mutant adg1 (Lin et al., 1988), and the ABA-biosynthetic mutant aba2 (Léon-Kloostersiel et al., 1996) after short- and long-term Suc treatment. In this comparison, a specific regulation is shown for s-APX. It is concluded that s-APX regulation is indirectly sugar-responsive via ascorbate biosynthesis.
59
60
Heiber et al. • s-APX Transcript Abundance Is Regulated by Ascorbate
Chlorophyll Levels after Long-Term Treatment After 10-day growth on MS medium supplemented either with 1–3% (w/v) Suc or 1% (w/v) Suc + 1% or 2% (w/v) Sor, the chlorophyll (chl) contents were very similar in all samples (Figure 2) reflecting acclimation and the mildness of the treatments.
Sugar Metabolite Levels and Energetization and Carbohydrate Redox Status after Long-Term Sugar Treatment
Ascorbate Levels and the Ascorbate Redox State after Long-Term Sugar Treatment Total ascorbate and reduced ascorbate contents were measured to determine the ascorbate pool size and to calculate the ascorbate redox state. The ascorbate level was 3.7-fold increased on 2% (w/v) Suc, 3.9-fold on 3% (w/v) Suc, and 1.6-fold on 1% (w/v) Suc + 1% (w/v) Sor and 1% (w/v) Suc + 2% (w/v) Sor compared to plant material grown on 1% (w/v) Suc (Figure 2), demonstrating a carbohydrate flux into the ascorbate pool. The ascorbate redox state was not affected (Figure 2).
Anthocyanin Levels after Long-Term Treatment In response to excess carbohydrate availability, Arabidopsis accumulated anthocyanins (Solfanelli et al., 2006). The anthocyanin contents were 2.73-, 4.35-, 1.32-, and 1.83-fold increased after 10-day growth on 2% and 3% (w/v) Suc and 1% (w/v) Suc supplemented with 1% or 2% (w/v) Sor, respectively (Figure 2).
Effect of Short-Term Sucrose Application on Arabidopsis thaliana To analyze short-term responses, the seedlings grown on 0.5% (w/v) Suc were either floated on liquid MS medium containing 0.5 or 1.5% (w/v) Suc for 24 h. For a set of 22 genes, the transcript
Effect of Ascorbate, Starch, and ABA-Biosynthesis on the Suc Effect The use of mutants enables short- and long-term studies with modified strains. The mutant vtc1 carries a mutation in GDPD-mannose pyrophosphorylase which catalyzes a final step in the main ascorbate biosynthesis pathway (Conklin et al., 1999). Adg1 has a point mutation in the small subunit of ADP-glucose pyrophosphorylase and is limited in chloroplast starch biosynthesis, while aba2 is restricted in ABA-biosynthesis (Rook et al., 2001; Cheng et al., 2002). Here, the mutant in short- and longterm acclimation response was compared to that of wildtype (wt) plants.
Long-Term Effect of Suc on Transcript Abundance in the Mutants vtc1, adg1, and aba2 After 10-day growth on 2% (w/v) Suc, the transcript levels of most of the analyzed genes were decreased in vtc1, adg1, and aba2 (Figure 4, top). Exceptions were s-apx and apl3, whose transcript levels were increased. The PET-E1, PET-E2, and PET-C transcript levels were barely altered. Consistently with the hypothesis that ABA signaling supports the sugar effect on the apl3 promoter (Rook et al., 2001), the transcript levels of APL3 were most induced in aba2. s-APX regulation resembled APL3 induction. s-APX and APL3 transcripts were similarly accumulated in adg1 as they were in wt, indicating that starch biosynthesis capacities have a minor effect on the induction. Using the vtc1 mutant distinguished s-APX and APL3 regulation: s-APX transcripts accumulated less in vtc1 than in wt. In contrast, the APL3 transcript levels increased in a wt-like manner. This demonstrates a specific sucrose-linked impact of ascorbate biosynthesis on s-APX regulation. APL3 transcript levels were (consistently with previous reports by Rook et al. (2001)) induced in wt in response to sorbitol and decreased when ABA signaling was affected. Similarly, the transcript levels of most other tested genes were slightly increased in adg1, decreased in aba2, and was hardly affected in vtc1. In contrast, s-APX levels were decreased in all four tested genotypes in response to sorbitol application (Figure 4, bottom). The effect was strongest in aba2 and vtc1,
Downloaded from http://mplant.oxfordjournals.org/ at University of California, San Diego on November 17, 2014
To study the effect of sugar feeding on the metabolic signature, DHAP (dehydroxyacetone phosphate), FruBP (fructose6-bisphosphate), 3PGA (3-phosphoglycerate), ATP (adenosine triphosphate), ADP (adenosine diphosphate), ascorbate, and anthocyanin levels were analyzed as indicators for the cellular carbohydrate, redox, and energy status. Most metabolite levels were only slightly changed between the treatments, demonstrating the mildness of the treatments (Figure 2). The ATP and ADP contents increased in response to Suc application. However, the ATP/ADP ratio, like the 3PGA/DHAP ratio, which reflects the carbohydrate redox status, and the redox status of the NADPH pool (NADPH+NADP+), and, consequently, the [ ATP ] [NADPH ] assimilatory force, (Dietz and Heber, 1989), [ ADP ][Pi ] NADP + ] were hardly affected, demonstrating that the metabolic control capacity was not exhausted in the seedlings (Figure 2).
abundance regulation factors were calculated from the transcript levels quantified for ACTIN-2 standardized cDNA samples. The transcript levels of all genes encoding chloroplast antioxidant enzymes, except again s-APX, and, of all tested genes, encoding proteins of the thylakoid membrane (LHCA5– PET-E2) and of RBC-S were decreased in response to Suc application (Figure 3). The transcripts of the sugar-inducible apl3 gene accumulated and that of the sugar-suppressible stp1 were less abundant. The transcript levels of the ROS-marker genes bap1 (BONZAI associated protein I; At3g61190) and fer1 (Ferritin-1 precursor; At5g01600) (op den Camp et al., 2003) were hardly affected, demonstrating the mildness of the conditions.
Heiber et al. • s-APX Transcript Abundance Is Regulated by Ascorbate
61
demonstrating the impact of ABA and ascorbate biosynthesis on the osmotic regulation of s-APX.
Short-Term Effect of Sucrose on Transcript Abundance in Arabidopsis in the Mutants vtc1, adg1, and aba2 For comparison of short-term responses, vtc1, adg1, and aba2 seedlings were grown for 9 d on 0.5% (w/v) Suc and
afterwards floated for 24 h either on 1.5% (w/v) Suc in MS medium or on 0.5% (w/v) Suc in MS medium. The transcript abundances relative to wt (Figure 3) were calculated from RT–PCR data of three independent plant sets. For most genes encoding chloroplast proteins, the transcript levels were decreased the most in vtc1, demonstrating a wide positive impact of ascorbate biosynthesis in regulation of chloroplast function via nuclear gene expression (Figure 5). The s-APX,
Downloaded from http://mplant.oxfordjournals.org/ at University of California, San Diego on November 17, 2014
Figure 1. Transcript Level Regulation in Response to Sucrose and Sorbitol. Transcript abundance levels of genes encoding chloroplast peroxiredoxins (2CPA, 2CPB, Prx-Q, Prx-IIE), chloroplast CuZn-superoxide dismutase 2 (CSD2), monodehydroascorbate reductase (MDHAR), and stromal and thylakoid-bound ascorbate peroxidase (s-APX and t-APX) in comparison to genes encoding proteins of the photosynthetic membrane (PET-E1, PET-E2, PET-M, PET-C, and LHCB2.2), carbohydrate-induced APL3 and RBC-S, and carbohydrate-repressed STP1 and constitutively expressed ACTIN-2 and UBQ11 in Arabidopsis seedlings grown for 10 d on MS plates supplemented with 1%, 2%, or 3% sucrose or 1% sucrose plus 1% or 2% sorbitol. RT–PCR analysis was performed in three biological replicates with at least two technical replicates. For better comparison, here, band intensities are shown for a characteristic data set after separation of the PCR products on ethidium bromide gels and the means ± standard deviation after qPCR are presented.
62
Heiber et al. • s-APX Transcript Abundance Is Regulated by Ascorbate
Downloaded from http://mplant.oxfordjournals.org/ at University of California, San Diego on November 17, 2014
Figure 2. Metabolic Indicators in 10-Day-Old Seedling Grown for 10 d on MS plates Supplemented with 1%, 2%, or 3% Suc or 1% Suc plus 1% or 2% Sor. The ascorbate content, ascorbate redox state, the concentration of dihydroxyacetone phosphate (DHAP) fructose-bisphosphate (FruBP) and 3-phophoglycerate (3PGA), the 3PGA/DHAP ratio, the ATP and ADP content, the ATP/ADP ratio, the redox state of the NADP+NADPH pool, the assimilarity force, the chlorophyll content, and the anthocyanin content were analyzed in 6–10 biological replicates.
Heiber et al. • s-APX Transcript Abundance Is Regulated by Ascorbate
63
MDHAR, and APL3 transcript levels were higher in vtc1 than in wt, indicating that ascorbate synthesis impacts one gene regulation. While the MDHAR transcript level was decreased less in vtc1 than in wt in response to higher sugar levels, the APL3 and s-APX levels were increased more (Figures 3 and 5). The relative effect was stronger for APL3 than for s-APX. Similarly to long-term sucrose feeding, limitations in starch biosynthesis (adg1) increased the APL3 levels more strongly and the s-APX levels less. The aba2-mutation had a promoting effect on s-APX regulation and a highly variable impact on APL3 transcript abundance. In contrast, for the other genes, the mutations did not affect gene regulation significantly or had a preferentially negative effect on transcript levels.
Ascorbate Levels in Arabidopsis wt and in the Mutants vtc1, adg1, and aba2 after Long- and Short-Term Treatment with Suc In all mutants, the ascorbate level was increased after 10-day growth on 2% (w/v) Suc, but not on 1% (w/v) Suc + 1% (w/v) Sor (Figure 6, top). The effect was similar in wt, adg1, and aba2, but weaker in vtc1. The redox state of the ascorbate pool was barely affected (Figure 6, top). After 9 d growth on 0.5% (w/v) Suc, the ascorbate levels were more consistent than after 10 d growth on 1% (w/v) Suc in wt, aba2, and adg1. Twenty-four hours after application of additional 1% (w/v) Suc, the ascorbate levels were slightly increased in wt, aba2, and adg1, but not in the ascorbate-biosynthetic
mutant vtc1 (Figure 6, bottom). The redox state of ascorbate was slightly lower on 0.5% (w/v) Suc than on 1% Suc (Figure 6, bottom). Application of additional 1% (w/v) Suc increased the redox state slightly in wt and significantly in aba2 and adg1, but not in vtc1 (Figure 6, bottom), demonstrating that overcoming sugar starvation promoted ascorbate redox control and indicating more subtle, but specific, signaling effects.
Light Regulation of Transcript Abundance and Ascorbate Availability To finally test the effect of photosynthetic activity in the context of carbohydrate regulation, plants were grown for 10 d on Suc or Sor media and were placed for 4 h under 50 μmol quanta m–2 s–1 and 600 μmol quanta m–2 s–1 or left under the growth conditions of 100 μmol quanta m–2 s–1. For transcripts, except PET-E1, PET-E2, and s-APX, an inhibitory effect of 2% Suc was observed compared to 1% Suc and 1% Suc + 1% Sor at all light intensities. While the transcript levels of the two pet-E genes did not respond, s-APX levels were increased in response to elevated Suc under all light regimes. For 2CPA, 2CPB, Prx-Q, and RBC-S, the gradual response was strongest at lower light intensities, and for t-APX, MDHAR, PRX-C, and STP1 at highest light intensity.
GeneMANIA and PlaNet Analysis of Transcript Abundance Co-Regulation Transcript abundance co-regulation analysis can provide insight into the gene regulation context. GeneMANIA (Zuberi et al., 2013) and PlaNet (Mutwil et al., 2011) use publicly
Downloaded from http://mplant.oxfordjournals.org/ at University of California, San Diego on November 17, 2014
Figure 3. Relative Transcript Amount Regulation in Response to Elevated Sucrose Levels. The relative transcript amount of genes encoding chloroplast antioxidant enzymes (2CPA–MDHAR), extra-plastidic antioxidant enzymes (CAT– APX2), proteins of the photosynthetic membrane (LHCA5–PET–E2), marker genes for the carbohydrate status (APL3–GDH2), and reactive oxygen levels (BAP1 and FER1) in 10-day-old seedlings after a 24-hour increase in the sucrose availability to 1% (w/v) to 2% (w/v) (n = 3). The asterisk shows statistical significance (α
RESEARCH ARTICLE
Linking Chloroplast Antioxidant Defense to Carbohydrate Availability: The Transcript Abundance of Stromal Ascorbate Peroxidase Is Sugar-Controlled via Ascorbate Biosynthesis a Former address: Plant Physiology and Biochemistry, Bielefeld University, Universitätsstraße 25, 33615 Bielefeld, Germany b Plant Physiology, Free University of Berlin, Dahlem Center of Plant Sciences, Königin-Luise-Str. 12–16, 14195 Berlin, Germany
ABSTRACT All genes encoding chloroplast antioxidant enzymes are nuclear-encoded and posttranscriptionally targeted to chloroplasts. The transcript levels of most of them decreased upon sucrose feeding like the transcript levels of many genes encoding components of the photosynthetic electron transport chain. However, the transcript abundance of stromal ascorbate peroxidase (s-APX; At4g08390) increased. Due to mild sugar application conditions, the plants kept the phosphorylation status of the ADP+ATP pool and the redox states of the NADPH+NADP+ and the ascorbate pools under control, which excludes them as signals in s-APX regulation. Correlation with ascorbate pool size regulation and comparison of transcript abundance regulation in the starch-biosynthetic mutant adg1, the ascorbate biosynthesis mutant vtc1, and the abscisic acid (ABA) biosynthetic mutant aba2 showed a link between sugar induction of s-APX and ascorbate biosynthesis. Key words: ascorbate; antioxidant; carbohydrate; chloroplast; gene expression; stromal ascorbate peroxidase.
Introduction Sugar availability regulates the expression of many genes for chloroplast proteins (Rook and Bevan, 2003; Rolland et al., 2006). While low sugar concentrations support seedling growth, excess carbohydrates antagonize greening and seedling development (Koch, 1999). Carbohydrate surplus inhibits the Calvin-Cycle (Macdonald and Buchanan, 1992) and promotes generation of reactive oxygen species (ROS) via feedback inhibition of photosynthetic electron transport (Rook et al., 2006a). Antioxidant enzymes and low-molecular-weight antioxidants, such as ascorbate (Asc) and glutathione (GSH), form a ROS-protective network (Asada, 1999; Dat et al., 2001; Dietz et al., 2002; Chang et al., 2009). Stromal and thylakoidbound ascorbate peroxidase (s-APX (At4g08390) and t-APX (At1g77490)) detoxify H2O2 on the expense of ascorbate (Asada, 1999). The co-substrate is regenerated by monodehydroascorbate and dehydroascorbate reductases (MDHAR and DHAR). In parallel, glutathione peroxidases (GPX) and peroxiredoxins (PRX) reduce peroxides via ascorbate-independent thiol-mediated pathways (Dietz et al., 2002). These enzymes are nuclear-encoded and posttranslationally targeted to the organelles by N-terminal transit peptides (Pitsch et al., 2010). Most proteins, like the four peroxiredoxins and t-APX, for
example, are exclusively targeted to chloroplasts. s-APX and MDHAR (At1g63940) can be alternatively targeted to chloroplasts and (the intermembrane space of) mitochondria (Chew et al., 2003), yet they display strong preference towards chloroplasts. To acclimate the chloroplast antioxidant protection upon (photo-)oxidative stress, the genes for chloroplast antioxidant enzymes respond to organellar signals. Organelle-to-nucleus signaling has been best studied for photosynthesis associated genes, like cab (encoding chlorophyll-a/b-binding proteins/light-harvesting complex proteins) and rbc-S (encoding the small subunits of ribulose-1,5-bisphosphate carboxylase/oxygenase). They are suppressed by carbohydrates (Arenas-Huertero et al., 2000), while apl3 (encoding a large subunit of ADP-glucose pyrophosphorylase; At4g39210) is induced by sugars (Rook et al., 2006b).
1 To whom correspondence should be addressed. E-mail margarete.baier@ fu-berlin.de, fax +49-30-838-51688, tel. +49-(0)30-838-53183
© The Author 2013. Published by the Molecular Plant Shanghai Editorial Office in association with Oxford University Press on behalf of CSPB and IPPE, SIBS, CAS. doi:10.1093/mp/sst154, Advance Access publication 7 November 2013 Received 25 June 2013; accepted 27 October 2013
Downloaded from http://mplant.oxfordjournals.org/ at University of California, San Diego on November 17, 2014
Isabelle Heibera, Wenguo Caib, and Margarete Baiera,1
Heiber et al. • s-APX Transcript Abundance Is Regulated by Ascorbate
RESULTS Optimization of the Experimental Set-Up In nature, the mucilage of the outer seed coat provides the first carbohydrate source upon germination. In experiments, most of the mucilage is washed off during seed sterilization. Here, Arabidopsis was screened for the optimal background sucrose (Suc) concentration prior to gene-expression analyses. In continuous moderate light (50 μmol quanta m–2 s–1), the seedling biomass increased on Suc-supplemented MS-media up to 0.8% (w/v) Suc (data not shown). With Suc concentrations higher than 1.6% (w/v), anthocyanins accumulated in hypocotyls and cotyledons indicating excess carbohydrates (Solfanelli et al., 2006). Suc concentrations higher than 2.5% (w/v) resulted in growth inhibition. Based on these
observations, 1% (w/v) Suc was chosen as optimal sugar concentration and 0.5% (w/v) Suc was used to establish moderate sugar starvation. Sorbitol (Sor) was used as an osmotic control. Compared to Suc, it is taken up and metabolized much less efficiently (Shabala and Lew, 2002; Gibson, 2005). Chemically, the osmolarities of Suc and Sor are similar. In planta, hydrolysis of Suc into glucose and fructose increases the osmolarity and the heterotrophic metabolism decreases it. The Sor concentration best suited as an osmotic control for Suc treatments was determined empirically based on biomass comparison. At 10 d, the fresh weights of seedlings grown on 1% (w/v) Suc supplemented with 0.8–1.3% (w/v) Sor were in the range of seedlings grown on 2% (w/v) Suc (data not shown). Finally, 1% Sor was chosen as control for 2% Suc application.
Long-Term Carbohydrate Effect in Arabidopsis Seedlings Long-Term Carbohydrate Effect on the Transcript Abundance of Nuclear-Encoded Chloroplast Proteins Long-term effects of Suc and Sor on the expression of genes encoding chloroplast antioxidant enzymes were studied in ACTIN-2-standardized cDNA samples by RT–PCR (Figure 1). As secondary control, the transcript level of ubiquitin-11 (UBQ11; At4g05050) was quantified. Relative to ACTIN-2 transcript levels, the UBQ11 transcript levels were not significantly changed by any of the treatments (Figure 1). APL3 (encoding a large subunit of ADP-glucose-pyrophosphorylase) served as a control for induction by Suc (Rook et al., 2001). RBC-S (encoding the small subunit of Ribulose-1,6-bisphosphatcarboylase/oxygenase; At5g38430) and STP1 (encoding a high-affinity monosaccharide/proton symporter; At1g11260) were included to monitor suppression by surplus amounts of carbohydrates (Figure 1; Koch, 1999; Sherson et al., 2000). This study focuses on regulation of chloroplast antioxidant enzymes. On 2% (w/v) Suc, the transcript levels of the four chloroplast peroxiredoxins, 2CPA (At3g11630), 2CPB (At5g06290), Prx-Q (At3g26060), and Prx-IIE (At3g52960), of CuZn-superoxide dismutase 2 (CSD2; At2g28190), MDHAR (At1g63940), t-APX (At1g77490), the Rieske protein PET-C (At4g03280), the photosystem-II antenna protein LHCB2.2 (At2g05070), and of the cytochrome b6f proteins PET-M (At2g26500) were decreased compared to 1% (w/v) Suc and the osmotic control containing 1% (w/v) Sor demonstrating a broad inactivation of genes encoding chloroplast proteins. On 3% (w/v) Suc, the transcript levels were below those observed on 1% (w/v) and 2% (w/v) Suc. On 1% (w/v) Suc supplemented with 2% (w/v) Sor, they were barely detectable (Figure 1). PET-E1 (At1g76100) and PET-E2 (At1g20340) levels, encoding plastocyanins, were unchanged irrespective of the treatment. The s-APX mRNA level was increased on 2% (w/v) Suc, but not on 1% (w/v) Suc supplemented with 1% (w/v) Sor demonstrating Suc-dependent induction.
Downloaded from http://mplant.oxfordjournals.org/ at University of California, San Diego on November 17, 2014
Screens for sugar signaling mutants demonstrated a strong, but diverse, crosstalk of ABA, sugar, and ROS/redox signaling (Rook et al., 2001; Gonzalez-Guzman et al., 2002; Staneloni et al., 2008). So far, little is known about the transcriptional regulation of genes encoding chloroplast antioxidant enzymes. Array studies and RT–PCR demonstrate that transcript abundances often increase only by mild stress, but are decreased upon severe stress due to general inactivation of chloroplast function (summarized in Baier et al. (2010)). Regulation of the 2-Cys peroxiredoxin-A gene (At3g11630) has been investigated in detail (Baier et al., 2004b): cis-acting motifs located within 200 bp upstream of the transcription initiation site mediate developmental responses. Photosynthetic redox signals and ABA-signals integrate antagonistically at a more distally located promoter region, designated as redoxbox. There, the APETALA-2-type transcription factor RAP2.4a induces 2cpa transcription upon moderate oxidative stress (Shaikhali et al., 2008). Upon severe stress, RAP2.4a oligomerizes. As a consequence, 2cpa transcription activity decreases (Shaikhali et al., 2008). Transcriptional regulation of other genes for chloroplast antioxidant enzymes has not been studied so far. Mutants impaired in redox regulation of the 2cpa promoter (Heiber et al., 2007) and knockout lines of the 2cpa redox-regulating transcription factor RAP2.4a (Shaikhali et al., 2008) show links between s-APX, t-APX, and 2CPA in redox regulation. To investigate the impact of the photosynthetically controlled parameters, carbohydrate availability, light, and ABA on the expression of nuclear genes for chloroplast antioxidant enzymes, here the transcript abundances were analyzed in Arabidopsis following light and sugar treatments. The responses were compared to those in the ascorbate-biosynthetic mutant vtc1 (Conklin et al., 1997), the starch-biosynthetic mutant adg1 (Lin et al., 1988), and the ABA-biosynthetic mutant aba2 (Léon-Kloostersiel et al., 1996) after short- and long-term Suc treatment. In this comparison, a specific regulation is shown for s-APX. It is concluded that s-APX regulation is indirectly sugar-responsive via ascorbate biosynthesis.
59
60
Heiber et al. • s-APX Transcript Abundance Is Regulated by Ascorbate
Chlorophyll Levels after Long-Term Treatment After 10-day growth on MS medium supplemented either with 1–3% (w/v) Suc or 1% (w/v) Suc + 1% or 2% (w/v) Sor, the chlorophyll (chl) contents were very similar in all samples (Figure 2) reflecting acclimation and the mildness of the treatments.
Sugar Metabolite Levels and Energetization and Carbohydrate Redox Status after Long-Term Sugar Treatment
Ascorbate Levels and the Ascorbate Redox State after Long-Term Sugar Treatment Total ascorbate and reduced ascorbate contents were measured to determine the ascorbate pool size and to calculate the ascorbate redox state. The ascorbate level was 3.7-fold increased on 2% (w/v) Suc, 3.9-fold on 3% (w/v) Suc, and 1.6-fold on 1% (w/v) Suc + 1% (w/v) Sor and 1% (w/v) Suc + 2% (w/v) Sor compared to plant material grown on 1% (w/v) Suc (Figure 2), demonstrating a carbohydrate flux into the ascorbate pool. The ascorbate redox state was not affected (Figure 2).
Anthocyanin Levels after Long-Term Treatment In response to excess carbohydrate availability, Arabidopsis accumulated anthocyanins (Solfanelli et al., 2006). The anthocyanin contents were 2.73-, 4.35-, 1.32-, and 1.83-fold increased after 10-day growth on 2% and 3% (w/v) Suc and 1% (w/v) Suc supplemented with 1% or 2% (w/v) Sor, respectively (Figure 2).
Effect of Short-Term Sucrose Application on Arabidopsis thaliana To analyze short-term responses, the seedlings grown on 0.5% (w/v) Suc were either floated on liquid MS medium containing 0.5 or 1.5% (w/v) Suc for 24 h. For a set of 22 genes, the transcript
Effect of Ascorbate, Starch, and ABA-Biosynthesis on the Suc Effect The use of mutants enables short- and long-term studies with modified strains. The mutant vtc1 carries a mutation in GDPD-mannose pyrophosphorylase which catalyzes a final step in the main ascorbate biosynthesis pathway (Conklin et al., 1999). Adg1 has a point mutation in the small subunit of ADP-glucose pyrophosphorylase and is limited in chloroplast starch biosynthesis, while aba2 is restricted in ABA-biosynthesis (Rook et al., 2001; Cheng et al., 2002). Here, the mutant in short- and longterm acclimation response was compared to that of wildtype (wt) plants.
Long-Term Effect of Suc on Transcript Abundance in the Mutants vtc1, adg1, and aba2 After 10-day growth on 2% (w/v) Suc, the transcript levels of most of the analyzed genes were decreased in vtc1, adg1, and aba2 (Figure 4, top). Exceptions were s-apx and apl3, whose transcript levels were increased. The PET-E1, PET-E2, and PET-C transcript levels were barely altered. Consistently with the hypothesis that ABA signaling supports the sugar effect on the apl3 promoter (Rook et al., 2001), the transcript levels of APL3 were most induced in aba2. s-APX regulation resembled APL3 induction. s-APX and APL3 transcripts were similarly accumulated in adg1 as they were in wt, indicating that starch biosynthesis capacities have a minor effect on the induction. Using the vtc1 mutant distinguished s-APX and APL3 regulation: s-APX transcripts accumulated less in vtc1 than in wt. In contrast, the APL3 transcript levels increased in a wt-like manner. This demonstrates a specific sucrose-linked impact of ascorbate biosynthesis on s-APX regulation. APL3 transcript levels were (consistently with previous reports by Rook et al. (2001)) induced in wt in response to sorbitol and decreased when ABA signaling was affected. Similarly, the transcript levels of most other tested genes were slightly increased in adg1, decreased in aba2, and was hardly affected in vtc1. In contrast, s-APX levels were decreased in all four tested genotypes in response to sorbitol application (Figure 4, bottom). The effect was strongest in aba2 and vtc1,
Downloaded from http://mplant.oxfordjournals.org/ at University of California, San Diego on November 17, 2014
To study the effect of sugar feeding on the metabolic signature, DHAP (dehydroxyacetone phosphate), FruBP (fructose6-bisphosphate), 3PGA (3-phosphoglycerate), ATP (adenosine triphosphate), ADP (adenosine diphosphate), ascorbate, and anthocyanin levels were analyzed as indicators for the cellular carbohydrate, redox, and energy status. Most metabolite levels were only slightly changed between the treatments, demonstrating the mildness of the treatments (Figure 2). The ATP and ADP contents increased in response to Suc application. However, the ATP/ADP ratio, like the 3PGA/DHAP ratio, which reflects the carbohydrate redox status, and the redox status of the NADPH pool (NADPH+NADP+), and, consequently, the [ ATP ] [NADPH ] assimilatory force, (Dietz and Heber, 1989), [ ADP ][Pi ] NADP + ] were hardly affected, demonstrating that the metabolic control capacity was not exhausted in the seedlings (Figure 2).
abundance regulation factors were calculated from the transcript levels quantified for ACTIN-2 standardized cDNA samples. The transcript levels of all genes encoding chloroplast antioxidant enzymes, except again s-APX, and, of all tested genes, encoding proteins of the thylakoid membrane (LHCA5– PET-E2) and of RBC-S were decreased in response to Suc application (Figure 3). The transcripts of the sugar-inducible apl3 gene accumulated and that of the sugar-suppressible stp1 were less abundant. The transcript levels of the ROS-marker genes bap1 (BONZAI associated protein I; At3g61190) and fer1 (Ferritin-1 precursor; At5g01600) (op den Camp et al., 2003) were hardly affected, demonstrating the mildness of the conditions.
Heiber et al. • s-APX Transcript Abundance Is Regulated by Ascorbate
61
demonstrating the impact of ABA and ascorbate biosynthesis on the osmotic regulation of s-APX.
Short-Term Effect of Sucrose on Transcript Abundance in Arabidopsis in the Mutants vtc1, adg1, and aba2 For comparison of short-term responses, vtc1, adg1, and aba2 seedlings were grown for 9 d on 0.5% (w/v) Suc and
afterwards floated for 24 h either on 1.5% (w/v) Suc in MS medium or on 0.5% (w/v) Suc in MS medium. The transcript abundances relative to wt (Figure 3) were calculated from RT–PCR data of three independent plant sets. For most genes encoding chloroplast proteins, the transcript levels were decreased the most in vtc1, demonstrating a wide positive impact of ascorbate biosynthesis in regulation of chloroplast function via nuclear gene expression (Figure 5). The s-APX,
Downloaded from http://mplant.oxfordjournals.org/ at University of California, San Diego on November 17, 2014
Figure 1. Transcript Level Regulation in Response to Sucrose and Sorbitol. Transcript abundance levels of genes encoding chloroplast peroxiredoxins (2CPA, 2CPB, Prx-Q, Prx-IIE), chloroplast CuZn-superoxide dismutase 2 (CSD2), monodehydroascorbate reductase (MDHAR), and stromal and thylakoid-bound ascorbate peroxidase (s-APX and t-APX) in comparison to genes encoding proteins of the photosynthetic membrane (PET-E1, PET-E2, PET-M, PET-C, and LHCB2.2), carbohydrate-induced APL3 and RBC-S, and carbohydrate-repressed STP1 and constitutively expressed ACTIN-2 and UBQ11 in Arabidopsis seedlings grown for 10 d on MS plates supplemented with 1%, 2%, or 3% sucrose or 1% sucrose plus 1% or 2% sorbitol. RT–PCR analysis was performed in three biological replicates with at least two technical replicates. For better comparison, here, band intensities are shown for a characteristic data set after separation of the PCR products on ethidium bromide gels and the means ± standard deviation after qPCR are presented.
62
Heiber et al. • s-APX Transcript Abundance Is Regulated by Ascorbate
Downloaded from http://mplant.oxfordjournals.org/ at University of California, San Diego on November 17, 2014
Figure 2. Metabolic Indicators in 10-Day-Old Seedling Grown for 10 d on MS plates Supplemented with 1%, 2%, or 3% Suc or 1% Suc plus 1% or 2% Sor. The ascorbate content, ascorbate redox state, the concentration of dihydroxyacetone phosphate (DHAP) fructose-bisphosphate (FruBP) and 3-phophoglycerate (3PGA), the 3PGA/DHAP ratio, the ATP and ADP content, the ATP/ADP ratio, the redox state of the NADP+NADPH pool, the assimilarity force, the chlorophyll content, and the anthocyanin content were analyzed in 6–10 biological replicates.
Heiber et al. • s-APX Transcript Abundance Is Regulated by Ascorbate
63
MDHAR, and APL3 transcript levels were higher in vtc1 than in wt, indicating that ascorbate synthesis impacts one gene regulation. While the MDHAR transcript level was decreased less in vtc1 than in wt in response to higher sugar levels, the APL3 and s-APX levels were increased more (Figures 3 and 5). The relative effect was stronger for APL3 than for s-APX. Similarly to long-term sucrose feeding, limitations in starch biosynthesis (adg1) increased the APL3 levels more strongly and the s-APX levels less. The aba2-mutation had a promoting effect on s-APX regulation and a highly variable impact on APL3 transcript abundance. In contrast, for the other genes, the mutations did not affect gene regulation significantly or had a preferentially negative effect on transcript levels.
Ascorbate Levels in Arabidopsis wt and in the Mutants vtc1, adg1, and aba2 after Long- and Short-Term Treatment with Suc In all mutants, the ascorbate level was increased after 10-day growth on 2% (w/v) Suc, but not on 1% (w/v) Suc + 1% (w/v) Sor (Figure 6, top). The effect was similar in wt, adg1, and aba2, but weaker in vtc1. The redox state of the ascorbate pool was barely affected (Figure 6, top). After 9 d growth on 0.5% (w/v) Suc, the ascorbate levels were more consistent than after 10 d growth on 1% (w/v) Suc in wt, aba2, and adg1. Twenty-four hours after application of additional 1% (w/v) Suc, the ascorbate levels were slightly increased in wt, aba2, and adg1, but not in the ascorbate-biosynthetic
mutant vtc1 (Figure 6, bottom). The redox state of ascorbate was slightly lower on 0.5% (w/v) Suc than on 1% Suc (Figure 6, bottom). Application of additional 1% (w/v) Suc increased the redox state slightly in wt and significantly in aba2 and adg1, but not in vtc1 (Figure 6, bottom), demonstrating that overcoming sugar starvation promoted ascorbate redox control and indicating more subtle, but specific, signaling effects.
Light Regulation of Transcript Abundance and Ascorbate Availability To finally test the effect of photosynthetic activity in the context of carbohydrate regulation, plants were grown for 10 d on Suc or Sor media and were placed for 4 h under 50 μmol quanta m–2 s–1 and 600 μmol quanta m–2 s–1 or left under the growth conditions of 100 μmol quanta m–2 s–1. For transcripts, except PET-E1, PET-E2, and s-APX, an inhibitory effect of 2% Suc was observed compared to 1% Suc and 1% Suc + 1% Sor at all light intensities. While the transcript levels of the two pet-E genes did not respond, s-APX levels were increased in response to elevated Suc under all light regimes. For 2CPA, 2CPB, Prx-Q, and RBC-S, the gradual response was strongest at lower light intensities, and for t-APX, MDHAR, PRX-C, and STP1 at highest light intensity.
GeneMANIA and PlaNet Analysis of Transcript Abundance Co-Regulation Transcript abundance co-regulation analysis can provide insight into the gene regulation context. GeneMANIA (Zuberi et al., 2013) and PlaNet (Mutwil et al., 2011) use publicly
Downloaded from http://mplant.oxfordjournals.org/ at University of California, San Diego on November 17, 2014
Figure 3. Relative Transcript Amount Regulation in Response to Elevated Sucrose Levels. The relative transcript amount of genes encoding chloroplast antioxidant enzymes (2CPA–MDHAR), extra-plastidic antioxidant enzymes (CAT– APX2), proteins of the photosynthetic membrane (LHCA5–PET–E2), marker genes for the carbohydrate status (APL3–GDH2), and reactive oxygen levels (BAP1 and FER1) in 10-day-old seedlings after a 24-hour increase in the sucrose availability to 1% (w/v) to 2% (w/v) (n = 3). The asterisk shows statistical significance (α
