Involvement of thioredoxin y2 in the preservation of leaf methionine sulfoxide reductase capacity and growth under high light

Authors

  • EDITH LAUGIER,

    1. CEA, DSV, IBEB, Lab Ecophysiol Molecul Plantes, Saint-Paul-lez-Durance, F-13108, France
    2. CNRS, UMR 7265 Biol Veget & Microbiol Environ, Saint-Paul-lez-Durance, F-13108, France
    3. Aix-Marseille Université, Saint-Paul-lez-Durance, F-13108, France
    Search for more papers by this author
    • Present address: Laboratoire de Biologie du Développement des Plantes, Bâtiment 178, SBVME, CEA-Cadarache, 13108 Saint-Paul-Lez-Durance Cedex, France;

  • LIONEL TARRAGO,

    1. CEA, DSV, IBEB, Lab Ecophysiol Molecul Plantes, Saint-Paul-lez-Durance, F-13108, France
    2. CNRS, UMR 7265 Biol Veget & Microbiol Environ, Saint-Paul-lez-Durance, F-13108, France
    3. Aix-Marseille Université, Saint-Paul-lez-Durance, F-13108, France
    Search for more papers by this author
    • Center for Redox Medicine, Division of Genetics, Department of Medicine, Brigham and Women s Hospital and Harvard Medical School, New Research Building, Room 435, 77 Avenue Louis Pasteur, Boston, MA 02115, USA.

  • AGATHE COURTEILLE,

    1. CEA, DSV, IBEB, Lab Ecophysiol Molecul Plantes, Saint-Paul-lez-Durance, F-13108, France
    2. CNRS, UMR 7265 Biol Veget & Microbiol Environ, Saint-Paul-lez-Durance, F-13108, France
    3. Aix-Marseille Université, Saint-Paul-lez-Durance, F-13108, France
    Search for more papers by this author
  • GILLES INNOCENTI,

    1. IBP, UMR CNRS 8618, Univ Paris-Sud, 91405, Orsay Cedex, France
    Search for more papers by this author
  • FRANÇOISE EYMERY,

    1. CEA, DSV, IBEB, Lab Ecophysiol Molecul Plantes, Saint-Paul-lez-Durance, F-13108, France
    2. CNRS, UMR 7265 Biol Veget & Microbiol Environ, Saint-Paul-lez-Durance, F-13108, France
    3. Aix-Marseille Université, Saint-Paul-lez-Durance, F-13108, France
    Search for more papers by this author
  • DOMINIQUE RUMEAU,

    1. CEA, DSV, IBEB, Lab Ecophysiol Molecul Plantes, Saint-Paul-lez-Durance, F-13108, France
    2. CNRS, UMR 7265 Biol Veget & Microbiol Environ, Saint-Paul-lez-Durance, F-13108, France
    3. Aix-Marseille Université, Saint-Paul-lez-Durance, F-13108, France
    Search for more papers by this author
  • EMMANUELLE ISSAKIDIS-BOURGUET,

    1. IBP, UMR CNRS 8618, Univ Paris-Sud, 91405, Orsay Cedex, France
    Search for more papers by this author
  • PASCAL REY

    Corresponding author
    1. CEA, DSV, IBEB, Lab Ecophysiol Molecul Plantes, Saint-Paul-lez-Durance, F-13108, France
    2. CNRS, UMR 7265 Biol Veget & Microbiol Environ, Saint-Paul-lez-Durance, F-13108, France
    3. Aix-Marseille Université, Saint-Paul-lez-Durance, F-13108, France
      P. Rey. Fax: +33 4 42 25 74 80; e-mail: pascal.rey@cea.fr
    Search for more papers by this author

P. Rey. Fax: +33 4 42 25 74 80; e-mail: pascal.rey@cea.fr

ABSTRACT

Methionine (Met) in proteins can be oxidized to two diastereoisomers of methionine sulfoxide, Met-S-O and Met-R-O, which are reduced back to Met by two types of methionine sulfoxide reductases (MSRs), A and B, respectively. MSRs are generally supplied with reducing power by thioredoxins. Plants are characterized by a large number of thioredoxin isoforms, but those providing electrons to MSRs in vivo are not known. Three MSR isoforms, MSRA4, MSRB1 and MSRB2, are present in Arabidopsis thaliana chloroplasts. Under conditions of high light and long photoperiod, plants knockdown for each plastidial MSR type or for both display reduced growth. In contrast, overexpression of plastidial MSRBs is not associated with beneficial effects in terms of growth under high light. To identify the physiological reductants for plastidial MSRs, we analyzed a series of mutants deficient for thioredoxins f, m, x or y. We show that mutant lines lacking both thioredoxins y1 and y2 or only thioredoxin y2 specifically display a significantly reduced leaf MSR capacity (–25%) and growth characteristics under high light, related to those of plants lacking plastidial MSRs. We propose that thioredoxin y2 plays a physiological function in protein repair mechanisms as an electron donor to plastidial MSRs in photosynthetic organs.

INTRODUCTION

Proteins constitute major targets for reactive oxygen species and undergo numerous types of oxidative modifications, most of them being irreversible (Davies 2005). Thus, fragmentation of the peptidic backbone chain and carbonylation in the side chain of amino acids are common consequences of oxidation, impairing protein structure and often leading to degradation (Davies 2005; Stadtman 2006). Cysteine (Cys) and methionine (Met) are highly prone to oxidative damage due to reactivity of the sulfur atom with reactive oxygen species. However, most oxidative modifications in these two residues are reversible through the action of enzymes possessing redox-active Cys and termed thiol reductases. Oxidized forms of Cys, like disulfide bridge or Cys sulfenic-acid, are reduced mainly by thioredoxins (Trxs) and also by glutaredoxins (Grxs) (Arnér & Holmgren 2000; Rouhier, Lemaire & Jacquot 2008; Tarrago et al. 2009b, 2010) and Met sulfoxide is reduced back to Met by methionine sulfoxide reductases (MSRs) (Brot et al. 1981; Grimaud et al. 2001). Met is oxidized to two diastereoisomers of Met sulfoxide (MetO), Met-S-O and Met-R-O, depending on the position of the oxygen atom on sulfur. Met-S-O and Met-R-O, are reduced back to Met by two types of MSRs, A and B, respectively, which fulfil the same biochemical function, but do not share any sequence relationship (Brot et al. 1981; Grimaud et al. 2001). An increasing amount of reports indicates a crucial role for MSRs in all living organisms during stress conditions, pathologies and ageing. For instance, insects and mammal cells knockout for msra or msrb expression are more susceptible to oxidative stress and display reduced lifespan (Moskovitz et al. 2001; Ruan et al. 2002; Cabreiro et al. 2008).

Unlike in non-photosynthetic organisms, MSRs are encoded by multigenic families in plants (Rouhier et al. 2006; Tarrago, Laugier & Rey 2009a). Arabidopsis thaliana possesses five msra and nine msrb genes encoding proteins distributed in various cell compartments. Substantial insight has been gained about the physiological functions of plant MSRs in relation with stress tolerance. Using transgenic Arabidopsis plants, Romero et al. (2004) reported a protective role of plastidial MSRA4 upon severe oxidative stress conditions. A more subtle function has been described for a cytosolic MSRA, termed MSRA2, as plants deficient in this isoform display reduced growth under short-day conditions. MSRA2 is presumed to repair oxidized proteins in the dark, thus allowing plants to save carbon resources (Bechthold, Murphy & Mullineaux 2004). Regarding plant MSRBs, the endoplasmic reticulum MSRB3 isoform is cold responsive and participates in the acclimation process leading to freezing tolerance (Bae et al. 2003; Kwon et al. 2007). In the last years, we carried out an extensive characterization of the two Arabidopsis plastidial MSRB isoforms, MSRB1 and MSRB2, which are preferentially expressed in photosynthetic organs. Using Arabidopsis knockout mutants, we showed that plants lacking both MSRB1 and MSRB2 display substantially reduced growth and impaired photosynthesis under high light or low temperature conditions (Laugier et al. 2010). Consistently, most proteins interacting with plastidial MRBs were found to participate in photosynthetic processes (Tarrago et al. 2012).

Most MSRs, like plant MSRA4 and MSRB2, possess two redox active Cys and are reduced by the Trx system through a classical disulfide exchange (Lowther et al. 2000; Kumar et al. 2002; Boschi-Muller, Gand & Branlant 2008; Tarrago et al. 2009b). Note that MSRB1 possesses only one redox-active Cys and is regenerated through the direct reduction of the Cys sulfenic-acid form by Grxs (Tarrago et al. 2009b) or by a plant-specific and unusual Trx, CDSP32, involved in the protection against oxidative damage (Broin et al. 2002; Rey et al. 2005; Tarrago et al. 2010). Plants are characterized by a tremendous variety of Trxs (Meyer, Reichheld & Vignols 2005), with, for example, at least 20 plastidial Trx-related proteins divided in several classes (f, m, x y, z, NTRc, CDSP32. . .). Several lines of evidence indicate that plastidial Trxs display in vitro specificity towards various types of substrates, participating in photosynthetic processes or in detoxification of organic peroxides (Collin et al. 2003, 2004; Navrot et al. 2006). With regard to MSRs, we previously showed that Trxs m and y are more efficient in providing electrons to plastidial MSRB2 in vitro (Vieira Dos Santos et al. 2007). At the present time, no data are available concerning the identity of physiological electron donors for plant MSRs among Trxs.

In the present study, we characterized Arabidopsis plants with strongly decreased or increased leaf MSR capacity due to modifications in the expression of msra4, msrb1 and msrb2 genes. In parallel, to gain insight about the identity of electron donors to MSRs in vivo, we investigated the leaf MSR capacity of mutants deficient in plastidial thioredoxins f, m, x or y and we analyzed their phenotype under the conditions impairing growth of mutants deficient in plastidial MSRs.

MATERIALS AND METHODS

Plant material and growth conditions

A. thaliana, ecotype Columbia (Col-0), plants were grown in control conditions under an 8 h photoperiod and a photon flux density of 190 µmol photons m–2 s–1. For the high-light regime, plants were grown from sowing under a 14 h photoperiod and a photon flux density of 500 µmol photons m–2 s–1. For both conditions, the temperature regime was 22 °C/18 °C (day/night), and the relative humidity was 55%. Low temperature treatment (10 °C during night and day) was carried out for 18 d in controlled environmental chambers on 3-week-old plants grown under control conditions.

A. thaliana stable transformation

To generate plants overexpressing msrb1 or msrb2, full-length cDNAs were cloned in the sense direction into the pKYL vector using HindIII and XbaI restriction sites. The vector includes the cauliflower mosaic virus 35S promoter and a kanamycin-resistance gene. For down-regulation of msra4 expression, the full-length cDNA was amplified using the primers allowing the addition of attB recombination sites (Supporting Information Table S1). The cDNA was cloned into a pDONR201 vector (Invitrogen, Carlsbad, CA, USA) and transferred into the binary GATEWAY destination vector pK2WG7,0 (Plant Systems Biology, VIB-Ghent University, Ghent, Belgium) (Karimi, Inzé & Depicker 2002). The pK2WG7,0 vector allows expression of an antisense cDNA under the control of the cauliflower mosaic virus 35S promoter and includes a kanamycin-resistance gene. Transformation using Agrobacterium tumefaciens C58 strain was performed as described by Clough & Bent (1998). Homozygous transgenic lines (T3) were produced and selected from resistance segregation assays as described in Laugier et al. (2010).

PCR and semi-quantitative RT-PCR analyses

Transformant lines and homozygous T-DNA insertion lines were verified using PCR analysis on genomic DNA with the Phire® Plant Direct PCR Kit (Finnzymes, Vantaa, Finland) and gene-specific primers (Supporting Information Table S1). Actin2 was used as a positive control for each PCR. The PCR programme was 95 °C for 5 min, 30 cycles (denaturation at 95 °C for 50 s, annealing at 55 °C for 50 s, extension at 72 °C for 2 min) and finally 72 °C for 7 min (GeneAmp, PCRSystem 2700, Applied Biosystems, Foster City, CA, USA). For RT-PCR analysis, total RNA was extracted from leaves or from roots (for the mutant deficient in Trx y1) using RNeasy Plant Mini Kit (Qiagen, Valencia, CA, USA) according to the manufacturer's instructions. Reverse transcription was performed using total RNA (500 ng), a SuperScript® III Reverse Transcriptase (Invitrogen) and an Oligo(dT)20 primer. RT-PCR assays were performed using the same gene-specific primers used for genotyping, except for the mutant lacking Trx y1 (Supporting Information Table S1).

Protein extraction, SDS–PAGE and Western analysis

Leaf samples were blended in liquid nitrogen, and the powder was resuspended to prepare soluble and membrane proteins as described in Vieira Dos Santos et al. (2005) and Gillet et al. (1998) respectively. The protein content was determined using the BC Assay Reagent (Interchim, Montluçon, France). Coomassie Brilliant Blue staining of protein gels was carried out to control protein loading. Proteins were separated using sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE) and electro-transferred onto a nitrocellulose membrane (Pall Corporation, Ann Arbor, MI, USA). Western analysis was carried out using primary antibodies raised in rabbit against AtMSRA4 (1:1250), AtMSRB1 (1:1000) or AtMSRB2 (1:1000). Bound AtMSR antibodies were detected using an anti-rabbit IgG alkaline phosphatase conjugate (Sigma-Aldrich, St Louis, MO, USA) diluted 1:10 000. Antibodies raised in rabbit against Lhcb1 were used, diluted 1:5000 and detected using the goat anti-rabbit ‘Alexa Fluor® 680’ IgG from Invitrogen. Bound antibodies were revealed at 680 nm using the ‘Odyssey Infrared Imager’ from Li-Cor (Lincoln, NE, USA). Western analysis of 2-Cys Prx abundance and redox status was performed as described in Rey et al. (2007).

MSR activity assay

Total MSR activity in leaf soluble extracts was assayed by monitoring the reduction of the synthetic substrate, dabsyl-MetO, in the presence of DTE (Vieira Dos Santos et al. 2005). After blending leaves and resuspension in extraction buffer, the content in soluble proteins was determined as above. The reaction mixture contained 15 mm HEPES pH 8, 10 mm MgCl2, 30 mm KCl, 20 mm DTE, 0.25 mm dabsyl-MetO and 30 or 300 µg soluble proteins. After incubation for 3 h at 37 °C, stopping using an ethanol : acetate buffer and centrifugation, a supernatant aliquot was loaded on a C18 reverse phase 3.5 µm, 3 × 50 mm column SunFireTM (Waters, Milford, MA, USA). High-performance liquid chromatography (HPLC) separation of dabsyl-MetO and dabsyl-Met was performed as described in Tarrago et al. (2009b).

Analysis of the chlorophyll content

One cm diameter leaf disks were taken from fully expanded mature leaves and immediately frozen in liquid nitrogen and stored at −80 °C until use. Leaf disks were crushed in 1 mL 80% acetone. After storing overnight in the dark at 4 °C and centrifugation (14 000 g, 10 min), the content in chlorophylls a and b was measured spectrophotometrically and calculated according to Lichtenthaler (1987).

RESULTS

Phenotype of plants knockdown for plastidial MSRs

Three MSR isoforms (MSRA4, MSRB1 and MSRB2) are located in plastids (Vieira Dos Santos et al. 2005). Due to the lack of T-DNA insertion mutant for msra4, we developed an antisense strategy to down-regulate msra4 expression in A. thaliana wild-type (Wt) plants and in the double mutant deficient in both msrb1 and msrb2, termed DM (Laugier et al. 2010). For both genetic backgrounds (Wt and DM), the presence in genomic DNA of the product of the MSRA4 antisense-construct (MSRA4 cDNA) was validated (Fig. 1a) and two homozygous lines exhibiting low MSRA4 protein abundance were selected for further characterization (Fig. 1b). Regarding Wt lines knockdown for msra4 (A4-13 and A4-16), Western analysis revealed a reduced MSRA4 abundance of ca. 75 and 65%, respectively. Selected DM plants knockdown for msra4 expression (DA4-2 and DA4-7) display protein abundance lowered by ca. 45 and 60% (Fig. 1b).

Figure 1.

Characterization of plants antisense for MSRA4 expression. (a) PCR analysis of plants antisense for msra4 expression using gene-specific primers to amplify msr genes or MSRA4 cDNA inserted in genomic DNA. Actin gene was used as control. (b) Western analysis of MSRB1, MSRB2 and MSRA4 abundance in leaf soluble proteins. Proteins (10 µg per lane for MSRA4, 25 µg for MSRB1 and MSRB2) were separated by SDS-PAGE. (c) MSR capacity in leaf extracts. MSR activity was measured in extracts prepared from plants grown under control conditions as described in Materials and Methods. Results are the average of three independent experiments using leaf pieces from four plants per genotype and per experiment. *, significantly different from the Wt value with P < 0.05 and **, significantly different from the DM value with P < 0.01 (t-test). Wt, wild type; A4-13 and A4-16, two independent lines carrying the MSRA4-antisense construct in the Wt genetic background; DM, double mutant deficient in the expression of both msrb1 and msrb2; DA4-2 and DA4-7, two independent lines carrying the MSRA4-antisense construct in the DM genetic background. MSR, methionine sulfoxide reductase.

The total MSR activity in leaf protein extracts from plants down-regulated for msra4 expression was determined. In A4-13 and A4-16 lines, the MSR activity is decreased by around 20 and 12% to ca. 42 and 47 pmol Met mg prot−1 min−1, respectively, compared with the Wt value (54 pmol Met mg prot−1 min−1) (Fig. 1c). These data are in agreement with the Western blot analysis showing a lower MSRA4 amount in the two A4-13 and A4-16 transgenic lines and validate them from a functional point of view. Regarding DM plants, the MSR activity is reduced by ca. 75% (Fig. 1c), consistently with previous data (Laugier et al. 2010). In DA4-2 and DA4-7 plants, further decreases were noticed, as MSR activity is reduced by 81 and 84%, respectively, compared with Wt (Fig. 1c). Altogether, these data clearly demonstrate that plastidial MSRs account for the greatest part of leaf MSR capacity.

We investigated the phenotype characteristics of plants knockdown for msra4. After 6 weeks of growth under control conditions (moderate light and short photoperiod) (Fig. 2a), the two MSRA4-antisense lines display rosette weight values similar to Wt, like DM lines knockdown for msra4 expression. No obvious difference has been observed regarding development and seed production in msra4 antisense plants, even in the msrb1/msrb2 genetic background (data not shown). Plants were then subjected from sowing to a high-light regime (500 µmol photons m−2 s−1, 14 h photoperiod). In this experiment (Fig. 2b), both MSRA4-antisense lines in the Wt background exhibit a significantly decreased rosette weight value (ca. 1.050 g) compared with Wt plants (1.345 g). The lack of both MSRBs also results in a reduced rosette weight value (1.054 g) as previously observed (Laugier et al. 2010), and down-regulation of msra4 expression amplifies this decrease, as both DA4-2 and DA4-7 lines are characterized by the lowest rosette weight values: 0.980 and 0.869 g, respectively (Fig. 2b).

Figure 2.

Growth of plants deficient in plastidial MSRs under control (a) and high-light (b) conditions. Rosettes were weighted from 6- and 3-week-old plants grown under control conditions (190 µmol photons m–2 s–1, 8 h photoperiod) or high light (500 µmol photons m–2 s–1, 14 h photoperiod), respectively. Data are means of four average values ± SD originating from independent experiments (12 plants per genotype and growth condition). *, ** and ***, significantly different from the Wt value with P < 0.05, P < 0.01 and P < 0.001, respectively (t-test). Wt, wild type; A4-13 and A4-16, MSRA4-antisense lines; DM, double mutant for msrb1 and msrb2; DA4-2 and DA4-7, MSRA4-antisense lines in the DM genetic background.

The DM mutant was previously reported to exhibit a pale phenotype under high-light conditions (Laugier et al. 2010). We thus investigated the chlorophyll composition in leaves of MSRA4-antisense lines (Table 1). The abundance of chlorophylls a and b in leaves of plants grown under control conditions is very similar in all lines. When grown under a high-light regime, an increase in chlorophyll a content by ca. 10% was observed in Wt, A4-13 and A4-16 plants compared with control conditions, whereas no change was noticed regarding chlorophyll b. In contrast, in plants deficient in both plastidial MSRBs (DM), significantly lower chlorophyll a and b contents were observed under high light compared with Wt values, with a particularly much lower chlorophyll b content (5.1 against 7.1 µg cm−2, respectively). DM plants knockdown for msra4 expression display chlorophyll content values similar to those measured in DM. Consequently, these three lines are characterized by chlorophyll a/b ratios ranging from 3.8 to 4.1, significantly higher than those measured in Wt and MSRA4-antisense plants (3.3 and 3.6, respectively, Table 1). Based on the preferential distribution of chlorophyll b in light-harvesting antennae, we investigated the abundance of one main Lhc protein belonging to photosystem II (PSII) antennae, Lhcb1, in lines antisense for msra4. Under high light conditions, the Lhcb1 amount is noticeably decreased in all lines compared with control conditions, but to a higher extent in MSR-deficient plants (Supporting Information Fig. S1a). A quantitative analysis showed that, compared with Wt, the Lhcb1 abundance is reduced by ca. 20 and 7% in A4-13 and A4-16 lines, respectively, and by 31, 45 and 37% in DM, DA4-2 and DA4-7 plants, respectively (Supporting Information Fig. S1b). Altogether, these data reveal that down-regulation of msra4 expression leads to reduced growth under a high-light regime, as observed for the mutant deficient in plastidial MSRBs, but does not alter chlorophyll composition.

Table 1. Chlorophyll composition in plants modified in the expression of genes encoding plastidial MSRs
Genotype ControlHigh light
Chl aChl b a/bChl aChl b a/b
  1. *, ** and ***, significantly different from the Wt value with P < 0.05, P < 0.01 and P < 0.001, respectively (t-test).

  2. Chlorophyll content is expressed in µg per cm2. Pigment analysis was performed on leaf disks from 6- and 3-week-old plants grown under control and high-light conditions, respectively. Data are means of values ± SD originating from 5 to 10 1 cm diameter leaf disks from independent plants for each genotype.

  3. a/b, Chl a/Chl b ratio; Wt, wild type; A4-13 and A4-16, MSRA4-antisense lines; DM, double mutant for both msrb1 and msrb2; DA4-2 and DA4-7, MSRA4-antisense lines in the DM genetic background; B1-8 and B1-10, lines overexpressing msrb1; B2-3 and B2-5, lines overexpressing msrb2; MSR, methionine sulfoxide reductase.

Wt20.0 ± 2.06.4 ± 0.83.1 ± 0.223.7 ± 2.07.1 ± 0.73.3 ± 0.2
A4-1321.0 ± 1.56.3 ± 0.53.3 ± 0.122.0 ± 0.46.2 ± 0.3*3.6 ± 0.2*
A4-1620.6 ± 0.56.7 ± 0.23.1 ± 0.123.3 ± 2.46.7 ± 0.93.5 ± 0.2
DM20.5 ± 1.66.3 ± 0.63.3 ± 0.119.5 ± 2.9**5.1 ± 0.9***3.9 ± 0.1***
DA4-219.7 ± 0.35.8 ± 0.03.4 ± 0.118.3 ± 0.9***4.5 ± 0.3***4.1 ± 0.1***
DA4-720.6 ± 1.06.3 ± 0.43.2 ± 0.120.4 ± 1.9**5.5 ± 0.4***3.8 ± 0.4**
B1-820.4 ± 1.46.8 ± 0.43.0 ± 0.118.8 ± 1.1***5.4 ± 0.6**3.5 ± 0.2
B1-1019.9 ± 0.16.7 ± 0.03.0 ± 0.018.5 ± 0.4***5.5 ± 0.1***3.4 ± 0.1
B2-319.9 ± 0.56.6 ± 0.13.0 ± 0.017.7 ± 1.6***5.5 ± 0.6**3.2 ± 0.1
B2-520.1 ± 0.96.8 ± 0.43.0 ± 0.019.0 ± 2.0**5.5 ± 0.7**3.5 ± 0.1

Phenotype of plants overexpressing plastidial MSRBs

To investigate whether MSR gain-of-function could be beneficial for plant growth under environmental constraints, Arabidopsis plants were transformed to overexpress either msrb1 or msrb2 genes. For each isoform, two homozygous lines, showing the presence of transgenes without introns (MSRB1 or MSRB2 cDNA) in genomic DNA (Fig. 3a), were selected on the basis of expression at the protein level. Western analysis revealed a higher MSRB1 abundance in both B1-8 and B1-10 lines compared with Wt (Fig. 3b). Regarding MSRB2, an increased protein amount was observed in B2-3 and B2-5 lines, the protein being much more abundant in the latter. In plants overexpressing MSRB1, the abundance of MSRB2 and MSRA4 is not modified compared with Wt. Similarly, MSRB1 and MSRA4 levels are not altered in plants overexpressing MSRB2 (Fig. 3b). The total MSR activity in the two lines overexpressing MSRB1 (ca. 180 pmol Met mg prot−1 min−1) is four times higher than in Wt (Fig. 3c). In plants overexpressing MSRB2, activities of ca. 380 and 3200 pmol Met mg prot−1 min−1 were recorded in B2-3 and B2-5 lines, respectively, in agreement with overexpression levels revealed by Western analysis. Altogether, these data demonstrate the possibility to raise plants with a very high MSR capacity (65-fold higher in B2-5 plants than in Wt).

Figure 3.

Characterization of Arabidopsis lines overexpressing MSRB1 or MSRB2. (a) PCR analysis of overexpressing lines using gene-specific primers to amplify MSRB genes or MSRB cDNAs inserted in genomic DNA. Actin gene was used as control. (b) Western analysis of MSRB1, MSRB2 and MSRA4 abundance in leaf soluble proteins. Proteins (20 µg per lane for MSRB1, 10 µg for MSRA4 and MSRB2) were separated using SDS–PAGE. (c) MSR capacity in leaf extracts. MSR activity was measured in extracts prepared from plants grown under control condition as described in Materials and Methods. Results are the average of three independent experiments using leaf pieces from four plants per genotype and per experiment. (d) Growth of plants overexpressing plastidial MSRBs under high-light conditions. Rosettes from 3-week-old plants were weighted. Data are means of average values ± SD originating from four independent experiments (12 plants for each genotype). Wt, wild type; B1-8 and B1-10, lines overexpressing msrb1; B2-3 and B2-5, lines overexpressing msrb2. MSR, methionine sulfoxide reductase.

Under control conditions, no difference was observed regarding growth of lines overexpressing either MSRB1 or MSRB2 compared with Wt (data not shown). Plants were subjected from sowing to the light regime impairing growth of mutants deficient in plastidial MSRs (long photoperiod and high-light intensity). In these conditions, lines overexpressing MSRB1 display similar rosette weights in the range of that measured for Wt plants (Fig. 3d). Regarding lines overexpressing MSRB2, slightly higher, but not significantly different, values of rosette weights were recorded compared with Wt (Fig. 3d). We also investigated the growth characteristics of plants overexpressing MSRB1 or MSRB2 at 10 °C, a temperature strongly delaying growth of plants lacking both plastidial MSRBs (Laugier et al. 2010), but we did not notice any significant change compared with Wt (data not shown).

Under control conditions, all lines overexpressing plastidial MSRBs display chlorophyll a and b contents similar to those measured in Wt (Table 1). When grown under high light, these lines, unlike Wt, exhibit contents reduced by ca. 5 and 20% for chlorophylls a and b, respectively, compared with control conditions. As a consequence, the content in both chlorophylls is significantly lower in these lines than in Wt under a high-light regime. When investigating the abundance of the main Lhc protein, Lhcb1, we observed similar protein amounts in leaves of Wt and overexpressing lines (Supporting Information Fig. S1c). These data reveal that lines exhibiting a highly increased MSR capacity show a reduced chlorophyll content, but no change in Lhcb1 abundance under high light conditions.

Leaf MSR capacity in Arabidopsis mutants lacking plastidial Trxs

Despite a strong increase in MSR capacity, we found that overexpression of plastidial MSRBs does not confer a beneficial effect in environmental conditions impairing the growth of mutants lacking these MSRs. It is worth mentioning that the measurement of MSR activity in plant extracts is performed in the presence of a large excess of exogenous reductant (DTE), which is able to provide reducing power to the physiological electron donors to MSRs present in extracts. This measurement thus represents the maximal MSR enzymatic capacity. We previously showed in in vitro assays using recombinant proteins that Arabidopsis MSRB2 is more efficiently reduced by some classes of Trxs, which exhibit a large variety in plastids (Vieira Dos Santos et al. 2007). These data prompted us to investigate whether mutants modified in the expression of potential physiological electron donors to plastidial MSRs among Trxs could exhibit modifications in MSR capacity. To this purpose, we got from public SALK and SAIL collections Arabidopsis mutant lines knockout for the expression of genes coding for Trxs f1, m1, m4, y1, y2 and x. All these lines were validated with regard to the presence of a T-DNA in the coding sequence and the absence of transcript using PCR on genomic DNA and RT-PCR, respectively (Supporting Information Fig. S2). Further, we generated a double mutant deficient in the expression of both trx y1 and trx y2 genes (data not shown). When measuring the maximal MSR capacity in leaf protein extracts, we observed substantial decreases in some mutant lines (Fig. 4a). Indeed, whereas the MSR capacity is not altered in the trx y1 mutant and not significantly decreased in trx f1 and trx x mutants compared with Wt, it is reduced by ca. 17% in lines lacking Trx m1 or Trx m4, and by 20 and 25% in plants lacking Trx y2 or both Trx y isoforms, respectively. The data gained on mutants deficient for either each Trx y isoform or for both of them clearly show that Trx y1 does not participate in the maintenance of leaf MSR capacity. This is very likely due to the fact that this isoform is not expressed in photosynthetic organs (Collin et al. 2004). In other respects, we performed Western blot experiments and did not notice any change in the abundance of MSRA4, MSRB1 and MSRB2 in plants lacking Trxs compared with Wt (Fig. 4b and data not shown), clearly showing that the changes in MSR capacity do not result from variations in MSR protein abundance.

Figure 4.

MSR capacity and abundance of plastidial MSRs in Arabidopsis mutants deficient in plastidial thioredoxins. (a) MSR capacity in leaf extracts. MSR activity was measured in extracts prepared from plants grown under control condition as described in Materials and Methods. Results are the average of three independent experiments using leaf pieces from four plants per genotype and per experiment. * and **, significantly different from the Wt value with P < 0.05 and P < 0.01, respectively (t-test). (b) Western analysis of MSRB1, MSRB2 and MSRA4 abundance in leaf soluble proteins. Proteins (10 µg per lane for MSRA4, 25 µg for MSRB1 and MSRB2) were separated by SDS–PAGE. Wt, wild type; trx f1, trx m1, trx m4, trx x, trx y1, trx y2 and trx y1/y2, mutant plants deficient in Trx f1, Trx m1, Trx m4, Trx x, Trx y1, Trx y2 and Trxs y1/y2, respectively. MSR, methionine sulfoxide reductase.

Phenotype of Arabidopsis mutants lacking Trxs x or y under high light conditions

We explored the growth characteristics of all Trx mutant lines and compared them to those of Wt and DM plants. Under control light conditions, all mutant lines deficient in Trxs f, m, x, y or both plastidial MSRBs display no growth defect and exhibit rosette weight and chlorophyll content similar to Wt (Supporting Information Fig. S3, data not shown). Under a high light regime, we did not observe any difference, in a preliminary experiment, either in the rosette weight or in the chlorophyll content for plants lacking Trx f1, m1, m4 or y1 compared with Wt, whereas those devoid of Trx y2 or of both Trx y isoforms exhibit a somewhat reduced growth and a lower chlorophyll content (Fig. 5a, Supporting Information Fig. S4). Deeper investigations were then carried out on a much larger number of plants, by performing five independent experiments and comparing the means of the average values, to characterize the growth of the two mutants lacking Trx x or both Trx y isoforms, as these two lines exhibit contrasted features regarding leaf MSR capacity, the latter displaying a significantly reduced MSR activity compared with Wt (–25%), but not the former (Fig. 4a). Very interestingly, when grown from sowing under long photoperiod and high light, plants lacking both Trxs y show, like plants deficient in MSRB1 and MSRB2, a significantly reduced rosette weight value (1.137 g) compared with Wt (1.323 g) (Fig. 5b). In contrast, the rosette weight of plants deficient in Trx x is not significantly different from the Wt value in these conditions. Trx x-deficient plants have a chlorophyll content very similar to that measured in Wt plants (ca. 24.0 µg cm−2), whereas in plants lacking Trxs y or MSRBs, the chlorophyll content (21.8 and 20.1 µg cm−2, respectively) is significantly lower than in Wt (Fig. 5c). Regarding the chlorophyll a/b ratio, values in the range of 3.5 were observed for Wt and plants deficient in Trx x or Trxs y (Fig. 5d), indicating comparable chlorophyll a and b proportions under high-light conditions. In other respects, we investigated the abundance and redox state of plastidial 2-Cys peroxiredoxin (2-Cys Prx), a well-known target of Trxs involved in the detoxification of organic peroxides. It was shown previously that both Trxs x and y can serve as efficient reducing substrates for 2-Cys Prx in vitro (Collin et al. 2003, 2004). Under high light, we did not notice any difference regarding the amount and redox status of this enzyme in plants deficient in plastidial MSRBs, Trx x or Trxs y compared with Wt (Fig. 5e), indicating that the observed phenotype is very likely not linked to this thiol peroxidase. Altogether, these data reveal that among the various plastidial Trx mutants tested, only lines lacking Trx y2 exhibit phenotype characteristics related to those observed for the mutant deficient in plastidial MSRBs.

Figure 5.

Characteristics of mutant plants deficient in thioredoxins y grown under high light conditions. (a) Plants are shown at the rosette stage after 3 weeks of growth under high-light conditions. (b) Rosette weight of 3-week-old plants grown under high light. Data are means of five average values ± SD originating from independent experiments (six plants for each genotype and each experiment). (c and d) Chlorophyll content and chlorophyll a/chlorophyll b ratio, respectively, in 3-week old plants grown under high light. Data are means of three average values ± SD originating from independent experiments (five 1 cm diameter leaf disks from independent plants for each genotype per experiment). ** and ***, significantly different from the Wt value with P < 0.01 and P < 0.001, respectively (t-test). (e) Amount and redox state of 2-Cys Prx in plants grown for 3 weeks under high-light conditions. Western analysis was carried out as described in Rey et al. (2007). Wt, wild type; DM, double mutant for msrb1 and msrb2; trx x, trx y1, trx y2 and trx y1/y2, mutant plants deficient in Trx x, Trx y1, Trx y2 and Trxs y1/y2, respectively. 2-Cys Prx, 2-Cys peroxiredoxin.

DISCUSSION

Physiological electron donors to plastidial MSRs among Trxs

In this work, we showed that lowering leaf MSR capacity via msra4 down-regulation results in deleterious effects regarding growth under a high-light regime similar to those reported on a mutant deficient for msrb1 and msrb2 genes (Laugier et al. 2010). These data indicate that both MSR types are essential for growth in high-light conditions. In parallel, we analysed Arabidopsis mutant lines deficient in various types of plastidial Trxs (f1, m1, m4, x, y1 and y2) likely to provide MSRs with electrons in vivo. Only the two lines lacking Trx y2 or both Trxs y1 and y2 showed noticeable difference compared with Wt when grown under a high-light regime. Very interestingly, these two Trx mutant lines displayed phenotype characteristics resembling those of mutants lacking plastidial MSRs, that is, reduced rosette weight and lower chlorophyll content under high light and long day (Fig. 5). The absence of phenotype for the mutant lacking Trx y1 is very likely linked to the low level of expression of the gene encoding this isoform in photosynthetic organs (Collin et al. 2004). In other respects, we measured the total leaf MSR capacity in leaf extracts from Trx mutants and noticed that mutants deficient in Trx m1, m4 and y2 exhibit substantial and significant decreases in their capacity to repair oxidized Met (Fig. 4). In full agreement, by carrying out in vitro experiments using recombinant proteins, we previously revealed that Trxs m (except m3) and y are preferential electron donors to MSRB2 (Vieira Dos Santos et al. 2007). Taken together, the phenotype similarity of the mutants deficient in plastidial MSRBs or in Trx y2, the substantially decreased MSR capacity in plants lacking Trx y2, and the data gained from in vitro assays (Vieira Dos Santos et al. 2007) give high credence to a physiological function for Trx y2 in the repair of oxidized proteins. Through its disulfide reductase activity, Trx y2 likely acts as a reductant of plastidial MSRs, which possess two redox active Cys, like MSRB2 and MSRA4. This work provides the first evidence of an in vivo function for the Trx y class. Note also that Trxs y are able to supply reducing power in vitro to peroxiredoxins, enzymes detoxifying organic peroxides (Collin et al. 2004; Navrot et al. 2006). The ability of Trxs y to reduce enzymes preventing damage in proteins and lipids indicates that they constitute essential actors in the protection against the consequences of environmental constraints. In this study, we showed that in high-light conditions, neither the 2-Cys Prx abundance, nor its redox state, was altered in lines devoid of Trx y2 or of plastidial MSRBs, indicating that the observed phenotype is not linked to this class of peroxiredoxin.

This work uncovers a physiological and specific function for a plant Trx type. In Arabidopsis, there are at least 40 proteins with a Trx domain (Meyer et al. 2005) and assays using recombinant proteins early revealed a specificity of some Trx classes towards various types of substrates involved in photosynthetic metabolism or in oxidative stress responses (Schürmann & Jacquot 2000; Collin et al. 2003). In contrast, only recent data unveiled physiological roles for these disulfide reductases. For instance, Trx h5 participates in the responses to a fungus-induced blight (Sweat & Wolpert 2007), Trx z regulates chloroplast development (Arsova et al. 2010), and Trxs m3 and h9 are involved in intercellular transport and communication processes, respectively (Benitez-Alfonso et al. 2009; Meng et al. 2010). Using an RNA-interference strategy, Chi et al. (2008) reported abnormal chloroplast development and impaired growth in rice plants knockdown for the expression of Trx m. In the present work, we did not notice any modification in the development and growth of Arabidopsis mutants lacking Trx m1 or Trx m4 under control or high-light conditions. This discrepancy might originate from the fact that Trx m isoforms fulfil redundant functions in Arabidopsis, but not in rice. Accordingly, three Trx m isoforms have been shown to display similar biochemical properties towards various types of substrates (Collin et al. 2003). Regarding Trxs y, our data clearly demonstrate that they participate in plant responses to high-light conditions. This Trx class would fulfil a role related to that of the unusual CDSP32 Trx, which prevents oxidative damage through the supply of electrons to peroxiredoxins and MSRs (Rey et al. 2005; Tarrago et al. 2010). Collectively, these data provide strong evidence of a specialization of some plant Trx types in processes linked to oxidative stress responses, as previously proposed (Vieira Dos Santos & Rey 2006).

Critical roles of electron donors for modifying MSR capacity

Despite a highly increased MSR capacity, we observed that overexpression of plastidial MSRBs is not associated with beneficial effects under the light regime impairing growth of down-regulated lines. Based on overexpression data, plastidial MSRA has been shown to display a protective role against very severe oxidative stress in Arabidopsis (Romero et al. 2004) and to participate in the tolerance to salt treatment in rice (Guo et al. 2009). These last data would suggest that increasing plastidial MSRA capacity is more efficient than increasing MSRB capacity to improve plant tolerance to stress. Nevertheless, taking into consideration the fact that plastidial MSRBs account for the major part of leaf MSR capacity (75 against 15% for MSRA4), we can also hypothesize that increasing the abundance of MSRB isoforms, already predominant in terms of capacity compared with others, might be useless without a concomitant increase in the pool of electron donors. Indeed, measurements of MSR activity are performed in the presence of a large excess of reductant (DTE) and reflect the maximal MSR capacity, but not the actual activity level in leaves, likely due to a limited pool of available reducing power in vivo. Consistently, when performing measurements in leaf extracts without reductant, no detectable activity was measured even in extracts from overexpressing lines (data not shown), confirming that supply of electrons is critical to sustain MSR activity. Therefore, the absence of protective effect of MSRB overexpression could result from a shortage in electron suppliers as hypothesized in other models, like animal and insect cells where overexpression of some MSR isoforms does not always lead to extended lifespan or improved stress resistance (Shchedrina et al. 2009; Zhao et al. 2010). In other respects, Oh et al. (2010) showed that gain-of-function of a cytosolic pepper MSRB in tomato results in increased resistance to oxidative stress and enhanced tolerance to Phytophthora infection, and Kwon et al. (2007) found that overexpression of an MSRB located in endoplasmic reticulum is associated with a higher tolerance to paraquat. Based on our measurements of MSR activity, these MSRB isoforms very likely represent a minor part in leaf MSR capacity. Therefore, we can presume that overexpression of non-plastidial MSRBs is not limited by a shortage in reductants, like for plastidial MSRBs, and thus much more beneficial for plants.

Roles of plastidial MSRs in plant growth under high light

Our data show that MSRA4, the unique plastidial MSRA isoform highly expressed in photosynthetic organs compared with other isoforms (Rouhier et al. 2006), accounts for a relatively low proportion of the MSR capacity in leaves compared with plastidial MSRBs, but clearly demonstrate that altogether the three isoforms are responsible for the major part (at least 85%) of leaf MSR capacity. Very interestingly, this result is in full agreement with the pioneer report on plant MSRs by Sanchez, Nikolau & Stumpf (1983), who reported based on cell sub-fractionation that 85% of leaf MSR capacity is localized in chloroplasts of barley leaves. Regarding the physiological function of MSRA4, we noticed that compared with Wt, MSRA4-antisense plants grown under a high-light regime display a decrease in rosette weight similar to that recorded for the mutant lacking both MSRBs (Fig. 2). These data indicate that both MSR types would fulfil related functions in the protection of plastidial structures, thus preserving plant growth during environmental constraints. Western analysis of Lhcb1 amount (Supporting Information Fig. S1) revealed that plastidial MSRs could be involved in the maintenance of light-harvesting antennae. However, MSRA-antisense and MSRB-deficient plants do not share strictly identical phenotype characteristics. Indeed, we noticed that under high-light conditions, chlorophyll content is almost not altered in MSRA-antisense lines, but significantly reduced in plants lacking both MSRBs. Therefore, the two MSR types likely fulfil specific functions. Accordingly, some evidence has been provided in the last years regarding distinct physiological roles for MSRAs and MSRBs in other organisms. In yeast, deletion of msra results in reduced viability, but not deletion of msrb (Koc et al. 2004), and in Aspergillus nidulans, a strain lacking MSRB is more sensitive to H2O2 than a strain lacking MSRA, whereas both mutants display a similar susceptibility to paraquat (Soriani et al. 2009). Several lines of evidence argue for a participation of plastidial MSRs in the regulation of photosynthetic antennae composition in relation to light environment. Under a high-light regime, lines deficient in MSRBs exhibit reduced amounts of chlorophyll and Lhcb1, and a higher Chl a/Chl b ratio. In these conditions, overexpression of plastidial MSRBs also results in a significant decrease in chlorophyll content, but with no change in Chl a/Chl b ratio and Lhcb1 content. Thus, surprisingly, both down- and up-regulation of plastidial msr gene expression lead to reduced chlorophyll content. We previously proposed that MSRB1 and MSRB2 could control the redox status of Met in the protein system targeting Lhc polypeptides to thylakoids membranes and that this system might be impaired in the absence of plastidial MSRBs (Laugier et al. 2010). The decrease in chlorophyll content in lines overexpressing MSRs might originate from changes in the chloroplastic redox status, an essential component in the regulation of chlorophyll biosynthesis and catabolism (Alboresi et al. 2011; Stenbaek & Jensen 2011), as Met in proteins has been proposed as an antioxidant able to scavenge reactive oxygen species after recycling by MSRs (Levine et al. 1996; Gustavsson et al. 2002). In other respects, Day and co-workers (2012) demonstrated in fission yeast that the total Trx pool is limited and that there is a competition between Trx substrates such as thiol-peroxidases and MSRs. In the case of MSRB overexpressing plants, the abnormal level of MSR is likely to divert some reducing power from Trxs and thus to impair the reduction of other Trx physiological targets such as Mg chelatase, the first enzyme of chlorophyll biosynthesis. In line with this hypothesis, it is worth mentioning that Luo et al. (2012) very recently reported that Mg chelatase is regulated by Trxs m and f in vivo and that pea plants silenced for these two Trx types display severely decreased chlorophyll content. Further investigations are needed to decipher how MSRs and Trxs act in concert within the complex plastidial antioxidant network to control redox homeostasis and chlorophyll composition as a function of environmental conditions.

ACKNOWLEDGMENTS

Financial support by Agence Nationale de la Recherche (ANR-Génoplante, Grant GNP05010G) to E.L. and by Région Provence-Alpes-Côte d'Azur to L.T. is acknowledged. We are very grateful to Dr Christina Vieira Dos Santos for her participation in the generation of transgenic lines, to P. Henri for valuable technical assistance, and to the Groupe de Recherche Appliquée en Phytotechnologie (CEA, IBEB, SBVME) for technical assistance with phytotrons and controlled growth chambers. We thank Dr L. Nussaume (CEA, IBEB, SBVME) for providing us the antibodies raised against Lhcb1.

Ancillary