The function of plant chloroplast redox systems in adaptation to light

Instituto de Bioquímica Vegetal y Fotosíntesis Departamento de Bioquímica Vegetal y Biología Molecular Universidad de Sevilla - CSIC The function of plant chloroplast redox systems in adaptation to light Trabajo realizado para optar al grado de Doctora por la Universidad de Sevilla por la licenciada Belén Naranjo Río-Miranda Septiembre, 2015 Directores de la Tesis Dr. Francisco Javier Cejudo Fernández Catedrático de Bioquímica y Biología Molecular Dra. Anna Marika Lindahl Científico Titular del Consejo Superior de Investigaciones Científicas AGRADECIMIENTOS El trabajo presentado en esta tesis ha sido realizado en el Instituto de Bioquímica Vegetal y Fotosíntesis (Universidad de Sevilla-CSIC) gracias a la concesión de una ayuda Predoctoral de Formación de Personal Investigador por el Ministerio de Economía y Competitividad (BES-2011-045543). Me gustaría agradecer a mis directores de tesis la posibilidad de haber realizado este trabajo. Al Dr. Javier Cejudo, la confianza depositada en mí habiéndome permitido formarme en su grupo, así como su valiosa aportación de ideas y su disponibilidad. A la Dra. Anna Lindahl, su incesante apoyo en todo este periodo, su implicación en los experimentos realizados y sus buenos consejos. A mis compañeros de laboratorio, tanto presentes, Mari Cruz, Juan, Vicky y Valle, como pasados, Manolo, Bea, Julia y Leo, que han hecho de éste un sitio agradable. Gracias Juan y Leo por todas las dudas resueltas y Vicky y Valle por vuestra compañía, risas y apoyo, que facilitan el trabajo. Agradecer también a Julia, Paula y Bea que me ayudaron a empezar en el IBVF. A las personas que he conocido en este centro y con quienes he compartido tantos desayunos, comidas y buenos momentos. Y a Leo y Sandy, que siempre han estado ahí. A la gente de París, donde tan a gusto estuve. Kathleen y Catherine por ayudarme con todo y por supuesto Anja, a quien agradezco también todos sus consejos y su preocupación por mí. A esas personas que he ido conociendo gracias a la ciencia en distintos momentos y que me han motivado a seguir. Carlos, Isa, M. Valle, Nerea, Blanca, Cristina… A Jr, con quien todo empezó y tanto he aprendido. A Laura, Hugo, Tasia, Nacho, María, Ana… mis amigos de siempre, que desde hace ya muchos años sé que están ahí y me alegra saber que puedo contar con ellos. A Jerome, esto no hubiese sido lo mismo sin él, por su paciencia y cariño. A mis abuelos, que tanto se han preocupado por mí. Y por supuesto, a mis padres y hermanos, por su paciencia y apoyo incondicional, que me ayudan a tomar decisiones y avanzar, gracias por ser como sois y estar siempre ahí. INDEX 7 INDEX FIGURES ........................................................................................................................ 11 TABLES ......................................................................................................................... 15 ABBREVIATIONS ........................................................................................................ 16 1.INTRODUCTION ....................................................................................................... 21 1.1.Plants and life ........................................................................................................ 21 1.2.The photosynthetic process in higher plants ......................................................... 21 1.2.1.Photochemical reactions: photosynthetic electron transport and photophosphorylation ............................................................................................. 22 Photosystem II .......................................................................................... 23 Cytochrome b6f ......................................................................................... 25 Photosystem I ............................................................................................ 26 ATP synthase ............................................................................................ 26 Alternative electron pathways .................................................................. 27 1.2.2.Carbon assimilation: The Calvin-Benson cycle............................................. 30 1.3.Light acclimation .................................................................................................. 32 1.3.1.Reactive oxygen species production and PSII photoinhibition ..................... 34 1.3.2.Energy dissipation during photosynthesis: Non-Photochemical Quenching . 36 Energy-dependent quenching ................................................................... 37 The xanthophyll cycle ............................................................................... 39 1.4.Chloroplast redox regulation ................................................................................ 42 1.4.1.The Ferredoxin-thioredoxin system ............................................................... 44 1.4.2.Thioredoxins .................................................................................................. 45 1.4.3.NADPH-thioredoxin reductase C .................................................................. 47 INDEX 8 2.OBJECTIVES .............................................................................................................. 51 3.MATERIALS AND METHODS ................................................................................. 57 3.1.Plant material and growth conditions .................................................................... 57 Generation of Arabidopsis mutants ........................................................... 57 3.2.RNA extraction and quantitative PCR analysis .................................................... 59 3.3.Protein extraction and Western blot analysis ........................................................ 60 Alkylation assay with MM(PEG)24 ........................................................... 61 3.4.Analysis of pigments ............................................................................................. 61 Chlorophylls .............................................................................................. 61 Xanthophylls ............................................................................................. 62 3.5.Determination of ABA content ............................................................................. 62 3.6.Measurements of photosystem II activity: chlorophyll a fluorescence ................. 63 Maximum quantum yield of PSII .............................................................. 64 Induction-recovery curve .......................................................................... 64 Electron transport rates .............................................................................. 65 Post-illumination fluorescence .................................................................. 65 3.7.Measurements of photosystem I activity: P700 absorbance ................................. 65 Induction-recovery curve .......................................................................... 67 3.8.Thermoluminescence ............................................................................................ 67 3.9.Electrochromic pigment absorbance shift ............................................................. 68 3.10.Determination of the rate of carbon assimilation ................................................ 69 3.11.Determination of starch content .......................................................................... 70 3.12.Electron paramagnetic resonance ........................................................................ 70 3.13.Cloning, expression and purification of recombinant proteins ........................... 71 3.14.NTRC in vitro activity......................................................................................... 73 INDEX 9 In vitro alkylation assays .......................................................................... 73 Determination of NADPH concentration ................................................. 73 3.15.Bioinformatic methods ....................................................................................... 73 4.RESULTS .................................................................................................................. 75 4.1.The role of NTRC in light acclimation and energy utilisation ............................. 77 4.1.1.The sensitivity of the ntrc mutant to high irradiances ................................... 77 4.1.2.Acclimation to high light: PSII energy dissipation ........................................ 81 Protection against excess light in plants lacking NTRC ........................... 81 High light acclimation in ntrc plants in relation to the 2-Cys Prx activity 84 Characterisation of the component of NPQ regulated by NTRC ............. 88 The redox state of the xanthophyll cycle enzymes VDE and ZE in vivo . 90 ABA content ............................................................................................. 94 The trans-thylakoid proton gradient in ntrc plants ................................... 95 Isolation and analysis of a ntrc-psbs double mutant ................................. 97 4.1.3.Effect of NTRC deficiency on other components of the photosynthetic electron transport chain ......................................................................................... 102 Photosystem I .......................................................................................... 102 The plastoquinone redox state ................................................................ 106 4.2.The function of the f-type thioredoxins in vivo ................................................... 109 4.2.1.Generation of plants lacking f-type Trxs ..................................................... 109 4.2.2.Growth of Trx f-deficient mutants ............................................................... 112 4.2.3.Effect of Trx f deficiency on photosynthetic performance .......................... 114 4.2.4.Redox state of Calvin-Benson cycle enzymes ............................................. 122 4.2.5.Starch content .............................................................................................. 125 4.2.6.Interactions between the redox systems of NTRC and f-type thioredoxins 126 INDEX 16 ABBREVIATIONS ABA Abscisic acid ADP Adenosine diphosphate ATP Adenosine triphosphate Ax Antheraxanthin cDNA Complementary DNA CEF Cyclic electron flow Chl Chlorophyll CDS Coding sequences CTAB Cetyltrimethylammonium bromide Cyt Cytochrome DNA Deoxyribonucleic acid DTT Dithiothreitol ECS Electrochromic shift EDTA Ethylenediaminetetraacetic acid EPR Electron paramagnetic resonance ETR Electron transport rate FAD Flavin adenine dinucleotide FBPase Fructose 1,6-bisphosphatase Fd Ferredoxin FNR Ferredoxin-NADP+ reductase FPLC Fast protein liquid chromatography FQR Ferredoxin-plastoquinone reductase FR Far red FTR Ferredoxin-thioredoxin reductase g Gram g Standard gravity h Hour HEPES 4‐ (2‐hydroxyethyl) ‐1‐piperazineethanesulfonic acid HL High light HPLC High-performance liquid chromatography HRP Horseradish peroxidase IPTG Isopropyl‐β‐thio galactopyranoside kDa Kilodalton l Litre LEF Linear electron flow LHC Light harvesting complex m Metre M Molar min Minute MMPEG Methyl-polyethylene glycol-maleimide NADPH Nicotinamide adenine dinucleotide phosphate NCBI National Center for Biotechnology Information NDH Nicotinamide adenine dinucleotide dehydrogenase NPQ Non-photochemical quenching NTR NADPH‐dependent thioredoxin reductase PAGE Polyacrylamide gel electrophoresis INDEX 17 PAM Pulse-amplitude modulation fluorometer PAR Photosynthetically active radiation PC Plastocyanin PCR Polymerase chain reaction PEF Pseudocyclic electron flow PET Photosynthetic electron transport Pi Inorganic phosphate PIF Post-illumination fluorescence pmf Proton motive force PQ Plastoquinone Prx Peroxiredoxin PSI Photosystem I PSII Photosystem II Phe Pheophytin qE Energy-dependent quenching qI Photoinhibition quenching qP Photochemical quenching qT State transition quenching RC Reaction centre RNA Ribonucleic acid RT-qPCR Real-time quantitative PCR ROS Reactive oxygen species s Second SD Standard deviation SDS Sodium dodecyl sulphate SE Standard error SP Saturation pulse T-DNA Transfer DNA TAIR The Arabidopsis Information Resource TCA Trichloroacetic acid TL Thermoluminescence Trx Thioredoxin v/v Volume/volume VDE Violaxanthin de-epoxidase Vx Violaxanthin ZE Zeaxanthin epoxidase Zx Zeaxanthin w/v Weight/volume wt Wild type ΔpH Proton gradient 4-POBN 4-pyridyl-1-oxide-N-tert-butylnitrone INDEX 18 NITROGENOUS BASES A Adenine C Cytosine G Guanine T Thymine AMINO ACIDS A Ala Alanine L Leu Leucine R Arg Arginine K Lys Lysine N Asn Asparagine M Met Methionine D Asp Aspartic acid F Phe Phenylalanine C Cys Cysteine P Pro Proline E Glu Glutamic acid S Ser Serine Q Gln Glutamine T Thr Threonine G Gly Glycine W Trp Tryptophan H His Histidine Y Tyr Tyrosine I Ile Isoleucine V Val Valine INTRODUCTION INTRODUCTION 21 1. INTRODUCTION 1.1. Plants and life Plants, along with algae and cyanobacteria, belong to the group of photosynthetic organisms. They are responsible for the current life conditions on Earth. Due to their ability to use light energy for reduction and fixation of carbon dioxide to carbohydrates, a process referred to as photosynthesis, plants are a source of energy and organic material for the rest of the organisms. Moreover, since the ultimate source of electrons is water, molecular oxygen is evolved and this photosynthesis is known as oxygenic. Oxygenic photosynthesis plays an essential role for the aerobic life on Earth. This process is critical for maintaining oxygen levels in the atmosphere and has allowed the development of the existing life forms. Atmospheric dioxygen, which is a byproduct of photosynthesis, is used in cellular respiration and other biological oxidation processes. In addition, photosynthetic organisms constitute an essential machinery for reducing the levels of carbon dioxide in the atmosphere. From the scientific point of view, these are some of the reasons why plants are interesting subjects of study. Moreover, understanding how this process, optimized over billions of years, works might contribute to the development of technology for renewable energy sources. 1.2. The photosynthetic process in higher plants During photosynthesis light energy is converted into chemical energy that is used to fix carbon dioxide into carbohydrates. Focusing on the oxygenic photosynthesis of higher plants, the process can be divided into two phases: - Photochemical reactions: photosynthetic electron transport and photophosphorylation. - Carbon assimilation: Calvin-Benson cycle. INTRODUCTION 22 These reactions take place in different parts of the chloroplast. Photochemical reactions occur in the thylakoid membranes, while carbon assimilation reactions are located in the stroma. 1.2.1. Photochemical reactions: photosynthetic electron transport and photophosphorylation The first steps of photosynthesis, the harvesting of light and its conversion into chemical energy, are carried out by four multisubunit membrane-protein complexes: photosystem I (PSI) and II (PSII), cytochrome b6f (Cyt b6f) and ATP synthase. PSI, PSII and Cyt b6f, besides other cofactors, are involved in the transport of electrons from water to NADP+ (linear electron transport or Z-scheme) with the associated production of a proton motive force (pmf) (Figure 1). The ATP synthase catalyses generation of ATP coupled to the proton gradient formed across the thylakoid membrane during photosynthetic electron transport (PET) (Figure 1). Figure 1. Linear electron transport and photophosphorylation. Photosystem I (PSI) and II (PSII), cytochrome b6f (Cyt b6f) and other elements such as plastoquinone (PQ) and plastocyanin (PC) molecules are involved in the transfer of electrons from H2O to NADP+ producing NADPH. The ATP synthase catalyses generation of ATP coupled to the proton gradient formed during photosynthetic electron transport (photophosphorylation). Light energy is captured by pigments in the light harvesting complexes (LHC) associated to PSI and PSII. INTRODUCTION 23 The entire process is light-dependent and, thus, several types of pigments bound to photosystems I and II allow the uptake of light energy by absorbing photons at different wavelengths in the visible part of the electromagnetic spectrum. The most abundant pigment is chlorophyll (Chl), a tetrapyrrole with a magnesium atom at the centre of the ring and a hydrophobic phytyl chain. In plants, there are two kinds of chlorophylls, a and b (Chl b is synthesized from the a-form by conversion of a methyl- to a formyl group), and their maximum absorbance in the red region of the visible spectrum are at 665 and 653 nm, respectively. Other pigments present in photosynthetic organisms are carotenoids that absorb light of lower wavelengths than chlorophylls, about 480 nm, and comprise the carotenes and the xanthophylls. The carotenoids act not only as accessory pigments harvesting light, but also as structural components having a protective role against photodamage. Photosynthetic pigments together with a group of various pigment-binding proteins constitute the antenna, or light-harvesting complex (LHC). Light energy captured is transferred to the reaction centres (RC), where it is converted to electrical energy through charge separation. Each photosystem is composed of the RC and the surrounding LHC. Depending on the maximum absorbance wavelength the RC chlorophylls are referred to as P700, in PSI, or P680, in PSII. The Cyt b6f complex provides a connection for transfer of electrons between both photosystems. Photosystem II PSII is a complex where the light-driven splitting of water takes place producing the concomitant reduction of the plastoquinone (PQ) pool. The PQs are membranesoluble electron carrier intermediates, which transfer reducing equivalents to the next membrane complex, Cyt b6f. The LHC of PSII (LHCII), responsible for light-absorption, contains chlorophylls a, b and carotenoids attached to different chlorophyll a/b-binding proteins (Lhcb proteins). The minor Lhcb proteins (CP24, CP26 and CP29) are present as monomers and bind approximately 12 Chl molecules. The major Lhcb proteins (Lhcb 1, 2 and 3) can form mixed trimers, which together bind 24 Chl a, 18 Chl b and 12 carotenoids per trimer (Nevo et al., 2012; Nelson and Junge, 2015). INTRODUCTION 24 The number of monomeric and trimeric LHCII associated to each reaction centre of PSII may vary, resulting in an extensive and flexible antenna supercomplex important not only for light absorption, but also for stacking of thylakoid membranes. In fact, LHCII is the most abundant membrane protein in nature. Moreover, the dynamic increases and decreases in its size, play an important role in the long-term adaptation to light intensity and protection of the photosynthetic apparatus against photooxidative damage (Nevo et al., 2012). The reaction centre of PSII consists of two homologous proteins, D1 and D2, associated with a dimer of Chl a (P680). This complex also contains pheophytin (Phe) and PQ molecules, QA and QB, as electron transport components and the cytochrome b559, which along with -carotene bound to D2, is involved in protection against photoinduced damage (Barber, 2006). The PSII core multiprotein complex consists of the PSII reaction centre with the addition of two chlorophyll a-containing proteins, CP43 and CP47, and a few nuclear-encoded subunits, such as PsbH and PsbS (Nixon et al., 2010). The D1 protein binds the cofactors of the active branch of the reaction centre and it is subjected to a rapid light-dependent turnover. Because of the oxidizing environment created by the excitation of P680 during illumination, D1 frequently suffers light-induced damage and needs to be continuously replaced by a newly synthesized protein. The electron transport starts when light induces a charge separation in PSII. The Chl a dimer present in the reaction centre is excited (P680*) by a photon and gives the excited electron to the first acceptor, Phe, which rapidly reduces the plastoquinone QA that subsequently gives the electron to QB, which is the final and more stable acceptor of PSII. After reducing Phe, the oxidised P680 (P680+) produced has a very high redox potential and is oxidising enough to take electrons from water with the concomitant release of oxygen. The cleavage of water that provides electrons and protons for the photosynthetic process, takes place in the thylakoid lumen of the chloroplast and it is carried out by the oxygen-evolving complex, which forms part of PSII. This complex is linked to P680 by Tyr Z, a redox-active tyrosine residue of the D1 protein. The splitting of water is a fourelectron process, so that absorption of four photons is necessary. For each molecule of INTRODUCTION 25 O2 released, four electrons are withdrawn and accumulated to reduce the chlorophyll molecule one by one. This phenomenon takes place at the Mn-Ca cluster of the complex, where four manganese atoms reduce P680+ going through five intermediate S- states (S0-S4), which is known as the S-state cycle (Barber, 2006). There are several extrinsic proteins (PsbO, PsbP, PsbQ) attached to the luminal side of PSII that play a protective role preventing reductants other than water from having access to the high potential molecules generated at the catalytic site of water splitting. Quinones are highly lipophilic molecules, allowing a rapid lateral diffusion within the lipid bilayer of the thylakoid membrane. The plastoquinone QB is loosely bound to PSII and after receiving two electrons from QA, it takes up two protons from the stroma and the resultant form, a plastoquinol (PQH2), is displaced from its binding site in PSII by another oxidised PQ. PQH2 leaves PSII and reduces the next complex, the Cyt b6f. Cytochrome b6f The Cyt b6f is an integral membrane protein complex with plastoquinolplastocyanin oxidoreductase activity. Electrons from plastoquinol are transferred to Cyt b6f and then to PSI with plastocyanin (PC) as electron transfer intermediate. This complex is uniformly distributed in the thylakoid membrane and provides a connection between PSII and PSI. It includes, beside the polypeptides cytochrome f and b6, with heme-binding sites, a third redox protein with a Fe2S2 prosthetic group (Rieske ironsulphur centre) coordinated by cysteine (Cys) and histidine (His) residues, a quinonebinding protein (subunit IV) and some peripheral proteins necessary for assembly and stability of the complex (Baniulis et al., 2008). Electrons from plastoquinol can follow two pathways at the Cyt b6f. The high potential pathway drives electrons through the Rieske iron-sulphur subunit and cytochrome f to PC, while the low potential pathway involves the Cyt b6 and acts as a Q-cycle, returning the second electron from plastoquinol to the PQ-pool, allowing the net transport of two protons from the stromal side to the lumen, which helps driving the ATP synthesis (Baniulis et al., 2008). Electrons from Cyt b6f are rapidly transferred to INTRODUCTION 32 The last phase consist of regeneration of acceptor molecules converting molecules of 3-carbon into 5-carbon intermediates. Ten enzymes are involved and one ATP is consumed for the production of three RuBP from five GAP. One triose represents the net gain of carbon to the plant. 1.3. Light acclimation Since plants are sessile organisms, their acclimation to the environment is essential to survive adverse conditions. In this context, plants have evolved several mechanisms to sense and to acclimate to different abiotic stresses, among which is the excess of light. Plants are continuously exposed to light of different qualities and quantities. Excess light means that the light absorbed by the plant exceeds its capacity for light utilization (Figure 4). In this situation, photosynthesis becomes saturated and chloroplasts accumulate elevated levels of ROS, which constitutes an important stress for cells (Foyer and Noctor, 2009). Indeed, highly reactive intermediates could cause photo-oxidative damage to proteins and inhibit photosynthesis (Aro et al., 1993; Li et al., 2009; Tyystjarvi, 2013). Figure 4. Photon absorption and light utilization. Under elevated photon flux (high light intensities), the photon absorption exceeds the capacity for light utilization and photosynthetic electron transport rates becomes saturated. Adapted from (Li et al., 2009). INTRODUCTION 33 Adaptive mechanisms to cope with high light exposure may involve control of the capacity for light absorption, as well as dealing with the light energy that has already been captured. To this end, plants respond at different levels of organization; whole organism level, via leaf movements, cellular level, via chloroplast movements or controlling their number, and molecular level (Ruban et al., 2012). At the molecular level, plants have a variety of mechanisms under excess light to sense directly a high incident photon flux, using photoreceptors such as cryptochromes, or indirectly, through biochemical and metabolic signals: decrease in thylakoid lumen pH, increase in the reduction state of PQ-pool and thiol groups in chloroplast proteins, production of ROS and perturbation in chlorophyll biosynthesis (Li et al., 2009). All of these signals are then translated into acclimation responses or regulatory mechanisms, such as changes in gene expression or energy dissipation, to protect against photodamage and photoinhibition. The protective mechanisms operate on different time scale, between minutes and days and may therefore be classified into long- and short-term acclimation responses. The long-term acclimation responses comprise regulation of photosynthetic components, activity of metabolic pathways and antioxidant defence, predominantly by modifying nuclear gene expression, which involves a retrograde signalling between chloroplast and nucleus (Dietz, 2015). Many genes related to light harvesting and antioxidant functions change their expression under high light conditions, such as ELIP or APX genes, which encode light-stress related proteins belonging to the LHC family and ascorbate peroxidase, respectively (Li et al., 2009; Dietz, 2015). The levels of ELIP and APX proteins increase in response to high irradiance, whereas LHCB is decreased. Long-term adaptation processes mainly lead to alterations of the antenna size depending on the incoming light intensities. The induction or repression of these excess light-dependent genes is indirectly regulated by the PQ redox state. The excess of light absorption by PSII leads to an overreduction of the PQ-pool that is sensed via Cyt b6f by kinases, such as STN7 and STN8, which are involved in phosphorylation of LHCII and PSII (Adamiec et al., 2008; Li et al., 2009). Moreover, the ROS generation associated to excess light also regulates gene expression, especially singlet oxygen and H2O2 trigger specific transcriptional responses (op den Camp et al., 2003; Vanderauwera et al., 2005; Li et al., 2009). Other INTRODUCTION 34 genes regulated by high light are those encoding transcription factors and enzymes involved in tetrapyrrole synthesis. Excess light might have different effects on the accumulation of chlorophyll precursors depending on specific conditions (Li et al., 2009). Short-term responses that allow plants to rapidly adapt to variations of light intensity are related with the dissipation of the excess energy absorbed by PSII as heat, a phenomenon known as non-photochemical quenching (NPQ). This subject will be discussed below (see 1.3.2.). 1.3.1. Reactive oxygen species production and PSII photoinhibition Excess light is associated with enhanced generation of ROS in chloroplasts such as hydrogen peroxide (H2O2), superoxide (O2-), hydroxyl radical (OH·) and singlet oxygen (1O2). The decrease in photosynthetic efficiency associated with PSII damage, is called photoinhibition. The major source of ROS generated in chloroplasts is the PET. Electron transfer to molecular oxygen at the acceptor side of PSI can drive the formation of superoxide radicals (Mehler reaction) and, in subsequent reactions, hydrogen peroxide and hydroxyl radicals (Foyer and Noctor, 2009). Moreover, the main ROS responsible for oxidative stress during photosynthesis is considered to be singlet oxygen produced in the reaction centre of PSII (Jahns and Holzwarth, 2012). The excited state of chlorophyll (Chl*) allows conversion of the excitation energy into an electrochemical potential via charge separation, but if the energy is not efficiently used, the excited chlorophyll can give rise to a triplet state (spontaneous spin change) that corresponds to a lower energy excited state (Figure 5) (Krieger-Liszkay et al., 2008; Ruban et al., 2012). Under excess light, when the PQ-pool is in a highly reduced state and electron donation from P680 to QA is inhibited (reaction centres are closed), the chlorophyll triplet state (3Chl*) can be generated also by charge recombination involving a back-flow of electrons from Phe- to P680+ (Krieger-Liszkay et al., 2008). The chlorophyll triplet state can react with molecular oxygen (O2) to produce singlet oxygen (1O2) (Figure 5), which is considered the most important ROS INTRODUCTION 35 responsible for photoinhibition. 1O2 can react with proteins, pigments, nucleic acids and lipids. Therefore, in PSII, 1O2 may damage the reaction centre protein D1 and also cause pigment bleaching (Tyystjarvi, 2013). However, 1O2 is also a signal affecting gene expression and enzyme activation reactions required for maintenance of the components of PSII (Krieger-Liszkay et al., 2008). In the antenna, where chlorophyll is more abundant than in reaction centres, the excess energy of the excited state is quenched by carotenoids that are close enough and avoid 1O2 production, in contrast to reaction centres where this is not possible because of the location of β-carotenes (Krieger-Liszkay et al., 2008). Figure 5. Chlorophyll a excitation and relaxation. Light energy absorbed by chlorophylls associated with PSII can be used to drive photochemistry or can be dissipated as chlorophyll fluorescence or heat, coming back the excited chlorophyll (1Chl*) to its basal state (Chl). If the energy is not efficiently used, 1Chl* can give rise to a triplet state (3Chl*) that may react with molecular oxygen (O2) to produce singlet oxygen (1O2). Adapted from (Muller et al., 2001). To avoid accumulation of damaged D1 protein and progressive loss of PSII activity, this protein is subject rapid turnover allowing the removal of the non-functional protein and the insertion of a newly synthesised copy (Aro et al., 1993; Krieger-Liszkay et al., 2008; Tyystjarvi, 2013). After triggering D1 degradation by 1O2 or even a longlived P680+ radical that oxidizes the nearest pigments (Ruban et al., 2012), PSII is disassembled and the inhibited PSII reaction centre moves to the stroma thylakoids leaving the LHC in the appressed grana regions (Tyystjarvi, 2013). Thereafter, the damaged D1 protein is degraded by FtsH and Deg proteases (Nixon et al., 2010) followed by synthesis de novo of a D1 protein that is inserted into the reaction centre INTRODUCTION 36 with cofactors and pigments during the reassembly process. The reconstituted and activated reaction centre moves back to the grana thylakoids (Krieger-Liszkay et al., 2008; Tyystjarvi, 2013). Under moderate light intensities, PSII activity stays constant since even if photoinhibition lowers the number of active PSII centres, the simultaneous repair through new synthesis of D1 protein, restores it. However, very high light intensities or strong light combined with other abiotic stresses, such as cold or low CO2 may lead to an increase in net photoinhibition (Tyystjarvi, 2013). 1.3.2. Energy dissipation during photosynthesis: Non-Photochemical Quenching NPQ is a short-term response by which plants, and also eukaryotic algae, dissipate excess absorbed light energy as heat (Szabo et al., 2005). It is necessary to balance the absorption and utilization of light energy. Therefore, under excess light, NPQ minimizes the potential for photooxidative damage, since it regulates the photosynthetic electron transport rate (ETR) (Muller et al., 2001; Szabo et al., 2005). This phenomenon may be observed as quenching of chlorophyll a fluorescence, which is not due to photochemistry. The light absorption in LHCs results in singlet-state excitation of a chlorophyll a molecule (1Chl*), which can return to the ground state by different pathways. Excitation energy can be re-emitted as chlorophyll fluorescence, it can be transferred to reaction centres to drive photochemistry, it can be de-excited by thermal dissipation processes or it can decay through formation of the triplet state (3Chl*) (Figure 5) (Muller et al., 2001). The energy, which is diverted to photochemistry or dissipated as heat, results in a fluorescence decrease, or quenching. Therefore, these processes are referred to as photochemical quenching, qP, and nonphotochemical quenching, NPQ (or qN), respectively. Both types of quenching favour a shorter lifetime of 1Chl* and consequently decrease the 3Chl* state and ROS formation (Muller et al., 2001). Using a fluorometer, we can monitor the fluorescence yield of chlorophyll a (see Materials and methods), and according to relaxation times, NPQ may be divided into three different components: energy- or ΔpH-dependent quenching (qE), state transition INTRODUCTION 37 (qT) and photoinhibition (qI). The relaxation times are seconds, minutes and hours, respectively. In higher plants, the state transition does not significantly contribute to NPQ, in contrast to algae. It consists of the uncoupling of LHCs from PSII mediated by phosphorylation. The STN7 kinase is involved in the phosphorylation of a fraction of LHCII proteins (Bellafiore et al., 2005) which leads to their redistribution and association with PSI, resulting in reduction of the PSII antenna size and fluorescence quenching. State transitions are favoured under conditions of limiting CO2 and high reduction level of the chloroplast (Szabo et al., 2005) and their function is related to the excitation energy balance between PSII and PSI. The mechanism of photoinhibition is an irreversible quenching that involves a long-term down-regulation of PSII. Excess light leads to enhanced photooxidative damage of the D1 protein that may exceed the rate of repair resulting in a net reduction of the photochemically active PSII centres which corresponds to an irreversible decrease in chlorophyll fluorescence (qI). Energy-dependent quenching is the major and most rapid NPQ component in plants. qE constitutes an important protective mechanism and plants with higher capacity to induce qE are more resistant to photoinhibition (Li et al., 2002; Roach and Krieger-Liszkay, 2012). Energy-dependent quenching Energy- or ΔpH-dependent quenching is a rapidly inducible and reversible energy dissipation mechanism. qE requires the acidification of the thylakoid lumen, the presence of the PSII S subunit (PsbS) and the xanthophyll zeaxanthin (Zx) (Figure 6). Protonation of PsbS and attachment of Zx to LHCII proteins is suggested to cause a conformational change that leads to quenching of excitation energy (Muller et al., 2001; Szabo et al., 2005). Under high light conditions, the enzyme violaxanthin de-epoxidase (VDE) converts violaxanthin (Vx) into Zx required for qE. Thereafter, if plants are transferred to darkness or low light conditions, Zx is converted back to Vx through zeaxanthin epoxidase (ZE) (Figure 6) (Szabo et al., 2005). Either Zx or Vx are associated with INTRODUCTION 38 trimeric LHCII, bound at the external V1 binding site (Ruban et al., 1999; Jahns and Holzwarth, 2012). The xanthophyll cycle will be described more in detail below. The PsbS protein, also known as the 22-kDa protein or CP22, is a member of the LHC family of proteins. PsbS is closely associated to the core of PSII and seems to act as a sensor of lumen pH. The protonated form of PsbS promotes switching of the PSIILHCII unit from a light-harvesting state containing Vx, present in low light and darkness, to an energy-dissipating state containing Zx, present in high light (Niyogi et al., 2005; Jahns et al., 2009). Figure 6. Energy-dependent quenching. The NPQ component dependent of energy, qE, is induced under high light conditions with the acidification of the thylakoid lumen, the presence of the PSII S subunit (PsbS) and the conversion of the xanthophyll violaxanthin (Vx) into zeaxanthin (Zx) by the action of violaxanthin de-epoxidase (VDE). The qE reversion is possible under low light conditions implying the action of zeaxanthin epoxidase (ZE) which converts Zx back into Vx. Since ΔpH is built up during PET, a sudden decrease in the luminal pH is an immediate signal of excess light. Therefore, variations in pH allow induction or reversal of qE within seconds after a change in light intensity, which is fast enough to cope with natural light fluctuations, to which plants are exposed throughout the day (Muller et al., 2001). INTRODUCTION 39 A decrease in luminal pH induces qE through protonation of PsbS and activation of VDE with the concomitant production of Zx. Arabidopsis thaliana mutants lacking VDE, npq1 mutants, have greatly reduced NPQ (Niyogi et al., 1998). Similarly, mutants lacking PsbS, npq4 mutants, have hardly any NPQ (Li et al., 2000). In contrast, mutants without ZE, npq2 mutants, accumulate Zx constitutively and present higher levels of NPQ at low and moderate light intensities and a more rapid induction of NPQ (Niyogi et al., 1998). The increase in photoinhibition observed in mutants with deficient NPQ, such as npq4 and even npq1, is indicative of the important role of qE to attenuate photodamage (Roach and Krieger-Liszkay, 2012). In the absence of PsbS, qE can be obtained artificially with a lower pH in vitro. Thus, PsbS is required to shift the pKa of qE to physiologically attainable levels of lumen pH (Johnson and Ruban, 2011). Moreover, the level of PsbS determines the extent of qE and there exists a correlation between capacity to induce qE and the amount of PsbS protein (Li et al., 2002). Zx is necessary for about 70% of the total NPQ and 80% of qE in Arabidopsis (Li et al., 2002). Thus, in the absence of Zx, qE seems to be possible through the action of lutein bound at the L1 site of the antenna (Crouchman et al., 2006; Johnson et al., 2009). Zx also plays a role in the direct quenching of energy from excited chlorophyll, which constitutes an important photoprotective function independent from its function in qE. Thus, Zx acts as antioxidant in the lipid phase or protein/lipid interphase of the thylakoid membrane, preventing lipid peroxidation (Havaux and Niyogi, 1999; Johnson et al., 2007; Jahns and Holzwarth, 2012), even when unbound to proteins (Havaux et al., 2007; Dall'Osto et al., 2010). The Zx epoxidase activity is gradually inhibited after increasing high light stress, which is correlated with the down-regulation of PSII activity. The resulting high levels of Zx enhance the formation of qE and retard its relaxation (Jahns and Holzwarth, 2012). The xanthophyll cycle Avoiding photoinhibition under excess light is important for plant survival but it is equally important to optimise light harvesting under light-limiting conditions to permit plant growth. The rapid interchange between violaxanthin and zeaxanthin bound to LHC, allows plants to adapt efficiently to changes in light intensity. The reversible conversion of violaxanthin, Vx, to zeaxanthin, Zx, via the intermediate antheraxanthin, INTRODUCTION 40 Ax, is known as the xanthophyll (or violaxanthin) cycle and two enzymes are involved: violaxanthin de-epoxidase, VDE, and zeaxanthin epoxidase, ZE (Figure 7) (Jahns et al., 2009). VDE catalyses the conversion of Vx into Zx through a two-step reaction in which two epoxy groups are removed from Vx, while ZE catalyses the opposite reaction, the conversion of Zx into Vx, adding two epoxy groups to Zx, also in two steps with Ax as intermediate (Figure 7). Both enzymes are lipocalin proteins located on the opposite sides of the thylakoid membrane: VDE is located in the thylakoid lumen and ZE on the stromal side (Jahns et al., 2009). Figure 7. The xanthophyll cycle. The xanthophyll cycle consists of the reversible conversion of violaxanthin (Vx) to zeaxanthin (Zx) via the intermediate antheraxanthin (Ax). Two enzymes are involved: violaxanthin de-epoxidase (VDE), in the thylakoid lumen, and zeaxanthin epoxidase (ZE), on the stromal side. Adapted from (Szabo et al., 2005). The activity of VDE is strictly controlled by the pH of the thylakoid lumen. The enzyme does not show any activity at pH values greater than 6.3 and shows maximum values at pH values less than 5.8 (Pfundel and Dilley, 1993). This ensures VDE activity and, subsequent Zx accumulation, only under saturating light intensities. Moreover, VDE requires ascorbate as cofactor, which is used as electron donor to the epoxide groups (Yamamoto et al., 1971) and the optimal amount of ascorbate depends on pH INTRODUCTION 41 (Bratt et al., 1995). Thus, the de-epoxidation reaction yields Zx, water and oxidised dehydroascorbate. VDE consists of a lipocalin protein with an N-terminal cysteine-rich region and a C-terminal region enriched in glutamyl residues, both with a high degree of sequence similarity between different plants (Figure 8) (Bugos and Yamamoto, 1996). An inhibitory effect of dithiothreitol (DTT), which reduces disulphide bonds, on VDE activity has been described (Yamamoto and Kamite, 1972). Zx accumulated at high light intensities is reconverted to Vx by ZE in low light or darkness. ZE is also a lipocalin protein with some conserved cysteines (Figure 8) (Bouvier et al., 1996; Marin et al., 1996) and therefore, potentially susceptible to be redox regulated. However, redox regulation of this enzyme has not been reported. Figure 8. Domain architectures and positions of cysteines in the Arabidopsis thaliana VDE and ZE. Sequence information was obtained from TAIR (http://www.arabidopsis.org). Chloroplast N-terminal transit peptides are labelled in green, 113 aa in VDE and 59 aa in ZE. VDE has a bipartite transit peptide, which targets the protein to the thylakoid lumen. In red are labelled the mature proteins, in blue, the FAD-binding domains and in yellow, the cysteines. In ZE, Cys are conserved in Oryza sativa and Cys are conserved in Chlamydomonas reinhardtii. NADPH, FAD, Fd and O2 are required as cofactors for ZE. In each reaction, one oxygen atom is incorporated into Zx using the reducing power of NADPH via reduced Fd and FAD and the optimal pH for ZE activity is 7.5 (Bouvier et al., 1996; Jahns et al., 2009). Apart from playing several roles in photoprotection, Zx is also an intermediate in the biosynthetic pathway of abscisic acid (ABA), a plant hormone which plays an INTRODUCTION 48 severe phenotype as compared to wild-type (wt) plants (Serrato et al., 2004). ntrc plants are stunted and chlorotic and this phenotype is dependent on the photoperiod, being more pronounced under short-day conditions. Moreover, ntrc shows lower CO2 fixation rates, especially at low light intensities, abnormal chloroplast structure and is hypersensitive to different abiotic stresses, such as drought, cold, high salinity and prolonged darkness (Perez-Ruiz et al., 2006; Lepisto et al., 2009). Figure 11. Proposed model for the reaction mechanism of NTRC. The minimal catalytic unit of NTRC is the dimer. Targets of NTRC are reduced via the Trx domain. Electrons are transferred from NTR to Trx domains through an inter-subunit pathway. Adapted from (Perez-Ruiz and Cejudo, 2009). A role in protection against oxidative stress has been attributed to NTRC through reduction of 2-Cys Prx, being the most efficient electron donor of this enzyme in vitro and in vivo compared to other Trxs, such as CDSP32 and Trx x (Perez-Ruiz et al., 2006; Pulido et al., 2010). This also implies a role for NTRC in redox signalling (Puerto-Galan et al., 2013). This function could be more important in darkness when NTRC is the only candidate as catalyst for disulphide reduction (Spinola et al., 2008). In addition, other enzymes can be reduced by NTRC implying this protein in the regulation of several chloroplast processes (Cejudo et al., 2012). This is also suggested from the severe growth phenotype of ntrc as compared to the almost wt phenotype of a mutant containing less than 5% of the 2-Cys Prx, Δ2cp (Pulido et al., 2010). NTRC is involved in the biosynthesis of tetrapyrroles through reduction of different enzymes, such as glutamyl-transfer RNA reductase 1 (GluTR1), Mgprotoporphyrin methyl transferase (CHLM) (Richter et al., 2013) and the CHLI subunit of the Mg-chelatase complex (Perez-Ruiz et al., 2014). Furthermore, the lower content INTRODUCTION 49 of starch in the ntrc mutant led to the identification of ADP-glucose pyrophosphorylase (AGPase) as an NTRC-regulated enzyme (Michalska et al., 2009; Lepisto et al., 2013). The redox state of AGPase was found to be altered in ntrc not only in leaf chloroplasts, but also in roots, suggesting the presence of NTRC also in heterotrophic tissues (Michalska et al., 2009). The presence of NTRC in plastids of non-photosynthetic tissues as well as in chloroplasts has recently been demonstrated. Interestingly, the selective expression of NTRC in leaves is sufficient to rescue the phenotype of the ntrc mutant with respect to growth and pigmentation, revealing the supreme importance of chloroplast redox regulation for plant growth (Kirchsteiger et al., 2012). The co-existence of two different thiol-based light-dependent redox-signalling systems in the chloroplast (Figure 9) is intriguing and raises a number of questions. Some of theses questions relate to the presumed substrate specificities and redundancies between these systems. In addition, putative interactions and even interdependence between the NAPDH-dependent NTRC system and Fd-FTR-dependent simple module Trx systems should be taken into account. The possible implication of chloroplast redox enzymes as signalling molecules in high light acclimation was postulated several years ago (Li et al., 2009), but has so far not been proven. Recently, the involvement of Trxs in adaptation of plants to fluctuating light conditions has also been considered (Nikkanen and Rintamaki, 2014). Therefore, the specific contributions of NTRC and FTR-dependent Trxs to the possible redox regulation of light acclimation mechanisms should be determined in order to understand how plants adapt to the natural variations in their light environment. OBJECTIVES OBJECTIVES 53 2. OBJECTIVES The aim of this thesis was to elucidate the possible implication of the two distinct chloroplast redox systems in adaptation of plants to changes in light conditions and the regulation of photosynthesis. To this end, the following specific objectives were addressed: 1. To investigate the possible role of NTRC in long- and short-term acclimation to different light intensities and in protection against excess excitation. 2. To establish the relevance of ftype Trxs for plant growth and their function in carbon assimilation. Part of this work has been performed during a three months stay in the laboratory of Dr. Anja Krieger-Liszkay at the Institute for Integrative Biology of the Cell (I2BC), Commissariat à l’Energie Atomique et aux Energies Alternatives (CEA) Saclay, France. Part of the results presented in this thesis has been included in the following manuscripts: - The chloroplast NADPH thioredoxin reductase C, NTRC, is required for efficient utilisation of light energy and control of photosynthetic electron transport in Arabidopsis. Belén Naranjo, Clara Mignée, Anja Krieger-Liszkay, Dámaso Hornero-Méndez, Lourdes Gallardo-Guerrero, Francisco Javier Cejudo and Marika Lindahl. Plant Cell & Environment (pending minor revisions). - Type-f thioredoxins function in short-term activation of carbon metabolism after a dark-light transition but are dispensable for growth in Arabidopsis thaliana. OBJECTIVES 54 Belén Naranjo, Antonio Díaz-Espejo, Marika Lindahl and Francisco Javier Cejudo. Journal of Experimental Botany (pending minor revisions). - Thioredoxin f1 and NADPH-dependent thioredoxin reductase C have overlapping functions in regulating photosynthetic metabolism and plant growth in response to varying light conditions. Ina Thormählen, Tobias Meitzel, Julia Groysman, Alexandra Öchsner, Edda von Roepenack-Lahaye, Belen Naranjo, Francisco Cejudo, and Peter Geigenberger. Plant Physiology, September 2015 (http://dx.doi.org/10.1104/pp.15.01122). MATERIALS AND METHODS MATERIALS AND METHODS 57 3. MATERIALS AND METHODS 3.1. Plant material and growth conditions Arabidopsis thaliana wild type and mutant plants (ecotype Columbia) were grown on soil in growth chambers (SANYO, Leicester, UK) under long-day (16 h light / 8 h darkness) or short-day (8 h light / 16 h darkness) conditions at a light intensity of 120 mol quanta m-2 s-1 or in a growth chamber FITOCLIMA 700 EDTU from Aralab (Rio de Mouro, Portugal) under a photoperiod of 10 h light / 14 h darkness at different light intensities: 50 (low light), 120 (normal light), 600 (moderately high light), or 1000 mol quanta m-2 s-1 (high light). Temperature, 22 ºC during the light and 20 ºC during the dark periods, and relative humidity (60%) were strictly controlled. Arabidopsis mutants used in this work, previously reported, were the ntrc knockout mutant (Serrato et al., 2004; Perez-Ruiz et al., 2006), the trxx and  2cp mutants (Pulido et al., 2010), and the trxf1 mutant (Perez-Ruiz et al., 2014). In addition, two new single mutants were isolated, trxf2 (GK-020E05-013161) and psbs (SALK_095156), which were used to generate the double mutants trxf1f2 and ntrc-psbs. Generation of Arabidopsis mutants A homozygous line GK-020E05-013161 with a T-DNA insertion in the TRX f2 gene (At5g16400.1) from Arabidopsis, termed the trxf2 mutant, was selected by PCR analysis of genomic DNA with the oligonucleotides described in Tables 1 and 2. This mutant was manually crossed with the trxf1 mutant. Seeds resulting from this cross were tested for heterozygosity of the T-DNA insertions in the TRX f1 and TRX f2 genes. Plants were then self-crossed and double homozygous lines, denoted trxf1f2 mutants, were identified in the progeny by PCR analysis of genomic DNA using oligonucleotides described in Tables 1 and 2. DNA extraction was performed with CTAB (Sigma) based on the protocol described by (Murray and Thompson, 1980). The same procedure was performed with the insertion line SALK_095156, which contains the T-DNA in the second exon of the PsbS gene (At1g44575). Plant genotypes were analysed with oligonucleotides PsbS LP - RP (wild type allele) and Lba1 - PsbS RP (mutant allele) described in Tables 1 and 2, and homozygous psbs MATERIALS AND METHODS 64 negligible. After switching off the light, depending on the recovery kinetics of the Fm values, it is possible to distinguish between the different components of NPQ: qI, qE and qT. Figure 12. Chlorophyll a fluorescence. Applying saturating light pulses, maximum chlorophyll fluorescence values (Fm in darkness and Fm’ in light) are obtained. Black and white bars indicate darkness and actinic light, respectively. The fluorescence quenching (Fm’ - F’) corresponds to the energy used in photochemistry, i.e. photochemical quenching (qP). The difference between Fm and Fm’, corresponds to the light energy absorbed that does not reach P680 (NPQ). The fluorescence maxima rapidly recovered during the subsequent dark period reflects the energy-dependent component of NPQ, qE, while the fraction remaining is the inhibitory component, qI. Maximum quantum yield of PSII The maximum quantum yield of PSII was assayed after incubation of plants in the dark for 30 min by calculating the ratio of the variable fluorescence, Fv, to maximal fluorescence, Fm, (Fv/Fm). Fm value was determined applying a saturation pulse of red light at 10000 mol quanta m-2 s-1 intensity and 0.6 s duration to the dark adapted leaf. Induction-recovery curve Induction-recovery curves were performed using red actinic light at the intensities specified for each experiment during 8 min. Saturating pulses of red light at 10000 mol quanta m-2 s-1 intensity and 0.6 s duration were applied every 60 s and MATERIALS AND METHODS 65 recovery in darkness was recorded for up to 12 min. The parameters Y(II), Y(NPQ) and Y(NO), corresponding to the respective quantum yields of PSII photochemistry, nonphotochemical quenching and non-regulated basal quenching (NO), were calculated by the DUAL-PAM-100 software according to the equations in (Kramer et al., 2004b). In addition, induction-recovery curves were performed using IMAGING-PAM M-Series Chlorophyll Fluorescence System (Walz, Effeltrich, Germany) to simultaneously determine various photosynthetic parameters in the complete aerial part of plants. Dark-adapted plants were illuminated with 81 mol quanta m-2 s-1 of actinic light during 360 s and subsequently, kept in darkness for 180 s. Saturating pulses were applied each 20 s. Electron transport rates Relative linear electron transport rates were measured in pre-illuminated plants applying stepwise increasing actinic light intensities up to 2000 mol quanta m-2 s-1. Values were calculated by the DUAL-PAM100 software. Post-illumination fluorescence Dark-adapted leaves (10 min in darkness) were subjected to a saturation pulse followed by 5 min of actinic light. Subsequently, post-illumination fluorescence (PIF) was measured during 5 min in darkness. The transient increase detected in chlorophyll fluorescence after actinic light illumination depends on QA reduction due to backflow of electrons from the stroma to the PQ-pool. PIF has been associated with CEF around PSI and is typically absent in mutants lacking the NDH-complex (Shikanai et al., 1998; Munekage et al., 2004). 3.7. Measurements of photosystem I activity: P700 absorbance The oxidised form of P700 displays a broad absorbance peak around 800-840 nm. Thus, it is possible to analyse its redox state monitoring changes in the absorbance MATERIALS AND METHODS 66 of leaves at 830 nm. The P700 signal represents the difference between the transmittance signals at 830 nm and 875 nm, respectively. In dark-adapted leaves, P700 is found reduced, the acceptor side of P700, i.e., the Calvin-Benson cycle and subsequent reactions, are de-activated. Applying far red (FR) light (730 nm), that is preferentially absorbed by PSI pigments and reduces Fd, P700 is oxidised. The maximal level of oxidation (Pm) is achieved by application of a red light saturation pulse (635 nm, 10000 µmol quanta m-2 s-1 during 0.6 s) in the presence of FR illumination. The signal a few ms after the SP, corresponds to P700 fully reduced (P0) (Figure 13). Figure 13. Oxidation of P700. In dark-adapted plants P700 is reduced. The maximal level of oxidation (Pm) is achieved by applying a saturating light pulse (SP) in the presence of far red (FR) (SP after 10 s of FR pre-illumination). Fully reduced P700 (P0) corresponds to the signal level milliseconds after this SP. An illuminated sample displays a P700 signal denoted by P. Applying SPs during actinic light illumination, strong P700 oxidation (Pm’) followed by full reduction (P0) is obtained. The difference between P and P0 corresponds to the fraction P700+ A, which reflects a donor deficit to P700 (ND). The difference between Pm and Pm’ corresponds to the fraction P700 A-, which reflects an acceptor deficit from P700 (NA). The difference between Pm’ and P corresponds to the fraction P700 A, which reflects the P700 pool able to perform charge separation, i.e., the PSI quantum yield, Y(I). Under actinic light P700 is reduced by electrons coming from the PQ-pool and applying SP its ability to become reduced and oxidised can be determined. SP induces strong P700 oxidation (Pm’) followed by full reduction (P0). The PSI quantum yield depends on the redox states of P700 and of the PSI acceptors (A). The difference MATERIALS AND METHODS 67 between Pm and Pm’ corresponds to the pool of P700 that cannot become oxidised because of a deficit at the acceptor side of PSI (P700 A-), which is referred to as nonphotochemical acceptor-side limited heat dissipation (NA) (Figure 13). The difference between P, the P700 signal under illumination, and P0 corresponds to the pool of P700 that cannot become reduced because of a donor deficit (P700+ A), which is referred to as non-photochemical donor-side limited heat dissipation (ND) (Figure 13). The PSI quantum yield, Y(I), is the fraction of open centres, when P700 is reduced and acceptors are oxidised (P700 A) and corresponds to the difference between Pm’ and P (Klughammer and Schreiber, 2008). Induction-recovery curve The redox state of photosystem I was monitored by following changes in absorbance at 830 nm versus 875 nm using the DUAL-PAM-100. Plants were kept in the dark for 30 min and, to probe the maximum extent of P700 oxidation, a red light SP of 0.6 s at 10000 µmol quanta m-2 s-1 was applied to leaves after 10 s of FR preillumination. Thereafter, absorbance traces were recorded during 5 min illumination with 126 µmol quanta m-2 s-1, 635 nm, red actinic light followed by 5 min darkness. Saturation pulses were applied every 20 s. The quantum yields of PSI photochemistry Y(I), donor side limitations Y(ND) and acceptor side limitations Y(NA) were based on saturating pulse analysis and calculated by the DUAL-PAM-100 software (Klughammer and Schreiber, 2008). 3.8. Thermoluminescence Thermoluminescence (TL) from PSII is delayed light emission by singlet excited chlorophylls stimulated by warming of samples which have been previously illuminated either by single turnover flashed or continuous FR light at 1°C. Illumination was performed at a relatively low temperature to stabilize the charge pair formed in the light. TL results from recombination of charge pairs between positively charged carriers on the donor side of PSII and electrons on the acceptor side of PSII that produce excited P680. Thus, the S2 and S3 oxidation states of the Mn cluster of the oxygen evolving MATERIALS AND METHODS 68 complex may recombine with the primary and secondary quinone electron acceptors of PSII, QA and QB, generating the S2QA-, S2QB- and S3QB- charge pairs (Ducruet, 2003). After a single turn-over flash, which induces one charge separation in each PSII centre, the B-band is typically observed. This band originates from S2/3QB− recombination and has a temperature maximum at about 35 °C. The afterglow band (AG-band) is typically induced by FR illumination and has a temperature maximum at about 45 °C. The AG-band is a B-type recombination that depends on the back transfer of electrons from stroma to the acceptor side of PSII (Ducruet, 2003; Ducruet and Vass, 2009). In this back electron transfer PQH2/QB- is generated. FR illumination excites mainly PSI but does also excite PSII so that higher oxidation states of the Mn cluster (S2/S3) are generated (Thapper et al., 2009). TL measurements were performed to probe the redox state of the PQ-pool in dark-adapted leaves according to the method described in (Roach and Krieger-Liszkay, 2012). The measurements were carried out with a home-built apparatus on leaf segments taken from plants dark adapted for 15 min. TL was excited with 1 single turnover flash or 30 s of FR at 1 °C. Samples were heated at a rate of 24 °C/min from 1 to 70 °C, and the light emission was recorded. Graphical and numerical data analyses were performed according to (Ducruet and Miranda, 1992). 3.9. Electrochromic pigment absorbance shift The movement of electrons and protons through the thylakoid membrane generates a proton motive force which comprises an electric field, or transmembrane potential, (Δψ) and a proton concentration gradient (ΔpH). The presence of the electric field has also an effect on the photosynthetic pigments modifying their spectrum (Bailleul et al., 2010). The "electrochromic shift" (ECS) signal measured at 520 nm has been shown to respond linearly to changes in trans-thylakoid potential and is mainly due to carotenoids and Chl b (Schreiber and Klughammer, 2009). The relaxation kinetic of ECS in darkness after illumination allows determination of the two components of pmf, Δψ and ΔpH (Takizawa et al., 2007). The dark relaxation of ECS is studied after photosynthetic steady-state conditions where proton influx into the lumen is precisely balanced by efflux (Figure MATERIALS AND METHODS 69 14). Therefore, dark relaxation kinetics of ECS were measured in leaves after 15 min of actinic illumination by red LEDs (650 nm) at intensities from 12 to 940 µmol quanta m- 2 s-1. The total rapid change in ECS (tens of ms after switch off the light), termed ECSt, should be proportional to the light-induced pmf (pmf in light minus pmf in dark). The rapid signal decline is followed by a biphasic signal increase to an apparent “darkbaseline”. This slowly-reversible (tens of s) phase of ECS decay is known as inverted ECS or ECSinv and is proportional to the light-induced ΔpH. The difference between ECSt and ECSinv therefore reflects the Δψ portion of light induced pmf (Takizawa et al., 2007). Figure 14. Dark-relaxation of electrochromic shift (ECS). After steady-state photosynthetic conditions, the rapid ECS signal decline (ECSt) is proportional to the light-induced pmf. This relaxation is slowly reversible and the increased signal is known as inverted ECS (ECSinv) which is proportional to the light-induced ΔpH. The difference between ECSt and ECSinv reflects the Δψ portion of light induced pmf. 3.10. Determination of the rate of carbon assimilation Net CO2 assimilation rate (AN) was measured using an open gas exchange system Li-6400 equipped with the 6400-40 leaf chamber fluorometer (LI-COR Biosciences, Lincoln, NE, USA). All measurements were conducted in dark-adapted leaves at 500 mol mol-1 of CO2, constant leaf temperature of 20 ºC and vapour pressure deficit between leaf and air below 1 kPa. Before the light, at an intensity of 70 mol quanta m-2 s-1, was turned on in the leaf chamber, AN was recorded for 30 min every 5 s. Then recording continued for 10 min or until equilibrium was reached. MATERIALS AND METHODS 70 3.11. Determination of starch content Starch content was determined in leaves of plants grown under long-day conditions harvested before flowering (22 to 26 days-old), essentially as described in (Lin et al., 1988). For starch extraction, leaves (100 to 200 mg fresh weight) harvested at the end of the day, were ground in liquid nitrogen and washed with 80% (v/v) ethanol in 10 mM HEPES pH 7.6, during 2 h at 80ºC and then washed with the same solution at room temperature until the tissue was free of pigments. Dry pellets, after centrifugation, were resuspended in 1 ml of 0.2 N KOH and heated at 100 °C for 30 min. After cooling, samples were centrifuged at 17000 g for 10 min and the supernatant was adjusted to pH 5.0 with 1 N acetic acid. An aliquot of 200 μl of this solution was used to determine the starch amount using the enzymatic method described by (Lin et al., 1988). 3.12. Electron paramagnetic resonance Electron paramagnetic resonance (EPR) spectroscopy allows direct detection of radicals in complex biological systems such as intact cells and tissue samples. EPR is a technique based on the absorption of electromagnetic radiation by a paramagnetic sample, i.e. a species with one or more unpaired electrons, when placed in a magnetic field. These absorptions only occur at very well defined frequencies (usually in the microwave region) and magnetic field combinations, which depend on the type of paramagnetic species present and its characteristics. Hence, different types of radicals (and also metal ions with unpaired electrons) will produce lines at very different regions of the scanned magnetic field and the signal is directly proportional to the concentration of the paramagnetic species (Hawkins and Davies, 2014). Spin trapping agents are used to facilitate the detection of reactive radicals present in low concentration. This method involves the addition of a spin trap, typically a nitrone or nitroso compound to give stable and detectable nitroxide radical adducts (Hawkins and Davies, 2014). To detect H2O2-derived hydroxyl radicals the spin trap 4-pyridyl-1-oxide-N-tert- butylnitrone (4-POBN) was used. Adding Fe-EDTA, OH· radicals are produced from H2O2 via the so-called Fenton reaction (Mubarakshina et al., 2010). Spin trapping assays were carried out using leaves from plants growing at 120 mol quanta m-2 s-1. A MATERIALS AND METHODS 71 piece of 15 - 20 mg of leaf per sample was infiltrated with a solution containing 50 mM 4-POBN, 4% ethanol, 2.5 µM Fe-EDTA, and 20 mM phosphate buffer pH 7.0. Infiltrated leaves in 250 l of the same solution were incubated during 1 h in growth light (120 mol quanta m-2 s-1), moderately high light (450 mol quanta m-2 s-1) or darkness. EPR spectra were recorded at room temperature in a standard quartz flat cell using an ESP-300 X-band (9.73 GHz) spectrometer (Bruker, Rheinstetten, Germany). The following parameters were used: 9.73 GHz microwave frequency, 100 kHz modulation frequency, 1 G modulation amplitude, 63 mW microwave power, 2 × 104 receiver gain, 40.96 ms time constant; number of scans, 2 (Mubarakshina et al., 2010; Roach and Krieger-Liszkay, 2012; Galzerano et al., 2014). 3.13. Cloning, expression and purification of recombinant proteins For the expression of proteins, the coding sequences (CDS) for VDE (At1g08550) and ZE (At5g67030), without putative transit peptides (113 and 59 amino acids, respectively), were amplified from the full-length cDNA by PCR using genespecific oligonucleotides, which included restriction sites for cloning into the pQE-30 vector (Qiagen Sciences, Germantown, MD, USA). The sequences of these oligonucleotides were as follows (Table 4). Table 4. Oligonucleotides used to amplify the complete coding sequences (CDS) of mature proteins VDE (At1g08550) and ZE (At5g67030). Oligonucleotides include restriction sites (underlined at the DNA sequence). Primer DNA Sequence (5’ - 3’) VDE Fw (BamH1) GAGGATCCGTTGATGCACTTAAA VDE Rv (PstI) AACTGCAGCCTGACCTTCCTGAT ZE Fw (BamH1) AAGGATCCGCGGCGACG ZE Rv (HindIII) GCAAGCTTAGCTGTCTGAAGTAA cDNA sequences amplified with oligonucleotides described in Table 4, using iProof High Fidelity enzyme (BIO-RAD), were ligated into the pGEMt vector MATERIALS AND METHODS 72 (Promega). Escherichia coli DH5α cells were transformed by heat shock (30 min on ice, 2 min at 42 ºC and 5 min on ice) with the ligation product, to obtain a high number of copies of the plasmid. The constructions were extracted with a plasmid DNA purification kit, DNA-Spin (Intron biotechnology), according to the manufacturer’s instructions. All cDNAs had both strands sequenced and were restricted with the corresponding enzymes, BamH1 and PstI or HindIII to introduce the VDE or ZE sequences into the pQE-30 expression vector. Proteins were overexpressed from pQE-30 in Escherichia coli XL1-Blue cells (transformed by heat shock), using 0.5 mM IPTG. VDE expression was induced during 4 h at 25 ºC and ZE, over-night at 25 ºC. After breaking cells using sonication or French press, the recombinant proteins, containing a His-tag at the N-terminus, were purified by nickel-affinity chromatography. ZE was purified using an ÄKTA prime FPLC system (GE Healthcare, Buckinghamshire, UK) with Hi-Trap chelating columns, 1 ml (GE-Healthcare), and 20 mM phosphate buffer pH 7.4 containing 200 mM NaCl and 10% glycerol. An elution gradient was applied from 40 to 500 mM imidazole (ZE eluted at ~ 250 mM imidazole). VDE was not soluble and therefore, was purified from inclusion bodies with 6 M urea in 20 mM Tris-HCl pH 8.0 and 500 mM NaCl buffer using Ni-NTA Agarose resin (Qiagen). Protein was eluted with 500 mM imidazole. Quantification of proteins was determined using the Bio-Rad protein assay. Purified recombinant rice NTRC was prepared as in (Serrato et al., 2004), but using nickelinstead of copper-affinity chromatography. Trx x and BAS1 (2-Cys Prx) were purified as in (Perez-Ruiz et al., 2006), and Arabidopsis Trx f1 according to (Perez-Ruiz et al., 2014). Escherichia coli cultures were grown in LB medium (10 g/l NaCl, 5 g/l yeast extract, 10 g/l tryptone). For solid media, LB was supplemented with 15 g/l agar and 100 µg/ml ampicillin was added for selection of strains containing pGEMt and pQE30 vectors. In some cases, 0.1 mM X-Gal was added to select positive colonies with pGEMt containing inserts and 2% glucose to repress pQE30 basal expression. MATERIALS AND METHODS 73 3.14. NTRC in vitro activity In vitro alkylation assays Purified recombinant proteins in 10 mM phosphate buffer pH 7.4 were incubated during 20 min at room temperature in different combinations and in the presence of CuCl2, H2O2, DTT or NADPH. Some of these proteins were previously treated with CuCl2 or DTT during 20 min at room temperature and then dialyzed using Amicon Ultra-0.5 filters (Millipore) to remove the reagents. After the reactions, proteins were precipitated using 10% TCA and then alkylated with 10 mM MM(PEG)24 as described above for tissue samples. Twenty l of alkylated protein samples were subjected to SDS-PAGE under reducing conditions and stained with Coomassie Brilliant Blue. Determination of NADPH concentration The reduction of 1 mM NADP+ to NADPH by 2, 5 or 10 M NTRC was determined following absorbance increase at 340 nm during 20 min in 250 l of 100 mM K3PO4 solution pH 7.0. One mM DTT or 0.1 mM DTT in the presence of 20 Trx f1 was used as electron donor. The concentration of NADPH was determined using the extinction coefficient for NAD(P)H at 340 nm (ε340 = 6220 M-1 cm-1). 3.15. Bioinformatic methods Sequences of genes and proteins were obtained from the Arabidopsis database TAIR (The Arabidopsis Information Resource) (http://www.arabidopsis.org) and the Universal Protein Resource, Uniprot (http://www.uniprot.org/). Oligonucleotides were designed using the Primer-BLAST application from the National Center for Biotechnology Information, NCBI (http://blast.ncbi.nlm.nih.gov/Blast.cgi). To analyse mutant plants, the T-DNA Express tool from Salk Institute was also used (http://signal.salk.edu/cgi-bin/tdnaexpress). The search for nucleotide or amino acid sequence similarity with the sequences available in databases was made using the BLAST application from NCBI RESULTS 80 Figure 17. Photoinhibition in the presence of lincomycin. Leaves from 5-weeks old plants grown at 120 mol quanta m-2 s-1 light intensity under short photoperiod, were infiltrated with 2 mM lincomycin or water and transferred to a light intensity of 450 mol quanta m-2 s-1. Photoinhibition was assayed by chlorophyll fluorescence as Fv/Fm in untreated (A) and lincomycin-treated (B) leaves. Prior to measurements, leaves were kept in the dark for 10 min. During light incubation, infiltrated leaves were floated on water or lincomycin solution. Each data point is the mean of the Fv/Fm values measured from 4 leaves and SDs are indicated by error bars. Under high light conditions the increase in formation of superoxide anion leads to enhanced production of hydrogen peroxide and hydroxyl radicals (Foyer and Noctor, 2009; Li et al., 2009). Taking into account that NTRC is an important source of electrons for one of the most abundant chloroplast peroxidases, the 2-Cys Prx (Moon et al., 2006; Perez-Ruiz et al., 2006; Alkhalfioui et al., 2007), we examined the content of peroxides in plants at their growth light intensity (120 mol quanta m-2 s-1) and after exposure to moderate high light (450 mol quanta m-2 s-1) or darkness. To this end, leaves infiltrated with 4-POBN/ethanol in the presence of Fe-EDTA were immersed in the same solution while illuminated during 1 h and EPR spectra of the 4-POBN/αhydroxyethyl adducts were recorded (Figure 18A). Notably, ntrc plants had 40% higher content of peroxide-derived hydroxyl radicals than the corresponding wild type controls at growth light intensity, but not at higher light intensity (Figure 18B). This would speak against a function of NTRC in peroxide detoxification at high irradiances. RESULTS 81 Figure 18. Hydrogen peroxide content in leaves from wild type and ntrc mutant plants. Hydrogen peroxide in wild type and ntrc mutant leaves from short-day plants after 1 h in the dark, D, growth light, GL (120 mol quanta m-2 s-1), and in high light, HL (450 mol quanta m-2 s-1). The presence of peroxide-derived hydroxyl radicals was shown by indirect spin trapping with 4-POBN/ethanol in the presence of Fe-EDTA. A) Typical electron paramagnetic resonance (EPR) spectra of the 4-POBN/α-hydroxyethyl adducts are shown. B) Relative content of hydroxyl radicals based on EPR signals from experiments, such as in (A). The signal from the ntrc mutant is expressed as % of the corresponding wt control sample ± SE. Darkness, n = 5; growth light, n = 8; high light, n = 4. Significantly different values according to Student’s t-test are marked with different letters (p < 0.05). 4.1.2. Acclimation to high light: PSII energy dissipation Protection against excess light in plants lacking NTRC To measure the ability of the ntrc mutant to adapt to a sudden increase in light intensity, chlorophyll fluorescence was monitored during illumination of dark-adapted leaves with actinic light followed by darkness (Figure 19). The intensity of the actinic light in this experiment, 278 mol quanta m-2 s-1, was about twice the growth light of these plants. The saturating light pulse-induced peaks of maximal fluorescence during illumination (Fm’) were smaller than the maximal fluorescence in the dark-adapted state (Fm) for both wild type and ntrc mutant plants (Figure 19). This shows that the mutant is capable of responding adequately to an increase in light intensity by inducing NPQ. RESULTS 82 During the subsequent dark period, fluorescence maxima were recovered in both wild type and mutant plants with similar kinetics (Figure 19). Figure 19. Chlorophyll fluorescence in wild type and ntrc mutant plants. Induction and recovery curves with 278 mol quanta m-2 s-1 intensity actinic light. Attached leaves of 5-weeks old plants grown at 120 mol quanta m-2 s-1 light intensity under a photoperiod of 10 h light and 14 h darkness were used. Following 30 min dark incubation and determination of F0 and Fm, actinic light was turned on and saturating pulses were applied every 60 s. After 8 min illumination, measurements were continued for another 10 min in the dark. White and black bars below graphs indicate periods of illumination with actinic light and darkness, respectively. At actinic light intensities lower than growth light intensity NPQ is normally not observed, except shortly after the onset of light due to transient acidification of the thylakoid lumen (Kalituho et al., 2007). This was confirmed in the fluorescence curves recorded for the wild type plants (Figure 20A). In contrast, the ntrc mutant displayed extensive NPQ during the entire illumination period, as revealed by the dramatic reduction of the Fm’ values (Figure 20A). This response is abnormal and distinguishes plants lacking NTRC from wild type plants. In order to compare quantitatively the photosynthetic performance of these plants during induction and recovery, the respective quantum yields of NPQ, PSII photochemistry and non-regulated, or basal, energy dissipation (NO) were analysed according to (Kramer et al., 2004a). At 75 mol quanta m-2 s-1 actinic light the fraction of total energy dissipated through NPQ in the ntrc mutant plants was as high as 0.5 under steady state photosynthesis, whereas the corresponding value for wild type plants was close to 0.05 (Figure 20B). Accordingly, RESULTS 83 the effective quantum yield of PSII in the light was more than four times lower in the mutant as compared to wild type plants, while NO did not differ (Figure 20B). Figure 20. PSII activity and NPQ in wild type and ntrc mutant plants. Attached leaves of 5-weeks old plants grown at 120 mol quanta m-2 s-1 light intensity under a photoperiod of 10 h light and 14 h darkness were used. A) Induction and recovery curves applying 75 mol quanta m-2 s-1 intensity actinic light. B) Quantum yields of NPQ, Y(NPQ), PSII photochemistry, Y(II), and non-regulated energy dissipation, Y(NO), based on experiments such as in (A). Each data point is the mean of the values from 4 plants and SDs are indicated by error bars. White and black bars below graphs indicate periods of illumination with actinic light and darkness, respectively. It should be noted that these experiments were performed using plants grown under short-day conditions with 10 h light and 14 h darkness. Since some features of the ntrc mutant phenotype depend on the day length during growth (Perez-Ruiz et al., 2006; Lepisto et al., 2009; Lepisto et al., 2013), we also measured chlorophyll fluorescence of plants grown under long-day (16 h light / 8 h dark) photoperiod. However, ntrc mutant RESULTS 84 plants grown under long photoperiods also displayed exceptionally high and persisting NPQ at low light intensities (Figure 21). Figure 21. NPQ in wild type and ntrc mutant plants grown under long-day conditions. Chlorophyll fluorescence was measured using 3-weeks old plants grown at 120 mol quanta m-2 s-1 light intensity under a photoperiod of 16 h light and 8 h darkness. Prior to measurements plants were kept in the dark for 30 min. Actinic light at 75 mol quanta m-2 s-1 was turned on and saturating pulses were applied every 60 s. After 8 min the actinic light was switched off and measurements were continued for another 10 min. White and black bars below graphs indicate periods of illumination with actinic light and darkness, respectively. The quantum yields of NPQ were calculated and each data point is the mean of the Y(NPQ) from 4 leaves and SDs are presented as error bars. Chlorophyll fluorescence imaging showed that the high-NPQ phenotype is particularly pronounced in the young expanding leaves of ntrc plants and that there is variegation within single leaves (Figure 22). High light acclimation in ntrc plants in relation to the 2-Cys Prx activity NTRC is known to be an important electron donor to chloroplast 2-Cys Prxs (Perez-Ruiz et al., 2006; Pulido et al., 2010) and, since ROS are involved in the regulation of various responses to excess light (Li et al., 2009), we addressed the relevance of this function for the control of NPQ. A severe knock-down mutant with minimal levels of 2-Cys Prxs, the  2cp mutant (Pulido et al., 2010), was examined for its capacity to induce NPQ as a function of light intensity and compared to ntrc and RESULTS 85 wild type plants. In addition, mutant plants devoid of the chloroplast Trx x (Pulido et al., 2010) were included in these experiments, since Trx x has been proposed as an alternative electron donor for 2-Cys Prxs (Collin et al., 2003). Figure 22. Fluorescence imaging. False-colour images representing Fv/Fm in wild type and ntrc mutant plants grown under short day conditions at 120 µmol quanta m-2 s-1 and the respective Y(NPQ) and Y(II) after 3 min of 81 mol quanta m-2 s-1 intensity illumination. Induction-recovery curves were performed using IMAGING-PAM Chlorophyll Fluorescence System (Walz). Fv/Fm values were determined in darkadapted plants. Subsequently, plants were illuminated with 81 µmol quanta m-2 s-1 of actinic light during 6 min and saturating pulses were applied each 20 s. Images of ntrc plants have been enlarged to facilitate viewing. Plants were grown at four different light intensities and fluorescence inductionrecovery curves were recorded using three actinic light intensities for each set of plants. The yields of NPQ for the  2cp and trxx mutants were similar to those of wild type plants (Figure 23), except  2cp plants grown at high light intensities, which had less NPQ under lower actinic light. In contrast, the yields of NPQ for the ntrc mutant were higher under all conditions tested (Figure 23). This indicates that the effect that NTRC exerts on NPQ is not related to reduction of 2-Cys Prx. To compare the photosynthetic performance of these plants, the relative rates of linear PET were analysed by chlorophyll fluorescence using gradually increasing intensities of actinic light up to 2000 mol quanta m-2 s-1 (Figure 24). Wild type plants RESULTS 86 adapted to higher light intensities had higher linear PET rates, but the electron transport for the ntrc mutant remained extremely slow at all intensities examined. The electron transport rates of plants deficient in Trx x and 2-Cys Prx were comparable to those of the wild type (Figure 24A). The yields of NPQ in these experiments were always higher for ntrc plants, particularly at lower light intensities (Figure 24B). In contrast, the  2cp plants grown at high light intensities displayed considerably less NPQ under low actinic light. Figure 23. NPQ in wild type and ntrc,  2cp and trxx mutant plants grown under different irradiances. Chlorophyll fluorescence was measured using attached leaves of 5-weeks old plants grown at 50, 120, 600 and 1000 mol quanta m-2 s-1 under a photoperiod of 10 h light / 14 h darkness. Plants were kept in the dark for 30 min prior to measurements. Three different actinic light intensities of 18, 75 and 278 mol quanta m-2 s-1 were used and saturating pulses were applied every 60 s. Quantum yields of NPQ, Y(NPQ), were calculated from saturating pulse analyses. White and black bars below graphs indicate periods of illumination with actinic light and darkness, respectively. Each data point is the mean of the Y(NPQ) from 4 plants and SDs are indicated by error bars. RESULTS 87 Figure 24. Linear electron transport and NPQ as function of light intensity in wild type and ntrc,  2cp and trxx mutant plants grown under different irradiances. Attached leaves of plants grown at 50, 120, 600 and 1000 mol quanta m-2 s-1 for 5 weeks under a photoperiod of 10 h light and 14 h darkness were used for chlorophyll fluorescence. A) Relative electron transport rates of PSII, ETR(II), were determined during stepwise increasing photosynthetically active radiation (PAR). B) The quantum yields of NPQ, Y(NPQ), were determined during the same curves than (A). Each data point is the mean of the values from 8 plants and SDs are presented as error bars. RESULTS 88 Characterisation of the component of NPQ regulated by NTRC As we have seen, the unusually high NPQ present in the ntrc mutant shows rapid kinetics characteristic for the energy-dependent component, qE, of NPQ (Figure 20). It is rapidly activated, remains constant during illumination and is rapidly reverted in darkness. Therefore, we have studied the factors contributing to the qE process in ntrc mutant plants. The PsbS protein is a subunit of PSII required for induction of the qE component of NPQ (Li et al., 2009), and transgenic plants overexpressing PsbS have higher NPQ (Li et al., 2002; Roach and Krieger-Liszkay, 2012). To search for a possible alteration of the PsbS levels, we examined the expression of PsbS in the ntrc mutant. However, the levels of the PsbS protein were similar in mutant and wild type plants (Figure 25). The presence of the xanthophyll zeaxanthin is another determinant for formation of NPQ, while violaxanthin is inactive in this process (Jahns et al., 2009; Jahns and Holzwarth, 2012). Therefore, we analysed the levels of VDE, the enzyme responsible for the conversion of violaxanthin to zeaxanthin, but found that the quantities of VDE did not differ between ntrc mutant and wild type plants (Figure 25). Similarly, ZE, which catalyses the conversion of zeaxanthin back to violaxanthin, was equally abundant in ntrc mutant and wild type plants (Figure 25). Hence, there is no evidence for changes in gene expression that could contribute to explaining the observed differences in NPQ. Figure 25. Levels of PsbS, VDE and ZE proteins in wild type and ntrc mutant plants. Proteins extracted from leaves of plants grown at 120 mol quanta m-2 s-1 light intensity were analysed by protein gel immunoblot. Numbers above each lane indicate the quantity of chlorophyll in µg. RESULTS 89 Although the amounts of the enzymes VDE and ZE do not differ between plants lacking NTRC and wild type plants, the possibility remains that their activities may be altered. Therefore, the levels of the xanthophyll cycle pigments violaxanthin, antheraxanthin and zeaxanthin were measured in wild type and ntrc mutant plants grown at 120 mol quanta m-2 s-1. Samples were collected at the end of the night and following the onset of light during a normal day/night cycle. As expected, wild type leaves displayed only a minor increase in the content of antheraxanthin and zeaxanthin 20 min after turning on the light, which reverted after 40 min of illumination (Figure 26). Then, as the light was turned off and after an additional 90 min of darkness, the original violaxanthin levels were nearly completely recovered in wild type leaves. In contrast, the ntrc mutant showed ten-fold higher levels of zeaxanthin throughout the 40 min of illumination (Figure 26). This implies that the high NPQ of ntrc plants at normal light intensities corresponds to the qE component and is a direct consequence of excess zeaxanthin and antheraxanthin, despite wild type levels of VDE and ZE. Therefore, NTRC seems to be regulating the activity of some of these enzymes involved in the xanthophyll cycle, either regulating their redox states or their activation through lumen acidification. Figure 26. Xanthophyll cycle pigments in wild type and ntrc mutant plants. Pigments were extracted from leaves of plants grown at 120 mol quanta m-2 s-1 light intensity and the levels of violaxanthin (Vx), antheraxanthin (Ax) and zeaxanthin (Zx) were determined. Leaves were harvested just before the onset of light in the morning (0 min) and after 20 and 40 min of illumination with normal light. Thereafter, plants were kept in the dark for another 90 min and samples were taken again (130 min). Values are the mean (n = 3)  SD of the xanthophyll levels determined in leaves from 3 plants. RESULTS 96 Figure 31. Electrochromic shift in wild type and ntrc plants at different light intensities. Plants grown at 120 µmol quanta m-2 s-1 under short-day photoperiod. Dark relaxation of ECS was measured in leaves after photosynthetic steady-state conditions, 15 min of actinic illumination by red LEDs (650 nm) at different photosynthetically active radiations (PAR). Intensities were from 12 to 940 µmol quanta m-2 s-1. A) The graph shows the recovery phase of the ECS decay, ECSinv, after the light-dark transition which is proportional to the trans-thylakoid proton gradient formed during photosynthesis. B) Zoom in the part of the curve (A) up to 100 mol quanta m-2 s-1 PAR. Each value is the mean  SE of ECSinv in 3 leaves. A relevant target for redox regulation involved in ΔpH formation, is the - subunit of the thylakoid ATP synthase, which is known to mediate light-induced activation of ATP synthesis (Hisabori et al., 2013). A possible deficiency in the reduction of the -subunit would lead to a build-up of the trans-thylakoid proton gradient and enhanced NPQ. In the dark this subunit was found in the oxidised form both in the wild type and the ntrc mutant (Figure 32). After 1 min of illumination the - subunit was fully reduced in the wild type and nearly completely reduced in the ntrc mutant plants (Figure 32). However, a small fraction of this protein remained oxidised in ntrc plants. Therefore, the -subunit of the ATP synthase appears to be reduced by RESULTS 97 NTRC in vivo, although there must be other enzymes more efficient than NTRC in reducing this subunit. Figure 32. The in vivo redox state of the ATP synthase  subunit. Total leaf proteins from wild type and ntrc mutant plants in the dark (0 min) and after 1, 5 and 10 min of illumination at 120 mol quanta m-2 s-1 intensity were extracted in the presence of 10% TCA to preserve the thiol redox state. Protein thiols were alkylated with 10 mM MM(PEG)24 prior to electrophoresis in a 10% polyacrylamide gel under non-reducing conditions and the  subunit of the ATP synthase was detected by immunoblot. red, reduced; ox, oxidised. Isolation and analysis of a ntrc-psbs double mutant Since the PsbS protein is necessary for induction of the qE component of NPQ (Li et al., 2000), a double mutant lacking both NTRC and PsbS presumably would be devoid of qE and, thus, would provide useful insight into the nature of the NPQ of the ntrc mutant. Furthermore, such a double mutant would permit establishing the specific impact of qE on the ntrc mutant phenotype. The Arabidopsis ntrc-psbs double mutant was obtained by manual crossing of the corresponding single mutants and the absence of the two proteins NTRC and PsbS was confirmed by Western blot analysis (Figure 33). RESULTS 98 Figure 33. Isolation of a psbs knockout mutant and ntrc-psbs double knockout mutant. A) The SALK_012208 line contains a T-DNA insertion in the ninth exon of the NtrC gene (At2g41680). The SALK_095156 line contains a T-DNA insertion in the second exon of the PsbS gene (At1g44575). Sequence information was obtained from TAIR (http://www.arabidopsis.org). B) The presence of the T-DNA insertions was tested by PCR using as template genomic DNA isolated from wild type plants, the insertion line SALK_095156 (psbs mutant) and from the cross between the homozygous psbs mutant and the ntrc knockout mutant. Oligonucleotide positions are indicated in (A) and detailed in Tables 1 and 2. C) The absence of NTRC and/or PsbS protein was verified in extracts from leaves by Western blot using antibodies against PsbS and NTRC. Samples corresponding to 2 g chlorophyll were loaded in each lane. When plants were grown at moderate light intensity (120 mol quanta m-2 s-1) for 10 weeks under short-day photoperiod (8 h light / 16 h darkness), the double mutant RESULTS 99 plants were larger and had higher chlorophyll content than the single ntrc mutant (Figure 34). The rosette weight and chlorophyll content of ntrc-psbs mutant plants were twice as high as compared to those of ntrc plants (Figure 34C and D). Figure 34. Growth and chlorophyll content of the ntrc-psbs double mutant. A) Plants grown for 10 weeks at 120 mol quanta m-2 s-1 light intensity under a photoperiod of 8 h light / 16 h darkness. B) The largest rosette leaves from representative plants of wild type, ntrc and the ntrc-psbs mutant. C) Fresh weight of rosettes from 10-weeks old plants. Each data point is the mean of 8 to 12 plants and SDs are presented as error bars. Significantly different values according to Student’s t-test are indicated with different letters (p < 0.01). D) Content of total chlorophylls a and b determined in leaf discs from 10-weeks old plants. Each data point is the mean of 12 discs from different plants. Significantly different values according to Tukey-test (ANOVA) are indicated with different letters (p < 0.01). Chlorophyll fluorescence imaging indicated that the ntrc-psbs mutant has low NPQ similar to the psbs simple mutant and the effective PSII quantum yield is higher in ntrc-psbs than in the ntrc mutant (Figure 35). Furthermore, less difference in the effective PSII quantum yield was observed between leaves of the double mutant than in ntrc (Figure 35). RESULTS 100 Figure 35. Fluorescence imaging of the ntrc-psbs double mutant. False-colour images representing Fv/Fm in wild type and ntrc mutant plants and the respective Y(NPQ) and Y(II) after 3 min of 81 mol quanta m-2 s-1 intensity illumination. Induction-recovery curves were performed using IMAGING-PAM Chlorophyll Fluorescence System (Walz). Fv/Fm values were determined in plants grown under short day conditions at 120 µmol quanta m-2 s-1. Subsequently, plants were illuminated with 81 µmol quanta m-2 s-1 of actinic light during 6 min and saturating pulses were applied each 20 s to obtain PSII quantum yields. Images of ntrc plants have been enlarged to facilitate viewing. Measurements of chlorophyll fluorescence showed that qE is missing in the psbs mutant and the ntrc-psbs double mutant (Figure 36A). Interestingly, the effective PSII quantum yield was higher in plants lacking both NTRC and PsbS than in plants lacking only NTRC (Figure 36B). The higher degree of basal non-regulated energy dissipation in the double mutant (Figure 36C) accounts for the lower effective PSII quantum yield as compared to the wild type and psbs simple mutant (Figure 36B). The linear electron transport rates at growth light and low light intensities were significantly higher in the ntrc-psbs double mutant than in the ntrc mutant (Figure 37). However, the double mutant was hypersensitive to high irradiances similarly to the psbs simple mutant (Figure 37). RESULTS 101 Figure 36. Photosystem II quantum yield of the ntrc-psbs double mutant. Chlorophyll fluorescence was measured with a pulse-amplitude modulation fluorometer using plants grown for 5 weeks at 120 mol quanta m-2 s-1 light intensity under a photoperiod of 10 h light / 14 h darkness. Plants were dark-adapted for 30 min before measurements. Induction-recovery curves were performed with actinic light of 75 mol quanta m-2 s-1 intensity and saturating pulses applied every 60 s. After 6 min of illumination, the actinic light was switched off and measurements were continued for another 6 min. White and black bars indicate periods of illumination with actinic light and darkness, respectively. A) Quantum yields of NPQ, Y(NPQ). B) Quantum yields of PSII photochemistry, Y(II). C) Quantum yields of basal non-regulated energy dissipation, Y(NO). Values are the means of 6 measurements ± SD. RESULTS 102 Figure 37. Electron transport rate of the ntrc-psbs double mutant. Relative linear electron transport rates ETR(II) were determined by chlorophyll fluorescence as a function of actinic light intensity in pre-illuminated leaves of plants grown for 5 weeks at 120 mol quanta m-2 s-1 light intensity under a photoperiod of 10 h light / 14 h darkness. Values are the means of 6 measurements ± SD. 4.1.3. Effect of NTRC deficiency on other components of the photosynthetic electron transport chain Photosystem I Since the lack of NTRC strongly influences the levels of NPQ and PSII activity, we proceeded to investigate the PSI activity. Various biosynthetic processes, such as starch synthesis, are deficient in the ntrc mutant (Michalska et al., 2009; Toivola et al., 2013) and, therefore, a shortage of electron acceptors for PSI might be expected. This, in turn, could lead to activation of NPQ because of a lower ATP consumption and a concomitant proton accumulation in the lumen (Kalituho et al., 2007). NTRC was previously reported to reduce and activate the small subunit of the ADP glucose pyrophosphorylase, a key enzyme in starch synthesis (Michalska et al., 2009). An Arabidopsis mutant lacking the regulatory small subunit of the ADP glucose pyrophosphorylase, aps1, is devoid of leaf starch and grows poorly, except under continuous light conditions (Ventriglia et al., 2008). However, aps1 shows a pattern of NPQ identical to wild type plants (Figure 38) indicating that the role of NTRC in starch synthesis is not relevant for the increased energy dissipation. RESULTS 103 Figure 38. NPQ in the aps1 mutant in comparison to wild type and ntrc plants. Chlorophyll fluorescence was measured using aps1, ntrc and wild type plants grown at 120 mol quanta m-2 s-1 light intensity. Prior to measurements plants were kept in the dark for 30 min. Actinic light at 75 mol quanta m-2 s-1 was turned on and saturating pulses were applied every 60 s. After 8 min the actinic light was switched off and measurements were continued for another 10 min. White and black bars indicate periods of illumination with actinic light and darkness, respectively. Each data point is the mean of the Y(NPQ) from 4 leaves and the error bars represent SDs. Measurements of the PSI activity in wild type and ntrc plants through inductionrecovery curves based on P700 absorbance changes showed that the redox state of P700 during illumination was affected in plants lacking NTRC (Figure 39). Analysis of the respective quantum yields of PSI activity, Y(I), donor side limitations, Y(ND), and acceptor side limitations, Y(NA), revealed the lower effective PSI quantum yield under growth light intensity in the ntrc mutant (Figure 40A). This turned out to be the result of limitations at the donor side of PSI in the electron transport chain during illumination (Figure 40B). In contrast, wild type and ntrc plants were indistinguishable with respect to acceptor side limitations (Figure 40C). Thus, the absence of NTRC leads to a deficiency in PSI activity affecting specifically the supply of electrons to this photosystem in the light. This could be a consequence of the high NPQ observed in the ntrc mutant that decreases the flux through the photosynthetic electron transport chain. However, the demand for electrons from PSI appears not to be altered. RESULTS 104 In order to know if the partial recovery of PSII activity observed in the double mutant ntrc-psbs (Figure 36B) led to a concomitant recovery of PSI, quantum yields of this photosystem were analysed in plants lacking PsbS. Thus, PSI activity was found to be partially restored in ntrc-psbs plants (Figure 41A), probably due to the relief of the donor side limitations observed in the ntrc plants (Figure 41B). A slight increase in acceptor side limitations (Figure 41C) counteracts the complete rescue of PSI activity in the double mutant. Thus, abolishing qE in plants lacking NTRC stimulates linear PET. Figure 39. Redox state of the P700 in wild type and ntrc mutant plants. The redox state of the PSI reaction centre P700 was determined measuring the absorbance of attached leaves at 830 and 875 nm using the DUAL PAM-100. Following the initial determination of maximal oxidation of P700 with a saturating pulse after 10 s of far red light, the actinic light at an intensity of 126 mol quanta m-2 s-1 was turned on and saturating pulses were applied every 20 s. After 5 min actinic light was switched off and measurements were continued for another 5 min. White and black bars below graphs indicate periods of illumination with actinic light and darkness, respectively. RESULTS 105 Figure 40. Activity of PSI in wild type and ntrc mutant plants. The redox state of the PSI reaction centre P700 was monitored through changes in absorbance at 830 nm versus 875 nm. Five-weeks old plants grown at 120 mol quanta m-2 s-1 light intensity were kept in the dark for 30 min prior the induction-recovery curve. Following the initial determination of maximal oxidation of P700 actinic light at an intensity of 126 mol quanta m-2 s-1 was turned on and saturating pulses were applied every 20 s. After 5 min the actinic light was switched off and measurements were continued for another 5 min. The quantum yields of PSI photochemistry, Y(I), donor side limitations Y(ND) and acceptor side limitations Y(NA) are based on saturating pulse analyses (Figure 39). White and black bars below graphs indicate periods of illumination with actinic light and darkness, respectively. Each data point is the mean of values from 6 plants and SDs are shown as error bars. RESULTS 112 4.2.2. Growth of Trx f-deficient mutants To determine the effect of Trxs f on plant growth, mutant lines were grown under short- and long-day conditions. The trxf1f2 double mutant, like the single mutants trxf1 and trxf2, showed wild type phenotype when grown under long-day photoperiod (Figure 46A), as confirmed by the weight of the rosette leaves (Figure 46B) and leaf chlorophyll content (Figure 46C) of mature plants. In contrast, short-day conditions slightly affected the growth of the double mutant (Figure 46D), as shown by the lower weight of rosette leaves (Figure 46E), though the chlorophyll content was unaffected (Figure 46F). Therefore, despite the central function previously attributed to Trxs f in redox regulation of chloroplast metabolism in vitro, these Trxs are dispensable for plant growth under short- and long-day photoperiod. Nevertheless, the retarded growth of the trxf1f2 double mutant under short-day photoperiod confirms the light-dependent participation of f-type Trxs in chloroplast photosynthetic metabolism. RESULTS 113 Figure 46. Characterization of Arabidopsis lines lacking Trxs f. A, D) Wild type and mutant lines were grown under long- and short-day conditions during 3 and 8 weeks, respectively. B, E) The weight of the rosette leaves was determined from 9 plants of long-day and 7 plants of short-day conditions, except for the trxf1f2 double mutant, which was determined from 14 plants. C, F) Chlorophyll content was determined from leaf discs (n = 6) and average values ± SD are represented. Letters indicate significant differences with the Tukey-test (ANOVA) and a confidence interval of 99%. In order to know if the lack of f-type Trxs was compensated for an increase in the amounts of other Trxs, which could explain the mild phenotype of the trxf1f2 mutant, the levels of transcripts of other plastidial Trxs were analysed by qRT-PCR. RESULTS 114 Only minor differences of the level of transcripts of genes encoding NTRC, m- and xtype Trxs were detected (Figure 47), indicating that the absence of f-type Trxs has not a significant effect on the expression of these Trxs. Figure 47. Level of NTRC, m- and x-type TRX gene transcripts in Trx f-deficient mutants. The content of transcripts of the different genes, as indicated, was determined by qRT-PCR. Total RNA was extracted from leaves of wild type, trxf1, trxf2 and trxf1f2 plants grown under long-day conditions during 3 weeks and harvested after 3 h illumination. The pairs of oligonucleotides used for cDNA amplification are indicated in Table 3, except for m-type Trxs that are described in (Barajas-Lopez Jde et al., 2011). Transcript levels were normalized to UBIQUITIN amplification and expressed as percent of the corresponding level in wild type plants. Determinations were performed three times and mean values ± SD are represented. 4.2.3. Effect of Trx f deficiency on photosynthetic performance To explore the function of f-type Trxs in photosynthetic metabolism, different photosynthetic parameters of plants lacking either or both Trx f enzymes were examined. Using chlorophyll fluorescence, the maximum PSII quantum yield was determined as the ratio of variable to maximal fluorescence (Fv/Fm) in dark-adapted leaves. The result revealed that the integrity of PSII was not affected in the single or double trxf1f2 mutants (Table 7). RESULTS 115 Table 7. Maximum PSII quantum yield in Trx f deficient mutants. The integrity of PSII was determined as variable fluorescence (Fv) to maximal fluorescence (Fm), Fv/Fm, in dark adapted leaves of 8-weeks old plants grown under short-day conditions. Each Fv/Fm value is the mean of 12 measurements from 6 different plants ± SD. Genotype wt trxf1 trxf2 trxf1f2 Fv/Fm 0.81 ± 0.01 0.82 ± 0.01 0.81 ± 0.01 0.82 ± 0.01 Effective PSII quantum yield, Y(II), and NPQ quantum yield, Y(NPQ), were determined through induction-recovery curves, during 8 min of illumination with actinic light followed by 10 min of darkness. The light intensity applied was 75 µmol quanta m- 2 s-1, which is slightly lower than the growth light intensity. Since NPQ is a loss of chlorophyll fluorescence that reflects adaptation mechanisms regulating the fraction of absorbed light that reaches the PSII reaction centre, it is typically observed under strong illumination. However, as shown previously (Figure 20A), following the onset of actinic light at low or moderate intensity, a brief peak of NPQ is normally observed in wild type plants due to transient acidification of the thylakoid lumen before activation of photosynthesis (Kalituho et al., 2007). Such an initial peak of NPQ was also found in the trxf2 mutant and a somewhat broader peak in the trxf1 mutant (Figures 48A and 49). In contrast, the double mutant displayed an extensive NPQ, which did not relax completely even after 8 min of illumination at 75 µmol quanta m-2 s-1 (Figures 48A and 49). As a result, the PSII effective quantum yield, Y(II), increased more slowly in the light and remained lower in the trxf1f2 double mutant and also in the trxf1 single mutant, although less pronounced (Figures 48A and 49). RESULTS 116 Figure 48. Non-photochemical quenching and photosystem II activity in wild type and Trx f deficient mutants. Chlorophyll a fluorescence of photosystem II was measured using a pulse-amplitude modulation fluorometer in attached leaves of plants grown for 8 weeks under short-day conditions. After incubation of plants during 30 min in darkness, an induction-recovery curve was performed to determine the quantum yields of non-photochemical quenching, Y(NPQ) (A), and photosystem II, Y(II) (B). During the 8 min induction period, 75 mol quanta m-2 s-1 intensity actinic light was applied. Thereafter, the actinic light was switched off and measurements were continued for another 10 min in the dark. Saturation pulses were applied every 60 s. Each data point is the mean of the values from 6 plants and SDs are indicated by error bars. White and black bars below graphs indicate periods of illumination with actinic light and darkness, respectively. RESULTS 117 Figure 49. Chlorophyll fluorescence of PSII. Typical chlorophyll a fluorescence traces from the experiments represented in Figure 48. The amplitude of the saturation pulses in the light period is proportional to the PSII activity, while the decrease in the fluorescence maxima achieved with the saturation pulses during illumination (Fm’) with respect to the fluorescence maxima in the dark (Fm) correspond to the NPQ. Light and dark periods are indicated by white and black bars below graphs. In agreement with these results, the relative linear photosynthetic electron transport rates of photosystem II (ETR(II)) were considerably lower in the double mutant at all light intensities examined (Figure 50A). In contrast, only a small but significant reduction in photosynthetic electron transport rates was observed in the trxf1 mutant, whereas the trxf2 mutant displayed wild type rates (Figure 50A). Notably, the yields of NPQ were higher in trxf1f2 plants, particularly at low light intensities (Figure 50B). RESULTS 118 Figure 50. Light-dependent linear photosynthetic electron transport and NPQ in wild type and Trx f deficient mutants. A) Relative linear electron transport rates of photosystem II, ETR(II), was measured in pre-illuminated attached leaves of 8-weeks old plants grown at 125 µmol quanta m-2 s-1 under short-day conditions. Chlorophyll fluorescence of photosystem II was determined using a pulse-amplitude modulation fluorometer. B) Quantum yields of NPQ, Y(NPQ), from the corresponding lighttitration curves in (A). Each data point is the mean of measures from 8 plants ± SD. PAR (photosynthetically active radiation). The effect of Trxs f on the control of energy quenching and photosynthetic yield might be through direct interaction with the photosynthetic apparatus or an indirect consequence of limited biosynthesis and demand for ATP, leading to lower luminal pH. The latter should be revealed by a lower rate of consumption of electrons from the photosynthetic electron transport. To test this possibility, we measured the PSI quantum yield, Y(I), based on P700 absorbance changes during an induction-recovery curve, in mutant and wild type plants (Figures 51 and 52). Indeed, the double mutant had lower PSI activity due to prolonged limitations on the acceptor side, Y(NA), after turning on the light (Figure 51). This suggests that the NPQ observed during illumination is a RESULTS 119 consequence of luminal acidification because of limited demand for ATP, since biosynthetic processes are not completely activated. Figure 51. Comparative analysis of photosystem I activity in wild type and Trx f deficient mutants. The redox state of P700 was inferred from changes in the absorbance at 830 nm versus 875 nm and monitored during 10 min. Dark adapted leaves from 8-weeks old plants under short-day photoperiod, were illuminated with a saturation pulse after 10 s of far red light to determine the maximum oxidation of P700. Subsequently, actinic light (126 µmol quanta m-2 s-1) was kept on during 5 min followed by another 5 min of darkness. White light saturation pulses were applied every 20 s. The quantum yields of PSI, Y(I), donor side limitations, Y(ND), and acceptor side limitations, Y(NA), are based on saturating pulse analysis. Data were collected from 6 plants and mean ± SD are represented. Light and dark periods are indicated by white and black bars below graphs. RESULTS 120 Figure 52. Absorbance of the oxidised form of photosystem I. Typical P700 absorbance traces at 830 nm from the experiments represented in Figure 51. The amplitude of the saturation pulses in the light period is proportional to the PSI activity, while the decreased of the maximal absorbance reached during illumination correspond to a deficit in electron acceptors of PSI, which consequently limits the ability of P700 oxidation. In contrast, the increase of the minimal absorbance is caused by a deficit in electron donors to PSI and therefore, a decrease in P700 reduction. Light and dark periods are indicated by white and black bars below graphs. The results inferred from chlorophyll a fluorescence and P700 absorbance measurements indicate retardation of the activation of the biosynthetic processes, which consume reducing equivalents derived from the photosynthetic electron transport. Therefore, we proceeded to study the response of the rate of CO2 fixation (AN) to illumination in plants lacking Trxs f. Despite the similar rates of carbon assimilation (AN) in wild type and mutant plants (Figure 53), the halftime for reaching the maximal CO2 fixation rate after turning on the light was more than twice as long in the double mutant as compared to wild type RESULTS 121 plants (Table 8). The response of the rate of CO2 fixation (AN) to illumination was more retarded in the trxf1 than in the trxf2 single mutant while the trxf1f2 double mutant displayed even slower kinetics (Table 8 and Figure 53). These results suggest the function of f-type Trxs in the rapid response of the chloroplast assimilatory metabolism to light. Figure 53. Net CO2 assimilation rate (AN) upon a dark-light transition in wild type and Trx f deficient plants. The rate of net CO2 assimilation was determined using an infrared gas analyser in short-day grown plants (7-weeks old) previously adapted to darkness. The vertical dashed line represents when actinic light of 75 mol quanta m-2 s- 1 was turned on. The kinetic of one leave of each line is represented (statistics are in Table 8). Table 8. Kinetics of response of net CO2 assimilation rate to light. t1/2 represents the time to achieve 50% of the final rate of CO2 (Figure 53). Data are the means ± SE of 6 measures from 7-weeks old plants grown under short-day conditions. The significant differences between mutant and the wild type plants are indicated by different letters (ANOVA Tukey-test, P<0.01). Genotype wt trxf1 trxf2 trxf1f2 t1/2 (s) 106.0 ± 2.3 a 187.9 ± 7.5 b 134.3 ± 9.5 a 237.2 ± 10.1 c RESULTS 128 Figure 60. Reduction of Trx f1 recombinant protein. Purified Trx f1 protein in their mature form (without transit peptide) was incubated as indicated above each lane during 20 min at room temperature in a 10 mM phosphate buffer pH 7.4. After reactions, proteins were precipitated with 10% TCA, alkylated with 10 mM MM(PEG)24 and electrophoresed under reducing conditions in a 15% polyacrylamide gel (5 µg of Trx f1 were loaded). Gels were stained with Coomassie Brilliant Blue. DTT and NADPH were used as reductants and CuCl2 as oxidant agent. red, reduced; ox, oxidised. Once established the electrophoretic mobility of reduced and oxidised Trx f1, we tested the effect of NTRC on the redox state of Trx f1 in vitro. In the presence of a low concentration of DTT (25 µM), which reduced Trx f1 (Figure 61), incubation with NTRC led to a decrease in the intensity of the upper band while the fast migrating band increased, thus, indicating oxidation of Trx f1 (Figure 61). Incubation in the presence of NTRC and NADPH, but not NTRC alone, led to reduction of Trx f1 in the absence of DTT (Figure 61). Taken together, these results suggest that NTRC and Trx f1 are able to exchange electrons at least in vitro. RESULTS 129 Figure 61. Redox exchange between Trx f1 and NTRC recombinant proteins. Purified proteins in their mature forms (without transit peptides) were incubated as indicated above each lane during 20 min at room temperature in a 10 mM phosphate buffer pH 7.4. After reactions, proteins were precipitated with 10% TCA, alkylated with 10 mM MM(PEG)24 and electrophoresed under reducing conditions in a 15% polyacrylamide gel (5 µg of Trx f1 were loaded). Gels were stained with Coomassie Brilliant Blue. DTT and NADPH were used as reductants and CuCl2 as oxidant agent. red, reduced; ox, oxidised. After establishing the redox interaction between NTRC and Trx f1, we tested the involvement of the two domains of NTRC in this interaction. Neither the NTR nor the Trx domains of NTRC were efficient in Trx f1 oxidation (Figure 62A). Thereafter, two mutant variants of NTRC, C140S and C380S, with Cys at the active site of the NTR and Trx domains of the enzyme, respectively, replaced by Ser, were tested for their efficiency in exchanging electrons with Trx f1. Figure 62B shows that both mutants were unable to affect the redox state of Trx f1, thus indicating that a fully functional NTRC is required for redox interaction with Trx f1. RESULTS 130 Figure 62. Reduction of Trx f1 recombinant protein by each domain of NTRC, NTR and Trx. Purified proteins in their mature forms (without transit peptides) were incubated as indicated above each lane during 20 min at room temperature in a 10 mM phosphate buffer pH 7.4. After reactions, proteins were precipitated with 10% TCA, alkylated with 10 mM MM(PEG)24 and electrophoresed under reducing conditions. A) Five µg of Trx f1 in a 15% polyacrylamide gel. B) Ten µg of NTRC in a 9% polyacrylamide gel. Gels were stained with Coomassie Brilliant Blue. DTT and NADPH were used as reductants and CuCl2 as oxidant agent. red, reduced; ox, oxidised. RESULTS 131 4.3. The ability of NTRC to reduce NADP+ The finding described in the previous section showing that NTRC is able to oxidise reduced Trx f1 suggests that NTRC is able to act as oxidant. This result raises the question of the fate of the electrons in this redox interchange. In this work we have performed preliminary experiments to address this issue. In the presence of a reductant such as DTT, at a concentration of 1 mM, NTRC efficiently catalysed the reduction of NADP+ in a concentration-dependent manner (Figure 63A). Moreover, NTRC was able to catalyse the reduction of NADP+ in the presence of Trx f1 reduced by a low concentration of DTT (0.1 mM) (Figure 63B). DTT alone at this lower concentration was not sufficient to support production of NADPH catalysed by NTRC. These results confirm the redox interaction of NTRC and Trx f1 and uncover the ability of NTRC to catalyse the reverse reaction. Figure 63. Production of NADPH catalysed by NTRC. The reduction of NADP+ to NADPH was monitored through the absorbance increase at 340 nm during 20 min. A) Different concentrations of NTRC in the presence of 1 mM DTT and 1 mM NADP+. B) Five M NTRC with DTT and Trx f1, and 1 mM NADP+. DISCUSSION DISCUSSION 135 5. DISCUSSION Photosynthesis is the process in which light energy is converted into chemical energy, ATP, and reducing power in the forms of reduced Fd and NADPH, which are used for chloroplast biosynthetic processes. Photosynthesis is essential for plants and, thus, it is carefully controlled in response to changes in light intensity. At lower light intensities, plants need an efficient absorption of energy, which is necessary for proper plant growth and development. However, the absorption of excess light leads to the production of ROS, which may be harmful (Li et al., 2009; Ruban et al., 2012). To prevent damage to the photosynthetic apparatus or photoinhibition caused by ROS (Krieger-Liszkay et al., 2008; Tyystjarvi, 2013), plants have evolved different mechanisms to cope with high light intensities (Li et al., 2009; Spetea et al., 2014; Dietz, 2015). These mechanisms operate at different time scales. Thus, at the molecular level, long-term acclimation mechanisms comprise mainly changes in the expression of genes encoding antenna proteins and enzymes with antioxidant functions. In contrast, short-term acclimation mechanisms are mostly based on post-translational protein modifications and may involve dissipation of the excess energy absorbed as heat through the process known as NPQ (Li et al., 2009; Dietz, 2015). In plants, the major component of NPQ is the energy-dependent quenching, qE, which allows a rapid and reversible response to light stress (Muller et al., 2001; Szabo et al., 2005; Baker, 2008). The ability to induce and revert acclimation mechanisms, such as qE, allows plants to survive and grow under the continuous light fluctuations that they face in nature. Trxs have been proposed to be redox sensors mediating plant responses to light stress, since their redox state is intimately linked to the photosynthetic electron transport and can exhibit significant changes depending on light intensity (Li et al., 2009; Nikkanen and Rintamaki, 2014). Beside the Fd-dependent Trxs, the chloroplast contains an NTR with a joint Trx domain at the C-terminus, which is termed NTRC. NTRC shows a high affinity for NADPH and, thus, allows the use of this nucleotide for chloroplast redox regulation (Spinola et al., 2008; Cejudo et al., 2012). Redox regulation based on disulphide-dithiol interchange would be a suitable regulatory mechanism in this context, since it allows the rapid and reversible adaptation of proteins function to environmental stimuli. Based on these premises, the hypothesis that redox regulation might be important in chloroplast acclimation to light was postulated. In this DISCUSSION 136 work, we have addressed this issue by analysing the effect of two important chloroplast redox systems, NTRC and f-type Trxs, in chloroplast performance in response to different light conditions. 5.1. The role of NTRC in plant acclimation to changes in light intensity The Arabidopsis thaliana ntrc knockout mutant displays pale leaves and retarded growth as compared to the wild type (Serrato et al., 2004) and this phenotype is particularly pronounced under short-day conditions (Perez-Ruiz et al., 2006; Lepisto et al., 2009). Moreover, Arabidopsis plants lacking NTRC have been reported to be hypersensitive to several kinds of abiotic stresses, such as high salinity and drought (Serrato et al., 2004), prolonged darkness (Perez-Ruiz et al., 2006) and heat (Chae et al., 2013). Under these conditions the already low growth rate of the ntrc mutant is further reduced in comparison to wild type plants (Serrato et al., 2004; Perez-Ruiz et al., 2006). In this context and given the possible redox control of light acclimation mechanisms (Li et al., 2009), we have investigated the putative role of NTRC in photoprotection. Previous results showing the implication of NTRC in abiotic stress tolerance and peroxide detoxification suggested that NTRC might also participate in adaptation to high light stress. Surprisingly, we found that plants lacking NTRC are not hypersensitive to high light stress. In contrast, the ntrc mutant tolerates light intensities up to 1000 mol quanta m-2 s-1, maintaining the proportion of their rosette size and chlorophyll content to those of wild type plants (Figure 15). This is in agreement with a recent report showing that growth of the ntrc mutant at 600 mol quanta m-2 s-1 was not further compromised (Toivola et al., 2013), and confirms that NTRC is not involved in molecular mechanisms controlling long-term acclimation to high light intensities. NTRC is the most relevant reductant of chloroplast 2-Cys Prxs (Moon et al., 2006; Perez-Ruiz et al., 2006; Alkhalfioui et al., 2007; Puerto-Galan et al., 2013). A double knockout mutant devoid of 2-Cys Prxs A and B was recently reported to be hypersensitive to photoinhibition upon a transfer from low to high irradiances (Awad et al., 2015), suggesting that plants devoid of NTRC might suffer increased photoinhibition. However, ntrc plants grown to adult stage at 120 mol quanta m-2 s-1 DISCUSSION 137 and transferred to 800 mol quanta m-2 s-1 light intensity did not show exacerbated photoinhibition (Figure 16). This indicates that NTRC is not required for short-term acclimation to high light and suggests that the 2-Cys Prx benefits from electron donors other than NTRC at high light intensities. In support of this notion, we found also that NTRC is not required for peroxide detoxification in high light. The hydrogen peroxide content in ntrc leaves after 1 h at 500 mol quanta m-2 s-1 was indistinguishable from that in wild type leaves (Figure 18), while hydrogen peroxide accumulated in plants devoid of 2-Cys Prxs under high light conditions (Awad et al., 2015). However, the absence of NTRC results in higher hydrogen peroxide levels under growth light and in darkness (Figure 18), in agreement with the protective role attributed to NTRC against oxidative stress as the main reductant of 2-Cys Prxs in normal light and especially, in darkness, when NTRC becomes the most relevant redox system using NADPH produced from sugars through the oxidative pentose phosphate pathway (Perez-Ruiz et al., 2006; Pulido et al., 2010). While exploring the light responses of the ntrc mutant a new distinctive feature was uncovered. At light intensities equal to the irradiance applied during growth, or even lower light intensities, these plants present drastically elevated levels of NPQ, which are retained throughout the illumination period. This characteristic is unique to the plants lacking NTRC and is not found in wild type plants or the  2cp and trxx mutants (Figure 23). NPQ is usually observed in wild type plants after a sudden increase in light intensity (Figure 19) that exceeds their capacity for utilization in metabolic processes (Li et al., 2009; Ruban et al., 2012). As expected, the high NPQ in plants lacking NTRC was accompanied by a decrease in PSII yield (Figure 20), which shows that these plants are not able to use the energy properly and this could be a factor contributing to the retarded growth of the ntrc mutant. Indeed, the increase of NPQ in ntrc plants was more pronounced in the youngest leaves (Figure 22), suggesting that down-regulation of NPQ by NTRC, to maximize light energy utilization, could be especially important during early leaf development. Moreover, it could affect the synthesis of chlorophylls since leaves with higher levels of NPQ are the most pale. To understand the influence of NTRC on NPQ, the factors contributing to different elements of dissipation of the light energy have been considered. NPQ DISCUSSION 144 shortage of electron donors for PSI. A further implication of this enhanced dissipation and slow electron transport is that less energy is available for biosynthesis. Hence, the low starch content in ntrc plants (Michalska et al., 2009; Toivola et al., 2013) may also reflect the modest production of photosynthates. Indeed, this view affords an explanation to several aspects of the ntrc mutant phenotype. Plants that are continuously starved for light energy should be undersized and grow slowly (Serrato et al., 2004; Perez-Ruiz et al., 2006; Lepisto et al., 2009). Furthermore, longer days should promote growth of such plants by adding to the quantity of total energy absorbed and this is also true for plants lacking NTRC (Perez-Ruiz et al., 2006; Lepisto et al., 2009; Lepisto et al., 2013). Here we show that more electrons can be forced through the LEF pathway in plants lacking both NTRC and PsbS than in plants lacking NTRC alone. This rescues part of the ntrc phenotype with respect to growth and pigmentation (Figure 34). Figure 64. Model of the effect of NTRC on photosynthetic electron flow. Under steady state photosynthetic conditions at low or moderate light intensity, plants lacking NTRC build up a larger pH gradient. Hence, the ntrc mutant has lower rates of linear electron flow, due to extensive dissipation of light energy as heat through qE, and PSI suffers shortage of electron donors. In addition, the PQs receive more reducing equivalents from the stroma. DISCUSSION 145 In conclusion, NTRC is required for down-regulation of qE-mediated energy dissipation and stimulation of LEF, which is essential for plant growth, particularly during early leaf development. NTRC is not involved in long- or short-term high light tolerance, nor is implied in peroxide detoxification at high light intensities. On the contrary, plants lacking NTRC are over-protected against light even at low and moderate irradiances leading to a state of permanent starvation for light energy. 5.2. The role of f-type Trxs in photosynthetic performance The f-type Trxs have been considered the main reductants responsible for the light-dependent regulation of enzymes involved in carbohydrate metabolism. Extensive biochemical and proteomic work has shown the central role of f-type Trxs in the reductive activation of most of the Calvin-Benson cycle enzymes. Indeed, in vitro studies have shown that Trxs f are the most efficient catalysts for disulphide reduction of several Calvin-Benson enzymes (Collin et al., 2003). The Arabidopsis f-type Trxs are also the most efficient in reducing NADP-MDH in comparison to the m-type Trxs, as well as the x- and y-type Trxs (Collin et al., 2003; Collin et al., 2004) and spinach Trx f is a better reductant for the enzyme α-glucan water dikinase than spinach Trx m (Mikkelsen et al., 2005). However, unless examined in vivo the true physiological function of each type of Trx could not be established and, until now, knockout mutants devoid of Trxs f have not been available. Two f-type Trxs are present in Arabidopsis. Both isoforms are very similar, and their functions might be redundant (Collin et al., 2003; Thormahlen et al., 2013). With the aim of analysing the function of these Trxs in vivo, in this study we have generated a double mutant of Arabidopsis, here termed trxf1f2, lacking both Trx f1 and Trx f2 (Figures 44 and 45). The growth of the single mutants, trxf1 and trxf2, was indistinguishable from that of the wild type plants. This was consistent with previous studies on trxf1 mutants which suggest that the low content of the remaining Trx f2, either alone or in conjunction with other plastidial redox systems, may be sufficient to support the redox regulation that allows plant growth (Thormahlen et al., 2013). However, plant growth was still similar to that of the wild type plants when both Trxs f were absent. Thus, the trxf1f2 double mutant displayed wild type phenotype when grown under long day conditions, as shown for f-type null mutants recently reported (Yoshida et al., 2015), but DISCUSSION 146 showed growth inhibition under short-day conditions (Figure 46). Therefore, despite the central function attributed for Trxs f in chloroplast redox regulation based on biochemical in vitro analyses, the in vivo approach reported here show that these Trxs are dispensable for plant growth at least under the standard long-day conditions. However, it should be noted that short-day conditions have a negative effect on growth of the double mutant indicating that the function of Trxs f becomes more relevant when the light is a limiting factor. Analysis of chlorophyll fluorescence shows that the integrity of PSII remains in the absence of either or both Trxs f (Table 5) and, therefore, the different mutants still have the same maximum capacity to perform photochemistry as wild type plants. However, the PSII effective quantum yield was lower in the trxf1f2 mutant after a darklight transition with a concomitant increase of NPQ that was maintained during a few minutes of illumination (Figures 48 and 49). As mentioned above, an initial increase of NPQ just after a dark-light transition, that relaxes within less than 2 min, is typical in wild type plants and reflects the light-dependent activation of metabolic reactions (Kalituho et al., 2007). In the double mutant, trxf1f2, this initial NPQ was extended during illumination, suggesting retarded activation of the ATP synthase, or subsequent reactions, in plants lacking Trxs f. Nevertheless, there are clear differences between plants lacking NTRC and plants lacking Trxs f. In contrast to the stable NPQ of the ntrc mutant, the energy dissipation observed in trxf1f2 plants starts relaxing already in the light. This suggests that the pH of the lumen begins to recover even during illumination and, therefore, that the carbon assimilation could work properly although its activation might be slower. However, the key to distinguish between the functions of NTRC and Trxs f was provided by measurements of PSI activity. The lower PSI activity in the trxf1f2 mutant proved to be caused by a high deficit of acceptors (Figures 51 and 52). This situates the main function of the Trxs f downstream PSI in the photosynthetic reactions, whereas the main function of NTRC, with a high deficit of PSI electron donors, is situated upstream PSI in the photosynthetic electron transport chain. The saturation of photosynthesis and the concomitant decrease of electron transport rate are intimately related to the capacity for CO2 assimilation and the subsequent metabolic reactions, which determine the availability of electron acceptors DISCUSSION 147 for PSI. Thus, the light-harvesting efficiency and PET are highly sensitive to changes in carbon assimilation. For example, sub-atmospheric levels of CO2, which result in limited regeneration of ADP and NADP+, lead to elevated NPQ and decreased effective PSII quantum yield at low or moderate light intensities (Kramer et al., 2004b; Takizawa et al., 2007). Similarly, inhibition of the Calvin-Benson cycle enzymes in vivo by iodoacetamide leads to higher NPQ and slower linear photosynthetic electron transport (Joliot and Alric, 2013). Notably, the qE component of NPQ ensures safe dissipation of the light energy absorbed and prevents excess excitation of PSII (Szabo et al., 2005; Baker, 2008; Ruban et al., 2012). Such a negative feedback control arisen from hampered CO2 fixation is the most likely explanation for the results concerning the activities of PSII and PSI in the trxf1f2 mutant. Indeed, the activation rate of CO2 fixation was also retarded in the double mutant after a dark-light transition (Table 6 and Figure 53). To address this issue, we proceeded to study the redox state in vivo of two Calvin-Benson cycle enzymes, FBPase and Rubisco activase, well-known in vitro targets of f-type Trxs (Schurmann and Buchanan, 2008). This analysis shows that reduction of both FBPase and Rubisco activase upon illumination was impaired in the trxf1f2 knockout mutant, which presented a slower reduction of both enzymes (Figures 54, 55 and 56). Nevertheless, despite the complete absence of Trx f, both enzymes became partially reduced during illumination. The light-dependent reduction of FBPase and Rubisco activase show very similar patterns in wild type plants on one hand and in Trx f-deficient plants on the other, suggesting common regulatory mechanisms for the two enzymes. This may be significant since both enzymes are involved in the pathway of CO2 fixation in the chloroplast. A relevant question concerning chloroplast redox regulation is the level of redundancy or specificity among the many different types of Trxs of this organelle. In this regard, the redox status in vivo of the Mg chelatase CHLI subunit from pea plants was affected by the simultaneous silencing of the TRX f and TRX m genes, but not by the silencing of the TRX f gene alone (Luo et al., 2012), showing the compensatory effect of Trx m on the regulation of this enzyme of the chlorophyll biosynthesis pathway. Therefore, m-type Trxs could substitute for f-type Trxs in the light-dependent DISCUSSION 148 redox regulation of the Calvin-Benson cycle enzymes. However, comparative analyses in vitro of a large number of the Arabidopsis plastid Trxs (f1, f2, m1, m2, m3, m4, x, y1 and y2) confirmed that only the f-type Trxs are capable of activating FBPase using recombinant enzymes (Collin et al., 2003; Collin et al., 2004). In addition, purified Rubisco activase has been shown to be readily reduced and activated by spinach Trx f, whereas spinach Trx m is unable to reduce this enzyme (Zhang and Portis, 1999). The results obtained in this study showing partial reduction of FBPase and Rubisco activase after a dark-light transition in the trxf1f2 mutant, attribute a specific function to the Trxs f that is not compensated for by other Trxs. In addition, the ability of the double mutant, trxf1f2, to partially reduce FBPase and Rubisco activase, indicates that another redox system is involved in the reduction of these enzymes, which until now have been considered specific targets of f-type Trxs. Therefore, in vitro studies showing the predominant function of Trxs f in redox regulation of the Calvin-Benson enzymes cannot be extrapolated to the physiological situation and show that the double mutant relies on alternative system(s), yet to be discovered, capable of limited reductive activation of these enzymes in response to light. In addition, an interesting observation regarding the redox state of FBPase and Rubisco activase is that these enzymes do not become fully reduced under the growth light intensity used routinely in this work even in wild type plants. This is in agreement with recent reports (Yoshida et al., 2014; Yoshida et al., 2015) showing that Arabidopsis plants display only partial reduction of the FBPase when illuminated at low light intensity. However, we observe complete reduction of the FBPase in wild type plants after 10 min illumination with higher light intensity, 500 mol quanta m-2 s-1 (Figure 54). A remarkable implication of these results is that plants that have been adapted to low light conditions have a significant pool of inactive, oxidised enzymes. A possible advantage of this seemingly wasteful synthesis of excess Calvin-Benson cycle enzymes is that after a sudden increase in photon flux leading to higher rates of photosynthetic electron transport, the ATP and NADPH generated would immediately be utilised for carbon dioxide fixation and, thus, would not pose a problem of light stress. Obviously, a prerequisite for this to occur is that there are sufficiently high amounts of Trxs f to catalyse the reductive activation of these enzymes. Moreover, we have analysed the changes of redox status of FBPase in light-dark transitions. Re- DISCUSSION 149 oxidation of FBPase in response to darkness is again a very rapid process, which is essentially completed in 5 minutes in wild type plants (Figure 26). The fact that reoxidation is faster in the mutants could indicate that, until exhausted, a pool of reduced Trxs f continues to catalyse reduction in the dark. It should be noted that plant chloroplasts contain two types of FBPase, termed FBPase I and II, but only one of them, FBPase I, is redox-regulated (Serrato et al., 2009). The result obtained after illumination with 500 mol quanta m-2 s-1, that shows FBPase fully reduced (Figure 54), indicates that the redox-insensitive form, FBPase II, is present in minor amounts in chloroplasts, in agreement with previous results (Rojas- Gonzalez et al., 2015). Despite the delayed response of carbon assimilation to light of plants devoid of Trxs f, these plants reached carbon assimilation rates similar to those of wild type plants (Figure 53). However, the content of starch was diminished in the trxf1f2 mutant by more than 30% with respect to the wild type (Figure 58). This is in agreement with previous reports showing that overexpression of Trx f, but not Trx m, enhanced starch accumulation and increased the content of sugars in tobacco leaves (Sanz-Barrio et al., 2013) and that Arabidopsis mutants knockout for Trx f1 display impairment of AGPase redox regulation and leaf diurnal starch turnover alterations (Thormahlen et al., 2013). The redox regulation of AGPase by Trx f, could be a direct cause for the low starch content in the trxf1f2 double mutant. In any case, the leaf starch accumulated during the day under a long-day light regime seems sufficient to endure the correspondingly short night and to support wild type growth rates (Figure 46). Only in short days the growth of the Trx f-null mutant was reduced by approximately one third (Figure 46). The genes encoding Trx f1 and Trx f2 isoforms are subject to different regulation in Arabidopsis. The expression of the TRX f1 gene responds to light, whereas the TRX f2 gene is under circadian control (Barajas-Lopez Jde et al., 2011). The differential pattern of expression of these genes might indicate different functions for the respective Trx f isoforms. However, all the photosynthetic parameters analysed in this study, such as photosynthetic electron transport and response to illumination of the rate of CO2 fixation (AN), as well as reduction of Calvin-Benson cycle enzymes and starch content, were more affected in the trxf1 than the trxf2 mutant, which displayed a pattern similar to that of the wild type plants, while the double mutant showed an additive effect. Most DISCUSSION 150 probably, these results reflect the higher content of Trx f1 (Figure 45C) and support a redundant function for both f-type Trxs in Arabidopsis. In summary, the in vivo approach undertaken in this work identifies the impact of Trxs f on photosynthetic performance in Arabidopsis, such as the rapid response of the rate of carbon assimilation and redox status of Calvin-Benson cycle enzymes upon a dark-light transition, and the control of photosynthetic electron transport. The fact that these parameters were impaired in plants devoid of Trxs f indicates that the functions of these Trxs are specific and are not compensated for by other Trxs or chloroplast redox systems. On the other hand, FBPase and Rubisco activase showed a significant level of reduction upon dark-light transition in the trxf1f2 double mutant indicating that additional chloroplast redox systems participate in light-dependent reduction of these Calvin-Benson cycle enzymes. This stands out against the well-established notion, based on biochemical in vitro analyses, of the almost exclusive role attributed to Trxs f in redox regulation of these enzymes. Surprisingly, the growth of plants devoid of Trxs f was indistinguishable from wild type plants when grown under standard conditions with long-day photoperiod, indicating that the function of Trxs f are dispensable for growth. However, the growth inhibition of the trxf1f2 null mutant under short-day conditions supports the idea that these Trxs may be essential for the continuous adaptations to variable irradiance under natural environmental conditions. 5.3. NTRC - Trx f interaction As stated above, the results revealing partial reduction of FBPase and Rubisco activase in the trxf1f2 double mutant indicate that additional redox systems participate in this regulation. Beside the Fd-dependent Trxs, the chloroplast also contains a redox system, NTRC, which depends on NADPH (Serrato et al., 2004). Several recent studies indicate a previously unpredicted relevance of NTRC in chloroplast redox regulation. Therefore, in search for an additional redox system contributing to the redox regulation of the Calvin-Benson enzymes we focused on NTRC. In this study, we present evidence showing that both Trx f and NTRC participate in regulation of photosynthesis, although with different functions. Trxs f are involved in regulation of electron acceptors, which act as sinks for the PET, while NTRC seem to DISCUSSION 151 exert an important effect in the control of the input of electrons from light energy into the PET chain. Interestingly, both FBPase and Rubisco activase, despite being Trx f targets were found to have impaired redox state in plants devoid of NTRC, suggesting a possible interaction between these redox systems. In line with this notion, it has been reported that enzymes such as 2-Cys Prx and AGPase are regulated by both redox systems, NTRC and Trx x (Pulido et al., 2010) or Trx f1 (Michalska et al., 2009; Lepisto et al., 2013; Thormahlen et al., 2013), respectively. The impairment of the redox state of FBPase and Rubisco activase in the ntrc mutant could have different reasons. First, this might be the indirect consequence of the lower electron transport rate resulting in defective regulation of the redox targets. Alternatively, impairment of the redox state of these enzymes in the ntrc mutant might suggest a direct participation of NTRC in the redox regulation of FBPase and Rubisco activase. Though not yet tested, this possibility is unlikely due to that fact that neither of these enzymes was identified as NTRC target in vitro (Ferrández, González and Cejudo, manuscript in preparation). Here, we have addressed an alternative possibility, which is the interaction between NTRC and Trx f. This was addressed by in vitro assays with purified enzymes. The results show that NTRC is able to oxidise reduced Trx f1, but the redox system is operative also in the opposite direction, that is, that NTRC is able to reduce oxidised Trx f1. Evidence in favour of the redox interaction between NTRC and Trx f1, which was based on alkylating assays (Figure 61) was further confirmed by the finding that NTRC is able to catalyse the reduction of NADP+ using electrons from DTT or reduced Trx f1 (Figure 63). In this regard, transgenic lines over-expressing each of the NTRC domains, or mutated versions of NTRC with one of the domains mutated, in the ntrc mutant background were recently reported to partially recover wild type phenotype (Toivola et al., 2013). These results are in agreement with the possible interaction of NTRC with other Trxs. However, our in vitro approach indicates that the conformation of NTRC is essential for a proper redox interaction with Trxs f (Figure 62). These results reflect the complexity of chloroplast redox regulation and suggest that many more mechanisms remain to be discovered. CONCLUSIONS