Plastid Terminal Oxidase (PTOX) Has the Potential to Act As a Safety Valve for Excess Excitation Energy in the Alpine Plant Species Ranunculus Glacialis L

Home , Plastid terminal oxidase, Plastoquinone

bs_bs_banner

Plant, Cell and Environment (2013) 36, 1296–1310 doi: 10.1111/pce.12059

Plastid terminal oxidase (PTOX) has the potential to act as a safety valve for excess excitation energy in the alpine plant species Ranunculus glacialis L.

CONSTANCE LAUREAU1, ROSINE DE PAEPE2, GWENDAL LATOUCHE1, MARIA MORENO-CHACÓN1, GIOVANNI FINAZZI3,4,5,6, MARCEL KUNTZ3,4,5,6, GABRIEL CORNIC1 & PETER STREB1

1Ecologie, Systématique et Evolution, Université Paris-Sud 11, UMR-CNRS 8079, Bâtiment 362, 91405 Orsay cedex, France, 2Institut de Biologie des Plantes, Université Paris-Sud 11, UMR-CNRS 8618, Bâtiment 630, 91405 Orsay cedex, France, 3Unité Mixte Recherche 5168, Laboratoire Physiologie Cellulaire et Végétale, Centre National Recherche Scientiﬁque, F-38054 Grenoble, France, 4Commissariat à l’Energie Atomique et Energies Alternatives, l’Institut de Recherches en Technologies et Sciences pour le Vivant, F-38054 Grenoble, France, 5Université Grenoble 1, F-38041 Grenoble, France and 6Institut National Recherche Agronomique, UMR1200, F-38054 Grenoble, France

ABSTRACT PSII; PFD, photon flux density; PQ, plastoquinone; PSI, photosystem I; PSII, photosystem II; PTOX, plastid terminal Ranunculus glacialis leaves were tested for their plastid ter- oxidase; qL, fraction of open PSII centres; RD, dark respira- minal oxidase (PTOX) content and electron flow to pho- tion, RL, light respiration; ROS, reactive oxygen species; torespiration and to alternative acceptors. In shade-leaves, Rubisco, ribulose 1·5-bisphosphate carboxylase/oxygenase; the PTOX and NAD(P)H dehydrogenase (NDH) content SDS–PAGE, sodium dodecyl sulphate–polyacrylamide gel were markedly lower than in sun-leaves. Carbon assimilation/ electrophoresis. light and Ci response curves were not different in sun- and shade-leaves, but photosynthetic capacity was the highest in INTRODUCTION sun-leaves. Based on calculation of the apparent specificity factor of ribulose 1·5-bisphosphate carboxylase/oxygenase Alpine plants growing at high altitude have to cope regularly (Rubisco), the magnitude of alternative electron flow unre- with extreme climatic events such as very high photon flux lated to carboxylation and oxygenation of Rubisco correlated densities (PFDs) combined with either low or high tempera- to the PTOX content in sun-, shade- and growth chamber- tures (Körner 1999; Laureau, Bligny & Streb 2011; Streb & leaves. Similarly, fluorescence induction kinetics indicated Cornic 2012). Combinations of high PFDs with temperature more complete and more rapid reoxidation of the plastoqui- extremes can lead to high rates of chloroplastic electron none (PQ) pool in sun- than in shade-leaves. Blocking elec- transport, which may induce the formation of reactive tron flow to assimilation, photorespiration and the Mehler oxygen species (ROS) and possible cellular destructions reaction with appropriate inhibitors showed that sun-leaves (Asada 1999; Ort & Baker 2002; Suzuki & Mittler 2006). were able to maintain higher electron flow and PQ oxidation. Numerous mechanisms to balance electron transport with The results suggest that PTOX can act as a safety valve in light absorption and carbon assimilation were demonstrated R. glacialis leaves under conditions where incident photon in several plant species, including the dissipation of excess flux density (PFD) exceeds the growth PFD and under con- absorbed energy as heat (Li et al. 2009) as well as the conditions where the plastoquinone pool is highly reduced. Such sumption of excess electrons by photorespiration (Ort & conditions can occur frequently in alpine climates due to Baker 2002; Streb et al. 2005; Streb & Cornic 2012), by the rapid light and temperature changes. Mehler reaction (Asada 1999; Ort & Baker 2002), by the malate valve (Scheibe & Dietz 2012) and electron cycling Key-words: alternative electron flow; photorespiration; pho- around photosystem I (PSI) to increase the proton gradient tosynthesis; stress tolerance. (Cornic et al. 2000). All these mechanisms can interact to keep the plastoquinone (PQ) pool oxidized. Their impor-

Abbreviations: AN, net carbon assimilation; Ca (Ci), external tance probably depends on the physiological state of the

(internal) CO2 partial pressure; DBMIB, 2,5-dibromo-3- plants and on the plant species. methyl-6-isopropyl-p-benzoquinone; EDTA, ethylenediami- Furthermore, several scavenging systems can keep the netetraacetic acid; ETRC (ETRO), electron flow to the ROS accumulation under control to prevent ROS-induced carboxylation (oxygenation) reaction of Rubisco; Fo (Fm), damage (Noctor & Foyer 1998;Asada 1999).While several of minimum (maximum) fluorescence yield; JT, total electron these protective mechanisms increase in response to low- flux at PSII; NDH, NAD(P)H dehydrogenase; NPQ, non- temperature acclimation (Ensminger, Berninger & Streb photochemical fluorescence quenching; FCO2, quantum 2012), their abundance is variable in different alpine species efficiency of CO2 assimilation; FPSII, quantum efficiency of and they are also widely distributed in plant species incapa- ble to survive alpine climates (Streb & Cornic 2012). Moreo- Correspondence: P. Streb. E-mail: [email protected] ver, the artificial decline of glutathione in two alpine plant 1296 © 2013 John Wiley & Sons Ltd PTOX in alpine plants 1297 species, one with strong (Soldanella alpina) and the other investigated: (1) did the electron flow through PTOX corre- with weak (Ranunculus glacialis) antioxidative protection late with PTOX content and could this activity contribute capacity, does not affect their susceptibility to low significantly to electron consumption? and (2) did PTOX temperature-induced photoinhibition (Laureau et al. 2011). content correlate with the sensibility to photoinhibition? The These findings raise the question whether alpine plants function of PTOX as a safety valve in relation to photores- balance excess excitation energy with an alternative, more piration under high-mountain stress conditions is discussed. specific safety valve, which is quantitatively less important in non-alpine plants. According to previous investigations, all MATERIAL AND METHODS alpine plant species investigated for plastid terminal oxidase (PTOX) contain high amounts of this protein. In addition, the Plant material and growing conditions PTOX protein content increases with the altitude where these plants were collected from (Streb et al. 2005; Laureau et al. R. glacialis (L) plants were collected in the French Alps at 2011).In lowland plants,the amount of PTOX is generally low approximately 2400 m altitude as described by Laureau et al. (Lennon, Prommeenate & Nixon 2003; Streb et al. 2005) and (2011).Whole plants were transferred into pots with soil from their capacity to accept excess photosynthetically generated the growing site in the first year and either further cultivated electrons was estimated to be insignificant (Ort & Baker 2002; at 2400 m altitude in full sunlight (sun-leaves), protected by Trouillard et al. 2012). However, in plant species acclimated to green nylon (high-density polyethylene) to reduce the inci- stress conditions like cold, heat, drought salinity and high dent PFD by 50% (shade-leaves) or transferred to a growing -2 -1 light,the amount of PTOX protein increases strongly,suggest- chamber at low altitude at 6 °C and 100 mmol m s PFD ing a potential involvement in stress tolerance (reviewed by (24 h). Measurements of mature leaves started the second McDonald et al. 2011; Sun & Wen 2011). In salt-stressed Thel- year and were repeated for at least two subsequent years. lungiella, up to 30% of photosynthetic electron flow was esti- Only leaves fully developed under the described conditions mated to be attributed to PTOX (Stepien & Johnson 2009). were used for the experiments. The PTOX protein can transfer electrons from PQ to molecular oxygen, thus forming H2O (Carol & Kuntz 2001). Microclimatic conditions PTOX is involved in chloroplast development. During carotenoid synthesis, desaturase reactions require PQ and are The leaf temperature and the PFD arriving at the leaf level driven by a redox chain in which PTOX is likely to reoxidize were recorded every 30 s during the first week after leaf reduced PQ (Carol & Kuntz 2001; Kuntz 2004). The lack of expansion for shade- and sun-leaves with a Campbell data PTOX leads to the variegated leaf phenotype of the Arabi- logger (CR23X; Campbell Scientific Inc., Utah, UT, USA) as dopsis mutant immutans and the tomato mutant ghost,in described by Laureau et al. (2011). For shade- and sun-leaves, which the bleached sectors originate early during chloroplast temperature and PFD were measured on three independent biogenesis as a consequence of a deficit in carotenoid content leaves. From these measurements, hourly mean leaf tempera- (Carol & Kuntz 2001). In mature leaves, PTOX is no longer ture and PFD during the day (0800–1800 h) and night (1800– needed for carotenoid synthesis, but it may act as a safety 0800 h) were calculated and shown in addition to the highest valve to keep the PQ pool oxidized and to prevent photooxi- and lowest measured values. dative damage under stress conditions (McDonald et al. 2011; Sun & Wen 2011; Streb & Cornic 2012). In Chlamydomonas Gas exchange and fluorescence measurements reinhardtii cells, two different PTOX proteins have been recently described, one is assumed to be involved in carote- Net CO2 uptake by leaves in combination with chlorophyll noid synthesis, while the second is supposed to play a role in fluorescence emission was measured using a LI-6400 (Li-Cor PQ oxidation during photosynthesis (Houille-Vernes et al. Inc., Lincoln, NE, USA) equipped with a leaf chamber fluor- 2011). In contrast, only a single copy gene of PTOX exists in ometer 6400–40 essentially as described by Priault et al. Arabidopsis leaves (Fu, Aluru & Rodermel 2009). However, (2006a). Oxygen evolution was determined with a Hansatech while in some transgenic plants with high PTOX content, oxygen electrode and a LD2 chamber (Hansatech Instru- enhanced PQ oxidation was shown (Joet et al. 2002), other ments Ltd, Kings Lynn, UK) as described by Streb, Feier- experimental evidence did either not confirm increased PQ abend & Bligny (1997). Chlorophyll fluorescence was pool oxidation (Rosso et al. 2006) or showed enhanced super- additionally measured with a PAM2000 (Walz, Effeltrich, oxide formation (Heyno et al. 2009). Germany) and a chlorophyll fluorescence imaging system R. glacialis leaves have one of the highest PTOX contents (Speedzeen, JBeamBio, Marsais, France). measured in higher plants so far. Growth at low altitude at For measurements with the LI-6400, intact plants in pots elevated temperature decreased the PTOX content and were transferred in the morning to the laboratory. After increased the susceptibility to photoinhibition (Streb et al. insertion of attached leaves into the leaf chamber (2 cm2), 2003, 2005). In the present investigation, growth conditions leaves were dark-adapted for at least 30 min to measure dark modulating the PTOX content of R. glacialis leaves were respiration, minimum fluorescence yield (Fo) and maximum tested. Leaves with high and low PTOX content were inves- fluorescence yield (Fm). Afterwards, leaves were acclimated tigated for their alternative electron flow to oxygen and to 300 mmol m-2 s-1 PFD. After stabilization of net carbon their potential PTOX activity. Two main questions were assimilation (AN) and stomatal conductance (approximately © 2013 John Wiley & Sons Ltd, Plant, Cell and Environment, 36, 1296–1310 1298 C. Laureau et al.

−1 30 min), AN/PFD [at 380 mmol mol external CO2 partial was less than 5% compared to values shown in the manu- -2 -1 pressure (Ca) ] and AN/internal CO2 partial pressure (Ci) (at script, corresponding to 9 mmol m s electrons at a JT of 750 mmol m-2 s-1 PFD) response curves in 21% oxygen and 191 mmol m-2 s-1 electrons.

1% oxygen were measured. Acclimation was repeated when From Supporting Information Fig. S1, an intercellular CO2 experimental conditions (gas, PFD) were changed. Gas compensation point (C*) of 47 mmol mol-1 was determined as exchange measurements were recorded after the signal stabi- described in Priault et al. (2006b), based on four independent lized (between 5 and 8 min after changing PFD or Ca). A measurements and which was identical in shade- and sun- saturating light flash, followed by darkness and far-red light leaves. The C* was used to calculate the apparent specific was given after gas exchange measurements to determine FЈ, factor of ribulose 1·5-bisphosphate carboxylase/oxygenase FmЈ and FoЈ.TheFo, Fm, FoЈ,FmЈ and FЈ nomenclature were (Rubisco) according to Laing, Ogren & Hageman (1974), -1 -1 taken from Baker (2008).The whole Li-Cor leaf chamber was using a solubility of 0.801 (l ¥ l ) for CO2 and 0.0298 (l ¥ l ) maintained in a second closed-thermostated chamber to avoid for O2 at 22.5 °C. any external gas exchange. Measurements were done at leaf == temperatures between 22 and 23 °C and at a leaf water pres- SOCapp ci 2* 84 (3) sure deficit of approximately 1 kPa in an atmosphere of 21 or 1% oxygen. The apparent specificity factor of 84 for Rubisco of R. glacialis leaves falls into the range of Rubisco specificity factors as calculated by in vitro measurements of several C3 species Calculations based on gas exchange and (von Caemmerer 2000). Possible errors of the determination chlorophyll fluorescence measurements of the intercellular CO2 compensation point on Sapp ci were estimated by varying C* by approximately 10%.A hypotheti- From chlorophyll fluorescence measurements, Fv/Fm ratios cal lower C* of 42, a value previously measured in Nicotiana of dark-adapted leaves, quantum efficiency of PSII (FPSII), sylvestris (Priault et al. 2006b) increased Sapp ci to 95 and a fraction of open PSII centres (qL) and non-photochemical hypothetical higher C* of 52 declined Sapp ci to 76. For further

ﬂuorescence quenching (NPQ) were calculated according to comparison, a Sapp ci of 70 was applied, which is lower than the

Baker (2008). From AN measurements, quantum efﬁciency of lowest Rubisco speciﬁcity factor reported in C3 plants in vitro

CO2 assimilation (FCO2) was calculated as: (von Caemmerer 2000).

Knowing Sapp ci, the ratio of the carboxylation/oxygenation ΦCO=− A RLeaf − absorption × PFD , 2 NL (1) velocity of Rubisco (Vc/Vo) could be estimated at varying Ci in an atmosphere of 21% oxygen (O). This estimation was where a leaf absorption of 0.8 in sun- and shade-leaves was applied at Ci molar ratios limiting AN as shown at 750 mmol measured as described previously (Streb et al. 2005) and light m-2 s-1 PFD (compare Fig. 3a with Supporting Information respiration (RL) corresponds to respiration in light. RL was Fig. S3B) and consequently also at higher PFDs and assum- estimated by the method of Laisk (1977), as described by ing that variation of Ci from the AN/PFD response curve did Priault et al. (2006b) for shade- and sun-leaves.A representa- not signiﬁcantly and differentially modify the mesophyll con- tive measurement is shown in Supporting Information ductance in sun- and shade-leaves. Fig. S1. A mean RL of 66% of dark respiration (RD) was determined in two independent experiments in sun- and in VVc o== S app ci() CO i ϕ (4) shade-leaves and used for calculating RL in every independent sample. RD values in 1% oxygen were found to be iden- The carboxylation velocity of Rubisco is then given by AN, tical to RD in 21% oxygen. respiration in light and CO2 liberated during photorespira-

FCO2 and FPSII measurements in 1% oxygen were used tion as described by von Caemmerer (2000). to calculate the relationship between both quantum efficien- =++ cies (Supporting Information Fig. S2). A regression line of VARcNL05. V o (5) FPSII = 8.303 FCO + 0.012 was determined for both shade- 2 Substituting Eqn 4 into Eqn 5, the carboxylation velocity is and sun-leaves and used to recalculate electron transport at given by: PSII (JT) in 21% oxygen from FCO2 as described by Streb et al. (2005). VARcNL=+()[(.)]105 −ϕ (6) =××Φ JCOPFDT 2 4 (2) Since the carboxylation reaction of Rubisco consumes four electrons for every CO , the electron transport at PSII Assuming that electron transport to redox reactions like 2 used for the carboxylation reaction of Rubisco could be nitrate and sulphate reduction is not influenced by the calculated. oxygen content, the recalculated electron transport rate (JT) has the advantage that the contribution of these electron- ETRCc= 4 V (7) consuming reactions was eliminated by this calculation. Pos- sible variations of the measured regression line on JT were The electron transport rate at PSII for the oxygenation reac- evaluated by changing the correction factor by 0.024 and the tion of Rubisco could be variable. In case that electrons slope by 0.3 units. The maximum calculated difference of JT liberated in the mitochondria for NADH reduction are used © 2013 John Wiley & Sons Ltd, Plant, Cell and Environment, 36, 1296–1310 PTOX in alpine plants 1299 for hydroxypyruvate reduction in peroxisomes, electron con- Inhibitor treatments sumption at the level of PSII would be four. In case electrons liberated in the mitochondria are used for oxygen reduction Leaves were incubated in small Petri dishes in 10 mL solution in mitochondria and electrons for hydroxypyruvate reduc- in the presence of 3 mm lincomycin, 100 mm glycolaldehyde tion are exported from chloroplasts via the malate valve, or 30 mm 2,5-dibromo-3-methyl-6-isopropyl-p-benzoquinone electron transport at PSII would be six electrons, but requires (DBMIB). Lincomycin-treated leaves were kept for 1 h in simultaneous activity of the malate valve. Therefore, the darkness before fluorescence measurements and light expo- activity of NADP-malate deshydrogenase was estimated sure. Leaves were kept in lincomycin during the whole experi- after acclimation to 1000 mmol m-2 s-1 PFD and found to be ment (Streb, Shang & Feierabend 1999). For treatment in the lower than required for an electron consumption of 6 mmol presence of glycolaldehyde and DBMIB, leaves were vacuum- m-2 s-1 electrons by photorespiration (Table 2). For this infiltrated for 1 min and further incubated for 10 min in dark- -2 -1 ness. Leaves were then removed from the inhibitor solution reason, a consumption of 4 mmol m s electrons for VO was assumed and could be written as: and kept on wet filter paper. The successful uptake of glycolaldehyde and DBMIB was verified by measuring the oxygen

ETROC= 4( V ϕ ) (8) evolution capacity in saturating CO2. No net oxygen evolution was observed after the inhibitor treatments. The experiments The calculated electron ﬂow to the carboxylation reaction of were repeated independently at least four times.

Rubisco (ETRC) and electron flow to the oxygenation reaction of Rubisco (ETRO) was compared to the total recalcu- Immunodetection lated JT at the level of PSII and defined as the alternative electron transport rate at PSII, which might be distributed to Leaves (1 g fresh weight) were frozen in liquid nitrogen at PTOX, to the Mehler reaction and to the malate valve. their growing site or after gas exchange measurements. Fluorescence measurements with the PAM2000 were done Frozen leaves were ground into a fine powder and suspended at 25 °C with detached leaves kept on moistened filter paper in extraction buffer (100 mm Tris-HCl, pH 8.0) containing or enclosed in plastic bags. For measurements of Fo and Fm, 10% sucrose, 0.2% b-mercaptoethanol and 500 mm Pefablock leaves were kept for 30 min in darkness. FoЈ was measured (Roche diagnostics GmbH, Mannheim, Germany). After after the light flash, switching off actinic light and short illu- 20 min stirring at 4 °C and centrifugation at 10 000 g for mination with far-red light. 20 min at 4 °C, the pellet was re-suspended in 200–600 mL extraction buffer containing 1% SDS and 1% Triton-X-100. Estimation of PTOX activity in vivo After further 20 min stirring at 20 °C, the centrifugation was repeated at 20 °C and the supernatant containing membrane PTOX activity in vivo was estimated as described by Trouil- proteins were further used. lard et al. (2012). Pots with R. glacialis plants were dark- Proteins (40–60 mg) were fractionated by denaturing acclimated for 30 min to inactivate Calvin cycle activity. The sodium dodecyl sulphate–polyacrylamide gel electrophoresis leaves were then illuminated with two consecutive light (SDS–PAGE) as described by Laemmli (1970) using 13% -2 -1 flashes (1 s duration of 150 mmol m s PFD green light) by acrylamide (w/v) gels, and transferred onto nitrocellulose varying the time between the two flashed from 5 to 300 s. membranes by electroblotting in Western transfer buffer Chlorophyll fluorescence signals were imaged with a com- (25 mm Tris, 192 mm glycine, 20% v/v methanol). Blotted mercial imaging system (Speedzeen). The area above the nitrocellulose membranes were blocked with 5% w/v non-fat chlorophyll fluorescence induction curve during the second milk powder and incubated with different antibodies. Poly- light exposure was used to estimate the rate of PQ reoxida- clonal rabbit antibody against Arabidopsis thaliana PTOX tion in the dark (Trouillard et al. 2012). This approach is was used as described by Cournac et al. (2000) at a final based on the following assumptions: during the first illumi- dilution of 1:2000. Other antibodies were diluted as: NDH-H nation, all the soluble electron carriers are reduced because polyclonal rabbit antibody to 1:3000 (Rumeau et al. 2005), the Calvin cycle is a major bottleneck for electron flow in FNR polyclonal rabbit antibody to 1:2000, CP43 polyclonal dark-adapted leaves. This promotes an increase of fluores- rabbit antibody to 1:3000, PSAB polyclonal rabbit antibody cence emission to the maximum level. During the dark time to 1:1000. Loading was on the basis of the same membrane between the two consecutive illuminations, oxidation of the protein content (40 mg). Immunodetection was performed electrons acceptors located downstream of PSI and PSII using the ECL Western blotting kit (Amersham Bioscience, occurs due to redox equilibration with the stromal electron Piscataway, NJ, USA) as recommended by the suppliers. sinks and to PTOX activity (Bennoun 1982). This leads to a Digital images of the intensity of the chemiluminescence partial fluorescence decrease.The second illumination results signal were analysed with the ImageJ free software (http:// therefore in a fluorescence rise (Fig. 5a,b), the shape of which rsbweb.nih.gov/ij/). Blots were repeated twice from extrac- is dependent on the concentration of the carriers reoxidized tions of three different leaves. in the dark (Joliot & Joliot 2002). Due to the linear relationship between the fluorescence yield and the rate of PSII Biochemical measurements photoreaction (Delosme, Joliot & Lavorel 1959), the area above the fluorescence induction curve quantifies this Leaves (0.3 g) were collected in the morning. Fresh leaves concentration. were extracted in 50 mm potassium phosphate (pH 7.5) or © 2013 John Wiley & Sons Ltd, Plant, Cell and Environment, 36, 1296–1310 1300 C. Laureau et al. frozen in liquid nitrogen, ground to a powder and suspended (a) in 2% metaphosphoric acid + 2mm ethylenediaminetetraace- Anti-PTOX 37 tic acid (EDTA) for ascorbate measurement as described by kDa Shade Sun Shade Sun Shade Sun GChamb Laureau et al. (2011). Leaves from the phosphate extract were used for deter- (b) mining the total soluble protein content and the activities of catalase, glycolate oxidase and ascorbate peroxidase as Shade-leaves Sun-leaves described by Streb et al. (1997). Relative PTOX content 53.7±18.1 143.5±18.6 Chlorophyll and carotenoid contents were measured in (c) 80% acetone extracts using the absorption coefficients for Shade Sun GChamb chlorophyll from Porra (2002).Totalcarotenoid contents were kDa estimated by absorption at 453 nm subtracting the chlorophyll 44- AnƟ-CP43 a + b content as described by Urbach, Rupp & Sturm (1983). 57- AnƟ-PSAB For NADP-malate deshydrogenase, leaves were illumi- 35- AnƟ-FNR nated for 1 h with 1000 mmol m-2 s-1 PFD white light and subsequently frozen in liquid nitrogen. Frozen leaves were 40- AnƟ-NDH-H ground in liquid nitrogen and suspended in an extraction 37- AnƟ-PTOX buffer containing 25 mm Hepes-KOH pH 7.5, 10 mm MgSO4, 1mm EDTA, 5 mm DTT, 1 mm PMSF, 5 mm polyvinylpyrro- (d) lidone and 0.05% Triton X 100 and measured as described by 250 Streb et al. (2005).

200 RESULTS 150 High PTOX content is not induced by 100

low temperature intensity Relative

Newly developed mature leaves from the growing chamber 50 were tested for their PTOX content and compared to mature 0 leaves collected at the natural growing site at 2400 m altitude. PSAB CP43 FNR NDH-H PTOX As shown in Fig. 1a,c, leaves from the growing chamber had no detectable PTOX content.This result shows that low tem- Figure 1. Immunoblot of PTOX in crude extracts of perature is not sufficient to induce high PTOX contents in Ranunculus glacialis leaves grown in the shade or in full sunlight R. glacialis leaves. (sun) at 2400 m altitude or in a growing chamber at low altitude at 100 mmol m-2 s-1 PFD and 6 °C (GChamb). The position of the proteins was verified by protein markers. (a) The same amount of Induction of high PTOX contents requires high membrane protein was loaded into every slot. The result of one extraction, which was loaded three times into separate slots (high PFD during leaf development altitude plants), is shown. (b) The mean relative intensity of the PTOX signal on the same protein basis of six independent extracts Shading at the natural growing site reduced the mean from shade- and sun-leaves. The statistical difference between absorbed PFD by approximately 50% and the maximum shade- and sun-leaves was less than P = 0.01. (c) Representative -2 -1 absorbed PFD by more than 1000 mmol m s , compared to immunoblot of PTOX, FNR, PSAB, CP43 and NDH-H. All sun-leaves (Table 1). However maximum PFD at the leaf immunodetections were done from the same set of extractions. The level was still as high as 1700 mmol m-2 s-1 in shade-leaves. experiment was repeated independently twice in shade- and three The mean leaf temperature was, in contrast, similar in shade- times in sun-leaves with similar results. (d) Mean values of the and sun-leaves, but leaf temperature extremes were slightly intensity of the immuno signal. The intensity was estimated from a digital image calculating the signal intensity from the same area higher in sun-leaves (Table 1). Only leaves developed under with ImageJ free software. continuous shading or in full sunlight were used for further investigations. nase (NDH) protein were only found in sun-leaves The PTOX content was significantly higher in sun- than in (Fig. 1c,d). Lower PSII contents in shade-leaves were accom- shade-leaves (Fig. 1a,b) showing that high PFDs are required panied by 30% lower maximum rates of JT, measured at to induce high PTOX contents in R. glacialis. PTOX contents −1 -2 -1 2000 mmol mol Ca and at a PFD of 2000 mmol m s ,as were compared to other photosynthetic protein complexes compared to sun-leaves (Table 2). involved in electron flow. Contents of FNR and PSAB (PSI) were similar in shade- and sun-leaves. Contents of CP43 Photosynthetic characteristics of shade- and (PSII) were higher in sun-leaves increasing the ratio of PSI/ sun-leaves PSII in shade- compared to sun-leaves (Fig. 1c,d). The ratio of PSII/PTOX was maintained in shade- and sun-leaves. In R. glacialis leaves developed in the shade had a slightly contrast, significant contents of the NAD(P)H dehydroge- reduced leaf thickness, a proportionally lower chlorophyll © 2013 John Wiley & Sons Ltd, Plant, Cell and Environment, 36, 1296–1310 PTOX in alpine plants 1301

Table 1. Light and temperature at the natural growing site of Ranunculus glacialis at the leaf level during the ﬁrst weak after leaf expansion of shade- and sun-leaves

Day T mean Night T mean Day T max Day T min Night T min Day PFD mean PFD max

Shade-leaves 18.1 °C 9.3 °C 26 °C 6 °C 3.2 °C 649 mmol m-2s-1 1783 mmol m-2s-1 Sun-leaves 19.0 °C 9.1 °C 33 °C 4 °C 2.2 °C 1231 mmol m-2s-1 2862 mmol m-2s-1

Mean, maximum (max) and minimum (min) leaf day (0800 h-1800 h) and night (1800 h-0800 h) temperatures (T) and PFDs during this period are shown.

a + b content and a lower chlorophyll a/b ratio compared to Alternative electron flux and assimilation as a sun-leaves. The chlorophyll/carotenoid ratio remained function of Ci unchanged (Table 2). The total soluble protein content was markedly lower in shade- compared to sun-leaves and, pro- Carbon assimilation in shade- and sun-leaves was measured portionally, the activities of catalase, ascorbate peroxidase as a function of internal CO2 molar ratio (Ci) at 21% and at -2 -1 and total NADP-malate deshydrogenase declined, although 1% oxygen and a PFD of 750 mmol m s (Fig. 3a). AN of most changes were not statistically different. Notably, the shade- and sun-leaves was not different as a function of Ci at activity of glycolate oxidase was not different in shade- either 21% or 1% oxygen.At low Ci, carbon assimilation was compared to sun-leaves and total ascorbate contents strongly stimulated in an atmosphere of 1% oxygen, demon- (reduced + oxidized) were significantly higher in shade- strating photorespiratory activity as expected in C3 plants. leaves (Table 2). The maximum photosynthetic activity and At high Ci, photorespiration was completely suppressed as the maximum total electron transport rate at high PFD and shown by nearly identical AN under 21 and 1% oxygen in high CO2 partial pressure in 21% oxygen was 30% lower in shade- and sun-leaves (Fig. 3a). shade- than in sun-leaves (Table 2). Assuming that photorespiration is not active in 1% oxygen The photosynthetic light response curve at atmospheric but that other oxygen consuming reactions remained unaf- −1 CO2 partial pressure (380 mmol mol Ca) was not signifi- fected, the difference of total electron flux measured in 21% cantly different in shade- and sun-leaves but even slightly oxygen and 1% oxygen (JT21% - JT1%) reflect the total electron lower in the latter. A strong decline of maximum AN and a flow to alternative electron acceptors using oxygen as final -2 -1 −1 light saturation at 500 mmol m s PFD was measured in acceptor. At Ci higher than approximately 400 mmol mol , leaves from the growing chamber (Fig. 2). Furthermore, this alternative electron flow was very similar in shade- values of JT, Ci, NPQ and qL did not differ markedly in and sun-leaves. At the highest applied Ci when photorespira- shade- und sun-leaves, but were altered in leaves from the tion is also suppressed in 21% oxygen, 20-22 mmol m-2 s-1 growing chamber as a function of PFD under atmospheric electrons were consumed by alternative electron flow conditions (Supporting Information Figs S3 & S4). (Fig. 3b) in shade- and sun-leaves and did not correlate to

Table 2. Characteristics of shade- and Sun-leaves Shade-leaves % Difference sun-leaves collected in the morning and their relative difference. Dry weight (DW) per leaf DW/area gm-2 28.9 Ϯ 1.9a 26.8 Ϯ 2.5a -7% area, total soluble protein content, chlorophyll Protein mgm-2 5166 Ϯ 386a 3748 Ϯ 557b -28% (Chl) a + b content and a/b ratio, Chl a + b mgm-2 385 Ϯ 37a 336 Ϯ 20a -13% chlorophyll/carotenoid ratio, catalase (Cat) Chl a / b 3.4 Ϯ 0.08a 3.1 Ϯ 0.1a -9% activity, ascorbate peroxidase (APX) activity, Chl/carotenoid 5.2 Ϯ 0.2a 5.2 Ϯ 0.3a 0% glycolate oxidase (G.O.) activity, total Cat mkat m-2 19.5 Ϯ 2.9a 13.5 Ϯ 1.8a -31% ascorbate content and maximum net carbon APX mkat m-2 54.3 Ϯ 9.0a 36.4 Ϯ 3.6a -33% assimilation (A ) and maximum total electron G.O. mkat m-2 8.7 Ϯ 1.0a 7.9 Ϯ 1.3a -10% N ﬂux at PSII (J ) at 2000 mmol mol−1 C , NADP-MDH initial mkat m-2 3.7 Ϯ 0.3a 4.5 Ϯ 0.7a +22% T a 2000 mmol m-2 s-1 PFD and 21% oxygen and NADP-MDH total mkat m-2 12.9 Ϯ 0.9a 9.8 Ϯ 0.6b -24% dark respiration (R ) NADP-MDH activation % 28.9 Ϯ 3.1a 43.4 Ϯ 4.8b +50% D Total ascorbate mmol m-2 6.7 Ϯ 0.7a 10.8 Ϯ 1.9b +61% -2 -1 a b AN max mmol m s 58.1 Ϯ 0.2 40.4 Ϯ 1.6 -31% -2 -1 a a JT max mmol m s 257 Ϯ 5 182 Ϯ 29 -29% -2 -1 a a RD mmol m s -2.9 Ϯ 0.4 -2.6 Ϯ 0.2 -11%

NADP-malate deshydrogenase activity (MDH initial and total) was measured after 1 h light acclimation to 1000 mmol m-2 s-1 PFD. All experiments were independently repeated for at least three times. Statistical signiﬁcant differences at the P < 0.1 level are indicated by different letters. © 2013 John Wiley & Sons Ltd, Plant, Cell and Environment, 36, 1296–1310 1302 C. Laureau et al.

shade-leaves. For comparison, alternative electron ﬂow was -2 -1 −1 20 calculated at 750 mmol m s PFD and 380 mmol mol Ca

from the AN/Ci and the AN/PFD response curve as described

15 earlier and additionally as the difference of JT measured at 21% O2 and at 1% O2. Mean alternative electron ﬂow was –1

s -2 -1

–2 approximately 20 mmol m s electrons higher in sun- than 10 in shade-leaves irrespective of the calculation used and cor-

mol m relating well with the different PTOX content (Table 3). m

N 5 A A quantitatively lower alternative electron ﬂow to oxygen

as calculated by JT - (ETRC + ETRO) could only be obtained 0 when calculation was repeated with a hypothetical higher intercellular compensation point, which results in a lower

Sapp ci. Applying a Sapp ci of 70 for sun-leaves at 750 mmol -2 -1 0 500 1000 1500 2000 m s PFD, reduced alternative electron flow to 60 mmol m-2 s-1 electrons (values taken from the A /PFD curve) PFD mmol m–2s–1 N compared to 70 mmol m-2 s-1 electrons as shown in Table 3. Figure 2. Light response curves of photosynthetic net carbon In conclusion, even possible significant variations in Sapp ci in assimilation (AN/PFD)ofRanunculus glacialis leaves grown in a sun-leaves only, resulted in higher electron flux to oxygen in growth chamber at low temperature (shaded symbols, n = 3) and of sun- compared to shade-leaves. shade- (black symbols, n = 7) and sun-leaves (white symbols, n = 6) Alternative electron flow might, beside PTOX, also −1 collected at 2400 m altitude at 380 mmol mol Ca and 21% oxygen. occur through the malate valve and to the Mehler reaction. The leaf temperature was maintained between 22 and 23 °C. Mean Therefore, initial NADP-malate-deshydrogenase activity was values and standard errors are shown. their strong difference in PTOX content. A higher difference 40 (a) of alternative electron flow between shade- and sun-leaves −1 may occur at Ci values lower than 400 mmol mol (Fig. 3b) but 30 ) requires estimating the photorespiratory electron flux under –1 s this condition. –2 20 mol m m ( N

Estimation of electron ﬂux to photorespiration A 10 and alternative electron ﬂux

Therefore, the apparent speciﬁcity factor of Rubisco was esti- 0 mated from Supporting Information Fig. S1 and was found to be identical in shade- and sun-leaves giving a value of (b) 175 60 Sapp ci = 84 (Eqn 3). This value allowed calculation of the carboxylation/oxygenation velocity of Rubisco (V /V )at 150 c o 50 varying Ci taken from the AN/PFD response curve (Support- )

–1 125 s –2 ing Information Fig. S3B). The ratio Vc/Vo was then used to 40 1% T J calculate Vc and Vo, using measured AN and RL according to 100 - 21%

mol m 30 T m Eqn 5 and 6 and the corresponding ETRC and to the ETRO of J ( 75 -2 -1 T Rubisco, assuming that 4 mmol m s electrons are con- J 20 50 sumed by either reaction (Fig. 4a,b). The maximum ETRO at -2 -1 -2 -1 2000 mmol m s PFD was around 40 mmol m s in both 25 10 plants and fitted roughly to the potential electron consump- 0 0 tion as calculated by measured glycolate oxydase activity 0 250 500 750 1000 1250 1500 1750 (Table 2), suggesting that every glycolate turnover consumes −1 Ci (mmol mol ) four electrons in the photorespiration pathway. The alternative electron flow was then calculated as the difference Figure 3. (a) Net carbon assimilation (AN) and (b) total electron of total electron transport (JT) (Supporting Information flux (JT) in 21% (circles) and 1% (squares) oxygen and the

Fig. S3A) and ETRC + ETRO. Surprisingly, a very high alter- difference of total electron flux between 21% oxygen and total native electron flow was determined in shade- as well as in electron flux in 1% oxygen (JT21% – JT1%) as a function of internal sun-leaves and maximum alternative electron flow peaked at CO2 partial pressure (Ci) in shade- (black symbols and grey dashed line) and sun- (white symbols and grey dotted line) leaves 1000 mmol m-2 s-1 PFD in shade- and at 1500 mmol m-2 s-1 of Ranunculus glacialis collected at 2400 m altitude. The leaf PFD in sun-leaves (Fig. 4c). temperature was maintained between 22 and 23 °C and the PFD -2 -1 At all PFDs higher than 500 mmol m s , alternative elec- at 750 mmol m-2 s-1. Mean values of at least six independent tron flow was always significantly higher in sun- than in experiments and standard errors are shown. © 2013 John Wiley & Sons Ltd, Plant, Cell and Environment, 36, 1296–1310 PTOX in alpine plants 1303

measured after 1 h light exposure to 1000 mmol m-2 s-1 PFD.

–1 40 The initial NADP-malate-deshydrogenase activity was s

–2 slightly, although not signiﬁcantly, higher in shade- than in 30 sun-leaves and could account for an electron consumption

mol m -2 -1 m

) of approximately 7 mmol m s electrons in sun- and 9 mmol – 20 m-2 s-1 electrons in shade-leaves, suggesting that two elec- (4e O 10 trons are consumed by the enzyme (Table 2). However, the

ETR activation state of NADP-malate-deshydrogenase was sig- 0 niﬁcantly higher in shade- than in sun-leaves (Table 2). Assuming the same apparent speciﬁcity factor of Rubisco 100

–1 and the same inhibition of respiration by light for leaves from s

–2 80 the growth chamber, as those measured in the alpine varie-

60 ties, the calculation was repeated. Leaves from the growth mol m

m chamber showed very low alternative electron ﬂux, which

C 40 was maximal at 500 mmol m-2 s-1 PFD and almost absent at ETR 20 the highest PFD applied. This conﬁrms that estimated alter-

–1 native electron ﬂow correlates well to PTOX protein con-

s 0 –2 tents in R. glacialis leaves (Fig. 4c). 80 mol m m ) – 60 PTOX activity as estimated by ) (4e

O ﬂuorescence imaging 40

+ ETR In order to conﬁrm higher PTOX activity in sun- compared C 20 to shade-leaves, a recently developed alternative approach to estimate PTOX activity was applied (see Material and - (ETR 0 T J

0 500 1000 1500 2000 Methods section). By illuminating dark-adapted leaves with

PFD mmol m–2s–1 inactivated Calvin cycle activity with two consecutive light ﬂashes, the area above the ﬂuorescence trace during the

Figure 4. Electron flow to the oxygenation (A: ETRO) and the second flash up to the Fm asymptote provides an estimate of carboxylation (B: ETRC) reaction of Rubisco and alternative the concentration of oxidized electron acceptors in the pho- electron flow (c) in shade- (black circles) and sun-leaves (white tosynthetic electron flow chain (Joliot & Joliot 2005, Houille- squares) and leaves grown in a growth chamber (shaded triangles) Vernes et al. 2011; Trouillard et al. 2012). A plot of the as a function of PFD in an atmosphere of 21% oxygen and evolution of this area as a function of the dark time between 380 mmol mol−1 CO .TheETR and ETR were calculated using 2 O C the two light pulses (Fig. 5b) shows a biphasic kinetic in both the apparent specificity factor of Rubisco, AN, Ci and RL as determined for shade- and sun-leaves, assuming that four electrons shade- and sun-leaves. While the first phase was identical in are consumed by the oxygenation and the carboxylation reaction. shade- and sun-leaves, the second one was significantly faster Alternative electron flow was then calculated as the difference of in sun-leaves. Following previous indications (Trouillard et al. JT and ETRO + ETRC. For leaves from the growth chamber, the 2012), we ascribe this second phase to the non-photochemical same apparent specificity factor and the same relative RL as for oxidation of the PQ pool. Consequently,we conclude that the plants from the alpine growing site were assumed. Mean values reoxidation of the PQ pool is faster in sun-leaves, likely from seven and six independent experiments for shade- and sun-leaves and three independent experiments for leaves from the because of the enhanced PTOX activity. Fitting the data of growth chamber and the standard errors are shown. Fig 5c indicates that the time constant of the slow oxidation

Table 3. Comparison of alternative electron Sun-leaves Shade-leaves ﬂow to oxygen between sun- and shade-leaves at a PFD of 750 mmol m-2 s-1 and at JT –(ETRC + ETRO) JT21% – JT1% JT –(ETRC + ETRO) JT21% – JT1% −1 380 mmol mol Ca n 6373 a a bc ab AN/PFD 70.1 Ϯ 4.9 70.7 Ϯ 1.3 48.2 Ϯ 6.4 51.4 Ϯ 14.9 n 4556 ac a b bc AN/Ci 61.5 Ϯ 5.1 69.6 Ϯ 4.4 48.2 Ϯ 2.6 47.2 Ϯ 5.4

Alternative electron flow in mmol m-2 s-1 electrons was estimated as total electron flow not used for the carboxylation and oxygenation reaction of Rubisco [JT –(ETRC + ETRO) ] or as the difference of total electron flow measured in an atmosphere of 21% or 1% oxygen

(JT21% – JT1%). Values were either taken from AN/PFD or AN/Ci response curves. Number of independent experiments (n). Mean values and standard errors are shown. Different letters indicate signiﬁcant differences at the 5% level. n, number of independent experiments. © 2013 John Wiley & Sons Ltd, Plant, Cell and Environment, 36, 1296–1310 1304 C. Laureau et al.

NPQ was highest in shade-leaves at higher PFD (Figs 6–8); 1.0 (a) (b) 1.0 (3) the incubation of leaves in glycolaldehyde blocked Rubisco activity but not electron ﬂux to the Mehler reaction 0.8 0.8 and to PTOX. Under these conditions JT was lowest in all samples (Fig. 6). In shade-leaves, J was signiﬁcantly lower 0.6 0.6 T than in sun-leaves and did not respond to increasing PFD.In

0.4 0.4 accordance, qL was lower in shade- than in sun-leaves but NPQ was higher (Figs 7 & 8); and (4) ﬁnally, electron trans- Sun-leaves Shade-leaves Fluorescence (relative units) Fluorescence (relative 0.2 0.2 units) Fluorescence (relative port at the cytochrome b6/f complex was blocked by DBMIB, thus allowing electron ﬂux to PTOX only. In the presence of

0.0 0.0 DBMIB, sun-leaves had higher JT than shade-leaves, which 0.00.20.40.60.8 0.0 0.2 0.4 0.6 0.8 Time (s) Time (s)

Control Plastic (c) 80 80 1.0

0.8 60 60 T T 0.6 J J 40 40

(relative units) (relative 0.4 20 20 0.2 Area above the fluorescence curve Area above 0 0 0.0 0 20 40 60 200 250 300 Glycolaldehyde DBMIB Time between two consequtive flashes (s) 80 80

Figure 5. Fluorescence induction kinetics measured in shade- and sun- intact plants. Plants were dark-acclimated for 30 min prior 60 60

to the experiment and then exposed as described in the Materials T T and Methods section. The intensity of the light used to excite J J 40 40 fluorescence was 150 mmol m-2 s-1 PFD. Examples of fluorescence induction curves in whole plants as a function of the duration of dark incubation (᭝ 5s,᭢15 s, • 30 s, ᭿ 60 s, ◊ 180 s) between two 20 20 consecutive illuminations in sun- (a) and shade- (b) leaves. (c) PTOX activity estimated from the bound area above the 0 0 fluorescence induction curve as a function of the dark period in 0 100 200 300 400 0 100 200 300 400 sun- (white squares) and shade (black circles) leaves. Mean values PFD (mmol m–2s–1) PFD (mmol m–2s–1) from four independent experiments with shade- and sun-leaves. Standard errors are relative to n = 45. Figure 6. Total electron flux (JT) calculated from fluorescence measurements in detached shade- (black) and sun- (white) leaves component is ~25 s in shade- and ~9 s in sun-leaves, confirm- of Ranunculus glacialis collected at 2400 m altitude. The oxygen evolution at saturating CO2 was measured before the experiment ing higher PTOX activity in sun-leaves. to verify leaf intactness. Leaves were then kept 30 min in darkness and shortly illuminated with far-red light. Afterwards dark Fo and Alternative electron flux in detached leaves Fm were determined. Leaves were then illuminated with increasing PFD for 30 min at each PFD, while giving a saturating In order to analyse alternative electron flux to oxygen under flash every 5 min. After the last flash, light was switched off and Ј steady-state conditions, shade- and sun-leaves were detached leaves were illuminated with far-red light to determine Fo .The temperature of the leaf clip, housing the leaf, was maintained at and the fluorescence parameters J , qL and NPQ were meas- T around 25 °C during the whole experiment. Control leaves were ured in four different conditions at four different low- kept on wet filter paper. ‘Plastic’ leaves were enclosed in air-tight intensity PFDs to mitigate strong photoinhibition: (1) control plastic bags before the fluorescence measurements. Leaves treated shade- and sun-leaves kept on moist filter paper had identical with 100 mm glycolaldehyde or 30 mm DBMIB were

JT, NPQ and qL under all PFDs applied (Figs 6–8) and vacuum-infiltrated for 1 min and further left for 10 min in the results were similar to those obtained with attached leaves inhibitor solution and subsequently transferred on wet filter paper. (Supporting Information Figs S3A & S4); (2) enclosing Successful inhibitor uptake was verified after the fluorescence measurement with an oxygen electrode at saturating CO . In all leaves into tightly closed plastic bags prevented gas exchange 2 cases, net oxygen evolution was zero. The FPSII was sufficiently and allowed electron flux to photorespiration and to alterna- high, even in the presence of DBMIB to calculate significant tive electron acceptors. Under these conditions JT and qL values of JT. The mean values of at least four independent were significantly higher in sun- than in shade-leaves, while experiments and the standard errors are shown. © 2013 John Wiley & Sons Ltd, Plant, Cell and Environment, 36, 1296–1310 PTOX in alpine plants 1305

DISCUSSION 1.0 Control Plastic 1.0 PTOX is induced by high light 0.8 0.8 R. glacialis leaves required very high PFDs during growth at ambient temperature with mean daytime PFDs above 0.6 0.6 1200 mmol m-2 s-1 and peak PFDs up to 2800 mmol m-2 s-1 to induce high PTOX protein contents (Fig. 1). Growth at low 0.4 0.4 temperature and low PFD was insufﬁcient to induce a high PTOX content. Reducing the mean growth PFD by 50% with 0.2 0.2 peak PFDs up to 1700 mmol m-2 s-1 markedly reduced the PTOX protein content (Fig. 1). This shows that light and not 0.0 0.0 temperature is the primary factor inducing PTOX in R. gla- 1.0 Glycolaldehyde DBMIB 1.0 cialis leaves. However, it remains to be investigated whether

0.8 0.8 a high mean daytime PFD is required to induce high PTOX contents, or whether this results from short periods of very

0.6 0.6 high PFD exposure. Nevertheless, the results are consistent with numerous reports showing that PTOX is induced under qL qL qL qL 0.4 0.4 various stress conditions, including cold, heat and salinity in combination with light (for review, see McDonald et al. 2011;

0.2 0.2 Sun & Wen 2011; Streb & Cornic 2012). According to the concept presented by Huner, Öquist & Sarhan (1998), acclimation responses are induced under conditions of high 0.0 0.0 0 100 200 300 400 0 100 200 300 400 excitation pressure, suggesting that light and additional PFD (mmol m–2s–1) PFD (mmol m–2s–1) environmental stress interacts.

Figure 7. Photochemical ﬂuorescence quenching (qL) calculated from ﬂuorescence measurements in detached shade- (black) and sun- (white) leaves of Ranunculus glacialis collected at 2400 m altitude. Conditions were exactly the same as described in Fig. 6. Control Plastic

4 4 peaked at 38 mmol m-2 s-1 electrons, while the capacity of shade-leaves was less than half as high (Fig. 6). In addition, 3 3 sun-leaves kept the oxidation state of PSII higher than shade- leaves, as indicated by a much higher qL (Fig. 7).While NPQ 2 2 was negligible under very low PFD, some signiﬁcant NPQ could be measured in sun-leaves at the highest PFD applied (Fig. 8). The successful application of glycolaldehyde and 1 1 DBMIB was conﬁrmed by measuring oxygen evolution at saturating CO2. After the application of both inhibitors, no 0 0 net oxygen exchange was measured. Glycolaldehyde DBMIB 4 4

Photoinhibition of shade- and sun-leaves 3 3 The previously mentioned treatments in the presence or absence of inhibitors or in plastic bags were done during 2 h NPQ NPQ NPQ NPQ 2 2 illumination (30 min for every PFD). At the end of this period, leaves were incubated in darkness and illuminated with far-red light before re-evaluating the Fv/Fm ratio, indi- 1 1 cating photoinhibition. Some photoinhibition was only measured in leaves treated with glycolaldehyde, but Fv/Fm ratios 0 0 were always slightly lower in shade- than in sun-leaves (Sup- 0 100 200 300 400 0 100 200 300 400 –2 –1 –2 –1 porting Information Fig. S5).The maximal photoinhibition of PFD (mmol m s ) PFD (mmol m s ) PSII was tested by exposing leaves in a solution containing Figure 8. Non-photochemical fluorescence quenching (NPQ) lincomycin to prevent D1 protein synthesis for up to 6 h to calculated from fluorescence measurements in detached shade- -2 -1 1000 mmol m s PFD. Under these conditions, photoinhibi- (black) and sun- (white) leaves of Ranunculus glacialis collected at tion of shade-leaves was significantly higher than in sun- 2400 m altitude. Conditions were exactly the same as described in leaves, showing higher photoprotection in the latter (Fig. 9). Fig. 6. © 2013 John Wiley & Sons Ltd, Plant, Cell and Environment, 36, 1296–1310 1306 C. Laureau et al.

2003). Similar to PTOX, the NDH-protein content was much higher in sun- than in shade-leaves (Fig. 1b). A coor- 100 dinated appearance of high PTOX and high NDH-protein contents were previously observed in R. glacialis (Streb et al. 2005) and in studies by Quiles and co-workers, present- 90 ing evidence for PTOX as safety valve (e.g. Quiles 2006; Diaz et al. 2007; Ibanez et al. 2010), but not in PTOX- overexpressing Arabidopsis (Rosso et al. 2006). 80 While AN/PFD and AN/Ci response curves did not differ % of control markedly in shade- and sun-leaves in either 21 or 1% oxygen, 70 AN as a function of PFD was even slightly lower in sun-leaves.

Fv/Fm This observation may be due to the fact that photosynthesis was not saturated in R. glacialis shade- and sun-leaves under

60 the applied experimental conditions. In fact, AN was highest at −1 -2 -1 2000 mmol mol Ca and 2000 mmol m s PFD (Table 2). In contrast, leaves from the growing chamber showed light satu- 0246ration already at 500 mmol m-2 s-1 PFD and marked differences in photochemical and non-photochemical fluorescence Incubation (h) quenching. Figure 9. Photoinhibition of PSII as measured by the Fv/Fm ratio in dark-adapted detached shade- (black) and sun- (white) PTOX as a potential electron acceptor leaves of Ranunculus glacialis collected at 2400 m altitude in the presence of lincomycin. Leaves were incubated for 1 h in darkness In order to demonstrate that different PTOX contents of sun- in small Petri dishes in the presence of 3 mm lincomycin. and shade-leaves corresponded to different rates of alterna- Subsequently, leaves were illuminated shortly with far-red light and tive electron flow unrelated to photorespiration, AN/Ci curves Fv/Fm at the beginning of the light exposure was measured. were analysed under conditions where photorespiration is Afterwards, leaves were illuminated with 1000 mmol m-2 s-1 PFD at 25 °C. After 1, 3 and 6 h, samples were taken, dark-adapted for largely absent, that is, under high Ci (Fig. 3). Assuming that 30 min, shortly illuminated with far-red light and Fv/Fm ratios electron flow to alternative electron acceptors apart from were determined. The results present means of at least four oxygen was similar in an atmosphere of 21 and 1% oxygen, independent experiments with standard errors. the difference of JT21% – JT1% should indicate alternative electron flow to oxygen. Such an alternative electron flow of -2 -1 Characteristics of R. glacialis shade- and approximately 20 mmol m s under non-photorespiratory sun-leaves conditions was taken previously as indicator of PTOX activity in R. glacialis leaves with high PTOX content (Streb In the present investigation, shade-leaves were compared to et al. 2005). While these previous results were confirmed, no sun-leaves. It should be noted that growth PFD for R. gla- major difference of alternative electron flow was measured cialis shade-leaves were untypically high (mean daytime between shade- and sun-leaves irrespective of their different PFD of more than 600 mmol m-2 s-1), which exceeds most PTOX content (Fig. 3b). However, assuming that photores- PFDs used for growth in a greenhouse, even for sun-plants. piration was identical in shade- and sun-leaves, Fig. 3b also However, shade- compared to sun-leaves showed several indicates that alternative electron flow might be much higher typical acclimation responses to low PFD (Walters 2005), under atmospheric CO2 in sun-leaves. A very similar activity concerning leaf thickness, chlorophyll a + b content, chloro- of photorespiration at 2000 mmol m-2 s-1 PFD was indicated phyll a/b ratio and RD. The soluble protein content and most by similar ETRO and a similar activity of glycolate oxidase enzyme activities were lower in shade- than in sun-leaves suggesting that photorespiration was at its maximum in sun- except for glycolate oxidase activity (Table 2). Since one and shade-leaves at the highest PFD applied. reported function of PTOX is its involvement in carotenoid To determine alternative electron flow, ETRC and ETRO synthesis in developing leaves (Carol & Kuntz 2001), it were estimated from the AN/PFD response curves under should be noted that the chlorophyll/carotenoid ratio was atmospheric conditions and compared to total electron flow the same in shade- and sun-leaves and therefore not related at PSII (JT). Results of this calculation suggest significantly to the PTOX content of mature leaves (Table 2). As higher alternative electron flow to oxygen in sun- compared described by Walters (2005), the ratio of PSI/PSII is higher to shade-leaves at PFDs higher than 500 mmol m-2 s-1 and the PSII content lower in shade- than in sun-leaves. This (Fig. 4c), in agreement with their different PTOX contents. was also the case for R. glacialis. Interestingly, the PTOX Moreover, comparing alternative electron flow from AN/PFD content paralleled the PSII content in shade- and sun-leaves, and AN/Ci curves as estimated by either JT - (ETRC + ETRO) suggesting that PTOX protein synthesis is coordinated or JT21% – JT1% obtained under the same experimental condi- with the PSII content in R. glacialis. In contrast, strongly tions, revealed a constantly higher alternative electron flow increased PTOX protein contents were measured in a of around 20 mmol m-2 s-1 electrons in sun- compared to tobacco PSBA deletion mutant (Baena-Gonzalez et al. shade-leaves and very similar results by the two independent © 2013 John Wiley & Sons Ltd, Plant, Cell and Environment, 36, 1296–1310 PTOX in alpine plants 1307 calculations (Table 3). The calculated alternative electron PTOX activity in vivo as estimated by flow remained even higher in sun- compared to shade-leaves light flashes when a low Sapp ci of 70 was assumed for sun-leaves only. The increased PTOX activity in sun- compared to shade- Assuming the same Sapp ci and RL in leaves from the growing chamber compared to values measured in the alpine varie- leaves was confirmed by estimating its activity using a fluo- ties, alternative electron flow was very low (Fig. 4c), further rescence approach (Bennoun 1982). Assuming that the confirming that calculated alternative electron flow corre- Calvin cycle was inactivated by dark acclimation and not lated to the PTOX content. activated by short light flashes, sun-leaves showed a higher However, the magnitude of the calculated alternative elec- oxidation state of the PQ pool after the first light flash. Esti- tron flow was exceptionally high and much higher than elec- mating that six PQ molecules (= 12 electrons) are associated tron consumption attributed to the Mehler reaction in with one PSII (Joliot, Lavergne & Béal 1992), the reoxidation rate constant of shade-leaves was estimated to be 0.048 s-1 several plant species under CO2-limiting conditions (Driever & Baker 2011). A quantitative lower alternative electron (12/25 s), that is, similar to results reported for tomato leaves (Trouillard et al. 2012). The reoxidation rate constant of sun- flow could be obtained, assuming a lower Sapp ci of 70, which leaves was estimated as 1.33 s-1 (12/9 s), which was much was chosen because it is lower than Sc/o values determined in vitro for various C3 plants (von Caemmerer 2000). Lowering higher than in shade-leaves but still markedly lower than in C. reinhardtii (Houille-Vernes et al. 2011). also Ci by 20%, as an approximation of mesophyll conductance-induced decline of CO2 in chloroplasts, reduced the maximum calculated electron flow in sun-leaves from Electron consumption by PTOX under 87 mmol m-2 s-1 electrons to 52 mmol m-2 s-1 electrons, which conditions of suppressed AN and is still higher than can be suggested for the Mehler reaction. photorespiration The malate valve, involved in redox homeostasis of the chloroplast (Scheibe & Dietz 2012), could not contribute signifi- In order to substantiate the different activity of PTOX in sun- cantly to eliminate these electrons in excess. At 1000 mmol and shade-leaves under steady-state conditions, fluorescence m-2 s-1 PFD, the malate valve could consume up to 9 mmol measurements were performed with detached leaves. Leaves m-2 s-1 electrons in shade-leaves and 7 mmol m-2 s-1 electrons were either enclosed in tight plastic bags to block any gas in sun-leaves. Even under full activation, the malate valve exchange but allowing photorespiration at the CO2 compen- could only eliminate up to 26 mmol m-2 s-1 electrons from sation point, or pretreated in the presence of glycolaldehyde chloroplasts of sun- and up to 20 mmol m-2 s-1 electrons from or DBMIB. Glycolaldehyde is an inhibitor of phosphoribu- chloroplasts of shade-leaves (Table 2). If part of this alterna- lokinase, depleting ribulose bisphosphate contents and thus tive electron flow is distributed to the Mehler reaction, blocking AN as well as photorespiration (Wiese, Shi & Heber several arguments suggest that the Mehler reaction would 1998). DBMIB blocks PQ oxidation by the Rieske centre of then be higher in shade- than in sun-leaves: (1) JT, qL and the cytochrome b6/f complex (Trebst 1980; Yan, Kurisu & NPQ were marginally higher in sun- than in shade-leaves Cramer 2006), therefore allowing only electron flow from (Supporting Information Figs S3 & S4), suggesting a similar PSII to PTOX (Joet et al. 2002). The successful application of redox state of the PQ pool and a similar amount of trans- both inhibitors was verified by a net oxygen exchange of zero, ported electrons, which was, however, slightly higher in demonstrating that for every O2 produced by PSII, another sun-leaves; (2) the activation state of NADP-malate- O2 was consumed while electron transport was active. -2 -1 deshydrogenase at 1000 mmol m s PFD was significantly Detached control leaves showed no difference of JT, qL and higher in shade- than in sun-leaves (Table 2), suggesting NPQ between shade- and sun-leaves at all PFDs applied and enhanced reduction of the chloroplast stroma in shade- results were consistent with those of attached leaves in the compared to sun-leaves despite higher AN; and (3) photoin- light response curve. Blocking gas exchange in plastic bags hibition under this condition was higher in shade- than in resulted in a marked decline of JT compared to control leaves sun-leaves (Fig. 9), suggesting that photoinhibition may be but in a higher JT with increasing PFD, a higher qL at high induced by superoxide formation. Superoxide formation at PFD and a lower NPQ in sun- compared to shade-leaves. In PSI can inactivate PSII (Krieger-Liszkay, Kos & Hideg 2011) the absence of photorespiration and carbon assimilation, in and methylviologen-treated R. glacialis leaves were very sus- glycolaldehyde-treated leaves, JT was lowest. However, a sig- ceptible to photoinhibition (Streb et al. 1998). Furthermore, nificant build up of the pH gradient is indicated by a rela- ascorbate, the key antioxidant of the water-water cycle tively high NPQ even at low PFD. Under this condition,

(Asada 1999), is nearly exclusively localized in chloroplasts NPQ was higher but JT and qL were lower in shade- than in of R. glacialis and oxidizes almost completely during the day sun-leaves (Figs 6–8). This is consistent with a higher PTOX at high altitude (Streb et al. 1997). Ascorbate contents were activity in sun- compared to shade-leaves, since the pH gra- significantly higher in shade- compared to sun-leaves and dient remains lower when electrons are consumed by PTOX may indicate a higher capacity for ascorbate oxidation in instead of transfer to O2 in the Mehler reaction. Furthermore, the water-water cycle in the former. All these arguments this result indicates that PTOX was active and not cyclic together with higher alternative electron flow indicate electron flow via NDH. If NDH would transfer electrons elevated electron consumption by PTOX in sun- compared back from PSI to PQ, sun-leaves should have a higher activ- to shade-leaves. ity, due to their much higher NDH content (Fig. 1c,d). If true, © 2013 John Wiley & Sons Ltd, Plant, Cell and Environment, 36, 1296–1310 1308 C. Laureau et al. qL in sun-leaves should be lower and NPQ higher than in Serge Aubert for their help to make experiments in the Alps shade-leaves, which was not the case. Finally, in the presence possible and the alpine laboratory of the Col du Lautaret for of DBMIB, qL and NPQ were lowest of all experimental working facilities and a great atmosphere. We appreciate the conditions but JT was markedly higher in sun- than in shade- helpful suggestions of an unknown reviewer. leaves. Up to 38 mmol m-2 s-1 electrons were transported in sun-leaves. In shade-leaves, the maximum JT remained at REFERENCES 14 mmol m-2 s-1 electrons. Furthermore, qL indicated a higher PSII oxidation state in sun-leaves and, at the highest PFD Asada K. (1999) The water-water cycle in chloroplasts: scavenging of active applied, NPQ was double as high in sun- compared to shade- oxygens and dissipation of excess photons. Annual Review of Plant Physi- ology & Plant Molecular Biology 50, 601–639. leaves. This is consistent with a higher JT to PTOX in sun- Baena-Gonzalez E., Allahverdiyeva Y., Svab Z., Maliga P., Josse E.-M., Kuntz leaves. It should be noted that under the applied conditions M., Mäenpää P. & Aro E.-M. (2003) Deletion of the tobacco plastid psbA no significant photoinhibition was observed in sun- and gene triggers an upregulation of the thylakoid-associated NAD(P)H dehy- shade-leaves except in the presence of glycolaldehyde. Pho- drogenase complex and the plastid terminal oxidase (PTOX). The Plant Journal 35, 704–716. toinhibition was, however, always higher in shade- than in Baker N.R. (2008) Chlorophyll fluorescence: a probe of photosynthesis in vivo. sun-leaves (Supporting Information Fig. S5) as well as in the Annual Review of Plant Biology 59, 89–113. presence of lincomycin (Fig. 9) and can be explained by Bennoun P. (1982) Evidence for a respiratory chain in the chloroplast. Pro- ceedings of the National Academy of Sciences of the United States of America enhanced superoxide formation at PSI in shade-leaves. 79, 4352–4356. von Caemmerer S. (2000) Biochemical Models of Leaf Photosynthesis. CSIRO Conclusions and implications for stress Publishing, Victoria, Australia, pp. 1–72. tolerance at high altitude and the contribution Carol P. & Kuntz M. (2001) A plastid terminal oxidase comes to light: implications for carotenoid biosynthesis and chlororespiration. Trends in Plant of PTOX Science 6, 31–36. Cornic G., Bukhov N.G., Wiese C., Bligny R. & Heber U. (2000) Flexible The induction of high PTOX content and associated high coupling between light-dependent electron and vectorial proton transport in PTOX activity by growth at high PFDs and at high altitude illuminated leaves of C3 plants. Role of photosystem I-dependent proton was demonstrated in this manuscript and distinguish R. gla- pumping. Planta 210, 468–477. Cournac L., Redding K., Ravenel J., Rumeau D., Josse E.M., Kuntz M. & cialis from experiments with Arabidopsis under non-stress Peltier G. (2000) Electron flow between photosystem II and oxygen in conditions. In Arabidopsis mutants, PTOX activity as safety chloroplasts of photosystem I-deficient algae is mediated by a quinol oxidase valve was not detectable or JT was even higher in double or involved in chlororespiration. Journal of Biological Chemistry 275, 17256– triple mutants lacking PTOX and cyclic electron flow (Rosso 17262. Delosme R., Joliot P. & Lavorel J. (1959) Sur la complémentarité de la fluo- et al.2006;Okegawa,Kobayashi & Shikanai 2010).This lack of rescence et de 1’émission d’oxygène pendant la période d’induction de la PTOX-safety-valve-function may be simply related to low photosynthèse. C.R. Academic Science (Paris) 249, 1409–1411. PTOX contents of ArabidopsisWT with a low catalytic capac- Diaz M., De Haro V., Munoz R. & Quiles M.J. (2007) Chlororespiration is ity (Ort & Baker 2002), while comparative PTOX quantifica- involved in the adaptation of Brassica plants to heat and high light intensity. Plant, Cell & Environment 30, 1578–1585. tion showed a much higher PTOX content in R. glacialis than Driever S.M. & Baker N.R. (2011) The water-water cycle in leaves is not a in other alpine species and even in PTOX-overexpressing major alternative electron sink for dissipation of excess excitation energy tomato (Streb et al.2005).In contrast,recent experiments with when CO2 assimilation is restricted. Plant, Cell & Environment 34, 837– 846. cold-acclimated Arabidopsis plants suggest higher electron Ensminger I., Berninger F. & Streb P. (2012) Response of photosynthesis to sink capacity mediated by PTOX of cold-acclimated plants low temperature. In Terrestrial Photosynthesis in a Changing Environment (Ivanov et al.2012).However,several results of Okegawa et al. (eds J. Flexas, F. Loreto & H. Medrano), pp. 272–289. Cambridge University (2010) cannot be explained by low PTOX capacity but Press, Cambridge, UK. Fu A., Aluru M. & Rodermel S.R. (2009) Conserved active site sequences in depends rather on complex, yet unidentified, regulation of Arabidopsis plastid terminal oxidase (PTOX). The Journal of Biological electron transport in the presence and absence of PTOX and Chemistry 284, 22625–22632. cyclic electron flow.The reported electron fluxes in this study, Heyno E., Gross C.M., Laureau C., Culcasi M., Pietri S. & Krieger-Liszkay A. (2009) Plastid alternative oxidase (PTOX) promotes oxidative stress when allow separation of electron flow to carboxylation, photores- overexpressed in tobacco. The Journal of Biological Chemistry 284, 31174– piration and alternative electron flux, which was shown to be 31180. mainly related to PTOX activity. Considering very high and Houille-Vernes L., Rappaport F., Wollmann F.-A.,Alric J. & Johnson X. (2011) rapid temperature and PFD changes in the alpine environ- Plastid terminal oxidase 2 (PTOX2) is the major oxidase involved in chlororespiration in Chlamydomonas. Proceedings of the National Academy of ment (Streb & Cornic 2012), PTOX may therefore be Sciences of the United States of America 108, 20820–20825. regarded as a safety valve enabling the plant to maintain high Huner N.P.A.,Öquist G. & Sarhan F. (1998) Energy balance and acclimation to electron transport rates and to avoid photoinhibition when light and cold. Trends in Plant Science 3, 224–230. other protective systems reducing electron transport are less Ibanez H., Ballester A., Munoz R. & Quiles M.J. (2010) Chlororespiration and tolerance to drought, heat and high illumination. Journal of Plant Physiology active. The function of the high NDH content in sun-leaves 167, 732–738. under these conditions remains to be determined but might Ivanov A.G., Rosso D., Savitch L.V., Stachula P., Rosembert M., Öquist G., contribute to maintain PSI electron flow. Hurry V. & Hüner N.P.A. (2012) Implications of alternative electron sinks in increased resistance of PSII and PSI photochemistry to high light stress in cold acclimated Arabidopsis thaliana. Photosynthesis Research 113, 191–206. ACKNOWLEDGMENTS Joet T., Genty B., Josse E.-M., Kuntz M., Cornac L. & Peltier G. (2002) Involve- ment of a plastid terminal oxidase in plastoquinone oxidation as evidenced We thank Dominique Rumeau and Anja Krieger-Liszkay for by expression of the Arabidopsis thaliana enzyme in tobacco. The Journal of antibodies against NDH-H, and FNR, Richard Bligny and Biological Chemistry 277, 31623–31630. © 2013 John Wiley & Sons Ltd, Plant, Cell and Environment, 36, 1296–1310 PTOX in alpine plants 1309

Joliot P. & Joliot A. (2002) Cyclic electron transfer in plant leaf. Proceedings of Streb P. & Cornic G. (2012) Photosynthesis and antioxidative protection in the National Academy of Sciences of the United States ofAmerica 99, 10209– alpine herbs. In Plants in Alpine Regions: Cell Physiology of Adaptation and 10214. Survival Strategies (ed. C. Lütz), pp. 75–97. Springer Wien, NewYork. Joliot P. & Joliot A. (2005) Quantification of cyclic and linear flows in plants. Streb P., Feierabend J. & Bligny R. (1997) Resistance to photoinhibition of Proceedings of the National Academy of Sciences of the United States of photosystem II and catalase and antioxidative protection in high mountain America 102, 4913–4918. plants. Plant, Cell & Environment 20, 1030–1040. Joliot P., Lavergne J. & Béal D. (1992) Plastoquinone compartmentation in Streb P., Shang W., Feierabend J. & Bligny R. (1998) Divergent strategies of chloroplasts. I. Evidence for domains with different rates of photo- photoprotection in high-mountain plants. Planta 207, 313–324. reduction. Biochimica et Biophysica Acta 1101, 1–12. Streb P., Shang W. & Feierabend J. (1999) Resistance of cold-hardened winter Körner C. (1999) Alpine Plant Life. Functional Plant Ecology of High Moun- rye leaves (Secale cereale L.) to photo-oxidative stress. Plant, Cell & Envi- tain Ecosystems. Springer Verlag, Berlin, Heidelberg. ronment 22, 1211–1223. Krieger-Liszkay A., Kos P.B. & Hideg E. (2011) Superoxide anion radicals Streb P., Aubert S., Gout E. & Bligny R. (2003) Reversibility of cold- and generated by methylviologen in photosystem I damage photosystem II. light-stress tolerance and accompanying changes of metabolite and antioxi- Physiologia Plantarum 142, 17–25. dant levels in the two high mountain plant species Soldanella alpina and Kuntz M. (2004) Plastid terminal oxidase and its biological significance. Planta Ranunculus glacialis. Journal of Experimental Botany 54, 405–418. 218, 896–899. Streb P., Josse E.-M., Gallouët E., Baptist F., Kuntz M. & Cornic G. (2005) Laemmli U.K. (1970) Cleavage of structural protein during the assembly of the Evidence for alternative electron sinks to photosynthetic carbon assimila- head of bacteriophage T4. Nature 227, 680–685. tion in the high mountain plant species Ranunculus glacialis. Plant, Cell & Laing W.A., Ogren W.L. & Hageman R.H. (1974) Regulation of soybean net Environment 28, 1123–1135.

photosynthetic CO2 fixation by the interaction of CO2,O2, and ribulose-1,5- Sun X. & Wen T. (2011) Physiological roles of plastid terminal oxidase in plant diphosphate carboxylase. Plant Physiology 54, 678–685. stress responses. Journal of Bioscience 36, 1–6. Laisk A.K. (1977) Kinetics of Photosynthesis and Photorespiration in C3 Suzuki N. & Mittler R. (2006) Reactive oxygen species and temperature Plants. Nauka, Moscow. stresses: a delicate balance between signaling and destruction. Physiologia Laureau C., Bligny R. & Streb P. (2011) The significance of glutathione for Plantarum 126, 45–51. photoprotection at contrasting temperatures in the alpine plant species Trebst A. (1980) Inhibitors in electron flow: tools for the functional and Soldanella alpina and Ranunculus glacialis. Physiologia Plantarum 143, structural localization of carriers and energy conservation sites. Methods in 246–260. Enzymology 69, 675–715. Lennon A.M., Prommeenate P. & Nixon P.J. (2003) Location, expression and Trouillard M., Shahbazi M., Moyet L., Rappaport F., Joliot J., Kuntz M. & orientation of the putative chlororespiratory enzymes, Ndh and IMMU- Finazzi G. (2012) Kinetic properties and physiological role of the plastoqui- TANS, in higher-plant plastids. Planta 218, 254–260. none terminal oxidase (PTOX) in a vascular plant. Biochimica et Biophysica Li Z., Wakao S., Fischer B.B. & Niyogi K.K. (2009) Sensing and responding to Acta 1817, 2140–2148. excess light. Annual Review of Plant Biology 60, 239–260. Urbach W., Rupp W. & Sturm H. (1983) Praktikum Zur Stoffwechselphysiolo- McDonald A.E., Ivanov A.G., Bode R., Maxwell D.P., Rodermel S.R. & Hüner gie Der Pflanzen. Thieme, Stuttgart, New York. N.P.A. (2011) Flexibility in photosynthetic electron transport: the physi- Walters R.G. (2005) Towards an understanding of photosynthetic acclimation. ological role of plastoquinol terminal oxidase (PTOX). Biochimica et Journal of Experimental Botany 56, 435–447. Biophysica Acta 1807, 954–967. Wiese C., Shi L.B. & Heber U. (1998) Oxygen reduction in the Mehler reaction Noctor G. & Foyer C.H. (1998) Ascorbate and glutathione: keeping active is insufficient to protect photosystem I and II of leaves against photoinacti- oxygen under control. Annual Review of Plant Physiology & Plant Molecu- vation. Physiologia Plantarum 102, 437–446. lar Biology 49, 249–279. Yan J., Kurisu G. & Cramer W.A. (2006) Intraprotein transfer of the quinone Okegawa Y., Kobayashi Y. & Shikanai T. (2010) Physiological links amoung analogue inhibitor 2,5-dibromo-3-methyl-6-isopropyl-p-benoquinone in the alternative electron transport pathways that reduce and oxidize plastoqui- cytochrome b6f complex. Proceedings of the National Academy of Sciences none in Arabidopsis. The Plant Journal 63, 458–468. of the United States of America 103, 69–74.

Ort D.R. & Baker N.R. (2002) A photoprotective role of O2 as an alternative electron sink in photosynthesis? Current Opinion in Plant Biology 5, 193– Received 24 May 2012; accepted for publication 18 December 2012 198. Porra R.J. (2002) The chequered history of the development and use of simul- SUPPORTING INFORMATION taneous equations for the accurate determination chlorophylls a and b. Photosynthesis Research 73, 149–156. Additional Supporting Information may be found in the Priault P., Fresneau C., Noctor G., De Paepe R., Cornic G. & Streb P. (2006a) online version of this article: The mitochondrial CMSII mutation of Nicotiana sylvestris impairs adjuste- ment of photosynthetic carbon assimilation to higher growth irradiance. Figure S1. Estimation of light respiration (RL) according to Journal of Experimental Botany 57, 2075–2085. Laisk (1977). A /C response curves at low C partial pressure Priault P., Tcherkez G., Cornic G., De Paepe R., Naik R., Ghashghaie J. & Streb N i i -2 -1 P. (2006b) The lack of mitochondrial complex I in a CMSII mutant of were measured at three different PFDs: 100 mmol m s Nicotiana sylvestris increases photorespiration through an increased internal (black), 300 mmol m-2 s-1 (grey) and 500 mmol m-2 s-1 (white). resistance to CO2 diffusion. Journal of Experimental Botany 57, 3195–3207. The intersection of the regression line of the three response Quiles M.J. (2006) Stimulation of chlororespiration by heat and high light intensity in oat plants. Plant, Cell & Environment 29, 1463–1470. curves is taken as RL. The experiment was repeated with Rosso D., Ivanov A.G., Fu A., et al. (2006) IMMUTANS does not act as a sun- and shade-leaves and for all measurements a mean RL of

stress-induced safety valve in the protection of the photosynthetic apparatus 66% of RD was determined. A representative estimation of of Arabidopsis during steady-state photosynthesis. Plant Physiology 142, sun-leaves is shown. 574–585. Rumeau D., Bécuwe-Linka N., Beyly A., Louwagie M., Garin J. & Peltier G. Figure S2. Relationship between quantum yield of CO2 (2005) New subunits NDH-M, -N, and -O, encoded by nuclear genes, are assimilation (FCO2) and quantum yield of PSII (FPSII)inan essential for plastid Ndh complex functioning in higher plants. The Plant Cell atmosphere of 1% oxygen in shade- (black symbols) and sun- 17, 219–232. Scheibe R. & Dietz K.J. (2012) Reduction-oxidation network for ﬂexible (white symbols) leaves of R. glacialis collected at 2400 m alti- adjustment of cellular metabolism in photoautotrophic cells. Plant, Cell & tude. Data from light and Ci response curves were taken.The 2 Environment 35, 202–216. regression line was: FPSII = 8.303 FCO2 + 0.012, r = 1.00. Stepien P. & Johnson G.N. (2009) Contrasting responses of photosynthesis to Figure S3. Light response curves of JT (A) and Ci (B) of salt stress in the glycophyte Arabidopsis and the halophyte Thellungiella: role of the plastid terminal oxidase as and alternative electron sink. Plant R. glacialis shade- (black symbols) and sun-leaves (white Physiology 149, 1154–1165. symbols) collected at 2400 m altitude and of leaves from the

© 2013 John Wiley & Sons Ltd, Plant, Cell and Environment, 36, 1296–1310 1310 C. Laureau et al. growing chamber (grey symbols) measured in an atmosphere of 21% oxygen and at a leaf temperature of 22–23 °C. The −1 of 21% oxygen and at a leaf temperature of 22–23 °C. The atmospheric CO2 partial pressure was 380 mmol mol . Mean −1 atmospheric CO2 partial pressure was 380 mmol mol . Mean values of three independent measurements (leaves from the values of three independent measurements (leaves from the growing chamber) or of at least 6 independent experiments growing chamber) or of at least 6 independent experiments leaves collected in the Alps) with standard errors are shown. (leaves collected in the Alps) with standard errors are shown. Figure S5. Fv/Fm ratios of dark-adapted detached shade- Figure S4. Light response curves of NPQ (A) and qL (B) of (black) and sun- (white) leaves of R. glacialis collected at R. glacialis shade- (black symbols) and sun-leaves (white 2400 m altitude after the measurements described in Fig. 6. symbols) collected at 2400 m altitude and of leaves from the After shutting off actinic illumination, leaves were accli- growing chamber (grey symbols) measured in an atmosphere mated for 30 min in darkness.