From Endoplasmic Reticulum to Mitochondria: Absenceof the Arabidopsis ATP Antiporter Endoplasmic ReticulumAdenylate Transporter1 Perturbs PhotorespirationW

ChristianeHoffmann,aBartolomePlocharski,a IlkaHaferkamp,aMichaelaLeroch,aRalphEwald,bHermannBauwe,b

Jan Riemer,c Johannes M. Herrmann,d and H. Ekkehard Neuhausa,1

a Department of Plant Physiology, University of Kaiserslautern, D-67663 Kaiserslautern, GermanybDepartment of Plant Physiology, University of Rostock, D-18059 Rostock, GermanycDepartment of Cell Biochemistry, University of Kaiserslautern, D-67663 Kaiserslautern, GermanydDepartment of Cell Biology, University of Kaiserslautern, D-67663 Kaiserslautern, Germany

The carrier Endoplasmic Reticulum Adenylate Transporter1 (ER-ANT1) resides in the endoplasmic reticulum (ER) membraneand acts as an ATP/ADP antiporter. Mutant plants lacking ER-ANT1 exhibit a dwarf phenotype and their seeds containreduced protein and lipid contents. In this study, we describe a further surprising metabolic peculiarity of the er-ant1mutants.Interestingly, Gly levels in leaves are immensely enhanced (263) when compared with that of wild-type plants. Glyaccumulation is caused by significantly decreased mitochondrial glycine decarboxylase (GDC) activity. Reduced GDC activityin mutant plants was attributed to oxidative posttranslational protein modification induced by elevated levels of reactiveoxygen species (ROS). GDC activity is crucial for photorespiration; accordingly, morphological and physiological defects iner-ant1 plants were nearly completely abolished by application of high environmental CO2 concentrations. The latterobservation demonstrates that the absence of ER-ANT1 activity mainly affects photorespiration (maybe solely GDC), whereasbasic cellular metabolism remains largely unchanged. Since ER-ANT1 homologs are restricted to higher plants, it is temptingto speculate that this carrier fulfils a plant-specific function directly or indirectly controlling cellular ROS production. Theobservation that ER-ANT1 activity is associated with cellular ROS levels reveals an unexpected and critical physiologicalconnection between the ER and other organelles in plants.

INTRODUCTION

Eukaryotic cells, particularly those of plants, are characterizedby a high number of intracellular compartments. Specific solutecarriers and channel proteins in the surrounding membranes arerequired for communication and metabolic embedding of therespective organelles. Because nucleotides fulfill multiple andimportant biological functions and because they are involved inalmost all metabolic processes, sufficient nucleotide import intoand export out of cells and organelles has to be guaranteed.Nucleotides represent basic molecules for DNA and RNA syn-thesis, they serve as cofactors in many enzymatic reactions,are crucial elements involved in intra- and extracellular signaltransduction, and are precursors of phytohormone production(Buchanan et al., 2000; Roux and Steinebrunner, 2007; Riederand Neuhaus, 2011). Moreover, ATP is by far the most importantnucleotide as it constitutes the main energy currency for cellularmetabolism and is required in many biosynthetic pathways.The size and charge of nucleotides renders free permeationof membranes impossible and necessitates specific transport

systems. So far, three different families of nucleotide trans-porting proteins have been identified on the molecular level: thenucleotide transporter (NTT) family, VNUT-type carriers (vesic-ular nucleotide transporters), and the mitochondrial carrier family(MCF) (Haferkamp et al., 2011).NTT proteins are restricted to plant plastids as well as to

certain intracellular parasites or endosymbionts with highly im-paired metabolic capacities. Most NTT-type transporters cata-lyze ATP uptake in exchange with ADP (plus phosphate) andhence are essentially involved in energy provision to the plastidor intracellular organism (Neuhaus et al., 1997; Emes andNeuhaus, 1998; Trentmann et al., 2008). Bacteria of the ordersChlamydiales and Rickettsiales (and diatom algae) possess ad-ditional, different NTTs with different specificities and transportmodes. These NTTs mediate, for example, NAD supply orproton-driven net uptake of purine and pyrimidine nucleotidesand compensate for the missing nucleotide and cofactor syn-thesis (Haferkamp et al., 2004)Mammalian VNUT-type carriers catalyze ATP import into vesi-

cles destined for fusion with the plasma membrane and hence areinvolved in ATP exocytosis. Extracellular ATP acts as an essentialand highly specific signal molecule stimulating various purinergicresponses (Sawada et al., 2008).The MCF group is composed of multiple phylogenetically and

structurally related carriers. MCF proteins mainly act in an an-tiport manner and mediate the transport of diverse substrates,such as nucleotides, organic acids, amino acids, phosphate, or

1 Address correspondence to neuhaus@rhrk.uni-kl.de.The author responsible for distribution of materials integral to the findingspresented in this article in accordance with the policy described in theInstructions for Authors (www.plantcell.org) is: H. Ekkehard Neuhaus(neuhaus@rhrk.uni-kl.de).W Online version contains Web-only data.www.plantcell.org/cgi/doi/10.1105/tpc.113.113605

The Plant Cell, Vol. 25: 2647–2660, July 2013, www.plantcell.org ã 2013 American Society of Plant Biologists. All rights reserved.

diverse cofactors. They are generally located in the inner mito-chondrial membrane. However, a few MCF members reside inother organelles, like plastids, peroxisomes, or the endoplasmicreticulum (ER).

The MCF of Arabidopsis thaliana comprises 58 carriers, and;50% of these proteins have been biochemically and/or phys-iologically characterized (Picault et al., 2004; Palmieri et al.,2011; Haferkamp and Schmitz-Esser, 2012). Phylogenetic in-vestigations revealed that functionally related MCF proteinsfrom different organisms generally form a specific cluster. Ac-cordingly, the affiliation to a certain subgroup allows postulationof the transport properties of biochemically uncharacterizedMCF members. Nucleotide transporting MCF-type carriers aresubdivided into different subclusters, like Mg-ATP/Pi carriers,adenylate carriers specific for CoA, phospho-adenosine phos-phate or phospho-adenosine phosphor sulfate, peroxisomaladenine nucleotide carriers (Linka et al., 2008) and NAD ex-changers, or plastidial brittle proteins involved in ADP-Glc/adenine nucleotide exchange or unidirectional adenylate export(Leroch et al., 2005; Kirchberger et al., 2007). A prominent MCFsubcluster of nucleotide transport proteins is composed of mi-tochondrial ADP/ATP carriers (AACs) from mammals, yeast, andplants. AACs catalyze ADP import in exchange with ATP andhence connect ATP synthesis in the matrix with ATP consumingreactions outside the mitochondrion (Aquila et al., 1987;Haferkamp et al., 2011). Two further but more distantly relatedplant specific carriers are also affiliated to the AAC subcluster:the recently identified plasma membrane–located adenylate car-rier plasma membrane-located adenylate transporter1 (PM-ANT1)as well as the ER-located carrier ER-ANT1.

PM-ANT1 catalyzes ATP export out of the plant cell and isinvolved in signaling processes required for controlled antherdevelopment (Rieder and Neuhaus, 2011), whereas ER-ANT1 isproposed to mediate energy provision to the ER of Arabidopsis(Leroch et al., 2008). ER-ANT1 was shown to act as an ATP/ADPantiporter when heterologously expressed in Escherichia coliand to exhibit a submillimolar affinity for both substrates.Arabidopsis ER-ANT1 loss-of-function mutants contain sub-stantially decreased protein and lipid levels in seeds and exhibita dwarf phenotype (Leroch et al., 2008). The metabolic changesare in line with the essential role of the ER in protein and storagelipid accumulation. The reported er-ant1 mutant phenotype in-cludes stunted growth, reduced chlorophyll contents, and analtered protein/lipid ratio in seeds.

We performed an extended physiological examination of theer-ant1 knockout plants and identified multiple phenotypicalalterations that are tightly associated to photorespiration.The observation that lack of an ER carrier affects the photo-respiratory pathway was unexpected. The cellular organizationof photorespiration and the proteins involved are well in-vestigated (Tolbert, 1997; Foyer et al., 2009; Bauwe et al., 2010;Bauwe et al., 2012). According to our current understanding,photorespiration essentially requires the physiological in-teraction of three different organelles, namely, chloroplasts,mitochondria, and peroxisomes (Foyer et al., 2009; Bauwe et al.,2010), and 25 different carrier proteins have been described toshuttle the required substrates, cofactors, and products be-tween these organelles (Eisenhut et al., 2013a). However, in

animals and yeast, a structural junction and physiological as-sociation of ER and mitochondria have been described (Kornmannet al., 2009; Kornmann and Walter, 2010; Michel and Kornmann,2012). Thus, our current study provides important evidence for aphysiological (reactive oxygen species [ROS]–mediated) interplaybetween the ER and mitochondria in plants.

RESULTS

ER-ANT1 Loss-of-Function Plants Accumulate Gly ina Light-Dependent Manner

ER-ANT1 was shown to be capable of ATP/ADP exchange andto reside in the membrane of the ER (Leroch et al., 2008). Thefact that direct homologs are missing in yeasts and animalssuggests that ER-ANT1 fulfills a plant-specific function and thatbasic energy provision to the ER is generally catalyzed bya different carrier type or system. To gain deeper insights intothe particular role of ER-ANT1, we determined primary metab-olites like selected phosphorylated intermediates representingkey intermediates of Suc and starch synthesis, of glycolysis, andof photosynthesis. In addition, we quantified Suc, Glc, Fru, andstarch as end products of photosynthesis, carboxylates as in-dicators for mitochondrial activities, and amino acids as C/Nintegrators (see Supplemental Table 1, Supplemental Methods 1and Supplemental References 1 online). From all of these 24individual intermediates, Gly levels of mutant plants showed byfar the highest alterations. Wild-type plants contain ;1.2 µmolg21 fresh weight (FW) Gly, whereas er-ant1 mutants exhibited26-fold higher Gly contents (Figure 1A; see Supplemental Table1 online).Quantification of metabolite concentrations in leaves har-

vested at different time points allowed the investigation of di-urnal changes in Gly content. During the course of the day, Glylevels of wild-type plants ranged from ;0.25 µmol g21 FW at theend of the night phase to ;1.2 µmol g21 FW after 8 h of illu-mination (Figure 1B). At the end of the dark, er-ant1 lines ex-hibited a Gly concentration of;2 µmol g21 FW, which increasedto more than 28 µmol g21 FW during 8 h of photosynthesis(Figure 1B). Accordingly, differences in Gly contents of wild-typeand mutant plants become more pronounced during the lightphase and hence are apparently associated to light-dependentprocesses. The decline in Gly contents of er-ant1-2 mutants inthe dark (more precisely, the substantial difference between themaximal and the minimal level) demonstrates that mutant plantsare in principle capable of Gly degradation.

Absence of ER-ANT1 Affects Photorespiration

Because Gly represents a key metabolite associated with pho-torespiration (Douce et al., 2001), we tested whether this light-dependent pathway is affected in er-ant1 plants. Physiologicaland morphological defects of many photorespiration mutantscan be compensated by elevating the environmental CO2 con-centration to ;1 to 2% (Somerville, 2001; Voll et al., 2006).Therefore, we compared growth of er-ant1 and wild-type plantsunder ambient (380 ppm, control) or permanently high CO2

2648 The Plant Cell

conditions (2%; Timm and Bauwe, 2013). Indeed, developmentand habitus of mutants were identical to the wild type in highCO2 and accordingly suppression of photorespiration compen-sated the dwarf phenotype of er-ant1 lines (cf. Figure 2A, 380ppm, and Figure 2B, high CO2).

To further pinpoint the postulated photorespiratory phenotypeof er-ant1 lines, we measured CO2 assimilation, the CO2 com-pensation points, and photosynthetic quantum efficiencies.These quantitative parameters of photosynthetic gas exchangeare inherently codetermined by photorespiratory metabolism(Bauwe et al., 2010; Timm and Bauwe, 2013). All gas-exchangemeasurements were performed with plants permanently grownin presence of elevated environmental CO2 (2%). Transfer tomoderate CO2 concentrations (380 ppm, ambient condition)resulted in substantially lower CO2 fixation capacities of the twoer-ant1 lines (3.1 and 4.0 µmol CO2 m22 s21) when compared

with that of wild-type plants (7.6 µmol m22 s21; Figure 3A).However, exposure to high external CO2 (2.000 ppm) revealedcomparable CO2 fixation rates of mutant (9.8 µmol m22 s21) andwild-type plants (11.5 µmol m22 s21; Figure 3B). This findingdemonstrates that particularly under ambient conditions er-ant1mutants are impaired in CO2 assimilation.The CO2 compensation point represents the external CO2

concentration at which photosynthetic CO2 fixation equals re-spiratory CO2 release. It was determined for each plant line bystepwise increases of the external CO2. In the wild-type control,the compensation point was reached at a CO2 concentration of;60 ppm (Figure 3C), which corresponds to values previouslyreported for Arabidopsis (Eisenhut et al., 2013b). Similar to otherphotorespiratory mutants (Timm et al., 2012), the er-ant1 plantsrequired significantly more CO2 (;90 ppm) to reach the equi-librium between photosynthetic CO2 fixation and (photo)re-spiratory CO2 release.In a further study, we investigated whether transfer from high

external CO2 to ambient conditions affects maximum photo-synthetic quantum efficiency of photosystem II (PSII) (given asthe Fv/Fm ratio) of the different plant lines. Prolonged exposureto environmental CO2 concentrations (380 ppm) reduced Fv/Fm

of the mutant lines, whereas wild-type plants showed no cor-responding alteration. Moreover, Fv/Fm of er-ant1 plants appar-ently decreased with increasing exposure time to low CO2 (1d compared with 5 d; Figure 3D). These data show that ex-tended exposure to photorespiratory conditions results in a re-duced photosynthetic efficiency of er-ant1 mutants probablycaused by increasing damage of the photosynthetic machinery.This is also in line with the fact that er-ant1 plants permanentlygrown in ambient CO2 show reduced chlorophyll contents whencompared with wild-type plants (Figure 2A; Leroch et al., 2008).

Figure 1. Determination of Gly Contents in Arabidopsis Wild-Type ander-ant1 Plants.

Gly contents in 4-week-old wild-type and er-ant1 knockout plants weredetermined by HPLC. Plants were grown under 2% CO2 and a light/darkcycle of 10/14 h for 3 weeks and subsequently shifted to ambient air.Samples were taken after 5 d of adaptation. Shown are mean values ofthree individual replicates, 6 SE. Asterisks indicate the significance levelbetween wild-type and er-ant1 knockout plants according to Student’s ttest (***P < 0.001).(A) Gly levels of wild-type and er-ant1 knockout plants at the end of thelight period.(B) Time course of Gly accumulation during illumination.

Figure 2. Typical Growth Phenotype of Wild-Type and er-ant1 KnockoutPlants.

Wild-type and er-ant1 plants were grown on soil for 4 weeks under dif-ferent CO2 conditions at a day/night cycle of 10/14 h.(A) Growth in ambient air (0.038% CO2). Col-0, Columbia-0.(B) Growth in 2% CO2.

Interaction of Endoplasmic Reticulum and Photorespiration 2649

In summary, these observations demonstrate that er-ant1 mu-tants, in all important aspects, clearly resemble plants with defectsin photorespiration (Somerville, 2001). First, suppression of pho-torespiration by high CO2 cures the stunted growth phenotype.Second, er-ant1 mutants display a distinctly higher CO2 com-pensation point than wild-type plants. Third, photorespiratoryconditions lead to reduced CO2 fixation in combination witha lower PSII maximum quantum efficiency. This clear diagnosis ofa photorespiratory phenotype of the er-ant1 mutants providesunexpected evidence for an interaction between ER metabolismand the complex process of photorespiration.

To figure out whether the lack of ER-ANT1 is in fact causative forthe observed phenotype, we performed control experiments byuse of the rice (Oryza sativa) ortholog Os-ER-ANT1. For this, weinvestigated the functional properties of the closest ER-ANT1homolog from rice (see Supplemental Figure 1 and SupplementalTable 2 online) by heterologous expression in E. coli and transportmeasurements with intact recombinant cells. Import studiesdemonstrated that Os-ER-ANT1 is able to transport ATP and ADP(see Supplemental Figure 2 online), as also known for ArabidopsisER-ANT1 (Leroch et al., 2008). The rice carrier Os-ER-ANT1 ex-hibits apparent affinities (Km) for ATP and ADP of 8.9 and 12.7 µM,

respectively (see Supplemental Figures 2A and 2B online) andhence possesses ;30-fold higher substrate affinities when com-pared with ER-ANT1 from Arabidopsis (Leroch et al., 2008). TheOs-ER-ANT1 gene has been introduced under control of a cauli-flower mosaic virus promoter into the homozygous Arabidopsisknockout line er-ant1-2 (see Supplemental Figures 3A to 3C, 4A,and 4B online). Growth analyses confirmed that developmentaldeficits of the er-ant1-2 line (under ambient CO2) were nearlycompletely compensated for by integration of the orthologous ricegene (see Supplemental Figures 4B and 4C online). These data,together with the already conducted detailed molecular andphysiological characterization of the two independent er-ant1mutant lines (Leroch et al., 2008), clearly demonstrate that ab-sence of ER-ANT1 is the reason for the observed photorespiratoryphenotype.

er-ant1 Mutants Exhibit Highly Reduced Gly DecarboxylaseActivity but Only Slightly Reduced Abundance of GlyDecarboxylase Components

With the aim to gain first insights into the interplay between ER andphotorespiration, we focused on the investigation of possible

Figure 3. Gas Exchange Measurements and Chlorophyll Fluorescence.

Wild-type and er-ant1 knockout mutants were grown under conditions of 2% CO2 in a light/dark cycle of 10/14 h for 4 weeks. CO2 assimilation ratesand compensation points were determined at a PFD of 1.000 µmol photons m22 s21 and at CO2 concentrations ranging from 0.038% to 0.2%. Datarepresent mean values of at least four individual replicates per line, 6 SE. Asterisks indicate the significance level between wild-type and er-ant1knockout plants according to Student’s t test (*P < 0.05, **P < 0.01, and ***P < 0.001).(A) Photosynthetic net CO2 uptake rates at 0.038% CO2. Col-0, Columbia-0.(B) Photosynthetic net CO2 uptake rates at 0.2% CO2.(C) CO2 compensation points.(D) Maximum quantum yield of PSII.

2650 The Plant Cell

factors causing Gly accumulation in er-ant1mutant plants (Figures1A and 1B). An important enzymatic step during photorespirationis the degradation of Gly to methylene tetrahydrofolate, CO2, andNH3 via the mitochondrial enzyme Gly decarboxylase (GDC;Douce et al., 2001; Bauwe et al., 2010). To test whether GDCactivity is affected in er-ant1 mutants, we performed enzyme ac-tivity assays with purified mitochondria isolated from the differentplant lines grown at ambient CO2. Mitochondria from the twoer-ant1 mutant lines showed ;40% GDC activity when comparedwith that of wild-type plants (set to 100%; Figure 4A). Interestingly,also under nonphotorespiratory growth conditions (2% CO2) GDCactivity in the er-ant1 lines was significantly reduced (36% iner-ant1-1 and 39% in er-ant1-2, respectively; Figure 4B).

Functional GDC requires four different proteins, namely, P-,T-, H-, and L-protein (Douce et al., 2001). Moreover, mitochondrialSer hydroxymethyltransferase (SHMT) is also important for adequate

in vivo GDC activity under photorespiratory conditions becauseit rapidly recycles methylene tetrahydrofolate to tetrahydrofolate(Douce et al., 2001). To gain insight into the relative abundance ofthese proteins associated with GDC function, we performed im-munoblot analyses of leaf extracts. Immunodetection with a set ofspecific antibodies raised against four of these five componentsrevealed that er-ant1 mutants show a reduced amount of theP-protein, whereas all other investigated GDC component proteinswere present at about identical levels in wild-type plants ander-ant1 mutants (Figures 4C and 4D). While we did not exactlydetermine leaf P-protein levels, the reduced GDC activities (Fig-ures 4A and 4B) correspond to the lower amounts of P-proteinat least to some degree.To check whether the lowered levels of P-protein (Figure 4D)

might be causative for the decreased GDC activity (Figures 4Aand 4B), we generated er-ant1 plants overexpressing either the

Figure 4. GDC Activity and Immunodetection of Mitochondrial Protein Abundance.

(A) GDC activity in mitochondria isolated from wild-type and er-ant1 mutants grown in ambient air. Col-0, Columbia-0.(B) GDC activity in mitochondria isolated from wild-type and er-ant1 knockout plants grown in 2% CO2. Mitochondria were isolated from green leaftissue, broken by sonication, and incubated in GDC assay buffer supplemented with [14C]-Gly as described in Methods. The 14CO2 released fromP-protein activity was trapped in 3 M KOH after 30 min, and the amount of radioactivity was quantified by scintillation counting. The data are means ofthree individual replicates, 6 SE. Asterisks indicate the significance level between wild-type and er-ant1 knockout plants according to Student’s t test(***P < 0.001).(C) Coomassie blue–stained SDS-PAGE. Ten micrograms of protein of total leaf extracts from wild-type and the two er-ant1 knockout plant lines wasloaded.(D) Immunoblots showing the protein abundance of GDC subunits and SHM.

Interaction of Endoplasmic Reticulum and Photorespiration 2651

P-protein from Flaveria pringlei (PFP5) or from the cyanobacteriumSynechocystis sp PCC6803 (SLR0293) under control of a consti-tutive 35S promoter (see Supplemental Figures 5A to 5F online).The corresponding genetic constructs (see Supplemental Figures5A and 5B online) were introduced into the er-ant1 backgroundand led to substantial levels of corresponding mRNAs (seeSupplemental Figure 5C online). Immunoblot analysis confirmedthat the representative overexpressing mutant lines exhibited notonly detectable mRNA levels, but also increased levels of P-protein(see Supplemental Figure 5D online). However, neither the mutant

line overexpressing the Flaveria gene nor the mutant line over-expressing the Synechocystis gene exhibited any rescue of thedwarf phenotype (see Supplemental Figures 5E and 5F online).

er-ant1 Mutants Show Elevated Levels of ROS andIncreased ROS Defense Capacity

Previous analyses revealed that GDC activity is distinctly lowerin er-ant1 than in the wild type, both at high and ambient CO2. Atleast in part this can be explained by the somewhat reduced

Figure 5. Histochemical Detection of ROS, Activity of ROS Scavenging Enzymes, and GSH Redox Status in Wild-Type and er-ant1 Plants.

(A) and (B) DAB staining for detection of H2O2 accumulation (A) and NBT staining for determination of superoxide accumulation (B). Wild-type ander-ant1 plants were grown in 2% CO2 for 4 weeks to ensure equal growth patterns. Prior to staining of the individual ROS species, plants were keptunder high CO2 conditions or shifted to ambient air for additional 5 d. The control solutions contained 10 mM ascorbate (DAB staining) or 10 units mL21

SOD (NBT staining). Col-0, Columbia-0.(C) Total catalase and SOD activity in Arabidopsis leaf extracts from plants grown under different CO2 conditions.(D) Levels of reduced and oxidized GSH and GSH redox status (ratio of GSH:GSSG). The data represent mean values of three individual replicates, 6 SE.Asterisks indicate the significance level between wild-type and er-ant1 knockout plants according to Student’s t test (*P < 0.05, **P < 0.01, and ***P < 0.001).

2652 The Plant Cell

amount of P-protein, which is the actual Gly decarboxylatingsubunit. Since GDC is known to be regulated by diverse factors,however, we hypothesized that the reduced amount of P-proteinis one but probably not the exclusive and perhaps not even themajor reason for the very strongly elevated Gly levels and thatother processes might also interfere with GDC activity.

Because GDC is highly sensitive to attack by ROS (Tayloret al., 2002; Palmieri et al., 2010), we investigated ROS pro-duction in er-ant1 mutant plants as another possible cause forGDC inhibition. For this, plants were first grown for 4 weeksunder high CO2 (to assure similar development; see Figure 2B)and subsequently cultivated for five additional days under eitherhigh CO2 (2%) or control CO2 levels (380 ppm). Leaves fromwild-type and mutant plants were collected 4 h after onsetof light and the relative abundances of two prominent ROSspecies, namely, H2O2 and the superoxide anion, monitored.Staining with 3,39-diaminobenzidine (DAB) allowed the detectionof H2O2, and nitro-blue tetrazolium chloride (NBT) was used toassess superoxide accumulation. Preincubation with ascorbateor the superoxide-degrading enzyme superoxide dismutase(SOD) confirmed that leaf staining was solely caused by H2O2

and superoxide, respectively (Figures 5A and 5B, control; seeSupplemental Figures 6A and 6B online for further independentreplicates of NBT-stained leaves). Comparison of the stainingintensity revealed that leaves of er-ant1 mutants containedhigher amounts of H2O2 and superoxide than those of wild-typeplants (Figures 5A and 5B; see Supplemental Figures 6A and 6Bonline). Interestingly, enhanced accumulation of these two ROSin mutant leaves not only became detectable after plant expo-sure to ambient CO2 but also occurred when plants were culti-vated under elevated environmental CO2 (Figures 5A and 5B).The increased H2O2 levels were directly confirmed via lumino-metric quantification. At ambient CO2, wild-type plants con-tained 200 nmol H2O2 g

21 FW, but both er-ant1-1 and er-ant1-2plants exhibited between 330 and 340 nmol H2O2 g21 FW (seeSupplemental Figure 6C online). When grown under high CO2,wild-type plants contained nearly identical H2O2 levels aspresent under ambient CO2, namely, 195 nmol H2O2 g21 FW.Similar to the situation in ambient CO2, growth of both er-ant1plant lines under high CO2 led again to increased H2O2 levelswhen compared with the wild type, namely, 285 nmol g21 FW iner-ant1-1 and 348 nmol g21 FW in er-ant1-2 (see SupplementalFigure 6D online).

ROS generation is connected to various physiological pro-cesses in higher plants (Foyer and Noctor, 2005), and differentreactions contribute to the defense machinery directed againstthese reactive compounds. Catalase and SOD are importantfactors controlling cellular ROS levels, and plants contain mul-tiple isoforms of these proteins (Willekens et al., 1994; Bowleret al., 1989). Exposure to ambient CO2 results in significantlyenhanced total catalase and SOD activities in er-ant1 knockoutlines when compared with wild-type plants, whereas under thenonphotorespiratory conditions of high environmental CO2,corresponding enzyme activities of mutant plants only margin-ally exceeded activities in the wild type (Figure 5C). Quantita-tively, photorespiration is the major ROS-producing pathway inilluminated leaves (Foyer et al., 2009). It is hence likely that ROSaccumulation in mutant plants is higher under ambient than

under elevated CO2 conditions. This difference cannot be seenin the staining analysis (Figures 5A and 5B) but is probablysufficient to significantly stimulate catalase and SOD activity.GSH is a prevalent reducing equivalent in the cell and repre-

sents a further factor controlling cellular ROS state (Noctor et al.,2011). Wild-type plants exposed to high CO2 contained GSHconcentrations of 201 nmol g21 FW, and identically treatedmutant lines showed slightly higher GSH levels (283 nmol g21

FW in er-ant1-1 and 260 nmol g21 FW in er-ant1-2; Figure 5D).Increase of GSH in er-ant1 mutants is accompanied by mod-erately decreased oxidized glutathione GSSG levels whencompared with that of the wild type, resulting in a marginallyaltered GSH pool (Figure 5D). However, when exposed to am-bient conditions, mutant lines exhibit strongly increased GSHlevels (619 nmol g21 FW in er-ant1-1 and 418 nmol g21 FW iner-ant1-2), whereas wild-type plants contain;243 nmol GSH g21

FW (Figure 5D). Moreover, apart from GSH also GSSG con-centrations of the er-ant1 lines were significantly increased (by;2.8 and 4.2 nmol g21 FW) when compared with wild-typevalues (Figure 5D). As a consequence, er-ant1 lines exhibit anelevated pool of reduced GSH, particularly under ambientconditions (;96.7% GSH in er-ant1-1 and 96.4% in er-ant1-2plants, respectively, versus 93.7% GSH in wild-type plants;Figure 5D).

GDC from er-ant1 Mutants Exhibits Oxidative Inhibition

Reactive oxygen and nitrogen species are important factorscausing oxidative alterations of the Cys thiol side chains inproteins. Cys side chains can be oxidized to disulfides, differentsulfur oxoacids, mixed disulfides with other thiols, or becomenitrosylated. ROS inhibit plant GDCs (Taylor et al., 2002),and posttranslational Cys modifications (S-glutathionylation/S-nitrosylation) are involved in the oxidative downregulation ofGDCs (Palmieri et al., 2010).As shown above, mutant plants lacking ER-ANT1 exhibit in-

creased ROS and reduced GDC activity. We applied the re-ducing agent DTT to check for an oxidative GDC processing andto examine whether the low GDC activity in er-ant1 mutants isdue to such protein modification (when compared with that ofthe wild-type enzyme). Enriched mitochondria from wild-typeplants and of one representative mutant line (er-ant1-2) weredisrupted and treated with DTT. Application of DTT stimulatedwild-type GDC threefold from 0.52 nmol min21 mg21 protein to;1.56 nmol min21 mg21 (Figure 6). Interestingly, the low GDCactivity of er-ant1-2 mutants was highly increased (to 1.94 nmolmin21 mg21 protein) by DTT and even slightly exceeded thecorresponding wild type (Figure 6). The extent of activity stim-ulation by DTT addition could be explained by an oxidativemodification of GDC in the er-ant1 mutants. This is consistentwith the observation that the reducing agent DTT reversed oxi-dation, which suggests that thiol-based modification accountsfor the highly reduced GDC activity in er-ant1 mutants.Glutathionylation is an important posttranslational modifica-

tion responsible for the reduction of GDC activity in plant mi-tochondria (Palmieri et al., 2010). Thus, to compare GDCglutathionylation of wild-type and er-ant1 plants, we investigatedactivity changes induced by deglutathionylation due to application

Interaction of Endoplasmic Reticulum and Photorespiration 2653

of the regenerating glutaredoxin system (Peltoniemi et al.,2006; Bedhomme et al., 2012). GDC activity of wild-type mi-tochondria was approximately threefold stimulated (from 0.52nmol min21 mg21 protein to 1.45 nmol min21 mg21 protein) byglutaredoxin treatment (Figure 6). However, GDC from er-ant1-2plants was hardly affected (reached 0.4 nmol min21 mg21 protein)and corresponds to only 27% of the observed wild-type GDCvalue (Figure 6). This finding implies that GDC from wild-typeplants is measurably downregulated by glutathionylation, whereasthe enzyme of er-ant1 plants is apparently inhibited by otherfactors.

ROS Accumulation Is Not Caused by Impaired ProteinSynthesis in the ER

ER-ANT1 was shown to reside in the ER (Leroch et al., 2008), andabsence of this carrier results in increased ROS production andfinally in downregulation of GDC activity. To investigate whetherROS production is due to basic malfunction of the ER, we analyzedprotein synthesis and folding in er-ant1 mutants. Accumulation ofunfolded proteins in the ER initiates the so-called unfolded proteinresponse (UPR), detectable by increased transcription of genesencoding ER-located chaperones (Lai et al., 2007). However, weobtained no indications for induction of UPR in er-ant1 plants. Infact, as also previously documented (Leroch et al., 2008), the levelsof mRNAs coding typical ER-located chaperones, namely, BIP1/2,BIP3, and calreticulin, were not enhanced but rather were evenlower in er-ant1 lines when compared with that of wild-type plants(see Supplemental Figure 7 online).

To investigate whether er-ant1 mutants are defective in UPR-associated signal transfer from the ER to the nucleus, we applied

the toxin tunicamyin, known to induce UPR reactions by blockingglycoprotein synthesis. After addition of tunicamycin, transcriptionof BIP1/2, BIP3, and calreticulin was highly stimulated in thewild type as well as in er-ant1 plants (see Supplemental Figure7 online).These results demonstrate that in er-ant1 mutants signal

transduction induced by the accumulation of unfolded proteinsin the ER lumen operates well and that basic metabolic pro-cesses in the ER apparently are not considerably altered. Con-clusively, ROS production in er-ant1 plants is not caused bydrastically impaired ER function. Moreover, massive failures inprotein synthesis would severely affect plant growth and de-velopment. However, the dwarf phenotype of er-ant1 mutantsoccurs solely under ambient CO2, whereas suppression ofphotorespiration by elevated CO2 results in mutant plants mor-phologically resembling the wild type. This indicates that onlyminor but important changes in the ER induce ROS production.

DISCUSSION

er-ant1 Mutants Exhibit a Photorespiratory Phenotype

Photorespiration occurs in cyanobacteria, algae, and at partic-ularly high rates in most higher plant species (Foyer et al., 2009;Bauwe et al., 2012). This metabolic process serves to recycle2-phosphoglycolate, which is unavoidably produced byribulose-1,5-bis-phosphate carboxylase/oxygenase, into 3-phosphoglycerate but also results in considerable rates ofCO2 release. Because photorespiration was long considereda wasteful process leading to reduced plant growth and cropyield, it is not surprising that several molecular-genetic strat-egies were pursued to possibly reduce photorespiratory CO2

losses (Kebeish et al., 2007; Maier et al., 2012). After its dis-covery in the 1950s, it turned out photorespiration is basedon a complex metabolic pathway spread over three cellularcompartments: chloroplasts, peroxisomes, and mitochondria(Tolbert, 1997; Foyer et al., 2009).In this study, we provide first evidence that a further organelle,

namely, the ER, is associated with photorespiration. We foundthat Arabidopsis mutants lacking the ER-located MCF carrierANT1 (er-ant1 loss-of-function lines) exhibit a very distinct pho-torespiratory phenotype. First, these mutants accumulate highamounts of Gly in the light. Second, CO2 compensation points areelevated relative to wild-type plants, whereas CO2 fixation ratesand PSII quantum efficiency are reduced. These features collec-tively result in chlorotic leaves and strongly impaired growth.Third, all these effects observed with the er-ant1 mutants in nor-mal air can be more or less completely reverted to a wild-type-likephenotype by plant growth at a high external CO2 concentration.These symptoms are characteristic of mutants defective in pho-torespiratory metabolism (Somerville and Ogren, 1981; Eisenhutet al., 2013b). The observation that er-ant1 plants exhibit alteredlevels of some other metabolites than Gly (see SupplementalTable 1 online) does not contradict this conclusion. It is generallyknown that limitation of carbon flow on photorespiration also af-fect levels of metabolites not directly involved in this pathway(Somerville, 2001; Timm et al., 2012; Eisenhut et al., 2013b).

Figure 6. Effects of Thiol-Modifying Agents on GDC Activity in Arabi-dopsis Wild-Type and er-ant1 Mitochondria.

Mitochondria from 6-week-old wild-type and er-ant1 knockout plantswere broken by sonication. Before starting the GDC assay, the sampleswere either pretreated with 25 mM DTT for 1 h or with a regeneratingglutaredoxin system for 15 min. The activity of the P-protein was de-termined by trapping the released 14CO2 in KOH and subsequent scin-tillation counting. The data represent mean values of three individualreplicates, 6 SE. Asterisks indicate the significance level between wild-type and er-ant1 knockout plants according to Student’s t test (*P < 0.05,**P < 0.01, and ***P < 0.001). Col-0, Columbia-0.

2654 The Plant Cell

The Photorespiratory Phenotype of er-ant1 Mutants IsCaused by Reduced GDC Activity

Investigation of the metabolite composition demonstrated thater-ant1 mutants massively accumulate Gly under photorespiratoryconditions (Figures 1A and 1B). This metabolic peculiarity waspreviously observed in Arabidopsis mutants lacking functionalGDC (Somerville and Ogren, 1982; Engel et al., 2007). GDC is anessential component of the mitochondrial part of photorespirationand catalyzes, together with SHMT, the formation of Ser from twoGly molecules. Analysis of individual and combined T-DNA in-sertion mutants for the GDC’s P-protein verified that, apart fromphotorespiration, the enzyme is also involved and indispensablefor basic plant metabolism (C1 metabolism; Engel et al., 2007).Arabidopsis plants completely lacking the P-protein die, evenunder nonphotorespiratory conditions (Engel et al., 2007). Bycontrast, individual knockout lines of either of the two P-proteingenes in Arabidopsis grow well in ambient air and do not show anyvisible difference to the wild type (Engel et al., 2007).

Enzyme assays revealed that er-ant1 mutants exhibit only;40% of wild-type GDC activity (Figure 4A). This residual GDCactivity is apparently sufficient to degrade Gly during the nightbut insufficient to cope with the high photorespiratory carbonflux during the light phase (Figure 1B). Inhibition of photorespi-ration by high CO2 results in normal growth and development ofthe er-ant1 mutants, demonstrating that the remaining GDCactivity readily meets the demands of basic C1 metabolism.

Oxidative Modification Mainly Causes the Reduced GDCActivity in er-ant1 Mutants

With the aim to unravel factors connecting ER metabolism andphotorespiration, we focused on the basis of GDC activity re-duction. It is imaginable that limited presence of one or moreGDC-associated proteins in er-ant1 mutants causes the re-duced in vivo activity. However, diverse subunits of the GDCholoprotein as well as the interacting mitochondrial enzymeSHMT exhibited similar abundance in wild-type and mutantplants (Figures 4C and 4D). Also, the amount of P-protein wasslightly reduced (Figure 4D); hence, activity reduction might beexplained, if at all, only to a very limited extent by loweredpresence of the GDC holoprotein (Figure 6). Moreover, the ob-servation that er-ant1 mutants overexpressing either the genecoding for the P-protein from Flaveria or Synechocystis showincreased P-protein abundance but no rescue of the dwarfphenotype (see Supplemental Figures 5A to 5F online) furtherspeaks against an influence of the decreased P-protein amounton total GDC activity in er-ant1 plants.

Thus, apart from a slightly reduced abundance of the P-protein,the vast majority of activity reduction was clearly attributed toa substantial GDC inhibition due to oxidative posttranslationalmodification. The GDC is highly susceptible to reactive oxygensince ROS accelerate carbonylation or S-glutathionylation, pro-cesses known to downregulate GDC activity (Taylor et al., 2002,2004; Palmieri et al., 2010; Lounifi et al., 2013). We believe thatthis type of GDC modification is caused by increased oxidativestress in er-ant1 plants as demonstrated by various indicators: (1)elevated H2O2 and superoxide levels, (2) stimulated catalase and

SOD activities, and (3) increased GSH levels (Figures 5A to 5D;see Supplemental Figure 6 online). The latter observation is,moreover, fully in line with the observation that oxidative stressper se stimulates Glu-Cys ligase activity provoking increasedGSH levels (Hicks et al., 2007) since this enzyme is rate limitingfor GSH biosynthesis.Carbonylation is an irreversible modification that targets pro-

teins to degradation (Rao and Moller, 2011; Lounifi et al., 2013).However, this type of modification is probably not responsiblefor GDC activity reduction in er-ant1 mutant plants. Althougher-ant1mutants harbored slightly lower amounts of the P-protein,the other GDC proteins were present at wild-type levels (Figure4D). Remarkably, the low GDC activity of er-ant1 mitochondriacould be fully restored by DTT treatment and then even exceededactivities of the correspondingly treated GDC from wild-type mi-tochondria (Figure 6). This finding pointed to an oxidative modi-fication of Cys residues.Cys side chains can undergo different states of oxidation. For ex-

ample, they are modified by S-thiolation (mainly S-glutathionylation)or S-nitrosylation, form intra- or intermolecular protein disulfidebridges, and are oxidized to sulfenic, sulfinic, and sulfonicacids. By use of a regenerating glutaredoxin assay, suitable toremove GSH residues from proteins (Bedhomme et al., 2012),we investigated the S-glutathionylation status of GDC fromwild-type plants and er-ant1 mutants. Several glutathionylatedCys residues have been identified in the GDC in vivo and wereshown to be responsible for reduced enzyme activity (Palmieriet al., 2010). We found that GDC activity of the wild type waspartially reduced due to glutathionylation, but the correspondingenzyme from mutant plants did not respond to the glutaredoxinassay (Figure 6).The latter observation does not fully exclude glutathionylation,

but we presently think that another type of oxidative alterationis responsible for the inhibition of GDC in er-ant1 plants. DTTcan reduce several modified thioles (Scheibe and Stitt, 1988;Kettenhofen et al., 2007), and GDC activity from er-ant1 plantswas totally restorable by application of reduced DTT (Figure 6).Therefore, activity reduction of GDC in er-ant1 lines is mostlikely caused by ROS induced oxidative alterations/damages,like disulfide bond formation absent in wild-type mitochondria(Figure 6). At this stage, it is important to mention that we cannotrule out that this type of modification overrides the inhibitoryimpact of a still possible S-glutathionylation in the GDC of mu-tant plants.The reduced GDC activity is apparently causative for the

photorespiratory phenotype of er-ant1 mutants. It was shown toresult from oxidative protein modification and, likely to a smallerdegree, less P-protein. Oxidative modification of GDC is fullyconsistent with an enhanced ROS production in er-ant1 mu-tants. Interestingly, er-ant1 mutants accumulated significantlyhigher amounts of H2O2 and superoxide not only under ambient,but also under high external CO2 concentrations when com-pared with wild-type plants (Figures 5A and 5B). However, otherphysiological and morphological defects emerge exclusively/mainly under photorespiratory conditions. These observationsindicate that the increased ROS level nearly exclusively affectsphotorespiration, whereas other metabolic processes remainrather unchanged or cope better with this situation.

Interaction of Endoplasmic Reticulum and Photorespiration 2655

The Missing Link: What Connects ER-ANT1 toROS Formation?

ER-ANT1 is unambiguously affiliated to with the ATP/ADP car-rier subgroup of the MCF. Moreover, it exhibits typical and es-sential protein motifs that are conserved among all AACs(Leroch et al., 2008) and therefore was primarily suggested torepresent a fourth AAC isoform in Arabidopsis. In fact, bio-chemical analyses verified that both recombinant ArabidopsisER-ANT1 and the rice ER-ANT1 protein mediate counterex-change of ATP and ADP (Leroch et al., 2008; see SupplementalFigures 2A and 2B online). Noteworthy, Arabidopsis ER-ANT1and its rice homolog exhibit a marked structural differenceto the AACs from Arabidopsis as its amino acid sequence isshorter because it lacks the mitochondrial targeting sequencecharacteristic for latter carriers. This feature already points toa different subcellular targeting and diverse methods clearlylocalized the carrier in the ER (Leroch et al., 2008).

It appears quite unlikely that ER-ANT1 acts as the main ATP/ADP carrier in this compartment. Of course ATP entry into theER is mandatory to enable various anabolic reactions within thisorganelle (Clairmont et al., 1992; Pimpl et al., 2006). However,the fact that ER-ANT1 homologs are restricted to the plantkingdom but energy provision to the ER is required in all eu-karyotes suggests that energy provision to this organelle ismainly/generally mediated by a different mechanism (Lerochet al., 2008; Haferkamp and Schmitz-Esser, 2012). This as-sumption is further supported by the observation that lack ofER-ANT1 especially interferes with photorespiration but notso markedly with basic cell metabolism. Thus, ER-ANT1 mightcontribute only to a limited extend to the overall ATP supply intothe ER; moreover, it cannot be ruled out that ER-ANT1 catalyzesthe transport of an additional so far not identified substrate.Regardless of its transport function, there is a clear metabolicconnection between ANT1 presence in the ER and ROS pro-duction; however, the mechanistic basis of this is still elusive.

It has frequently been reported that uncontrolled processes inthe ER, for example, accumulation of unfolded proteins, caninduce ROS production (Santos et al., 2009). However, er-ant1plants do not show typical molecular symptoms associated witha UPR because the expression of genes coding for ER-locatedchaperones is not enhanced and also the UPR signal trans-duction pathway was proven to operate well (see SupplementalFigure 7 online). Accordingly, enhanced ROS production iner-ant1 mutants is apparently not associated with UPR-relatedprocesses. We suggest that the ER of er-ant1 mutants exhibitsa subtle physiological impairment because main characteristicER functions, like protein synthesis and folding, proceed with novisible alterations and because the majority of the observedsymptoms in mutants can be cured by permanent growth ina high CO2 environment. The resulting ROS (H2O2 and super-oxide) accumulating in er-ant1 mutants (Figures 5A and 5B; seeSupplemental Figures 6A and 6B online) most likely derive froma modified mitochondrial metabolism, which is a result of im-paired ER processes. This assumption is based on the generalproperties of ROS because H2O2 is highly polar, and both H2O2

and superoxide are extremely reactive molecules, giving thema limited capacity to diffuse across the cell (Apel and Hirt, 2004).

Interestingly, the proposed primary (and so far unresolved)metabolic changes become detectable by the increased ROSproduction occurring in mutants grown under ambient air butalso in mutants cultivated under high CO2 conditions (Figures 5Aand 5B; see Supplemental Figures 6A and 6B online). GDC issuggested to be much more sensitive to oxidative stress thanother enzymes (Taylor et al., 2002); in fact, comparably slightoxidative stress in er-ant1 plants grown under high CO2 (Figures5B to 5D) already caused GDC inhibition (40% residual activity;Figure 4B). However, excessive oxidative modification and ac-tivity impairment of GDC is solely problematic during photores-piration and not under nonphotorespiratory conditions (Figures 2Aand 2B). Therefore, we propose that er-ant1 plants exhibit a cu-mulative defect that becomes potentiated by high carbon fluxthrough photorespiration. In this context, er-ant1 plants resembletypical photorespiratory mutants lacking one of the enzymescritical for photorespiratory metabolism (Bauwe and Kolukisaoglu,2003; Bauwe et al., 2010).

Conclusion

Our data indicate that the Arabidopsis carrier ER-ANT1 is re-quired to prevent uncontrolled ROS generation and, thus, toguarantee maintenance of photorespiration. Absence of ER-ANT1 causes essential but subtle and therefore so far un-resolved physiological changes in the ER that lead to ROSaccumulation. The increased ROS level promotes oxidativemodification and inhibition of the enzyme GDC under bothphotorespiratory and nonphotorespiratory conditions. However,impaired GDC activity is only detrimental under photorespiratoryconditions that further potentiate defects in er-ant1 mutants.In animal and fungal cells, the bidirectional communication

between ER and mitochondria has been observed for a longtime and is required for important processes like Ca2+ signaling,lipid metabolism, intracellular energy provision, and even pro-grammed cell death (Kim et al., 2006; de Brito and Scorrano2010; Michel and Kornmann, 2012). Our studies shed un-expected light on the ER/mitochondria communication by a sofar unknown interaction of a plant-specific ER resident carrier(Leroch et al., 2008) and a plant-specific metabolic pathway. Inthe future, more elaborate analyses of the er-ant1 mutants arerequired to completely unravel ER-ANT1–associated metabolicalterations in the ER and the basis of the observed ROSproduction.

METHODS

Plant Material and Growth Conditions

All studieswere performedwithArabidopsis thalianawild-type plants (ecotypeColumbia-0) and T-DNA insertion mutants for er-ant1-1 (Salk_043626) ander-ant1-2 (Salk_023441) as described earlier (Leroch et al., 2008). Prior togermination, seeds were incubated for 2 d in the dark at 4°C on standardizedED73 soil (Weigel andGlazebrook, 2002). Plants were grown at 22°C and 120µmol quanta m22 s21 in a 10-h-light/14-h-dark regime either at ambient CO2

levels (380 ppm CO2), or at high CO2 levels (20,000 ppm CO2; Timm andBauwe, 2013) in Percival plant climate chambers (CLF Plant Climatics). Formeasurements under ambient conditions, plants were shifted to normalair for 1 week prior to the experiment.

2656 The Plant Cell

RNA Isolation, Generation of cDNA, and Quantitative RT-PCR

Total RNA was prepared from Arabidopsis leaf tissue using the NucleoSpinRNA Plant Kit (Macherey-Nagel). Quantitative RT-PCR was performedas described (Leroch et al., 2005) using a MyIQ-Cycler (Bio-Rad) and IQSYBR Green Supermix (Bio-Rad), according to the manufacturer’s in-structions. The geneAt5g60390, encoding the elongation factor 1a, wasused for quantitative normalization. The respective primers are listed inSupplemental Table 3 online.

Protein Isolation and Immunodetection of GDC Subunits

For the isolation of total protein, 100 mg of leaf tissue was homogenized inliquid nitrogen using a mortar and pestle and extracted in 200 mL of buffermedium (50 mM HEPES-KOH, pH 7.6, 10 mM NaCl, 5 mM MgCl2, 100 mMsorbitol, and 1 mM phenylmethylsulfonyl fluoride). The homogenate wascentrifuged (4°C, 10 min, 20,000g), and the total protein content was de-termined (Bradford, 1976). For immunological analysis, 10 mg of total leafprotein was separated by SDS-PAGE and stained with Coomassie BrilliantBlueR 250 (Laemmli, 1970) or blotted onto a polyvinylidenfluoridemembraneaccording to standard protocols. Immunoblots were decorated with primaryantibodies raised against SHM or the GDC proteins P, T, and H.

Quantification of Metabolites and Enzyme Activities of CAT and SOD

For the determination of metabolites and enzyme activities, Arabidopsisleaf samples were harvested in the middle of the light phase and ground inliquid nitrogen usingmortar and pestle. Quantification of Gly by HPLCwasperformed according to a routine method in our lab (Jung et al., 2009)using 80% ethanol (v/v) as extraction medium. GSH was quantified fromneutralized HCl extracts according to the 5,59-dithiobis-C2-nitrobenzoicacid-based cycling assay (Queval and Noctor, 2007) using a TecanInfinite-200 plate reader. For determination of CAT and SOD activities,100 mg of frozen plant material was resuspended in 500 mL PBS (137 mMNaCl, 2.7 mM KCl, 10 mM Na2HPO4, 2 mM KH2PO4, pH 7.4, and 1 mMphenylmethylsulfonyl fluoride) and centrifuged to remove cell debris (4°C,10min, 20.000g), and total protein content of the supernatantwas determined(Bradford, 1976). CAT and SOD activity measurements were conductedaccording to Jacobo-Velazquez et al. (2011). CAT activity was determined byadding 5 to 15mL of sample toCAT assay buffermedium (0.05MKH2PO4, pH7.4, and 0.01MH2O2). The decrease in absorbance at 240 nmwasmonitoredwith a Hitachi U-2000 spectrophotometer. SOD activity was determined witha NBT-based assay using a Tecan Infinite-200 plate reader.

Detection of ROS

Histochemical stainings were performed as described (Lee et al., 2002;Reiser et al., 2004) with the following modifications: To detect the ac-cumulation of superoxide, leaves from wild-type and er-ant1 plants werevacuum-infiltrated for 10 min with 0.1 mg mL21 NBT in 25 mM HEPESbuffer, pH 7.6. The control solution additionally contained 10 units mL21

SOD and 10 mM MnCl2. Leaf samples were left for 2 h at room tem-perature in the dark and were subsequently destained by incubation at60°C in 95% ethanol (v/v) for 30 min.

The accumulation of H2O2 was visualized by vacuum infiltration ofleaves in a solution containing 0.1 mg mL21 DAB in 10 mM MES, pH 6.0,and KOH. The control solution contained additionally 10 mM ascorbate.After infiltration, the leaves were incubated for 16 h in the dark and de-stained by boiling in lactic acid:glycerol:ethanol (1:1:3) for 5 to 10 min.

Gas Exchange and Chlorophyll Fluorescence Measurements

Gas exchange measurements and fluorescence measurements wereperformed with a gas exchange and chlorophyll fluorescence system

(GFS-3000; Walz). All measurements were performed with 4-week-oldwild-type and er-ant1 plants grown under high CO2 conditions. Net CO2

assimilation rates were determined at ambient CO2 levels (380 ppm) orelevated CO2 levels (2.000 ppm), a PFD of 1.000 µE, a leaf temperature of25°C, 60% humidity, and an air flow of 750 µmol m22 s21. For the de-termination of CO2 compensation points, CO2 concentrations varied from40 to 400 ppm. Themaximum quantum yield of PSII (Fv/Fm) was quantifiedaccording to Harbinson et al. (1989) before and after shifting the plants toambient air for 1 week (Timm et al., 2011).

Isolation of Mitochondria and Determination of GDC Activity

Isolation of a crude mitochondrial extract from 10 g of Arabidopsis leaveswas performed according to a standard protocol (Keech et al., 2005).Mitochondria were washed twice in buffer medium (0.3 M Suc, 10 mMTES, 2 mM EDTA, and 10 mM KH2PO4, pH 7.5), resuspended in 0.6 mL ofGDC assay medium (0.36 M Suc, 20 mMMOPS, pH 7.2, 8 mM KCl, 4 mMNaH2PO4, 4 mMMgCl2, and 0.1% BSA), and broken by sonication for 33

30 s on ice. The total protein content was determined according toBradford (1976) after another centrifugation step (4°C, 5 min, 16,000g).GDC activity of the supernatant was quantified by measuring the de-carboxylation of [14C]-Gly (Somerville and Ogren, 1982). One hundredmicrograms of mitochondrial protein were incubated with 100 mL of GDCassay buffer supplemented with 8 mM Gly (4 µCi mL21 [14C]-Gly) in 1.5-mL reaction tubes. The effects of DTT were analyzed after incubation ofthemitochondrial protein with 25mMDTT for 1 h on ice before starting theassay. The effects of deglutathionylation were analyzed using a re-generative glutaredoxin system (Peltoniemi et al., 2006; Bedhomme et al.,2012). Samples were incubated in GDC assay buffer containing 0.05 unitsmL21 glutaredoxin, 0.02 units mL21 GSH reductase, 1 mM GSH, and50 µM NADPH as final concentration (all compounds were provided bySigma-Aldrich) and incubated for 15 min at room temperature beforestarting the assay. Released [14C]-CO2 was captured in small reactionvessels containing 100 mL 3 M KOH. The reaction was stopped after30 min by injecting 100 mL of 2 M HCl into the reaction mixture witha syringe through the closed lid. The hole in the lid was subsequentlysealed with grease and the vials were stored for one night to allow totalabsorption of radioactively labeled CO2. The trapped radioactivity wasquantified by scintillation counting (Perkin-Elmer Tricarb 2810 TR).

Accession Numbers

Sequence data from this article can be found in the Arabidopsis GenomeInitiative or GenBank/EMBL databases under the following accessionnumbers:ArabidopsisER-ANT1 (At5g17400),ArabidopsisAAC1 (At3g08580),Arabidopsis AAC2 (At5g13490), Arabidopsis AAC3 (At4g28390), ArabidopsisEF-1a (At1g07930), Arabidopsis BiP1 (At5g28540), Arabidopsis BiP2(At5g42420), Arabidopsis BiP3 (At1g09080),ArabidopsisCRT1 (At1g56340),ArabidopsisCRT2 (At1g09210), rice ER-ANT1 (Os11g0661300), putative riceAAC1/2 (Os02g0718900), and putative rice AAC3 (Os05g0302700).

Supplemental Data

The following materials are available in the online version of the article.

Supplemental Figure 1. Alignment of the Predicted Amino AcidSequence of Rice ER-ANT1 and Related MCF Carriers from Arabi-dopsis and Rice.

Supplemental Figure 2. Heterologous Expression of Rice ER-ANT1 inE. coli Cells and Substrate Dependency of Adenine Nucleotide Uptake.

Supplemental Figure 3. Molecular Characterization of 35S:Os-er-ant1Plants by PCR.

Supplemental Figure 4. Complementation of er-ant1-2 KnockoutPlants with the Closest Ortholog from Rice, Os-ER-ANT1.

Interaction of Endoplasmic Reticulum and Photorespiration 2657

Supplemental Figure 5. Overexpression of pfp5 and slr0293 iner-ant1 Knockout Plants.

Supplemental Figure 6. Further Evidence for Enhanced ROS Accu-mulation in Wild-Type and er-ant1 Knockout Plants by NBT Stainingand Determination of H2O2 in Leaf Extracts.

Supplemental Figure 7. Transcript Levels of ER Chaperone Genes inResponse to ER Stress.

Supplemental Table 1. Selected Metabolites of Wild-Type ander-ant1 Knockout Plants under Ambient CO2 Conditions.

Supplemental Table 2. Comparison between the Amino Acid Se-quences of Arabidopsis ER-ANT1 and Related MCF Carriers fromArabidopsis and Rice.

Supplemental Table 3. Primers Used for Screening, Cloning, andqRT-PCR Studies.

Supplemental Methods 1. Detailed Procedures for Supplemental Data.

Supplemental References 1. References for Supplemental Figuresand Methods.

ACKNOWLEDGMENTS

Work in the laboratories of H.E.N. was financially supported by theDeutsche Forschungsgemeinschaft (Reinhard Koselleck-Grant). Work inthe lab of H.E.N., I.H., J.M.H., and J.R. was further supported by theLandesschwerpunkt “Membrantransport” (http://www.uni-kl.de/rimb).We thank Mohammad Hajirezaei, IPK Gatersleben, Germany, for quanti-fications of some metabolites.

AUTHOR CONTRIBUTIONS

C.H. conducted most experiments. B.P. and M.L. cloned the rice ER-ANT1and conducted uptake experiments on E. coli. R.E. conducted immunoblotexperiments and established mitochondria purification protocols. I.H., J.R.,J.M.H., and H.B. supported development of experimental strategies.H.E.N. wrote the article.

Received May 13, 2013; revised June 21, 2013; accepted June 28, 2013;published July 16, 2013.

REFERENCES

Apel, K., and Hirt, H. (2004). Reactive oxygen species: Metabolism,oxidative stress, and signal transduction. Annu. Rev. Plant Biol. 55:373–399.

Aquila, H., Link, T.A., and Klingenberg, M. (1987). Solute carriersinvolved in energy transfer of mitochondria form a homologousprotein family. FEBS Lett. 212: 1–9.

Bauwe, H., Hagemann, M., and Fernie, A.R. (2010). Photorespiration:Players, partners and origin. Trends Plant Sci. 15: 330–336.

Bauwe, H., Hagemann, M., Kern, R., and Timm, S. (2012).Photorespiration has a dual origin and manifold links to centralmetabolism. Curr. Opin. Plant Biol. 15: 269–275.

Bauwe, H., and Kolukisaoglu, U. (2003). Genetic manipulation ofglycine decarboxylation. J. Exp. Bot. 54: 1523–1535.

Bedhomme, M., Adamo, M., Marchand, C.H., Couturier, J., Rouhier, N.,Lemaire, S.D., Zaffagnini, M., and Trost, P. (2012). Glutathionylation ofcytosolic glyceraldehyde-3-phosphate dehydrogenase from the model

plant Arabidopsis thaliana is reversed by both glutaredoxins andthioredoxins in vitro. Biochem. J. 445: 337–347.

Bowler, C., Alliotte, T., De, L.M., Van, M.M., and Inze, D. (1989). Theinduction of manganese superoxide dismutase in response tostress in Nicotiana plumbaginifolia. EMBO J. 8: 31–38.

Bradford, M. (1976). A rapid and sensitive method for the quantitationof microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: 248–254.

Buchanan, B.B., Gruissem, W., and Jones, R.L. 2000. Biochemistryand Molecular Biology of Plants. (Rockville, MD: American Societyof Plant Physiology).

Clairmont, C.A., De Maio, A., and Hirschberg, C.B. (1992). Translocationof ATP into the lumen of rough endoplasmic reticulum-derived vesiclesand its binding to luminal proteins including BiP (GRP 78) and GRP 94.J. Biol. Chem. 267: 3983–3990.

de Brito, O.M., and Scorrano, L. (2010). An intimate liaison: Spatialorganization of the endoplasmic reticulum-mitochondria relationship.EMBO J. 29: 2715–2723.

Douce, R., Bourguignon, J., Neuburger, M., and Rebeille, F. (2001).The glycine decarboxylase system: A fascinating complex. TrendsPlant Sci. 6: 167–176.

Eisenhut, M., Pick, T.R., Bordych, C., and Weber, A.P. (2013a).Towards closing the remaining gaps in photorespiration - Theessential but unexplored role of transport proteins. Plant Biol.(Stuttg.) 15: 676–685.

Eisenhut, M., et al. (2013b). Arabidopsis A BOUT DE SOUFFLE isa putative mitochondrial transporter involved in photorespiratorymetabolism and is required for meristem growth at ambient CO2

levels. Plant J. 73: 836–849.Emes, M.J., and Neuhaus, H.E. (1998). Metabolism and transport in

non-photosynthetic plastids. J. Exp. Bot. 48: 1995–2005.Engel, N., van den Daele, K., Kolukisaoglu, U., Morgenthal, K.,

Weckwerth, W., Parnik, T., Keerberg, O., and Bauwe, H. (2007).Deletion of glycine decarboxylase in Arabidopsis is lethal undernonphotorespiratory conditions. Plant Physiol. 144: 1328–1335.

Foyer, C.H., Bloom, A.J., Queval, G., and Noctor, G. (2009).Photorespiratory metabolism: Genes, mutants, energetics, andredox signaling. Annu. Rev. Plant Biol. 60: 455–484.

Foyer, C.H., and Noctor, G. (2005). Redox homeostasis andantioxidant signaling: a metabolic interface between stress perceptionand physiological responses. Plant Cell 17: 1866–1875.

Haferkamp, I., Fernie, A.R., and Neuhaus, H.E. (2011). Adeninenucleotide transport in plants: much more than a mitochondrialissue. Trends Plant Sci. 16: 507–515.

Haferkamp, I., and Schmitz-Esser, S. (2012). The plant mitochondrialcarrier family: Functional and evolutionary aspects. Front. Plant Sci. 3: 2.

Haferkamp, I., Schmitz-Esser, S., Linka, N., Urbany, C., Collingro,A., Wagner, M., Horn, M., and Neuhaus, H.E. (2004). A candidateNAD+ transporter in an intracellular bacterial symbiont related toChlamydiae. Nature 432: 622–625.

Harbinson, J., Genty, B., and Baker, N.R. (1989). Relationshipbetween the quantum efficiencies of photosystems I and II in pealeaves. Plant Physiol. 90: 1029–1034.

Hicks, L.M., Cahoon, R.E., Bonner, E.R., Rivard, R.S., Sheffield, J.,and Jez, J. (2007). Thiol-based regulation of redox-active glutamate-cysteine ligase from Arabidopsis thaliana. Plant Cell 19: 2653–2661.

Jacobo-Velazquez, D.A., Martinez-Hernandez, G.B., Del, C.R.,Cao, C.M., and Cisneros-Zevallos, L. (2011). Plants as biofactories:Physiological role of reactive oxygen species on the accumulation ofphenolic antioxidants in carrot tissue under wounding and hyperoxiastress. J. Agric. Food Chem. 59: 6583–6593.

Jung, B., Flörchinger, M., Kunz, H.H., Traub, M., Wartenberg, R.,Jeblick, W., Neuhaus, H.E., and Möhlmann, T. (2009). Uridine-

2658 The Plant Cell

ribohydrolase is a key regulator in the uridine degradation pathwayof Arabidopsis. Plant Cell 21: 876–891.

Kebeish, R., Niessen, M., Thiruveedhi, K., Bari, R., Hirsch, H.J.,Rosenkranz, R., Stabler, N., Schönfeld, B., Kreuzaler, F., andPeterhänsel, C. (2007). Chloroplastic photorespiratory bypassincreases photosynthesis and biomass production in Arabidopsisthaliana. Nat. Biotechnol. 25: 593–599.

Keech, P., Dizengremel, P., and Gardeström, P. (2005). Preparationof leaf mitochondria from Arabidopsis thaliana. Physiol. Plant. 124:403–409.

Kettenhofen, N.J., Broniowska, K.A., Keszler, A., Zhang, Y., andHogg, N. (2007). Proteomic methods for analysis of S-nitrosation. J.Chromatogr. B Analyt. Technol. Biomed. Life Sci. 851: 152–159.

Kim, R., Emi, M., and Tanabe, K. (2006). Role of mitochondria as thegardens of cell death. Cancer Chemother. Pharmacol. 57: 545–553.

Kirchberger, S., Leroch, M., Huynen, M.A., Wahl, M., Neuhaus,H.E., and Tjaden, J. (2007). Molecular and biochemical analysisof the plastidic ADP-glucose transporter (ZmBT1) from Zea mays.J. Biol. Chem. 282: 22481–22491.

Kornmann, B., Currie, E., Collins, S.R., Schuldiner, M., Nunnari, J.,Weissman, J.S., and Walter, P. (2009). An ER-mitochondria tetheringcomplex revealed by a synthetic biology screen. Science 325: 477–481.

Kornmann, B., and Walter, P. (2010). ERMES-mediated ER-mitochondriacontacts: Molecular hubs for the regulation of mitochondrial biology.J. Cell Sci. 123: 1389–1393.

Laemmli, K. (1970). Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature 227: 680–685.

Lai, E., Teodoro, T., and Volchuk, A. (2007). Endoplasmic reticulumstress: Signaling the unfolded protein response. Physiology(Bethesda) 22: 193–201.

Lee, B.H., Lee, H., Xiong, L., and Zhu, J.K. (2002). A mitochondrialcomplex I defect impairs cold-regulated nuclear gene expression.Plant Cell 14: 1235–1251.

Leroch, M., Kirchberger, S., Haferkamp, I., Wahl, M., Neuhaus,H.E., and Tjaden, J. (2005). Identification and characterization ofa novel plastidic adenine nucleotide uniporter from Solanum tuberosum.J. Biol. Chem. 280: 17992–18000.

Leroch, M., Neuhaus, H.E., Kirchberger, S., Zimmermann, S.,Melzer, M., Gerhold, J., and Tjaden, J. (2008). Identification ofa novel adenine nucleotide transporter in the endoplasmic reticulumof Arabidopsis. Plant Cell 20: 438–451.

Linka, N., Theodoulou, F.L., Haslam, R.P., Linka, M., Napier, J.A.,Neuhaus, H.E., and Weber, A.P. (2008). Peroxisomal ATP import isessential for seedling development in Arabidopsis thaliana. PlantCell 20: 3241–3257.

Lounifi, I., Arc, E., Molasiotis, A., Job, D., Rajjou, L., and Tanou, G.(2013). Interplay between protein carbonylation and nitrosylation inplants. Proteomics 13: 568–578.

Maier, A, and Fahnenstich, H, von Caemmerer, S., Engqvist, M.K.,Weber, A.P., Flügge, U.I., and Maurino, V.G. (2012). Transgenicintroduction of a glycolate oxidative cycle into A. thaliana chloroplastsleads to growth improvement. Front. Plant Sci. 3: 38.

Michel, A.H., and Kornmann, B. (2012). The ERMES complex andER-mitochondria connections. Biochem. Soc. Trans. 40: 445–450.

Neuhaus, H.E., Thom, E., Möhlmann, T., Steup, M., andKampfenkel, K. (1997). Characterization of a novel eukaryotic ATP/ADP translocator located in the plastid envelope of Arabidopsisthaliana L. Plant J. 11: 73–82.

Noctor, G., Queval, G., Mhamdi, A., Chaouch, S., and Foyer, C.H. (2011).Glutathione. The Arabidopsis Book. 9: e0142, doi/10.1109/tab.0142.

Palmieri, F., Pierri, C.L., De Grassi, A., Nunes-Nesi, A., and Fernie,A.R. (2011). Evolution, structure and function of mitochondrialcarriers in plants. Plant J. 66: 161–181.

Palmieri, M.C., Lindermayr, C., Bauwe, H., Steinhauser, C., andDurner, J. (2010). Regulation of plant glycine decarboxylase bys-nitrosylation and glutathionylation. Plant Physiol. 152: 1514–1528.

Peltoniemi, M.J., Karala, A.R., Jurvansuu, J.K., Kinnula, V.L., andRuddock, L.W. (2006). Insights into deglutathionylation reactions.Different intermediates in the glutaredoxin and protein disulfideisomerase catalyzed reactions are defined by the gamma-linkagepresent in glutathione. J. Biol. Chem. 281: 33107–33114.

Picault, N., Hodges, M., Palmieri, L., and Palmieri, F. (2004). Thegrowing family of mitochondrial carriers in Arabidopsis. TrendsPlant Sci. 9: 138–146.

Pimpl, P., Taylor, J.P., Snowden, C., Hillmer, S., Robinson, D.G.,and Denecke, J. (2006). Golgi-mediated vacuolar sorting of theendoplasmic reticulum chaperone BiP may play an active role inquality control within the secretory pathway. Plant Cell 18: 198–211.

Queval, G., and Noctor, G. (2007). A plate reader method for themeasurement of NAD, NADP, glutathione, and ascorbate in tissueextracts: Application to redox profiling during Arabidopsis rosettedevelopment. Anal. Biochem. 363: 58–69.

Rao, R.S., and Moller, I.M. (2011). Pattern of occurrence andoccupancy of carbonylation sites in proteins. Proteomics 11: 4166–4173.

Reiser, J., Linka, N., Lemke, L., Jeblick, W., and Neuhaus, H.E.(2004). Molecular physiological analysis of the two plastidic ATP/ADPtransporters from Arabidopsis thaliana. Plant Physiol. 136: 3524–3536.

Rieder, B., and Neuhaus, H.E. (2011). Identification of an Arabidopsisplasma membrane located ATP transporter important for antherdevelopment. Plant Cell 23: 1932–1944.

Roux, S.J., and Steinebrunner, I. (2007). Extracellular ATP: Anunexpected role as a signaler in plants. Trends Plant Sci. 12: 522–527.

Santos, C.X., Tanaka, L.Y., Wosniak, J., and Laurindo, F.R. (2009).Mechanisms and implications of reactive oxygen species generationduring the unfolded protein response: Roles of endoplasmic reticulumoxidoreductases, mitochondrial electron transport, and NADPH oxidase.Antioxid. Redox Signal. 11: 2409–2427.

Sawada, K., Echigo, N., Juge, N., Miyaji, T., Otsuka, M., Omote, H.,Yamamoto, A., and Moriyama, Y. (2008). Identification of a vesicularnucleotide transporter. Proc. Natl. Acad. Sci. USA 105: 5683–5686.

Scheibe, R., and Stitt, M. (1988). Comparison of NADP-malatedehydrogenase activation, QA reduction and O2 evolution in spinachleaves. Plant Physiol. Biochem. 26: 473–481.

Somerville, C.R. (2001). An early Arabidopsis demonstration. Resolvinga few issues concerning photorespiration. Plant Physiol. 125: 20–24.

Somerville, C.R., and Ogren, W.L. (1981). Photorespiration-deficientmutants of Arabidopsis thaliana lacking mitochondrial serinetranshydroxymethylase activity. Plant Physiol. 67: 666–671.

Somerville, C.R., and Ogren, W.L. (1982). Mutants of the cruciferousplant Arabidopsis thaliana lacking glycine decarboxylase activity.Biochem. J. 202: 373–380.

Taylor, N.L., Day, D.A., and Millar, A.H. (2002). Environmental stresscauses oxidative damage to plant mitochondria leading to inhibituion ofglycine decarboxylase. J. Biol. Chem. 277: 42663–42668.

Taylor, N.L., Day, D.A., and Millar, A.H. (2004). Targets of stress-induced oxidative damage in plant mitochondria and their impact oncell carbon/nitrogen metabolism. J. Exp. Bot. 55: 1–10.

Timm, S., and Bauwe, H. (2013). The variety of photorespiratoryphenotypes - Employing the current status for future researchdirections on photorespiration. Plant Biol. (Stuttg.) 15: 737–747.

Timm, S., Florian, A., Jahnke, K., Nunes-Nesi, A., Fernie, A.R., andBauwe, H. (2011). The hydroxypyruvate-reducing system inArabidopsis: Multiple enzymes for the same end. Plant Physiol. 155:694–705.

Timm, S., Mielewczik, M., Florian, A., Frankenbach, S., Dreissen,A., Hocken, N., Fernie, A.R., Walter, A., and Bauwe, H. (2012).

Interaction of Endoplasmic Reticulum and Photorespiration 2659

High-to-low CO2 acclimation reveals plasticity of the photorespiratorypathway and indicates regulatory links to cellular metabolism of Arabidopsis.PLoS ONE 7: e42809.

Tolbert, N.E. (1997). The C2 oxidative photosynthetic carbon cycle.Annu. Rev. Plant Physiol. Plant Mol. Biol. 48: 1–25.

Trentmann, O., Jung, B., Neuhaus, H.E., and Haferkamp, I. (2008).Non-mitochondrial ATP/ADP transporters accept phosphate asthird substrate. J. Biol. Chem. 283: 36486–36493.

Voll, L.M., Jamai, A., Renne, P., Voll, H., McClung, C.R., andWeber, A.P. (2006). The photorespiratory Arabidopsis shm1 mutantis deficient in SHM1. Plant Physiol. 140: 59–66.

Weigel, D., and Glazebrook, J. 2002. Arabidopsis: A Laboratory Manual.(Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press).

Willekens, H., Langebartels, C., Tire, C., Van, M.M., Inze, D., and Van,C.W. (1994). Differential expression of catalase genes in Nicotianaplumbaginifolia (L.). Proc. Natl. Acad. Sci. USA 91: 10450–10454.

2660 The Plant Cell

DOI 10.1105/tpc.113.113605; originally published online July 16, 2013; 2013;25;2647-2660Plant Cell

Bauwe, Jan Riemer, Johannes M. Herrmann and H. Ekkehard NeuhausChristiane Hoffmann, Bartolome Plocharski, Ilka Haferkamp, Michaela Leroch, Ralph Ewald, Hermann

Endoplasmic Reticulum Adenylate Transporter1 Perturbs Photorespiration ATP AntiporterArabidopsisFrom Endoplasmic Reticulum to Mitochondria: Absence of the

 This information is current as of February 16, 2021

 

Supplemental Data /content/suppl/2013/07/08/tpc.113.113605.DC1.html

References /content/25/7/2647.full.html#ref-list-1

This article cites 64 articles, 29 of which can be accessed free at:

Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists