Complex I Dysfunction Redirects Cellular and Mitochondrial Metabolism in Arabidopsis

Complex I Dysfunction Redirects Cellular andMitochondrial Metabolism in Arabidopsis1[W][OA]

Marie Garmier2, Adam J. Carroll, Etienne Delannoy, Corinne Vallet, David A. Day,Ian D. Small, and A. Harvey Millar*

Australian Research Council Centre of Excellence in Plant Energy Biology M316 (M.G., A.J.C., E.D., C.V.,I.D.S., A.H.M.) and School of Biomedical, Biomolecular, and Chemical Sciences M310 (A.J.C.), University ofWestern Australia, Crawley, Western Australia 6009, Australia; and Australian Research Council Centre ofExcellence in Plant Energy Biology, School of Biological Sciences, University of Sydney, New South Wales2006, Australia (D.A.D.)

Mitochondrial complex I is a major avenue for reduced NAD oxidation linked to oxidative phosphorylation in plants.However, the plant enzyme has structural and functional features that set it apart from its counterparts in other organisms,raising questions about the physiological significance of this complex in plants. We have developed an experimental model inwhich rotenone, a classic complex I inhibitor, has been applied to Arabidopsis (Arabidopsis thaliana) cell suspension cultures inorder to dissect early metabolic adjustments involved in cell acclimation to mitochondrial dysfunction. Rotenone induced atransitory decrease in cellular respiration (0–4 h after treatment). Cell respiration then progressively recovered and reached asteady state at 10 to 12 h after treatment. Complex I inhibition by rotenone did not induce obvious oxidative stress or cell deathbut affected longer term cell growth. Integrated analyses of gene expression, the mitochondrial proteome, and changes inprimary metabolism indicated that rotenone treatment caused changes in mitochondrial function via alterations in specificcomponents. A physical disengagement of glycolytic activities associated with the mitochondrial outer membrane wasobserved, and the tricarboxylic acid cycle was altered. Amino acid and organic acid pools were also modified by rotenonetreatment, with a marked early decrease of 2-oxoglutarate, aspartate, and glutamine pools. These data demonstrate that, inArabidopsis cells, complex I inhibition by rotenone induces significant remodeling of metabolic pathways involving themitochondria and other compartments and point to early metabolic changes in response to mitochondrial dysfunction.

Complex I (NADH:ubiquinone oxidoreductase; EC1.6.5.3) is a major entry point into the mitochondrialelectron transport chain (ETC) of reductant generatedwithin the mitochondrial matrix. Concomitant protontranslocation by complex I is coupled to mitochondrialoxidative phosphorylation, generating ATP. ComplexI is composed of over 40 different subunits in eukary-otes with a native mass of 800 to 1,000 kD (Brandt

et al., 2003). It has a large hydrophobic arm integral tothe inner mitochondrial membrane that contains mostof the mitochondrially encoded subunits and a hy-drophilic arm that protrudes into the matrix andcontains the NADH-binding site and most of the Fe-Sclusters.

As complex I represents the sole entry point ofelectrons from matrix NADH to the electron transportchain in mammals, mutation, damage, or inhibition ofthe complex has a profound effect on cellular energetics(DiMauro and Schon, 2003). Mutations in complex Isubunits are linked to a variety of serious diseases,such as Leber’s heredity optic neuropathy and leuko-dystrophy, and complete absence of complex I is con-sidered lethal. Rotenone, the most potent member ofthe rotenoids, a family of isoflavonoids produced byLeguminosae plants, has become a classic inhibitorof complex I and a valuable tool to mimic complexI-associated disorders in mammalian systems (Ayalaet al., 2007). Its mode of action has been extensivelystudied in mammals, including the impact of rotenoneinhibition on mitochondria due to reverse electronflow through the ubiquinone pool from succinate dehy-drogenase, leading to superoxide production fromcomplex I and associated oxidative damage to cellcomponents (Kussmaul and Hirst, 2006). Recent stud-ies have shown that chronic exposure of rodents torotenone causes Parkinson-like degenerative syndrome,

1 This work was supported by the Australian Research Council(grant no. CE0561495 to A.H.M., I.D.S., and D.A.D.) through theCentre of Excellence program. M.G. was the recipient of an AustralianResearch Council Linkage International fellowship (no. LX0560236),A.J.C. was the recipient of a Grains Research and DevelopmentCorporation Postgraduate Award, I.D.S. was supported as a WesternAustralia Premier’s Fellow, and A.H.M. is an Australian ResearchCouncil Australian Professorial Fellow (award no. DP0771156).

2 Present address: Institut de Biotechnologie des Plantes, Uni-versite Paris-Sud 11, CNRS, UMR 8618, Batiment 630, 91405 Orsaycedex, France.

* Corresponding author; e-mail [email protected] author responsible for distribution of materials integral to the

findings presented in this article in accordance with the policydescribed in the Instructions for Authors (www.plantphysiol.org) is:A. Harvey Millar ([email protected]).

[W] The online version of this article contains Web-only data.[OA] Open Access articles can be viewed online without a sub-

scription.www.plantphysiol.org/cgi/doi/10.1104/pp.108.125880

1324 Plant Physiology, November 2008, Vol. 148, pp. 1324–1341, www.plantphysiol.org � 2008 American Society of Plant Biologists www.plant.org on January 2, 2016 - Published by www.plantphysiol.orgDownloaded from

Copyright © 2008 American Society of Plant Biologists. All rights reserved.

http://www.plantphysiol.org/

http://www.plant.org

suggesting a link between the disease and complex Ifunction (Betarbet et al., 2000).

Key differences exist between animal and plantcomplex I, as the role of the complex in plants iscomplicated by a series of specific peripheral enzy-matic associations and the presence of complexI respiratory bypasses. For instance, the final enzymeof ascorbate synthesis in plants, converting galactolac-tone to ascorbate, is physically associated with com-plex I (Heazlewood et al., 2003; Millar et al., 2003), asare a series of carbonic anhydrase-like proteins and arange of plant-specific proteins of unknown function(Heazlewood et al., 2003, Perales et al., 2005, Meyeret al., 2007). Complex I is also part of a supercomplexwith complex III, in CI:CIII(2) and CI(2):CIII(4) con-figurations (Eubel et al., 2004; Dudkina et al., 2005).Plants contain bypasses of complex I in the form oftype II nonproton pumping NADH and NADPHdehydrogenases, located on both sides of the innermembrane, that can deliver electrons to ubiquinonefrom the matrix or from the intermembrane spaceNAD(P)H pool (Rasmusson et al., 2004). These type IINAD(P)H dehydrogenases are rotenone insensitiveand, consequently, plant respiration can continue inthe presence of the poison (Roberts et al., 1995; Meloet al., 1996), but they do not translocate protons and donot generate ATP.

Plants containing complex I mutations exist and areviable, presumably due to the presence and activationof the type II bypass dehydrogenases that allow alter-native means of oxidizing the matrix NADH pool,albeit with a lowered efficiency of coupled oxidativephosphorylation. The best studied case has been theCMSII (for cytoplasmic male sterile) mutant of tobacco(Nicotiana sylvestris), which harbors a deletion in themitochondrial nad7 gene encoding the NAD7 subunitof complex I (Gutierres et al., 1997; Pineau et al., 2005).CMSII plants lack rotenone-sensitive complex I activ-ity and structure and show a stable inherited pheno-type involving delayed germination and development,light-dependent cytoplasmic male sterility, decreasedphotosynthetic efficiency, modified light acclimationresponses, and altered organic acid and amino acidpools and antioxidant defenses (Sabar et al., 2000;Dutilleul et al., 2003a, 2003b, 2005). Interestingly, thismutant also shows enhanced expression of the cyanide-insensitive alternative oxidase (AOX) and enhancedresistance to ozone damage and tobacco mosaic virusinfection, suggesting a complicated link between mi-tochondrial metabolism, cellular redox regulation, andplant stress tolerance mechanisms (Dutilleul et al.,2003b; Vidal et al., 2007).

However, it is extremely difficult to use the com-parison of wild-type and mutant phenotypes to un-derstand the short-term events resulting from changesin complex I activity in plants. Mutants and wild typesrepresent distinct steady states separated by an un-known series of events and exhibit pleiotropic effectsthat do not necessarily portray direct roles of complexI or specific responses to complex I loss. Study of the

early events associated with loss of complex I activitynot only presents an opportunity to understand thetransition from one metabolic state to another but alsoto uncover elements in the pathway of mitochondrionto nucleus signaling that must occur to initiate theseevents. In this study, we have used rotenone to inhibitcomplex I function in an Arabidopsis (Arabidopsisthaliana) cell suspension and have followed cell re-sponses. By combining proteomic, transcriptomic, andmetabolomic analyses, we provide a detailed view ofhow cells manage mitochondrial dysfunction. Com-plex I inhibition by rotenone led to the induction ofalternative respiratory pathways; thus, overall cellularrespiration was maintained. Transcript, protein, andmetabolite analyses revealed complex metabolic ad-justments, with a disengagement of mitochondriafrom glycolysis.

RESULTS

Rotenone Transiently Inhibits Cell Respiration andAffects Long-Term Cell Growth

A heterotrophic cell suspension culture of Arabi-dopsis was grown in the dark until its midlog growthphase and treated with either 40 mM rotenone dis-solved in methanol (0.25%, v/v) or methanol (to0.25%, v/v) as a control. At specific time intervals,cell culture samples were frozen and kept for protein,RNA, and metabolite analyses, whereas fresh sampleswere taken for respiration, cell viability, and growthanalyses. Rotenone significantly affected cell respira-tion over the first 4 h of treatment, decreasing respi-ration rates by 25% to 45% (Fig. 1A, phase I).Respiration recovered to the control value over thefollowing 4 h, and no significant change was observedover the next 20 h (Fig. 1A, phase II). To assess thelongevity of the impact of rotenone on respiration, therespiratory rate of cells treated for 48 h with rotenonewas measured with or without a further rotenoneaddition (Supplemental Fig. S1, A and B), revealingthat the pretreated cells were resistant to furtherrotenone treatments. To determine if the cell superna-tant still contained rotenone for a substantial period,supernatant at 6, 24, and 48 h after rotenone treatmentwas harvested and used to inhibit fresh cells that hadnot been previously rotenone treated. In each case, thesupernatants inhibited respiration, and this rate couldnot be further inhibited by further rotenone additions(Supplemental Fig. S1C). The longer term impact ofrotenone treatment on cell growth and viability wasalso assessed over 7 d. Rotenone inhibited cell growthby nearly 10% after 16 h and by a further 20% after 48 hand up to 7 d (Fig. 1B). Measurement of cell viabilityby propidium iodide staining showed no significantdecrease in cell viability at 7 d after rotenone treatmentwhen compared with methanol treatment (Fig. 1C).This suggests that the cells were growing, albeit moreslowly, after rotenone treatment but that they were not

Early Metabolic Adjustments to Complex I Dysfunction

Plant Physiol. Vol. 148, 2008 1325 www.plant.org on January 2, 2016 - Published by www.plantphysiol.orgDownloaded from




senescing or undergoing programmed cell death at asignificantly higher rate than methanol-treated or un-treated cells. These conclusions were reinforced by theanalysis of the ATH1 GeneChip data from the samples,which showed significant decreases in transcripts en-coding histones (MapMan BIN 28.1.3), cell cycle andcell division proteins (BINs 31.3 and 31.2), and ribo-somal proteins (BIN 29.2.2), while no significant mod-ifications (P . 0.05) in transcript levels of BINs forstress-related genes were detected (BIN 20; Supple-mental Table S1). Thus, although whole cell respirationrapidly recovered from phase I inhibition, rotenonetreatment affected long-term cell culture growth. Res-piration as well as transcriptional and metabolic mod-ifications induced by rotenone were then furtherdissected in phases I and II of the response.

Respiration Acclimation to Rotenone (Phase II) Is

Associated with the Induction of AlternativeRespiratory Pathways

We have shown previously that transcripts for re-spiratory bypasses are induced in the first hours fol-lowing rotenone treatment of these cell cultures(Clifton et al., 2005). To confirm this, we carried outquantitative reverse transcription (RT)-PCR assays fortranscripts coding for alternative external (NDB1 andNDB2) and internal (NDA2) NAD(P)H dehydroge-nases as well as AOX1a, an isoform of AOX (Fig. 2A).AOX is present in the inner membrane of plantmitochondria and accepts electrons directly from theubiquinone pool and reduces oxygen to water, thusbypassing the cytochrome c oxidase (COX) pathwaythrough complexes III and IV. Transcript levels ofNDA2, NDB2, and AOX1a all showed a modest in-crease at 3 h and were more significantly induced at12 h after rotenone treatment, when whole cell respi-ration had recovered from phase I inhibition. Micro-array data confirmed these results and also showedan induction of UCP1 (At3g54110) and AOX1d (At1g32350)transcripts at 12 h as well as a significant induction ofthe mitochondrial alternative pathways as a functionalgrouping (see BIN 9.4 in Supplemental Table S1 andBIN 6 in Supplemental Table S2). These data confirmthat the rotenone treatment induced the expression ofcomplex I bypasses.

After 16 h, when the respiratory rate had recoveredand reached an equilibrium (Fig. 1A, phase II), wholecell respiration was further dissected using respiratoryinhibitors (Supplemental Fig. S2A). Total oxygen con-sumption rate, after addition of carbonyl cyanide

Figure 1. Treatment with rotenone caused a rapid inhibition of respi-ratory rate and a long-term lowering of cell dry weight (DW) withoutaffecting cell viability. A, Oxygen consumption of intact cells measuredat 1, 2 4, 8, 16, and 32 h after treatment with methanol (0.25% [v/v];white squares, dashed line) or 40 mM rotenone in methanol (0.25%[v/v]; black squares). Values correspond to means 6 SE of triplicatedeterminations. B and C, Cells were collected at time intervals (hours ordays) from untreated cell suspensions or suspensions treated withmethanol (0.25%, v/v) or 40 mM rotenone. B, Ratio of cell dry weight ofmethanol-treated cells (MET) versus untreated cells (white columns)and rotenone-treated (ROT) versus methanol-treated cells (gray col-umns). Values are means 6 SE from six independent experiments. Dryweight of rotenone-treated cells was significantly decreased at all timepoints compared with methanol treatment at P , 0.01. C, Cell viabilitydetermined by propidium iodide staining of untreated cells (C; white

columns), methanol-treated cells (M; gray columns), and rotenone-treated cells (R; black columns). The proportion of dead, fluorescentcells compared with the total number of cells was determined with anepifluorescence microscope, and data are expressed as percentage cellviability. At least 300 cells were scored per treatment, time point, andflask, with two aliquots per flask and two flasks analyzed per exper-iment. Data are means 6 SD from two independent experiments.

Garmier et al.

1326 Plant Physiol. Vol. 148, 2008 www.plant.org on January 2, 2016 - Published by www.plantphysiol.orgDownloaded from




m-chlorophenylhydrazone to remove adenylate control,was unchanged by rotenone treatment when ex-pressed on a cell dry weight basis. Addition of 1 mM

n-propyl gallate (nPG) to inhibit the AOX pathwayshowed that maximal respiration through the COXpathway was unchanged by rotenone treatment (Sup-plemental Fig. S2A). Blue-native PAGE (BN-PAGE)

separation of respiratory complexes and detection ofNADH dehydrogenase activity revealed that complexI was present and active in mitochondria isolated fromrotenone-treated cells (Supplemental Fig. S2B), consis-tent with the evidence that rotenone is a reversibleinhibitor and is lost during mitochondrial isolation(Singer, 1979). These gels also showed that the other

Figure 2. Up-regulation of respiratory bypasses of complex I during cell respiratory acclimation to rotenone treatment. A,Quantitative RT-PCR expression profiles of genes encoding external (NDB1 [At4g28220] and NDB2 [At4g05020]) and internal(NDA2 [At2g29990]) alternative NAD(P)H dehydrogenases and an isoform of AOX (AOX1a [At3g22370]). Arabidopsis cellsuspension cultures were treated with methanol (0.25% [v/v]; white columns) or 40 mM rotenone in methanol (0.25% [v/v]; blackcolumns). At time intervals (0, 3, and 12 h after treatment), cells were collected and transcript levels were quantified. Foldinduction in transcript levels was calculated as ratio of the transcript abundance of the studied gene to the transcript abundanceof actin2 (At3g18780). Asterisks indicate fold induction values that are significantly different (P , 0.01) in rotenone-treatedsamples compared with methanol samples. Error bars correspond to the SE of biological triplicate determinations. B, Oxygenuptake measurements of mitochondria isolated from cell suspensions at 16 h after treatment with methanol (white columns) or40 mM rotenone (black columns). Respiration rates were recorded at 25�C on 200 mg of mitochondrial protein with a Clarkoxygen electrode as described in ‘‘Materials and Methods.’’ Added substrates, cofactors, and inhibitors are indicated on the xaxis. Values are means 6 SE from at least five independent experiments. Statistically different values (P , 0.01) between methanoland rotenone treatments are indicated by asterisks. C, Left, Immunoblots of AOX proteins in mitochondria purified from cells at16 h after treatment with methanol (C) or 40 mM rotenone (Rot). Equal amounts (50 mg) of mitochondrial proteins were loaded ineach lane. Two independent sets of mitochondrial protein isolates as shown were loaded onto the same SDS-PAGE gel underreducing conditions (with DTT). Right, Immunoblots of AOX proteins in rapidly extracted membrane fractions from cellsuspension cultures at 16 h after treatment with methanol (C) or rotenone (Rot). Cells were ground in an extraction buffer (2),supplemented with the reducing agent DTT or the oxidizing agent diamide (Dia). Replicate gels were stained with CoomassieBrilliant Blue in order to check protein loading. Data shown are representative of three independent experiments.






classic mitochondrial OxPhos complexes were presentand had similar total protein stain intensity in bothcontrol and rotenone-treated samples. This is consistentwith the fact that there were scarcely any transcriptionalresponses to rotenone treatment by any of the mito-chondrial (Supplemental Fig. S3) or nuclear genes en-coding components of the OxPhos complexes (Table I).

Addition of 1 mM KCN to whole cells to inhibit theCOX pathway revealed that the maximal respiratoryrate via the AOX pathway had doubled during rote-none treatment. Succinate-dependent respiratory ratesof mitochondria isolated from the rotenone-treatedand methanol control cells mirrored this selectiveinduction of AOX by rotenone (Supplemental Fig.S2C). Using malate and Glu to provide substratesfor complex I and internal NADH dehydrogenasesthrough the tricarboxylic acid (TCA) cycle, we ob-served no change in the respiratory capacity and nochange in the degree of rotenone sensitivity (Supple-mental Fig. S2D), consistent with the relatively revers-ible nature of rotenone inhibition in vivo and itsremoval during mitochondrial isolation. Rates of ex-ternal NADH oxidation by mitochondria isolated fromrotenone-treated cells were not significantly differentfrom those in control cell mitochondria, but again theyshowed that AOX activity was greatly enhanced byrotenone treatment (Fig. 2B). Western blots of isolatedmitochondria using antibodies to AOX confirmed thatAOX protein accumulated in mitochondria after rote-none treatment (Fig. 2C), consistent with the increasedcapacity of the alternative pathway. The in vivo redoxstate of AOX was investigated by performing rapidtotal membrane isolations from control and rotenone-treated cells according to Noguchi et al. (2005). AOX

consists of a homodimer of 35-kD monomers cova-lently linked by a disulfide bridge. Cell membraneswere quickly isolated in native conditions or in reduc-ing conditions (extraction buffer supplemented withdithiothreitol [DTT]) or in oxidizing conditions (buffersupplemented with diamide). AOX proteins wereimmunodetected after separation on nonreducinggels. Increased levels of AOX protein were seen onthe total membrane western blots after rotenone treat-ment, and these were predominantly in the reduced,active form (Fig. 2C). No obvious oxidation of AOXproteins was observed in the cells, indicating that notonly was AOX protein induced by rotenone treatmentbut that it was likely to be in the fully activatable form.

Altogether, these data confirm that rotenone treat-ment induced an up-regulation of complex I respi-ratory bypasses [internal and external NAD(P)Hdehydrogenases] and the AOX pathway, but withoutaffecting the cytochrome pathway. Transcriptionalchanges for these components were significant at 12 hafter treatment, suggesting that alternative pathwayinduction is involved in cell respiratory acclimation torotenone during phase II (Fig. 1A).

Rotenone Affects Ascorbate Abundance But Not ItsRedox Poise and Does Not Affect the Glutathione Pool

Complexes I and III are primary sites of mitochon-drial reactive oxygen species (mtROS) formation. InArabidopsis cells, inhibition of complex III by anti-mycin A affected the redox status of mitochondria andled to a global oxidative stress (Sweetlove et al., 2002).In contrast, in the CMSII tobacco complex I mutant,loss of complex I function is not associated with

Table I. Changes in expression of nucleus-encoded genes for mitochondrial proteins following 3 and12 h of rotenone treatment

A set of 556 transcripts for nucleus-encoded mitochondrial proteins were extracted from ATH1microarray data and placed in 18 functional groups. Significant changes in the expression of each groupare reported based on Benjamini-Hochberg correction of P values calculated by the Wilcoxon rank sumtest. NS, P . 0.05. Details of genes in each category are given in Supplemental Table S2.

Category3 h 12 h

P Regulation P Regulation

Complex I NS – NS –Complex II NS – NS –Complex III NS – NS –Complex IV NS – NS –ATP synthase NS – NS –Alternative pathways NS – 5.83E-05 UpProtein import and fate NS – 3.33E-05 UpHeat shock proteins and chaperonin NS – 3.33E-05 UpStress-responsive proteins NS – NS –TCA cycle NS – NS –General metabolism 6.89E-04 Up/down 3.30E-07 UpTranslation 0.01836 Down NS –Signaling and structure NS – 3.30E-07 UpCarrier and transporter NS – 0.028826 Up/downDNA replication and transcription NS – 1.46E-04 UpPhotorespiration and C1 metabolism 2.41E-04 Down NS –Unknown proteins NS – NS –

Garmier et al.





elevated global ROS levels and oxidative stress. In thisplant, a retrograde signal, possibly involving mtROS,between mitochondria and the nucleus triggers in-creased antioxidant activities and maintains whole cellredox balance (Dutilleul et al., 2003b). To assess thestress status of Arabidopsis cells following rotenonetreatment, the content and redox state of two majorsoluble antioxidants, ascorbate and glutathione, wereanalyzed. Ascorbate (total and reduced forms) andglutathione (total and oxidized forms) were deter-mined using spectrophotometric methods at 1, 4, 8, 16,and 48 h after inhibitor treatments (Fig. 3). Totalascorbate (oxidized and reduced) level was signifi-cantly lowered (by 24%) during the first hour ofrotenone treatment, and the ascorbate content contin-ued to decrease over 16 h to less than half of the controllevel but recovered by 48 h to reach a similar level inboth inhibitor-treated and mock-treated samples. Incontrast, rotenone had little effect on the redox poise ofthe ascorbate pool. The glutathione pool, on the otherhand, changed little in the first 24 h of rotenonetreatment. Over the following 48 h, total glutathioneincreased significantly, mainly due to accumulation ofthe reduced form (Fig. 3). Taken together, these resultssuggest that the rotenone-treated cells were not underoxidative stress, and this was broadly supported bythe absence of transcriptional changes for classicalantioxidant machinery (Supplemental Table S1). How-ever, there does appear to be a link between ascorbatecontent and rotenone treatment. We have reportedpreviously that rotenone inhibits ascorbate synthesisin isolated plant mitochondria (Millar et al., 2003),presumably because galactonolactone dehydrogenaseis physically associated with complex I.

Rotenone Induces Changes inMitochondrial Biogenesis

In order to study the effect of complex I inhibition onother mitochondrial functions, we performed a quan-titative comparison of mitochondrial protein samplesisolated from control and rotenone-treated cells usingDIGE CyDye fluorophore technology (Fig. 4). Spotsthat increased or decreased more than 1.5-fold repro-ducibly over three independent experiments wereselected and analyzed using liquid chromatography-tandem mass spectrometry, and the identities of theproteins are summarized in Table II. We also looked atthe responses of transcripts for a broad set of 556nucleus-encoded components of the mitochondrialproteome to 3 and 12 h of rotenone treatment (TableI; Supplemental Table S2).

Two mitochondrial proteins increased in abundancefollowing rotenone treatment: formate dehydrogenaseand arginase (Table II). Three well-known mitochon-drial proteins decreased in abundance following rote-none treatment: two belong to the respiratory chain(matrix processing peptidase [MPP] b-subunit andSDH6) and one is part of the translational apparatus(elongation factor Tu). SDH6 is a plant-specific isoform

of a subunit of respiratory complex II, but a catalyticrole has not been identified for this protein (Millaret al., 2004) and there was no change in succinate-dependent respiration of isolated mitochondria (Sup-plemental Fig. 2B), indicating that no clear alteration inSDH function occurred. The MPP b-subunit proteinspot that decreased more than 2-fold following rote-

Figure 3. Redox status of cells in ascorbate and glutathione contentover 48 h after rotenone treatment. White columns, Control treatment(methanol, 0.25% [v/v]); black columns, treatment with 40 mM rote-none in methanol (0.25% [v/v]). Values are means 6 SE from threeindependent experiments. A, Total ascorbate (Asc) content expressed asnmol ascorbate g21 cell fresh weight (FW). The decrease in ascorbatecontent between rotenone-treated cells and control cells was signifi-cant at P , 0.01 at 1, 8, and 16 h after treatment. B, There was nosignificant change in the reduced to oxidized ascorbate ratio. C and D,There was no significant change in the total glutathione (GSH) content(C), but the reduced to oxidized glutathione ratio (D) was significantlyincreased (P , 0.01) in rotenone-treated cells compared with controlcells at 4 and 48 h. Dashed lines on the columns indicate levels of theoxidized forms of ascorbate and glutathione.






none treatment is the main spot for this protein onisoelectric focusing (IEF)-SDS-PAGE gels (Millar et al.,2001). Our previous studies have shown that Arabi-dopsis cell cultures have a lowered in vitro importcapacity following rotenone treatment (Lister et al.,2004), that in pea (Pisum sativum) plants import effi-ciency into mitochondria is decreased by a variety ofabiotic environmental stresses that also cause a de-crease in abundance of the MPP b-subunit (Tayloret al., 2003, 2005), and that MPP has a high controlcoefficient in the control of the rate of import in rice(Oryza sativa) during mitochondrial biogenesis (Howellet al., 2007). Thus, this loss of MPP b-subunit may bean important contributor to the lowered import ratesafter rotenone treatments (Lister et al., 2004).

The microarray analysis showed that none of 556known mitochondrial components changed transcriptabundance significantly at 3 h (P . 0.05 after falsediscovery rate correction); however, by 12 h, 98 of the556 had changed, 89 by increasing in abundance(Supplemental Table S2). Analysis of these changesin functional BINs showed a significant induction ofcomponents of the alternative pathway, protein importand fate, signaling and structure, general metabolism,heat shock proteins, mitochondrial DNA replicationand transcription, and MAM33 glycoproteins after 12 h.

Transcripts of genes encoding proteins involved intranslation and C1 metabolism, on the other hand,were slightly down at 3 h (Table I).

Rotenone Treatment Induces a Major Shift in CellularEnergy Metabolism

Seven other protein spots that decreased in abun-dance by 1.5- to 5.5-fold in isolated mitochondriafollowing rotenone treatment (Fig. 4; Table II) wereidentified as major cytosolic enzymes of primarycarbon metabolism. Notably among these were iso-forms of aldolase, glyceraldehyde-3-P dehydrogenase(GAPDH), enolase, and triosephosphate isomerase, allof which have been shown to selectively bind to plantmitochondria as a functional glycolytic unit on theouter membrane (Giege et al., 2003). These enzymesare postulated to function as a membrane-boundmetabolon, feeding glycolytic products to mitochon-dria. Our measurements of GAPDH and aldolaseactivities in mitochondrial fractions isolated from therotenone-treated cells also showed significant de-creases in their activity (Fig. 5). To determine if thechanges observed simply reflected decreases in thetotal cytosolic pool of these enzymes, we comparedactivity assays for GADPH and aldolase in whole cell

Figure 4. Differential gel electrophoresis of changesin the mitochondrial proteome after 16 h of rotenonetreatment. Quantitative comparison of mitochondrialsamples isolated from control cells (methanol treat-ment, 0.25% [v/v]) or 40 mM rotenone-treated cellswith differential 2D-DIGE IEF/SDS-PAGE CyDyetechnology. Proteins from each treatment (50 mg)were bound to different fluorescent dyes, Cy2 (redfluorescence) for control samples and Cy5 (greenfluorescence) for rotenone samples, and run througha standard procedure of IEF (pI 3-10NL)/SDS-PAGE.Gels were scanned using the Typhoon Trio VariableMode Imager. Protein spots that reproducibly changedin abundance following rotenone treatment wereselected for further analysis (numbered arrows atbottom). Top, Black and white images generated bythe Typhoon Imager. Bottom, Superimposition of gelsscanned with the fluorescence mode. Yellow spots,Both dyes present in equal abundance; green spots, up-regulated spots; red spots, down-regulated spots.Three independent sets of mitochondrial isolatesand a set of three gels were run and analyzed usingDyCyder quantitation software. Protein identificationwas performed from Coomassie Brilliant Blue-stainedgels loaded with a mix of 150 mg of proteins for eachtreatment. Molecular weight scales are indicated atleft of each image.

Garmier et al.





Tab

leII

.Pro

tein

sfr

om

mit

och

ondri

alsa

mple

sfo

und

tova

ryin

abundan

cefo

llow

ing

16

hof

rote

none

trea

tmen

t

The

iden

tity

ofpro

tein

sw

asdet

erm

ined

by

MS/

MS.

the

pre

dic

ted

mole

cula

rw

eigh

t(M

W)an

dpIofth

em

atch

edpro

tein

and

the

gelsa

mple

are

show

nal

ong

wit

hth

eM

OW

SEsc

ore

,num

ber

of

pep

tides

mat

ched

by

tandem

mas

ssp

ectr

a,an

dth

eper

centa

geco

vera

geofth

em

atch

edse

quen

ce.Pro

tein

spots

wit

hat

leas

t1.5

-fold

chan

gear

epre

sente

das

incr

ease

d(1

)or

dec

reas

ed(2

)by

rote

none

trea

tmen

t.Sp

otn

um

ber

ssh

ow

nco

rres

pond

topro

tein

spots

inFi

gure

4.N

o.E

xp,N

um

ber

ofe

xper

imen

tsin

whic

hth

ispro

tein

was

det

ecte

dfr

om

tota

loft

hre

eex

per

imen

tsper

form

ed.X

,The

pro

tein

has

not

bee

ndet

ecte

din

this

exper

imen

t.

Cat

egory

Spot

No.

Ara

bid

opsi

s

Gen

om

e

Init

iati

veN

o.

Iden

tity

Score

Per

centa

ge

Cove

rage

No.

of

Pepti

des

Mat

ched

MW

Expec

ted

MW

Mat

ched

pI

Expec

ted

pI

No.

Exp

(DIG

E)

Rat

io

Rote

none/

Contr

ol

Rat

io

Exper

imen

t

1

Rat

io

Exper

imen

t

2

Rat

io

Exper

imen

t

3

Dec

reas

ed

Know

n

mit

och

ondri

al

1A

t3g0

2090.1

Mit

och

ondri

al

pro

cess

ing

pep

tidas

eb

-subunit

(MPP)

796,

819

47,

53

25,

29

59,1

23

67,0

00

6.3

6.3

32

2.6

22.1

12

2.7

32

3.0

3

2A

t4g0

2930.1

Elonga

tion

fact

or

Tu645,

758,

835

42,

51,

50

24,

29,

25

49,3

79

50,0

00

6.2

56

32

2.1

21.6

22.8

42

1.7

9

3A

t1g0

8480.1

Succ

inat

e

deh

ydro

genas

e

subunit

6(S

DH

6)

151

26

615,8

03

15,0

00

6.2

88.2

32

1.9

23.2

22

1.3

52

1.2

4A

t5g1

4590.1

IDH

P-3

isoci

trat

e

deh

ydro

genas

e

1,1

51,

1,2

75

59,

64

40,

45

54,1

62

55,0

00

8.0

56.4

32

1.5

21.6

72

1.4

62

1.4

7

5A

t2g0

5710.1

Aco

nit

ate

hyd

rata

se197,

386,

118

6,

12,3

8,

15,

3108,1

33

85,0

00

6.7

26.8

22

2.2

22.5

92

1.8

4X

Gly

coly

sis

and

cyto

soli

c

carb

on

met

aboli

sm

6A

t3g5

2930.1

Fru-b

isP

aldola

se894,

1,0

01

64,

70

33,

35

38,5

16

45,0

00

6.0

56.6

32

3.6

23.5

52

42

3.1

7

7A

t3g0

4120.1

Gly

cera

ldeh

yde-

3-P

deh

ydro

genas

e

(GA

PC

)

674,

561

65,

55

31,

25

36,8

91

45,0

00

6.6

27.2

32

2.7

23.3

52

2.7

62

1.9

1

8A

t2g3

6530.1

Enola

se642,

558,

472

48,

43,

40

22,

25,

19

47,6

89

65,0

00

5.5

45.5

32

1.7

21.4

62

2.0

22

1.6

9A

t3g5

5440.1

Trio

sephosp

hat

e

isom

eras

e

510,

574

42,

53

16,

19

27,1

52

29,0

00

5.3

95.3

32

2.3

22.9

42

2.6

52

1.2

10

At1

g04410.1

MD

H-1

mal

ate

deh

ydro

genas

e

510,

412

44,

38

19,

14

35,5

48

45,0

00

6.1

16.5

32

3.5

24.3

22

2.7

12

3.5

5

11

At5

g11670.1

MO

X-5

ME

(AtN

AD

P-M

E2)

269

13

864,3

72

70,0

00

6.0

16.3

22

3.4

24

22.8

5X

12

At1

g65930.1

IDH

P-2

isoci

trat

e

deh

ydro

genas

e

640,

793,

1,1

28

50,

48,

59

23,

29,

36

45,7

17

60,0

00

6.1

36.8

32

3.7

25.5

72

3.4

22.1

4

Oth

ers

13

At5

g09810.1

Act

in7/a

ctin

2(A

CT7/2

)479,

654

44,

54

18,

27

41,7

09

55,0

00

5.3

15

32

2.0

22.5

62

2.0

52

1.4

14

At1

g62740.1

Stre

ss-i

nduci

ble

pro

tein

417,

817

35,

49

18,

32

64,4

79

80,0

00

5.8

56.1

32

3.2

24.2

23.3

62

2.1

6

15

At3

g09440.1

Hea

tsh

ock

70-k

D

pro

tein

3(H

SC70-3

)

214,

246

16,

15

9,

10

71,1

03

80,0

00

4.9

74.4

32

3.8

24.5

24.3

72

2.6

2

16

At3

g17820.1

Gln

synth

etas

e(G

S1)

120,

288

12,

18

6,

12

38,5

70

50,0

00

5.7

25.5

32

2.9

22.2

92

3.9

12

2.5

5

17

At1

g62380.1

1-A

min

ocy

clopro

pan

e-

1-c

arboxy

late

oxi

das

e

252,

227

29,

22

9,

836,1

60

45,0

00

4.9

84.4

22

2.9

23.0

32

2.8

5X

18

At3

g52880.1

Monodeh

ydro

asco

rbat

e

reduct

ase

234,

379

27,

32

11,

16

46,4

58

55,0

00

6.4

16.8

32

4.6

24.1

72

5.5

72

4.0

8

Incr

ease

d

Know

n

mit

och

ondri

al

19

At4

g08900.1

Arg

inas

e392,

443

28,

32

12,

13

37,3

21

50,0

00

6.1

15.7

31

2.3

12.8

91

2.1

51

1.8

1

20

At5

g14780.1

Form

ate

deh

ydro

genas

e

644,

604,

682

48,

43,

49

29,

25,

30

42,3

83

50,0

00

7.1

26.6

31

1.8

11.7

61

1.8

61

1.6

7






extracts with those in isolated mitochondrial samples(Fig. 5). This showed that the loss of these enzymeactivities in mitochondrial samples following rotenonetreatment occurred against a backdrop of net increasesin both enzyme activities in total cellular extracts; thatis, it was their association with mitochondria thatchanged. Transcripts for the enzymes of the glycolyticpathway also increased (Supplemental Table S1), con-sistent with these increased whole cell enzymaticactivities (Fig. 5).

Three other cytosolic carbon metabolism enzymeswere also decreased in mitochondrial extracts: NAD-malate dehydrogenase, NADP-malic enzyme (ME),and NADP-isocitrate dehydrogenase (ICDH; Table II);all three of these are directly linked to the import andexport of carbon skeletons in mitochondria. Cytosolicmalate dehydrogenase and NAD-ME are involved inthe connections between malate, OAA, and pyruvateprovision to mitochondria and thus can be consideredas extensions of the glycolytic pathway, while cytosolicICDH is involved in converting isocitrate exportedfrom mitochondria to 2-oxoglutarate as a carbon skel-eton for nitrogen assimilation. By analogy with theclaims made for the functional association of classicalglycolytic enzymes with mitochondria, these proteinsmay also be associated in a functional manner, con-necting mitochondrial metabolism with the broadercellular metabolism. The apparent dissolution of theselinkages may indicate a degree of disengagement ofmitochondria from cytosolic carbon metabolism.

Quantitative profiles of the abundance of the majormetabolites from rotenone-treated cells over the first 24 hof treatment were generated using gas chromatography-mass spectrometry (GC-MS) analysis of derivatizedcompounds from methanol-soluble cell extracts. A seriesof 48 significant changes were recorded in response tothe treatment in a time-dependent fashion. These in-cluded 18 metabolites that significantly increased and30 metabolites that significantly decreased in abun-dance during at least one point of the time series. InFigure 6, a heat map of the abundance of major com-pounds unambiguously identified by comparison withlibrary standards and classified into amino acids, or-ganic acids, sugars, and other compounds is presented.Within each grouping, the metabolites are orderedaccording to their degree of change at the 16-h timepoint, where most of the measurements reported inFigures 2 to 5 were made.

The first significant changes recorded were theapproximate halving of 2-oxoglutarate concentrationand the approximate 2.5-fold increases in Trp and Lyswithin 1 h of rotenone treatment. Interestingly, thedecrease in 2-oxoglutarate was sustained for at least 16h, while Trp and Lys returned to control levels within 3and 6 h, respectively. These changes were rapidlyfollowed by decreases in Gln, Glu, Asp, and Asn, allsignificantly reduced within 3 to 6 h. These rapidchanges are consistent with a slowing in TCA cyclefunction. At 12 to 16 h, there was evidence of decreasesin a variety of TCA cycle intermediates from thedecarboxylating portion of the cycle, namely citrate,aconitate, isocitrate, and of course 2-oxoglutarate,while there were trends to increases in abundancefor fumarate and malate in the other side of the cycle.The activity of TCA cycle enzymes, on the other hand,showed either no decrease or a slight increase (TableIII), suggesting that the slowing of the TCA cycle couldbe caused by an increase in the redox poise of thematrix NADH pool upon inhibition of complex I byrotenone. Pyruvate, 2-oxoglutarate, isocitrate, and ma-late dehydrogenases are all sensitive to the redox poiseof the NADH pool and dramatically decrease theiractivity as the NADH to NAD ratio increases (forreview, see Noctor et al., 2007). Such changes willdecrease the availability of 2-oxoglutarate for trans-amination to Glu and its related amino acids and ofoxaloacetate (OAA) for transamination to the Asp-derived amino acids (Fig. 7). As we are unable toreliably measure OAA by this method, it is not pos-sible to be sure whether the Asp changes were due tochanges in OAA availability or a greater demand forAsp (see below). As time progressed, these changesand the metabolic slowing of the TCA cycle appear tohave had a variety of effects on other metabolite pools(Figs. 6 and 7), and some of the later ones can be linkedto transcriptional changes observed in our microarrayanalysis.

Accumulation of carbon intermediates at the end ofthe glycolytic pathway led to the accumulation of Ala(first recorded at 6 h) and the pyruvate-derived

Figure 5. Glycolytic enzyme activities in whole cells versus mitochon-dria after rotenone treatment. Measurements of GAPDH and Fru-1,6-bisP aldolase activities on isolated mitochondria and whole cellextracts were taken at 16 h after methanol (0.25% [v/v]; white columns)or 40 mM rotenone (black columns) treatments. Values are expressed asnmol NADH min21 mg21 mitochondrial proteins or total proteins(whole cell extracts). Values are means 6 SE from three to fourindependent experiments. Statistical differences in rotenone treatmentcompared with control (methanol) treatment are indicated by differentletters (level of significance: a, P , 0.01; b, P , 0.05).

Garmier et al.





branched chain amino acids Leu and Val (first recordedat 12 h). By 6 h, there were already net increases in thepyruvate-derived fermentation product lactate. This isconsistent with transcriptional increases in the com-ponents of the glycolytic pathway (BINs 4.9 and 5) andbranched chain amino acid degradation pathways(BINs 13.2.4.1 and 13.2.4.4; Supplemental Table S1)and increased whole cell glycolytic enzymatic activi-ties at 16 h (Fig. 5).

Phosphoenolpyruvate-derived amino acids and com-pounds in the shikimate pathway decreased from3 h (Fig. 6) with the exception of Trp, which first in-creased at 1 h before dropping over the time course.Despite the decrease in Asp and Asn, increases werenoted in the homoserine branch of the Asp family ofamino acids, notably in Ile, Met, and 1-aminocyclo-propane carboxylic acid, while aminoadipic acid inLys catabolism decreased by half within 16 h. Losses inboth the Asp and Glu pools might be expected to alterthe biosynthetic functions of the cell, and this can beseen in the decrease in allantoin from the purinesynthesis pathway. Decreases in Glu were consistentwith decreases in Orn, and the concomitant loss of Aspappears to have lowered the urea cycle and decreasedurea abundance by 24 h. As Glu is also the precursorfor Pro and thus 4-Hyp, the decrease of the lattercompound was also likely to have stemmed from thelowering of TCA cycle activity and 2-oxoglutarateavailability (Fig. 7). Late decreases (12–24 h) were alsorecorded in hexose phosphate pools that either fed, orwere fed by, changes in several Glc-derived sugars.However, hexoses have largely reverted to controllevels by 24 h.

DISCUSSION

The aim of this work was to identify the earlyresponses of Arabidopsis cells involved in acclimationto complex I dysfunction. The long-term consequencesof complex I loss have been studied in several mutantplants, but it is difficult to unravel the initial eventsinvolved in the modification of primary metabolism inthese plants.

We showed that rotenone treatment induced a two-phase response of Arabidopsis cells: phase I, from 0 to4 h after treatment, characterized by a strong inhibition

Figure 6. Time course of metabolite measurements. Quantitative GC-MS analysis of metabolite changes at different times (0, 1, 3, 6, 12, 16,and 24 h) after treatment of cells with 40 mM rotenone. In parallel,

methanol treatment (0.25% [v/v]) was performed as a control. Data areexpressed as the ratio of rotenone to methanol signal values andstandardized according to the yield of the ribitol internal standard.Ratios that increased significantly (P , 0.05; n 5 5) in response to thestress treatment are highlighted in blue, while ratios that decreasedsignificantly (P , 0.05; n 5 5) are highlighted in red; boldface valuesindicates the fold change. Signals that showed no significant change(P . 0.05) are shown in black, while ratios that were significantlychanged (P , 0.05) by less than the 20% threshold are shown asboldface values in a black box. Color intensities are related to metab-olite response intensities, with more strongly responsive signals high-lighted in brighter tones. Three independent experiments wereperformed. Data from one representative experiment are shown (withsamples analyzed from five independently treated flasks of cells).






of cellular respiration by rotenone; and phase II, from4 to 32 h, characterized by a progressive recovery ofcell respiration to initial rates (Fig. 1A). Phase II wasassociated with transcriptional and posttranscrip-tional induction of alternative respiratory pathways,complex I bypasses [NAD(P)H dehydrogenases], andAOX, suggesting that these processes were part of cellacclimation to complex I inhibition. Up-regulation ofalternative pathways is a classic adaptive response ofplants lacking functional complex I, such as the to-bacco CMSII mutant or the Arabidopsis otp43 mutant(Sabar et al., 2000; Falcon de Longevialle et al., 2007),and has already been reported after treatment ofArabidopsis cell cultures with ETC inhibitors (Cliftonet al., 2005).

The use of chemical inhibitors raises the possibilityof pleiotrophic effects, as both mitochondria and chlo-roplasts can contain rotenone targets. We used a con-centration of 40 mM because previous experimentshave optimized this concentration for cell culturegrowth and survival (Lister et al., 2004; Clifton et al.,2005) and because at least low micromolar concentra-tions are needed for maximal complex I inhibition inplant mitochondria (Rasmusson and Møller, 1991).Reports of the impact of rotenone on chloroplastelectron transport are largely restricted to concentra-tions in the 200 mM to 1 mM range (Igambardiev et al.,1997; Corneille et al., 1998; Ikezawa et al., 2002). Thetime course data and the molecular data are consistentwith a response emanating from mitochondrial ETCinhibition.

From our data, changes in enzyme and protein werecorrelated with changes in metabolite and transcrip-tional profiles over the first 24 h of rotenone treatment.These changes included alterations of specific path-ways of mitochondrial electron transport, the source ofreductant in the matrix and cytosol, and the disen-gagement of mitochondria from glycolysis and relatedcarbon metabolism. Although most of the changes intranscript, protein, and metabolite levels were ob-

served during phase II ‘‘acclimation’’ (6–24 h), somechanges occurred in phase I (1–3 h; Fig. 6), precedingmeasurable changes in nuclear gene transcription.

Alteration of Mitochondrial Metabolism

Loss of complex I appeared to lead to a degree ofdisconnection between glycolysis, the TCA cycle, andthe electron transport chain. We saw this through aseries of cellular changes that reflect this disconnectionand that could also be interpreted as attempts toreconcile this problem through the enhancement ofalternative metabolic pathways. Even with the avail-ability of time series data, it has been relatively diffi-cult to neatly place these events in series, but it is clearthat progressive changes occur in only a few hours thatlead to a new steady state appearing within 10 to 12 h.The early decreases in 2-oxoglutarate and Gln/Glu areconsistent with slowing of the TCA cycle due to a lossof a major entry point for NADH, complex I. Thisappears to set in motion a series of changes in metab-olism that alter the carbon sources used to drive respi-ration and also increase capacity to bypass complex I asan entry point, in a complex and interrelated manner.

The loss of glycolytic enzymes from mitochondria(Table II; Fig. 5) suggests a degree of physical disen-gagement of mitochondria from glycolysis. The ob-served accumulation of glycolytic end products andassociated amino acid pools, together with decreasesin early TCA cycle-derived products, suggests that thisphysical disengagement leads to a degree of functionaldisengagement. This is consistent with the recentevidence provided by Graham et al. (2007), who ob-served that the degree of association of glycolyticenzymes with Arabidopsis mitochondria was corre-lated with respiratory rates and showed a repartition-ing of enzyme pools between mitochondria and thecytosol. Giege et al. (2003) and Graham et al. (2007)provided strong evidence that mitochondrially asso-ciated glycolytic enzymes form a functioning glyco-lytic pathway that can support mitochondrial TCAcycle flux and electron transport chain activity, prob-ably by providing pyruvate through substrate chan-neling to the mitochondria. If this disconnection ofmitochondria from a source of reducing substrate wasallowed to take its course, it would most likely havedetrimental effects for the cell. Instead, we observedthat this disconnection of glycolysis is followed bycompensation. Alternative NADH dehydrogenasesthat can use reductants from the cytosol and thematrix are up-regulated (Fig. 2A). In addition, newdehydrogenases, like formate dehydrogenase, accu-mulate in mitochondria (Table II) and the branchedchain amino acid degradation pathways are induced(Supplemental Table S1, BIN 13), providing newsources of matrix NADH independent of the TCAcycle. Enhanced branched chain amino acid catabolismprovides another bypass of complex I, as the electrontransfer flavoprotein delivers electrons from this path-way directly to ubiquinone (Ishizaki et al., 2006).

Table III. Difference in the specific activity of enzymes inmitochondria from control and rotenone-treated cells

Aconitate hydratase (aconitase), NAD-ME, NAD- and NADP-depen-dent ICDH, PDC, and OGDC activities were measured on 25 to 50 mgof mitochondrial proteins. Mitochondria were isolated from cells at16 h after treatment with methanol (0.25% [v/v]; control) or 40 mM

rotenone in methanol (0.25% [v/v]). Asterisks indicate significantchanges at P , 0.05 between mitochondrial activities from methanol-or rotenone-treated cells. Values are means 6 SE of data from fourindependent experiments.

SampleControl

Mitochondria

Rotenone-Treated

Mitochondria

nmol min21 mg21 protein

Aconitase 350 6 55 721 6 160*NAD-ME 112 6 12 77 6 8*NAD-ICDH 119 6 14 100 6 9NADP-ICDH 160 6 12 140 6 16PDC 111 6 32 82 6 23OGDC 629 6 90 558 6 107

Garmier et al.





The substrates for these compensatory pathwaysmay, in fact, be provided through the disconnection ofglycolysis from mitochondria. As pyruvate cannotenter mitochondria or is not rapidly used in mito-chondria, it can be used for the synthesis of Leu andVal, which are key substrates of the branched chaindegradation pathway. There is clear transcriptionalevidence for the induction of branched chain catabo-lism components located both within the mitochondria(Supplemental Table S2, BIN 12) and more generally inthe cell (Supplemental Table S1, BIN 13) and clearmetabolomic data for the increased availability ofthese amino acids (Figs. 6 and 7). Increased glycolyticflux disconnected from the mitochondria also leads tofermentation and lactate formation, shown by both thetranscript and metabolite analysis (Fig. 6; Supplemen-tal Table S1). This raises the potential for enhancedcytosolic NADH levels that could act as the substratefor the external rotenone-insensitive NADH dehydro-genases that are induced by rotenone treatment.

Formate oxidation generates NADH in the matrix,and formate could be provided as a consequence ofglycolytic interruption, either from a fermentationaldehyde product or via the predicted pyruvate-lyasereaction. Alternatively, it could be generated viathe Met salvage pathway/Yang cycle, as Met and1-aminocyclopropane carboxylic acid are two of themore rapidly increased metabolites and Met was theonly metabolite to significantly increase and thensignificantly decrease during the 24-h time course(Fig. 6). Interestingly, formate dehydrogenase is alsoelevated after disruption of a plant-specific subunit ofcomplex I, which leads to a substantial loss of assem-bly of the respiratory complex (Perales et al., 2005).

Retrograde Signaling duringMitochondrial Dysfunction

Details of the signal transduction pathways thatalter nuclear gene expression upon mitochondrial

Figure 7. Cartoon of metabolite changes in biochemical pathways after rotenone treatment. The data from Figure 6 were layeredonto a metabolic pathway cartoon and annotated by the timing of the first significant change in metabolite abundance afterrotenone treatment.






dysfunction remain elusive in plants. Mitochondrialdysfunction leads to changes in a large number ofmetabolites, and a single one or a combination ofseveral might be involved in the signaling process.

Redox-based signaling is often raised as a keycomponent in mitochondria-nucleus communication(Noctor et al., 2007; Rhoads and Subbaiah, 2007). Thecomplex III inhibitor, antimycin A, has been shown toinduce an increase in cellular ROS accumulation, lead-ing to oxidative stress in Arabidopsis cells (Maxwellet al., 1999), and has various regulatory effects onmitochondrial and nuclear gene expression (Sweetloveet al., 2002; Clifton et al., 2005; Vidal et al., 2007). Theinduction of AOX1a by antimycin A has long beenlinked to increased mtROS production by complex IIIinhibition (Saisho et al., 1997; Maxwell et al., 1999). Inour experiments, AOX induction by rotenone maysuggest an increase in mtROS, although it did notaffect whole cell redox status, as ascorbate and gluta-thione redox state was maintained (Fig. 3). In tobaccocomplex I mutants, AOX transcript and protein levelsare also constitutively high but are also not associatedwith elevated global ROS levels and oxidative stress(Dutilleul et al., 2003b). It is possible that changes incarbon metabolism trigger the induction of AOX un-der these circumstances, as shown for soybean (Glycinemax) suspension cells (Djajanegara et al., 2002). Thelack of oxidative stress induced by rotenone is in starkcontrast to the impact of rotenone on mammalian cellsin initiating ROS production and oxidation of the GSHpool (Betarbet et al., 2000; Kussmaul and Hirst, 2006;Ayala et al., 2007).

Interestingly, many of the components transcrip-tionally induced by rotenone belong to the machineryinvolved in mitochondrial transcription, translation,ETC organization/assembly, and protein fate (Table I;Supplemental Table S2), showing that cells respond byincreasing their capacity for mitochondrial biogenesisand ETC protection. Rotenone also induced HSP60s,HSP70s, and HSP90-related proteins, which act aschaperones and ensure the correct function of proteinsby preventing the aggregation of denatured proteins,by refolding of stress-denatured proteins, and by as-sisting in the rapid assembly of the oligomeric proteinstructures. HSP60 and HSP70 have also been reportedto be involved in Arabidopsis cell tolerance to heatstress and to respiration deficiency (Kuzmin et al.,2004; Rikhvanov et al., 2007). In the NCS2 mutant ofmaize (Zea mays), complex I impairment is also corre-lated with an increase in the expression of genesencoding a mitochondrial HSP60 and mitochondrialand cytosolic HSP70s (Kuzmin et al., 2004). It wassuggested that the decrease in ETC deficiency due toAOX induction decreased the mitochondrial trans-membrane potential, and this in turn activated signal-ing response(s), leading to increased expression of hspgenes. It was proposed that cytosolic HSP wouldcontrol the cellular redox state and the integrity ofmitochondria, while mitochondrial HSP would protectthe respiratory complexes and sustain their assembly. A

retrograde signal, other than a ROS-dependent one,was suggested to be involved in such a response.

Many stress-responsive genes are repressed duringnormal plant development, and stress-induced ex-pression could be due to the unbinding of a repressorrather than the binding of an activator molecule. Giventhis, the initial signal might also be loss of a normallypresent component rather than either the extraordi-nary accumulation of a primary metabolite or thegeneration of a specific ‘‘stress signaling’’ component.In this context, 2-oxoglutarate, Asp, and Gln poolsrapidly responded to rotenone inhibition, decreasingsignificantly by 1 h (Fig. 6) and preceding transcrip-tional responses of genes for nucleus-encoded mito-chondrial components (Fig. 2A; Table I). Consistentand prolonged elevation of other metabolic compo-nents (Fig. 6) did not occur significantly until somehours later, and in many instances this could havebeen caused by the loss of these primary metabolites.In yeast, evidence has accumulated that 2-oxoglutarateand Glu are probable signals in retrograde regulationof mitochondrial function. Glu is a potent repressor ofthe retrograde-dependent expression of genes that leadsto 2-oxoglutarate formation (notably peroxisomal andmitochondrial citrate synthase and mitochondrialaconitase and ICDH), so loss of Glu leads to enhancedexpression of these enzymes in a pathway allowingtruncated TCA cycle function during respiratory defi-ciency (Liu and Butow, 1999, 2006).

Links between the Effects of Short-Term Complex IDeficiency and the Phenotypes of Complex I Mutants

Complex I dysfunction has been reported to havevarious effects on plant growth and development.Rotenone treatment did not alter cell viability butreduced growth (Fig. 1). A similar growth reductionwas seen in suspension cell cultures generated fromArabidopsis plants with a knockout of a g-carbonicanhydrase-like complex I subunit (Perales et al., 2005).In the tobacco CMSII line, absence of complex I iscorrelated with partial male sterility, reduced germi-nation, and slow growth (Gutierres et al., 1997; Sabaret al., 2000). More severe phenotypes are reported forthe otp43 mutation (leading to a loss of mitochondrialnad1 expression) in Arabidopsis, which, while malefertile, is severely affected in seed development andplant growth rate (Falcon de Longevialle et al., 2007).In NCS2 maize plants, complex I mutation causeslethality during kernel development, and it was onlypossible to maintain them as heteroplasmic lines(Karpova et al., 2002). Our rotenone treatments con-firm that complex I deficiency is not lethal for plantcells; however, the ensuing metabolic adjustments arenot sufficient to sustain a high rate of growth and mayprove unviable during specific points in the life cycle,such as seed set and germination.

As observed after rotenone treatment, AOX and/oralternative dehydrogenase pathways are increased in

Garmier et al.





leaves of different complex I mutant plants, at the levelof gene expression, protein accumulation, or respira-tion rates (Sabar et al., 2000; Karpova et al., 2002;Falcon de Longevialle et al., 2007). In CMSII plantslacking complex I activity, leaf respiration was notdecreased in the dark when malate and pyruvate wereused as substrates, suggesting that TCA cycle activityis sustained by alternative dehydrogenase and com-plex II activities (Dutilleul et al., 2003a; Sabar et al.,2000). Leaf respiration is even slightly increased, sug-gesting that CMSII mitochondria may have increasedelectron flux through the COX pathway in order tosustain ATP production (Dutilleul et al., 2003a; Vidalet al., 2007). This is consistent with the recovery ofsteady-state respiratory rates in response to rotenoneover a 12- to 16-h period.

The metabolite profile of CMSII leaves is enriched inamino acids with low carbon to nitrogen ratios anddepleted in both starch and 2-oxoglutarate (Dutilleulet al., 2005). Interestingly, the deficiency in 2-oxoglutaratein the CMSII plants occurred despite higher citrate andmalate levels and increased NAD-ICDH capacity thanin the wild type. Oxoglutarate content was also sig-nificantly reduced following rotenone treatment,showing that it is a common response to complex Ideficiency (Figs. 6 and 7). Moreover, this change wasone of the earliest changes recorded, suggesting that itmay play a key role in metabolic adjustment and thesignaling of ETC dysfunction. The shift in abundanceto compounds with low carbon-nitrogen ratios in CMSIIcould be seen in the alteration of the 2-oxoglutarate-Glu-Gln and Asp-Asn ratios (Dutilleul et al., 2005).In this plant, a 2-fold increase in total free aminoacids has been reported, most notably in Arg, whereascarbohydrate abundance, such as Suc and Glc, wasreduced. We reported here different consequences ofrotenone complex I inhibition on the carbon-nitrogenratio. Indeed, although a decrease in carbohydratecontent was detected and Arg was increased, we alsoobserved decreased levels for some amino acids, es-pecially Glu, Gln, Asp, and Asn, and for organic acids(Figs. 6 and 7). These contrasting effects betweenrotenone and mutation-induced complex I inhibitionmay have different explanations. First, a decrease insome amino acids may correspond to an early, transi-tory response to complex I inhibition and may laterevolve to a readjustment correlated with acclimationto a new steady state, but proving this would needmuch more data. Second, it is difficult to compareArabidopsis and N. sylvestris, as these two plantsbelong to different plant families, requiring differentenvironmental conditions to grow and thus havingdifferent metabolism. Third, rotenone was applied onheterotrophic Arabidopsis cell suspension culturesgrown in the dark, whereas the CMSII plants wereanalyzed in light conditions. As complex I has beenshown to be involved in photosynthesis performance(Dutilleul et al., 2003a), it is thus highly probable thatrepercussion of complex I inhibition is different in thedark and in the light.

CONCLUSION

Using rotenone inhibition in Arabidopsis cell cul-tures, we have shown that plant cells respond to theloss of complex I function by readjusting their electrontransport properties and general cellular carbon andnitrogen metabolism to minimize oxidative stress andallow cell survival. This provides insights into themolecular and metabolic phenotypes of complex Imutants and also highlights very early changes inmetabolite concentrations that may function in retro-grade signaling to the nucleus upon changes in mito-chondrial function.

MATERIALS AND METHODS

Arabidopsis Cell Culture, Plant Growth,and Treatments

A heterotrophic cell suspension culture of Arabidopsis (Arabidopsis thaliana

ecotype Landsberg erecta) was maintained on Murashige and Skoog basal salt

medium as described by Sweetlove et al. (2002). Cells was routinely grown at

22�C under the light (100 mE m22 s21) on an orbital shaker (130 rpm) and

subcultured every 7 d by transferring a volume (about 20 mL) corresponding

to 3.5 g of fresh cell weight to 100 mL of fresh Murashige and Skoog medium

in 250-mL flasks. Cell cultures used in the following experiments were then

grown in the dark (22�C) with shaking (130 rpm). Cell cultures (4 d after

subculture) were treated with 40 mM (v/v) rotenone. Rotenone was purchased

from Sigma-Aldrich, and stocks were made in methanol (final volume of

methanol, 0.25% [v/v] in 120 mL of cell culture). Methanol (final volume,

0.25% [v/v]) was also applied on cell cultures as a control. Cells were collected

at intervals by filtration and washed with fresh Murashige and Skoog

medium. Fresh and dry weights of 10-mL aliquots of cell suspension were

determined. Cell viability was also checked by staining aliquots with

propidium iodide (2 mg mL21 final). The proportion of dead cells was counted

using an epifluorescence microscope (excitation, 546 nm; emission, 590 nm).

At least 300 cells were scored per treatment, time point, and flask.

Transcript Analysis

Total RNA from treated and control samples was extracted using the

Qiagen Plant RNeasy kit, and genomic DNA was removed using DNA-free

DNase (Ambion). Complete removal of both mitochondrial and nuclear DNA

was checked by PCR on diluted RNA prior to RT with random primers

(SuperScript III; Invitrogen). Quantitative PCR was performed with the

LightCycler 480 real-time PCR system (Roche) using the LightCycler 480

SYBR Green 1 Master Mix and the primer sets detailed in Supplemental Table

S3, originally developed by Clifton et al. (2005) and Falcon de Longevialle

et al. (2007). Microarray ATH1 chip hybridizations were conducted following

Affymetrix instructions using the same RNA samples as for the quantitative

RT-PCR. All experiments were done on three biological replicates. Microarray

data analysis was conducted using AVADIS (Strands). Arrays were normal-

ized using PLIER (for Probe Logarithmic Intensity Error). A stringent false

discovery rate correction was applied to P values when individual fold

changes were studied (Nettleton, 2006) but not when genes were studied in

functional groups. The latter was conducted using MapMan (Thimm et al.,

2004; http://gabi.rzpd.de/projects/MapMan/) with the Ath_AFFY_TAIR7

mapping file (Supplemental Table S1) or the functional groups defined in

Supplemental Table S2. A Benjamini-Hochberg correction was applied to the P

values calculated by the Wilcoxon rank sum test.

Isolation of Mitochondria from ArabidopsisCell Culture

Mitochondria were isolated and purified by differential centrifugations

followed by two Percoll step gradients according to the method of Millar et al.

(2001). After two final washes in Suc buffer (0.3 M Suc and 10 mM TES-NaOH,

pH 7.5) and centrifugations at 24,000g, 4�C, for 15 min, mitochondria were

resuspended in a small volume of Suc buffer. Mitochondrial proteins were

quantified using the Coomassie Plus Kit (Pierce). Aliquots of 50, 150, and 500






mg of mitochondrial proteins were immediately snap frozen in liquid nitrogen

and stored at 280�C until use. Respiration rates were determined on the

remaining fresh mitochondria.

Measurements of Oxygen on Whole Cells

and Mitochondria

Oxygen consumption was measured using a Clark-type oxygen electrode

(Hansatech Instrument) in 1 mL of reaction medium. Respiration on whole

Arabidopsis cell culture was performed at 22�C by suspending 300 mL of cell

culture in 700 mL of fresh cell medium. Total respiration was measured after

the addition of the uncoupler carbonyl cyanide m-chlorophenylhydrazone (4

mM). Subsequent addition of 1 mM nPG to inhibit the AOX pathway or 1 mM

KCN to inhibit the COX pathway allowed measurements of maximal activity

of the COX and AOX pathways, respectively. Respiration rates on isolated

mitochondria were determined at 25�C on 200 mg of mitochondrial proteins

suspended in the reaction medium (0.3 M Suc, 5 mM KH2PO4, 10 mM TES, 10

mM NaCl, 2 mM MgSO4, and 0.1% bovine serum albumin [w/v], pH 7.2).

Different substrates, cofactors, and inhibitors were successively added to the

reaction medium to modulate oxygen consumption by mitochondria. Total

respiration was measured in the presence of succinate (10 mM) and ATP (0.5

mM). Maximal capacity of the AOX pathway was measured after successive

additions of the complex III inhibitor myxothiazol (2.5 mM), activators of AOX,

pyruvate (10 mM), and the reductant DTT (5 mM), and finally the AOX

inhibitor nPG (0.5 mM). Respiration rate through the COX pathway was

measured in the presence of nPG (0.5 mM) and the complex IV inhibitor KCN

(1 mM). Total respiration was also measured in the presence of Glu (10 mM)

and malate (10 mM), which provide substrates for complex I and internal

NADH dehydrogenases through the TCA cycle. In this case, the pH of the

reaction medium was adjusted to 6.8. To Glu and malate were added

coenzyme A (12 mM), thiamine pyrophosphate (0.2 mM), NAD1 (2 mM), and

ADP (0.3 mM) in order to initiate electron transport. Rotenone (5 mM), nPG (0.5

mM), and KCN (1 mM) were then added. The capacity of mitochondria to use

external NADH was measured by additions of NADH (1 mM), CaCl2 (0.1 mM),

and ADP (0.3 mM) and then myxothiazol (2.5 mM), DTT (5 mM)/pyruvate (10

mM), and nPG (0.5 mM). Mitochondrial integrity was measured by following

oxygen uptake after the addition of ascorbate (10 mM), cytochrome c (25 mM),

and Triton X-100 (0.05%, w/v). In all experiments, it was consistently better

than 90%.

One-Dimensional SDS-PAGE and

AOX Immunodetection

In order to study AOX content and redox state in treated cells, one-

dimensional SDS-PAGE and western blot analysis were performed on mito-

chondrial proteins or on membrane proteins quickly extracted from whole

cells. For mitochondrial proteins, 50 mg of proteins was solubilized in sample

buffer (2% SDS, 125 mM Tris-HCl, 10% glycerol, 10% b-mercaptoethanol, and

0.002% bromphenol blue, pH 6.8) and heated at 95�C for 3 min before loading

onto a polyacrylamide gel. For isolated cell membranes, 0.5 g of cells was

ground at 4�C in 1 mL of an extraction buffer (0.45 M mannitol, 50 mM TES,

0.5% bovine serum albumin, 0.5% polyvinylpyrrolidone-40, 2 mM EGTA, 20

mM ascorbate, and a proteinase inhibitor tablet [Roche], pH 8.0) as described

by Noguchi et al. (2005). The homogenate was centrifuged for 5 min at 1,100g,

and the resulting supernatant was centrifuged for 5 min at 10,000g. The pellet

was resuspended in 200 mL of sample buffer (without b-mercaptoethanol),

and 50 mL was loaded onto a polyacrylamide gel. Cell samples were also

ground in extraction buffer including DTT (50 mM) that reduces AOX in its

monomer form or diamide (30 mM) that oxidizes AOX in its dimer form, and

membranes were resuspended in sample buffer including DTT or diamide.

Proteins were separated according to standard protocols using 12% (w/v)

polyacrylamide and 0.1% (w/v) SDS gels. For immunodetection, separated

proteins were transferred onto a polyvinylidene difluoride membrane (Immobilon-P;

Millipore) and incubated with a primary monoclonal antibody raised against

Sauromattum guttatum AOX (Elthon et al., 1989). A chemiluminescence detec-

tion system linked to horseradish peroxidase was used as a secondary

antibody (GE Healthcare). Images of the membrane were recorded using a

Luminescent Image Analyzer (LAS 100; Fuji).

BN-PAGE and Complex I Activity Assay

BN-PAGE was performed according to the method described by Jansch

et al. (1996). Mitochondrial proteins (500 mg) were solubilized with digitonin

(5% [w/v] digitonin, 150 mM acetate, 10% glycerol, and 30 mM HEPES-HCl,

pH 7.4), and after 20 min at 4�C, Serva Blue G (5%, v/v) was added to the

supernatant. The sample was then loaded onto a two-layer polyacrylamide gel

with a stacking gel and a separation gel (respectively 4% and 4.5%216%

acrylamide:bis-acrylamide [33:1] gels). Gels were stained with colloidal

Coomassie Brilliant Blue G250 or directly used for in-gel complex I activity

assay.

Complex I (NADH dehydrogenase) activity was measured after protein

separation by BN-PAGE according to Zerbetto et al. (1997). The gel was

washed for 5 min, four times, with distilled water and incubated in the

reaction medium (0.14 mM NADH, 1.22 mM nitroblue tetrazolium, and 0.1 M

Tris-HCl, pH 7.4). When the complex was visible as a dark-blue color band,

the reaction was stopped by transferring the gel in 40% methanol:10% acetic

acid (v/v).

Two-Dimensional Fluorescence DifferenceGel Electrophoresis

The two-dimensional fluorescence difference gel electrophoresis (2D-

DIGE) technology was used to allow comparison on one gel of mitochondrial

protein contents from two different samples: mitochondrial proteins from

control cells (treated with methanol) and mitochondrial proteins from rotenone-

treated cells. Proteins of each sample were first individually labeled with

CyDye fluorophores according to the manufacturer’s instructions (GE Health-

care). Proteins were precipitated by incubating purified mitochondria over-

night at 220�C in acetone. After centrifugation at 20,000g, 4�C for 20 min,

pellets were resuspended in 10 mL of DIGE lysis buffer (8 M urea, 4% CHAPS,

and 40 mM Tris, pH 8.5) and centrifuged again at 12,000g, 4�C for 10 min in

order to remove unsolubilized material. Fifty micrograms of proteins of each

sample (proteins from methanol- and rotenone-treated cells) were stained

with a different CyDye (Cy-2 or Cy-5) by addition of 400 pmol of dye (freshly

diluted in dimethylformamide) and incubation on ice in the dark for 30 min.

The reaction was stopped by incubating samples for 10 min on ice in the dark

with 10 nmol of Lys. An equal volume (12 mL) of DIGE lysis buffer with 22 mM

DTT was added to each sample. Each of the labeled protein samples was

mixed, and rehydration buffer (8 M urea, 4% CHAPS, 0.5% 3–10 nonlinear

immobilized pH gradient buffer, 18 mM DTT, and 0.001% [w/v] bromphenol

blue) was added to give a final volume of 450 mL. The mix was loaded onto a

strip of 24 cm in length with an immobilized nonlinear pH gradient of 3 to 10

(Immobiline DryStrip; GE Healthcare). Rehydration of the strips and the first

IEF dimension electrophoresis were performed on an IPGphor Unit (GE

Healthcare) using the following settings: 12 h at 30 V (rehydration step), 1 h at

500 V, 1-h gradient from 500 to 1,000 V, 1-h gradient from 1,000 to 3,000 V, 2-h

gradient from 3,000 to 8,000 V, and 5 h at 8,000 V. After IEF, strips were

incubated for 15 min in an equilibration buffer (6 M urea, 2% SDS, 26% glycerol,

65 mM DTT, 0.001% [w/v] bromphenol blue, and 50 mM Tris-HCl, pH 8.8) and

then for 15 min in an equilibration buffer containing iodoacetamide (6 M urea,

2% SDS, 26% glycerol, 135 mM DTT, 0.001% [w/v] bromphenol blue, and 50

mM Tris-HCl, pH 8.8). The equilibrated strips were then loaded on top of a 12%

acrylamide gel. Following separation, gels were scanned using the Typhoon

Trio Variable Mode Imager at a resolution of 100 (pixel size) with the

photomultiplier tube set to 500 V. Proteins were visualized using the Typhoon

Scanner Control software (version 5.0) and processed (quantification) using

DeCyder 2-D Differential Analysis software version 6.5 (GE Healthcare). In

order to get statistical significance from these experiments, three sets of

proteins (control and ‘‘rotenone’’ proteins) from three independent experi-

ments were labeled and submitted to electrophoresis. Standard gels were also

used, and precipitation, IEF, and SDS-PAGE were run in parallel with labeled

samples. These standard gels were loaded with a mix of 150 mg of each protein

sample (control and rotenone). After electrophoresis, proteins were visualized

by colloidal Coomassie Brilliant Blue G250 staining (Sweetlove et al., 2002).

The aims of these gels were to check profiles between labeled and unlabeled

samples and to allow the identification of proteins by excising gel spots from

colloidal Coomassie Brilliant Blue-stained gels.

Identification of Proteins by Liquid

Chromatography-Tandem Mass Spectrometry

Gel spots to be analyzed were cut from colloidal Coomassie Brilliant Blue-

stained gels. Proteins were digested with trypsin (trypsin sequencing grade;

Roche Diagnostic) according to Sweetlove et al. (2001). Prior to mass spec-

trometry, peptides were resuspended in 5% acetonitrile/0.1% formic acid (v/v).

Garmier et al.





Samples were analyzed with an Agilent 1100 series capillary liquid chroma-

tography system and an Agilent Technologies XCT Ultra IonTrap with an

electrospray ionization source equipped with a low-flow nebulizer in positive

mode controlled by Chemstation (Rev B.01.03 [204]; Agilent Technologies) and

MSD Trap Control version 6.0 (Build 38.15) software (Bruker Daltonik).

Samples were loaded with the Agilent 1100 series capillary liquid chroma-

tography system onto a 0.5- 3 50-mm C18 (5 mm, 100 A) reverse-phase column

(Higgins Analytical) with a C18 OPTI-GUARD guard column (Optimize

Technologies) at 10 mL min21 equilibrated with 5% acetonitrile and 0.1%

formic acid under a regulated temperature of 50�C. Peptides were eluted from

the C18 reverse-phase column at 10 mL min21 into the XCT Ultra by a 9-min

acetonitrile gradient (5%260%) in 0.1% formic acid at 50�C. Initial ion

detection utilized a mass range of 200 to 1,400 mass-to-charge ratio (m/z)

with a scan mode set to 8,100 m/z per second. Ions reaching an intensity of

80,000 counts per second were sent to tandem mass spectrometry (MS/MS)

analysis, and two precursors were selected for the initial MS scan. MS/MS

conditions employed SmartFrag for ion fragmentation (scan range, 70–2,200

m/z). Resulting MS/MS spectra were exported from the Data Analysis for the

LC/MSD Trap version 3.3 (Build 149) software package (Bruker Daltonik)

using default parameters for AutoMS(n) and compound Export. For protein

identification, data generated from MS/MS spectra were blasted against The

Arabidopsis Information Resource (release 6) Arabidopsis database using the

Mascot search engine (Matrix Sciences). Searches were conducted with

parameters set at error tolerances of 61.2 for MS and 60.6 for MS/MS and

with ‘‘max. missed cleavages’’ set to 1. Results were filtered using ‘‘standard

scoring,’’ ‘‘max. number of hits’’ set to 20, and ‘‘significance threshold’’ at

P , 0.05.

TCA Cycle and Glycolysis Activity Measurements

Activities were measured on purified mitochondria according to methods

optimized for plant mitochondrial activities: aconitase (EC 4.2.1.3), NAD-ME

(EC 1.1.1.39), pyruvate dehydrogenase complex (PDC), and 2-oxoglutarate

dehydrogenase complex (OGDC) as described by Millar et al. (1999); NAD-

and NADP-dependent ICDH (EC 1.1.1.41 and 1.1.1.42) as described by Millar

et al. (1999) and Rasmusson and Møller (1990). Fru-1,6-bisP aldolase (EC

4.1.2.13) and GAPDH (EC 1.2.1.12) activities were measured on isolated

mitochondria and whole cell extracts as described by Bardey et al. (2005).

Triton X-100 (0.05%, v/v) was added to reaction media in order to solubilize

membranes. PDC, OGDC, NAD-ME, NAD(P)-ICDH, and aconitase activities

were assessed on 25 to 50 mg of proteins, and GAPDH and Fru-1,6-bisP

aldolase were assessed on 100 and 200 mg, respectively.

Metabolite Extraction and GC-MS Analysis

Metabolites were extracted from frozen Arabidopsis cells according to a

method adapted from Roessner-Tunali et al. (2003). Frozen cells (approxi-

mately 50 mg) were ground to a fine powder (without thawing) in a ball mill

and extracted with metabolite extraction medium (85% [v/v] HPLC-grade

methanol [Sigma], 15% [v/v] untreated MilliQ water, and 100 ng mL21 ribitol)

for 15 min at 65�C with 1,400 rpm shaking. Tubes were centrifuged for 10 min

at 20,000g and 4�C in order to pellet cell debris. Supernatants (50 mL) were

dried for 2 h in a SpeedVac system. Prior to GC-MS analysis, polar metabolites

were derivatized. A first derivatization was performed by addition of 20 mL of

methoxyamine hydrochloride (stock solution of 20 mg mL21 in pyridine) on

the dry metabolite pellet and incubation for 90 min at 30�C with shaking (1,400

rpm). A second derivatization was performed by addition of 30 mL of

N-methyl-N-(trimethylsilyl)trifluoroacetamide (Sigma Aldrich) and incubation

for 30 min at 37�C with shaking (1,400 rpm). Ten microliters of a retention time

standard mixture (0.29% [v/v] n-dodecane, 0.29% [v/v] n-pentadecane, 0.29%

[w/v] n-nonadecane, 0.29% [w/v] n-docosane, 0.29% [w/v] n-octacosane,

0.29% [w/v] n-dotriacontane, and 0.29% [w/v] n-hexatriacontane dissolved in

anhydrous pyridine) was then added to each tube, and the samples were

transferred to GC-MS vials for analysis.

Derivatized metabolite samples were analyzed on an Agilent GC/MSD

system composed of an Agilent GC 6890N gas chromatograph fitted with a

7683B Automatic Liquid Sampler and 5975B Inert MSD quadrupole MS

detector (Agilent Technologies). The gas chromatograph was fitted with a

0.25-mm i.d., 0.25-mm film thickness, 30-m Varian FactorFour VF-5ms capil-

lary column with 10 m integrated guard column (Varian). GC-MS run con-

ditions were essentially as described for GC-quadrupole-MS metabolite

profiling on the Golm Metabolome Database Web site (http://csbdb.mpimp-golm.

mpg.de/csbdb/gmd/analytic/gmd_meth.html; Kopka et al., 2005). Samples

were injected into the split/splitless injector operating in splitless mode with

an injection volume of 1 mL, purge flow of 50 mL min21, purge time of 1 min,

and a constant inlet temperature of 300�C. Helium carrier gas flow rate was

held constant at 1 mL min21. The GC column oven was held at the initial

temperature of 70�C for 1 min before being increased to 76�C at 1�C min21 and

then to 325�C at 6�C min21 before being held at 325�C for 10 min. Total run

time was 58.5 min. Transfer line temperature was 300�C. MS source temper-

ature was 230�C. Quadrupole temperature was 150�C. Electron impact ion-

ization energy was 70 eV, and the MS detector was operated in full scan mode

in the range 40 to 600 m/z with a scan rate of 2.6 Hz. The MSD was pretuned

against perfluorotributylamine mass calibrant using the ‘‘atune.u’’ autotune

method provided with Agilent GC/MSD Productivity ChemStation Software

(version D.02.00 SP1; Agilent Technologies). Raw GC-MS data preprocessing

and statistical analysis were carried out using in-house Metabolome-Express

software (version 1.0). Detailed methods are supplied as Supplemental

Materials and Methods S1.

Ascorbate and Glutathione Assays

After collection, cells were snap frozen in liquid nitrogen and stored at

280�C. Frozen cells were ground to a fine powder with liquid nitrogen and

homogenized in 1 M perchloric acid (with a cell biomass:perchloric acid ratio

of 1:5 [w/v]). After centrifugation for 15 min at 15,000g and 4�C, a volume of

0.1 mL of 0.12 M NaH2PO4 (pH 5.6) was added to 0.5 mL of supernatant and

the pH was adjusted to a value between 5 and 6 with 2.5 M K2CO3. Insoluble

KClO4 was removed from samples by centrifugation at 15,000g for 5 min at

4�C. Ascorbate and glutathione were determined in the same supernatant by

spectrophotometry as described by Dutilleul et al. (2003b).

Statistical Analysis

ANOVA was performed with StatBox 6.6 software (Grimmersoft). Data are

expressed as means 6 SE or SD as specified.

Supplemental Data

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

Supplemental Figure S1. Continuity of rotenone inhibition during the

time course of cell treatment.

Supplemental Figure S2. Analysis of respiratory rates of whole cells and

isolated mitochondria and analysis of mitochondrial OxPhos com-

plexes by BN-PAGE after rotenone treatment.

Supplemental Figure S3. Transcriptional response of mitochondria-

encoded genes for mitochondrial components following rotenone

treatment.

Supplemental Table S1. Microarray analysis of the transcription response

of nucleus-encoded genes to rotenone treatment.

Supplemental Table S2. Microarray analysis of the transcription response

of nucleus-encoded genes for mitochondrial components to rotenone

treatment.

Supplemental Table S3. Primer sequences used in the quantitative

RT-PCR experiments.

Supplemental Materials and Methods S1. Detailed methodology of

GC-MS data analysis.

ACKNOWLEDGMENTS

From the Australian Research Council Centre of Excellence in Plant

Energy Biology, we thank Holger Eubel, Alison M. Winger, and Etienne H.

Meyer for assistance with BN-PAGE and 2D-DIGE technologies and Joshua L.

Heazlewood for advice and assistance with peptide mass spectrometry.

Received July 5, 2008; accepted September 5, 2008; published September 10,

2008.






LITERATURE CITED

Ayala A, Venero JL, Cano J, Machado A (2007) Mitochondrial toxins and

neurodegenerative diseases. Front Biosci 12: 986–1007

Bardey V, Vallet C, Robas N, Charpentier B, Thouvenot B, Mougin A,

Hajnsdorf E, Regnier P, Springer M, Branlant C (2005) Characteriza-

tion of the molecular mechanisms involved in the differential produc-

tion of erythrose-4-phosphate dehydrogenase, 3-phosphoglycerate kinase

and class II fructose-1,6-bisphosphate aldolase in Escherichia coli. Mol

Microbiol 57: 1265–1287

Betarbet R, Sherer TB, MacKenzie G, Garcia-Osuna M, Panov AV,

Greenamyre JT (2000) Chronic systemic pesticide exposure reproduces

features of Parkinson’s disease. Nat Neurosci 3: 1301–1306

Brandt U, Kerscher S, Drose S, Zwicker K, Zickermann V (2003) Proton

pumping by NADH:ubiquinone oxidoreductase: a redox driven con-

formational change mechanism? FEBS Lett 54: 9–17

Clifton R, Lister R, Parker KL, Sappl PG, Elhafez D, Millar AH, Day DA,

Whelan J (2005) Stress-induced co-expression of alternative respiratory

chain components in Arabidopsis thaliana. Plant Mol Biol 58: 193–212

Corneille S, Cournac L, Guedeney G, Havaux M, Peltier G (1998) Reduc-

tion of the plastoquinone pool by exogenous NADH and NADPH in

higher plant chloroplasts: characterization of a NAD(P)H-plastoquinone

oxidoreductase activity. Biochim Biophys Acta 1363: 59–69

DiMauro S, Schon EA (2003) Mitochondrial respiratory-chain diseases. N

Engl J Med 348: 2656–2668

Djajanegara I, Finnegan PM, Mathieu C, McCabe T, Whelan J, Day DA

(2002) Regulation of alternative oxidase gene expression in soybean.

Plant Mol Biol 50: 735–742

Dudkina NV, Eubel H, Keegstra W, Boekema EJ, Braun HP (2005)

Structure of a mitochondrial supercomplex formed by respiratory-chain

complexes I and III. Proc Natl Acad Sci USA 102: 3225–3229

Dutilleul C, Driscoll S, Cornic G, De Paepe R, Foyer CH, Noctor G (2003a)

Functional mitochondrial complex I is required by tobacco leaves for

optimal photosynthetic performance in photorespiratory conditions

and during transients. Plant Physiol 131: 264–275

Dutilleul C, Garmier M, Noctor G, Mathieu C, Chetrit P, Foyer CH, De

Paepe R (2003b) Leaf mitochondria modulate whole cell redox homeo-

stasis, set antioxidant capacity, and determine stress resistance through

altered signaling and diurnal regulation. Plant Cell 15: 1212–1226

Dutilleul C, Lelarge C, Prioul JL, De Paepe R, Foyer CH, Noctor G (2005)

Mitochondria-driven changes in leaf NAD status exert a crucial influ-

ence on the control of nitrate assimilation and the integration of carbon

and nitrogen metabolism. Plant Physiol 139: 64–78

Elthon TE, Nickels RL, McIntosh L (1989) Monoclonal antibodies to the

alternative oxidase of higher plant mitochondria. Plant Physiol 89: 1311–1317

Eubel H, Heinemeyer J, Braun HP (2004) Identification and characteriza-

tion of respirasomes in potato mitochondria. Plant Physiol 134: 1450–1459

Falcon de Longevialle AF, Meyer EH, Andres C, Taylor NL, Lurin C,

Millar AH, Small ID (2007) The pentatricopeptide repeat gene OTP43 is

required for trans-splicing of the mitochondrial nad1 intron 1 in Arabi-

dopsis thaliana. Plant Cell 19: 3256–3265

Giege P, Heazlewood JL, Roessner-Tunali U, Millar AH, Fernie AR,

Leaver CJ, Sweetlove LJ (2003) Enzymes of glycolysis are functionally

associated with the mitochondrion in Arabidopsis cells. Plant Cell 15:

2140–2151

Graham JW, Williams TC, Morgan M, Fernie AR, Ratcliffe RG, Sweetlove

LJ (2007) Glycolytic enzymes associate dynamically with mitochondria

in response to respiratory demand and support substrate channeling.

Plant Cell 19: 3723–3728

Gutierres S, Sabar M, Lelandais C, Chetrit P, Diolez P, Degand H, Boutry

M, Vedel F, de Kouchkovsky Y, De Paepe R (1997) Lack of mitochon-

drial and nuclear-encoded subunits of complex I and alteration of the

respiratory chain in Nicotiana sylvestris mitochondrial deletion mutants.

Proc Natl Acad Sci USA 94: 3436–3441

Heazlewood JL, Howell KA, Millar AH (2003) Mitochondrial complex I

from Arabidopsis and rice: orthologs of mammalian and fungal com-

ponents coupled with plant-specific subunits. Biochim Biophys Acta

1604: 159–169

Howell KA, Cheng K, Murcha MW, Jenkin LE, Millar AH, Whelan J

(2007) Oxygen initiation of respiration and mitochondrial biogenesis in

rice. J Biol Chem 282: 15619–15631

Igambardiev AU, Zhou G, Malmberg G, Gardestrom P (1997) Respiration

of barley protoplasts before and after illumination. Physiol Plant 99: 15–22

Ikezawa N, Ifuku K, Endo T, Sato F (2002) Inhibition of photosystem II of

spinach by the respiration inhibitors piericidin A and thenoyltrifluo-

roacetone. Biosci Biotechnol Biochem 66: 1925–1929

Ishizaki K, Schauer N, Larson TR, Graham IA, Fernie AR, Leaver CJ

(2006) The mitochondrial electron transfer flavoprotein complex is

essential for survival of Arabidopsis in extended darkness. Plant J 47:

751–760

Jansch L, Kruft V, Schmitz UK, Braun HP (1996) New insights into the

composition, molecular mass and stoichiometry of the protein com-

plexes of plant mitochondria. Plant J 9: 357–368

Karpova OV, Kuzmin EV, Elthon TE, Newton KJ (2002) Differential

expression of alternative oxidase genes in maize mitochondrial mu-

tants. Plant Cell 14: 3271–3284

Kopka J, Schauer N, Krueger S, Birkemeyer C, Usadel B, Bergmuller E,

Dormann P, Weckwerth W, Gibon Y, Stitt M, et al (2005) GMD@CSB.

DB: the Golm Metabolome Database. Bioinformatics 21: 1635–1638

Kussmaul L, Hirst J (2006) The mechanism of superoxide production by

NADH:ubiquinone oxidoreductase (complex I) from bovine heart mi-

tochondria. Proc Natl Acad Sci USA 103: 7607–7612

Kuzmin EV, Karpova OV, Elthon TE, Newton KJ (2004) Mitochondrial

respiratory deficiencies signal up-regulation of genes for heat shock

proteins. J Biol Chem 279: 20672–20677

Lister R, Chew O, Lee MN, Heazlewood JL, Clifton R, Parker KL, Millar

AH, Whelan J (2004) A transcriptomic and proteomic characterization

of the Arabidopsis mitochondrial protein import apparatus and its

response to mitochondrial dysfunction. Plant Physiol 134: 777–789

Liu Z, Butow RA (1999) A transcriptional switch in the expression of yeast

tricarboxylic acid cycle genes in response to a reduction or loss of

respiratory function. Mol Cell Biol 19: 6720–6728

Liu Z, Butow RA (2006) Mitochondrial retrograde signaling. Annu Rev

Genet 40: 159–185

Maxwell DP, Wang Y, McIntosh L (1999) The alternative oxidase lowers

mitochondrial reactive oxygen production in plant cells. Proc Natl Acad

Sci USA 96: 8271–8276

Melo AMP, Roberts TH, Møller AM (1996) Evidence for the presence of

two rotenone-insensitive NAD(P)H dehydrogenases on the inner sur-

face of the inner membrane of potato tuber mitochondria. Biochim

Biophys Acta 1276: 133–139

Meyer EH, Heazlewood JL, Millar AH (2007) Mitochondrial acyl carrier

proteins in Arabidopsis thaliana are predominantly soluble matrix pro-

teins and none can be confirmed as subunits of respiratory complex I.

Plant Mol Biol 64: 319–327

Millar AH, Eubel H, Jansch L, Kruft V, Heazlewood JL, Braun HP (2004)

Mitochondrial cytochrome c oxidase and succinate dehydrogenase

complexes contain plant specific subunits. Plant Mol Biol 56: 77–90

Millar AH, Hill SA, Leaver CJ (1999) Plant mitochondrial 2-oxoglutarate

dehydrogenase complex: purification and characterization in potato.

Biochem J 343: 327–334

Millar AH, Mittova V, Kiddle G, Heazlewood JL, Bartoli CG, Theodoulou

FL, Foyer CH (2003) Control of ascorbate synthesis by respiration and its

implications for stress responses. Plant Physiol 133: 443–447

Millar AH, Sweetlove LJ, Giege P, Leaver CJ (2001) Analysis of the

Arabidopsis mitochondrial proteome. Plant Physiol 127: 1711–1727

Nettleton D (2006) A discussion of statistical methods for design and analysis

of microarray experiments for plant scientists. Plant Cell 18: 2112–2121

Noctor G, De Paepe R, Foyer CH (2007) Mitochondrial redox biology and

homeostasis in plants. Trends Plant Sci 12: 125–134

Noguchi K, Taylor NL, Millar AH, Lambers H, Day DA (2005) Response of

mitochondria to light intensity in the leaves of sun and shade species.

Plant Cell Environ 28: 760–771

Perales M, Eubel H, Heinemeyer J, Colaneri A, Zabaleta E, Braun HP

(2005) Disruption of a nuclear gene encoding a mitochondrial gamma

carbonic anhydrase reduces complex I and supercomplex I 1 III2 levels and

alters mitochondrial physiology in Arabidopsis. J Mol Biol 350: 263–277

Pineau B, Mathieu C, Gerard-Hirne C, De Paepe R, Chetrit P (2005)

Targeting the NAD7 subunit to mitochondria restores a functional

complex I and a wild type phenotype in the Nicotiana sylvestris CMS II

mutant lacking nad7. J Biol Chem 15: 25994–256001

Rasmusson AG, Møller IM (1990) NADP-utilizing enzymes in the matrix

of plant mitochondria. Plant Physiol 94: 1012–1018

Rasmusson AG, Møller IM (1991) Effects of calcium ions and inhibitors on

internal NAD(P)H dehydrogenases in plant mitochondria. Eur J Bio-

chem 202: 617–623

Garmier et al.





Rasmusson AG, Soole KL, Elthon TE (2004) Alternative NAD(P)H dehy-

drogenases of plant mitochondria. Annu Rev Plant Biol 55: 23–39

Rhoads DM, Subbaiah CC (2007) Mitochondrial retrograde regulation in

plants. Mitochondrion 7: 177–194

Rikhvanov EG, Gamburg KZ, Varakina NN, Rusaleva TM, Fedoseeva IV,

Tauson EL, Stupnikova IV, Stepanov AV, Borovskii GB, Voinikov VK

(2007) Nuclear-mitochondrial cross-talk during heat shock in Arabi-

dopsis cell culture. Plant J 52: 763–778

Roberts TH, Fredlund KM, Møller IM (1995) Direct evidence for the

presence of two external NAD(P)H dehydrogenases coupled to the

electron transport chain in plant mitochondria. FEBS Lett 373: 307–309

Roessner-Tunali U, Urbanczyk-Wochniak E, Czechowski T, Kolbe A,

Willmitzer L, Fernie AR (2003) De novo amino acid biosynthesis in

potato tubers is regulated by sucrose levels. Plant Physiol 133: 683–692

Sabar M, De Paepe R, de Kouchkovsky Y (2000) Complex I impairment,

respiratory compensations, and photosynthetic decrease in nuclear and

mitochondrial male sterile mutants of Nicotiana sylvestris. Plant Physiol

124: 1239–1250

Saisho D, Nambara E, Naito S, Tsutsumi N, Hirai A, Nakazono M (1997)

Characterization of the gene family for alternative oxidase from Arabi-

dopsis thaliana. Plant Mol Biol 35: 585–596

Singer TP (1979) Mitochondrial electron-transport inhibitors. Methods

Enzymol 55: 454–462

Sweetlove L, Mowday B, Hebestreit HF, Leaver CJ, Millar AH (2001)

Nucleoside disphosphate kinase III is localised to the inter-membrane

space in plant mitochondria. FEBS Lett 508: 272–276

Sweetlove LJ, Heazlewood JL, Herald V, Holtzapffel R, Day DA, Leaver

CJ, Millar AH (2002) The impact of oxidative stress on Arabidopsis

mitochondria. Plant J 32: 891–904

Taylor NL, Heazlewood JL, Day DA, Millar AH (2005) Differential impact

of environmental stresses on the pea mitochondrial proteome. Mol Cell

Proteomics 4: 1122–1133

Taylor NL, Rudhe C, Hulett JM, Lithgow T, Glaser E, Day DA, Millar AH,

Whelan J (2003) Environmental stresses inhibit and stimulate different

protein import pathways in plant mitochondria. FEBS Lett 547: 125–130

Thimm O, Blaesing O, Gibon Y, Nagel A, Meyer S, Kruger P, Selbig J,

Muller LA, Rhee SY, Stitt M (2004) MAPMAN: a user-driven tool to

display genomics data sets onto diagrams of metabolic pathways and

other biological processes. Plant J 37: 914–939

Vidal G, Ribas-Carbo M, Garmier M, Dubertret G, Rasmusson AG,

Mathieu C, Foyer CH, De Paepe R (2007) Lack of respiratory chain

complex I impairs alternative oxidase engagement and modulates redox

signaling during elicitor-induced cell death in tobacco. Plant Cell 19:

640–655

Zerbetto E, Vergani L, Dabbeni-Sala F (1997) Quantification of muscle

mitochondria oxidative phosphorylation enzymes via histochemical

staining of blue native polyacrylamide gels. Electrophoresis 18:

2059–2064






Complex I Dysfunction Redirects Cellular and Mitochondrial Metabolism in Arabidopsis

Documents

Transcript of Complex I Dysfunction Redirects Cellular and Mitochondrial Metabolism in Arabidopsis