central metabolic responses to ozone and herbivory affect photosynthesis and stomatal ... ·...

Central Metabolic Responses to Ozone and HerbivoryAffect Photosynthesis and Stomatal Closure1[OPEN]

Stefano Papazian, Eliezer Khaling, Christelle Bonnet, Steve Lassueur, Philippe Reymond, Thomas Moritz,James D. Blande*, and Benedicte R. Albrectsen*

Umeå Plant Science Centre, Department of Plant Physiology, Umeå University, 90187 Umeå (S.P., B.R.A.);Department of Forest Genetic and Plant Physiology, Swedish University of Agricultural Sciences, 90183 Umeå,Sweden (T.M.); Department of Environmental and Biological Sciences, University of Eastern Finland, FIN–

70211 Kuopio, Finland (E.K., J.D.B.); and Department of Plant Molecular Biology, University of Lausanne,1015 Lausanne, Switzerland (C.B., S.L., P.R.)

ORCID IDs: 0000-0003-2538-8702 (S.P.); 0000-0001-5472-7297 (C.B.); 0000-0002-3341-6200 (P.R.); 0000-0002-4258-3190 (T.M.);0000-0001-6822-0649 (J.D.B.); 0000-0002-9337-4540 (B.R.A.).

Plants have evolved adaptive mechanisms that allow them to tolerate a continuous range of abiotic and biotic stressors. Troposphericozone (O3), a global anthropogenic pollutant, directly affects living organisms and ecosystems, including plant-herbivore interactions.In this study, we investigate the stress responses of Brassica nigra (wild black mustard) exposed consecutively to O3 and the specialistherbivore Pieris brassicae. Transcriptomics and metabolomics data were evaluated using multivariate, correlation, and networkanalyses for the O3 and herbivory responses. O3 stress symptoms resembled those of senescence and phosphate starvation, while asequential shift from O3 to herbivory induced characteristic plant defense responses, including a decrease in central metabolism,induction of the jasmonic acid/ethylene pathways, and emission of volatiles. Omics network and pathway analyses predicted a linkbetween glycerol and central energy metabolism that influences the osmotic stress response and stomatal closure. Furtherphysiological measurements confirmed that while O3 stress inhibited photosynthesis and carbon assimilation, sequential herbivorycounteracted the initial responses induced by O3, resulting in a phenotype similar to that observed after herbivory alone. This studyclarifies the consequences of multiple stress interactions on a plant metabolic system and also illustrates how omics data can beintegrated to generate new hypotheses in ecology and plant physiology.

Under natural conditions, plants are exposed con-tinuously to abiotic and biotic stresses. When studied inthe laboratory, individual stresses trigger a variety ofmolecular, cellular, and physiological responses (Rejebet al., 2014; Suzuki et al., 2014). However, the wayplants protect themselves in nature cannot be predicted

on the basis of responses to individual stresses alone,because combined stresses may elicit antagonistic,neutral, or synergistic effects (Rizhsky et al., 2002, 2004;Mittler, 2006; Pandey et al., 2015). Thus, there is in-creasing interest in plant responses to multiple stressconditions. However, while several studies have ex-amined multiple abiotic factors (Suzuki et al.., 2014),biotic factors usually are limited to pathogen infection(Sharma et al., 1996; Xiong andYang, 2003;Anderson et al.,2004; Prasch and Sonnewald, 2013; but see Atkinson andUrwin, 2012; Atkinson et al., 2013).

Global warming encompasses a range of interrelatedabiotic phenomena, including increases in the Earth’saverage temperature and changes in the atmosphere’scontent of greenhouse gasses such as methane, carbondioxide (CO2), and ozone (O3). Burning fossil fuels re-leases nitrogen oxides and hydrocarbons, which in thepresence of sunlight react to form tropospheric O3, themost significant atmospheric pollutant in terms ofphytotoxicity (Ludwikow and Sadowski, 2008; Renautet al., 2009; Van Dingenen et al., 2009). Depending onthe duration and intensity of the exposure, O3 can in-duce a range of responses that are typically deleteriousto plant fitness, disturbing photosynthetic processes,energy and carbon metabolism, cellular detoxification,and transpiration (Bagard et al., 2008; Dizengremelet al., 2008; Fares et al., 2010; Goumenaki et al., 2010;

1 This work was supported by the European Science Foundation,which funded the EUROCORES program A-BIO-VOC, by nationalfunds from the Vetenskapsrådet (grant no. VR/ESF 324–2011–787 toB.R.A. and T.M.), by the Academy of Finland (grant nos. 141053 and251898 to J.D.B.), and by the Swiss National Science Foundation(grant no. 31VL30_134414 to P.R.).

* Address correspondence to [email protected] and [email protected].

The author responsible for distribution of materials integral to thefindings presented in this article in accordance with the policy de-scribed in the Instructions for Authors (www.plantphysiol.org) is:Benedicte R. Albrectsen ([email protected]).

J.D.B. conceived the project; E.K., P.R, C.B., S.P., and B.R.A. assis-ted in design of the research; J.D.B., S.P., and E.K. performed theexperiments; S.P., C.B., and E.K. performed the analyses; S.L. pro-vided technical assistance to C.B.; J.D.B., P.R., and T.M. supervisedthe analyses; S.P. analyzed the data; B.R.A., C.B., and P.R. contrib-uted statistical analyses; S.P., J.D.B., and B.R.A. wrote the article withcontributions from all the authors.

[OPEN] Articles can be viewed without a subscription.www.plantphysiol.org/cgi/doi/10.1104/pp.16.01318

Plant Physiology�, November 2016, Vol. 172, pp. 2057–2078, www.plantphysiol.org � 2016 American Society of Plant Biologists. All Rights Reserved. 2057 www.plantphysiol.orgon June 29, 2020 - Published by Downloaded from

Copyright © 2016 American Society of Plant Biologists. All rights reserved.

http://orcid.org/0000-0003-2538-8702

http://orcid.org/0000-0001-5472-7297

http://orcid.org/0000-0002-3341-6200

http://orcid.org/0000-0002-4258-3190

http://orcid.org/0000-0001-6822-0649

http://orcid.org/0000-0002-9337-4540

mailto:[email protected]



http://www.plantphysiol.org


http://www.plantphysiol.org/cgi/doi/10.1104/pp.16.01318


Salvatori et al., 2015). While acute exposure to high O3levels can rapidly induce cell death and chlorosis(Ashmore, 2005; Kangasjärvi et al., 2005; Vainonen andKangasjärvi, 2015), negative yield responses are notalways correlated with the severity of symptoms inleaves, and exposure to O3 can affect metabolic pro-cesses before any visible injury (Long and Naidu, 2002;Dizengremel et al., 2009; Sawada and Kohno, 2009;Pinto et al., 2010).

Global climate change also is predicted to increasedamage to plants caused by herbivory, via both directeffects on the herbivore behavior and survival and in-direct effects on the host plant condition (Bale et al.,2002; Fuhrer, 2003; Ditchkoff et al., 2009; Lindroth,2010; Khaling et al., 2015). Negative effects of herbivoryand defoliation on plant fitness include the systemicdown-regulation of photosynthesis and reduced CO2assimilation (Zangerl et al., 2002; Hui et al., 2003; Ralphet al., 2006; Tang et al., 2006). Thus, to protect againstherbivores, plants allocate resources between growthand defense (Koricheva, 2002; Schwachtje and Baldwin,2008; Firn and Jones, 2009; Havko et al., 2016). Brassicaceaeplants fine-tune their defenses by inducing phytohormonesignaling, which activates cross talk between the jasmonicacid (JA) and ethylene (ET) pathways in response to leafchewing by herbivores (De Vos et al., 2005; Pieterse et al.,2012). As a specialist herbivore of Brassicaceae, the largewhite butterfly Pieris brassicae is an important pest ofBrassica nigra (wild blackmustard). P. brassicae ovipositionon thehost plant is influencedbyglucosinolates (sulfur- andnitrogen-containing glucosides) that function as chemicalcues for thebutterfly (Faheyet al., 2001; Petersen et al., 2002;Halkier and Gershenzon, 2006; Textor and Gershenzon,2008).Young caterpillars primarily feedonmature leavesofB. nigra, but after the third instar, theymove to fresh tissueswithhigher glucosinolate content, suchasflowers andbuds(Smallegange et al., 2007). Therefore, descriptions of plant-herbivore interactions frequently emphasize specialized(secondary)metabolism (Simmonds, 2003; vanDamet al.,2004;Poelmanetal., 2010;Boeckler et al., 2011;Lof et al., 2013;Onkokesung et al., 2014). However, there is an extensivefunctional overlap between the growth- and defense-relatedaspects of plant metabolism, and reconfiguration at the levelof primary and energymetabolism can play a central role inthe processes of stress tolerance, signal transmission, anddirect defense (Both et al., 2005; Rolland et al., 2006;Schwachtje and Baldwin 2008; Fernandez et al., 2010).

Themetabolomics approach has been applied in variousareas of plant biology and ecology (Sardans et al., 2011;Weckwerth, 2011), where it has helped in understandingthe regulation of pathways during plant-herbivore inter-actions (Jansen et al., 2008; Misra et al., 2010) and plantstress responses (Shulaev et al., 2008; Nakabayashi andSaito, 2015). What typically distinguishes metabolomicsfrom targeted analyses or generalmetabolic profiling is theambition to integrate with other omics sciences (Fiehn,2002; Barah and Bones, 2015). Combinedwith the guilt-by-association principle (Bino et al., 2004; Saito et al.,2008), omics analyses allow the prediction of unknowngene andmetabolite functions. For instance, the systematic

metabolomics approach has helped to determine the bio-synthetic regulation of glucosinolates by MYB transcrip-tion factors (Hirai et al., 2007; Sønderby et al., 2007) and tomodel the costs of glucosinolate biosynthesis in terms ofprimarymetabolism (Bekaert et al., 2012). This data-driven(top-down) strategy can guide the development of new hy-potheses about regulatory and metabolic pathways, whichmay subsequently predict the emergence of certain pheno-types (Fukushima et al., 2009; Saito and Matsuda, 2010).

Here, we describe the use of integrated transcriptomicsand metabolomics to investigate the responses of B. nigrato sequential O3 and P. brassicae stress. We found thatshifts from abiotic to biotic stress responses differentiallyregulated glycerol and energymetabolic networks. In ourmodel, these pathways were associated functionally withphotosynthesis and mitochondrial activity and were fur-ther predicted to involve physiological responses relatedto osmotic stress tolerance and stomatal closure. Our re-sults suggest that these central processes play an impor-tant role in the plant’s ability to adapt to sequential abioticand biotic stresses. In a later experiment, we measuredphotosynthesis and gas exchange to assess the impactof the stresses on B. nigra physiology. These findingsconfirmed a negative effect of P. brassicae herbivory onthe plant’s ability to regulate stomatal closure andtranspiration in response to O3.

RESULTS

P. brassicae was subjected to three stress scenarios: O3fumigation at 70 nL L21 for 5 d (O); herbivore feeding byfirst instarP. brassicae caterpillars for 24h (P); andexposure toO3 followed by herbivory (OP). Leaf samples were sharedfor omics analyses (Fig. 1, experiment 1). The transcriptomicsscreening was based on Arabidopsis (Arabidopsis thaliana)CATMA version 4 whole-genome microarrays (Sclep et al.,2007), while themetabolomics screening combined gaschromatography-mass spectrometry (GC-MS), liquidchromatography-mass spectrometry (LC-MS), and GC-MSanalysis of volatile organic compounds (VOCs) collected viadynamic headspace sampling (Supplemental Data Sets S1and S2). The omics data were first examined separately andthen integrated via network and Gene Ontology (GO)analyses. Analysis of themolecular andmetabolic responsesprompted the development of a model suggesting that theregulatory dynamics associated with O3 are opposed tothose induced by herbivory. This hypothesis was tested in afollow-up experiment (Fig. 1, experiment 2) that measuredphysiological variables including chlorophyll levels, pho-tosynthesis rates, and gas exchange for the same stressconditions (O, P, and OP) and during long-term (16-d) O3exposure (OL).

Transcriptome Responses

Hierarchical cluster and GO analyses of 970 differ-entially expressed genes (P , 0.05; Fig. 2, A and B)separated the O3 stress treatment (O) from the herbi-vore stress (P) and the sequential stress (OP). The on-tological category exhibiting the strongest response

2058 Plant Physiol. Vol. 172, 2016

Papazian et al.

www.plantphysiol.orgon June 29, 2020 - Published by Downloaded from Copyright © 2016 American Society of Plant Biologists. All rights reserved.

http://www.plantphysiol.org/cgi/content/full/pp.16.01318/DC1



after GO enrichment was energymetabolism, includingphotosynthesis and the mitochondrial electron trans-port chain (ETC; Fig. 2, A and B). Overall, changes ingene expression induced in Owere highly homogenous(Pearson’s coefficient, r = 0.6–0.8), whereas changesinduced in P and OP treatments exhibited a weakerintracluster correlation (r = 0.2–0.5; Fig. 2C). MapManpathway analysis confirmed that the responses for eachtreatment were consistent across several biologicallevels (Fig. 2D), and Venn diagrams based on log2 foldchanges of 0.585 or greater (for P , 0.05) indicated adominance of up- and down-regulated genes in O, withthe least impact in OP (Fig. 2E).Genes involved in light harvesting and carbon assimi-

lation were down-regulated in all stress scenarios butparticularly in O (Fig. 2, D and F), which also resulted inup-regulation of the nonphotochemical quenching geneNPQ1 (Fig. 2F). Other primary processes, such as aminoacid and carbohydrate metabolism, generally showed op-posite patterns of regulation forO3 and herbivory (Fig. 2, Dand G). In the sequential treatment (OP), B. nigra activateddefense responses that resembled a response to herbivorestress alone (P; Fig. 2, H and I): lipoxygenases LOX2 andLOX3 (Felton et al., 1994; Halitschke and Baldwin, 2003),the trypsin inhibitorWSCP (Zavala and Baldwin, 2004;Boex-Fontvieille et al., 2015), and the mitogen-activatedprotein kinase MPK3 (Pitzschke and Hirt, 2009). Nota-bly, genes involved in stress responses and phytohor-mone signalingwere differentially regulated betweenO3stress alone and the herbivore treatments (e.g. WRKYs,MYC2, and ERF2), and only in O were genes associatedwith abiotic stress responses induced, such as drought(MYB44), senescence (EIN3), and phosphate starva-tion (RNS1 and PT2/PHT1.4; Fig. 2, H and I).

Metabolome Responses

A multivariate analysis of the metabolome profileexplained 65% of the total metabolic variation and 95% ofthe treatment effects (partial least squares-discriminantanalysis [PLS-DA]; Fig. 3A; Supplemental Fig. S1; Supplemental Table S1). Both O3 and herbivore stress causeda shift from the basal metabolic state of the untreatedplants, inducing unique profiles as single treatments inOand P. However, the effect of herbivory largely over-shadowed that of O3 in the sequential treatment OP, as in-dicated by the swarm overlap with P (Fig. 3A). Despite theimportance of several metabolite pools in the multivariatemodel (Fig. 3B; Supplemental Table S2; 70 compounds forvariable importance for projection [VIP] scores. 1), singleunivariate effects were confirmed only for a subset of these(Fig. 3C; ANOVA posthoc Tukey’s tests).

Opposite effects of O3 and herbivory were observed onpools of primary metabolites (sugars, amino acids, andorganic acids), which increased in O and decreased in Pand OP (Fig. 3B; Supplemental Table S2). However, acommon response to these treatments was a 3-fold in-duction of g-aminobutyric acid (GABA; Fig. 3, B and C),which partly correlatedwith its precursor a-ketoglutarate(Fig. 3B).Most noticeably, glycerol levels inO increased to155% of the steady-state value (Fig. 3, B and C) but wererestored in the sequential treatment OP. A similar trendwas observed for ethanolamine, anothermetabolite of theglycerophospholipid pathway and a component of lipidmembranes (Fig. 3C). Phenolic compounds such as fla-vonols and cinnamic acid ester derivatives (Lin et al.,2011; Shao et al., 2014) increased in all treatments.

Glucosinolate levels were stable or reduced (withglucoraphanin the only exception), whereas VOCs

Figure 1. Design of omics experiments, hypothesis generation, andvalidation via physiological evidence. For experiment 1, 5-week-oldB. nigra plants were exposed to abiotic and/or biotic stresses: O3 fumigation at 70 nL L21 for 5 d (O), herbivore feedingwith 30 first instarP. brassicae caterpillars for 24 h (P), sequential stress ofO3 followedby herbivore feeding (OP), or no treatment as controls (C). To obtain acomprehensive understanding of metabolome and transcriptome responses to the treatments, leaves of equal developmental stage weresampled from the same plants from which volatile emissions (VOCs) had been obtained. Statistical and network analyses combinedomics data in a model that connected the metabolic responses with physiological adaptation to the multiple stress condition. B. nigraresponses were further evaluated in experiment 2, where physiological parameters were measured for the same stress treatments (O, P,and OP) and for a long-term exposure to O3 at 70 nL L21 for 16 d (OL). The initial hypothesis was thus verified through validation of theomics model, which predicted differential physiological responses of photosynthesis, carbon assimilation, and stomatal regulation.

Plant Physiol. Vol. 172, 2016 2059

Omics Multiple Stress Responses in Brassica nigra








Figure 2. Transcriptome responses in B. nigra under multiple O3 and herbivory stress treatments: O3 stress (O), herbivory byP. brassicae (P), and both stresses sequentially (OP). A, Heat map showing statistically significant (P # 0.05) changes in theexpression of 970 genes in different samples from theO, P, andOP treatments, relative to control plants. Red and blue indicate up-and down-regulation, respectively. The subfigure at right shows the GO enrichment of specific biological processes in eachcluster visible in the heatmap, ranked (I–III) according to the relative abundance of genes associatedwith the biological process inquestion within the cluster. The color scheme used here matches that used in B. B, All of the GO processes that were significantlyenriched in all the gene clusters, ordered by their relative abundance (for all 970 genes). C, Heat map of r as calculated for eachsample pair from the corresponding hierarchical clustering; darker shades of blue indicate stronger correlations. D, MapMan


Papazian et al.



were emitted upon herbivore damage (P and OP), in-cluding glucosinolate derivatives and green leaf vol-atiles (GLVs; Fig. 3, B and C; Supplemental Table S2).Two unidentified secondary compounds, describedpreviously by Khaling et al. (2015; i.e. 421 [M-H]2,mass-to-charge ratio [m/z] 485.13, and 381 [M-H]2,m/z349.15), also were induced, confirming their impor-tance for both O and OP (Fig. 3C).

Omics Integrative Network Correlation Analysis

Transcriptomics and metabolomics profiles were in-tegrated in a scale-free correlation network (Fig. 4; fortopology, see Supplemental Fig. S2). The network was

dominated by assortative high-degree nodes (hubs),with major hubs involved in processes such as stresssignaling, cellulose biosynthesis, chloroplast activity,and stomatal regulation (Table I). Several primary me-tabolites clustered around the central region of the net-work, while distinct modules of secondary metabolitesclustered at the periphery (i.e. glucosinolates, flavonolglucosides, hydroxycinnamic acid derivatives, andVOCs; Fig. 4). Most glucosinolates were connected toCYP71, a cytochrome involved in herbivore-inducedresponses and nitrile formation (At5g25120/At5g25180;Bennett et al., 1993; Irmisch et al., 2014). Glucobrassicinand neoglucobrassicin (indolics) clustered directly withVOCs (nitriles and GLVs) and correlated positively with

Figure 2. (Continued.)biological functions (bins) showing the effect of the stress treatments on gene expression relative to amino acidmetabolism, bioticand abiotic stress, cell wall biosynthesis, photosynthesis, and the Calvin cycle. E, Venn diagrams (MapMan generated) comparinggenes up- and down-regulated in response to the treatments for log2 fold-change thresholds greater than 0.585 (corresponding toP # 0.05). F to I, Univariate analysis relevant to the regulation of central metabolism and stress responses: processes of photo-synthesis and light harvesting (F), carbohydrate metabolism (G), biotic and abiotic stress (H), and response and signaling (I).Relative transcript accumulation is reported as log2 values. Significance relative to the controls (zero level) is reported as Student’st test values as follows: *, P # 0.05; **, P # 0.01; ***, P # 0.001; and ****, P # 0.0001. Error bars indicate SE. Treatments areindicated by color: O3 (O) in white, herbivory by P. brassicae (P) in gray, and sequential treatment (OP) in black.

Figure 3. Metabolome responses in B. nigra under multiple O3 and herbivory stress treatments: O3 stress (O), herbivory byP. brassicae (P), sequential treatment (OP), and untreated plants (C). Leaf-bound metabolites were analyzed by LC-MS and GC-MS,while VOCs were collected via headspace and analyzed by GC-MS. Five latent variables (LVs) cumulatively explained 65% of thetotal variation in X (R2X[c]) and 95% of the treatment response (R2Y[c]), with 44% total model predictability (Q2[c]). Model statisticsandmetabolite identities are presented in Supplemental Tables S1 and S2. A, PLS-DA score plot of first and second LVs, LV[1] versus LV[2]. For score plots relative to LV[3], see Supplemental Figure S1. B, PLS-DA loading plot showing important metabolites for themodel(VIP score . 1). a-KG, a-Ketoglutarate; TCA, trichloroacetic acid. C, ANOVAs with posthoc Tukey’s comparisons to test treatmenteffects for selected compounds (different letters indicate different response means). n.d., Not detected. Error bars indicate SE.









WRKY40 (involved in indolic glucosinolate biosynthesisand GLV emissions; Schön et al., 2013; Mirabella et al.,2015) as well asWRKY46 and CYP707A3 (both involvedin abscisic acid [ABA] metabolism; Saito et al., 2004; Liuet al., 2012; Geilen and Böhmer, 2015; Fig. 4C). The her-bivory response (P) linked to glycerol via four nodes,including two genes coding for flavin monooxygenases,NOGC1 (At1g62580) and FMO (At1g12200), whichnegatively correlatedwith glycerol (r =20.86,P, 0.001)and positively correlated with each other (r = 0.85, P ,0.001; Fig. 4, A and B). Because MYB44 linked to thesame module and negatively correlated with NOGC1(r = 20.91, P , 0.0001), MYB44 positively correlatedwith glycerol (r = 0.68, P = 0.006; Fig. 4B). Notably, bothMYB44 and NOGC1 are involved in the osmotic stressresponse and the regulation of stomatal closure.

Responses Relative to Energy and GlycerolMetabolic Networks

Genes associated with energy metabolism (includingboth the photosystems and the mitochondria) were themost heavily affected by the stress treatments (Fig. 2, Aand B). A PLS-DAmodel for these energy genes (which

were functionally connected in AraNet; Fig. 5A)strongly separated the stress treatments O, P, and OP,explaining 54% of the variation in gene expression and99% of the treatment effects (Fig. 5B; SupplementalTable S4). In response to O3 stress, the mitochondrialETC complex II succinate dehydrogenase subunitSDH1-1 (At5g66760) and the mitochondrial superoxidedismutase MSD1 (At3g10920) were both up-regulated(Fig. 5C; Supplemental Table S5). In the general modelfor expression of all 970 genes (Fig. 5, D and E;Supplemental Tables S6 and S7), SDH1-1 againstrongly described the effect of O3 (O and OP) and waspositively correlatedwithALDH7B4 (At1g54100; r = 0.65,P = 0.005; Fig. 5, E and F), an aldehyde dehydrogenaseinvolved in glycerol metabolism.

Visualization of the entire glycerolipid pathway in theKyoto Encyclopedia of Genes and Genomes (KEGG)/KaPPA-View4 confirmed the up-regulation in O ofALDH7B4 (P = 0.005; Supplemental Fig. S5) and of theadjacent aldo-keto-reductase AKR4C10 (At2g37790;P , 0.05), which, together with ALDH7B4, reversiblyconverts glycerate into glyceraldehyde and glycerol (forpaths, see Supplemental Figs. S6 and S7). Interestingly,both AKR4C10 and ALDH7B4 are involved in oxidative

Figure 4. Omics integrative network cor-relation analysis formultiple stress responsesin B. nigra. Gene expression profiles for970 genes (P , 0.05) were correlated withthe metabolome profile (156 compounds)for all the experimental treatments involvingO3 stress and P. brassicae herbivory. Theresulting graph was rendered as a networkusing Cytoscape. Edges connect nodeslinkedby r values of 0.85 or greater; positiveand negative correlations are represented byblue and red lines, respectively. Nodes cor-respond to genes (white squares/diamonds)and metabolites (colored circles). The her-bivore variable for the effect of P. brassicaetreatment is indicated by the boldface letterP inside awhite square. A,Network of gene-to-gene, gene-to-metabolite, andmetabolite-to-metabolite correlations. B, Expansionof the gene-to-metabolite subnet high-lighting the effect of P on glycerol andcoexpressed genes (MYB44, the flavinmonoxygenases NOGC1 and FMO, anLRR-RK receptor, and a SecY protein). C,VOCs subnet, connecting central metabo-lism to secondary metabolism (glucosino-lates, hydroxycinnamic acid derivatives,and flavonol glucosides) via WRK40,WRK46, and CYP707A3.


Papazian et al.









and osmotic stress tolerance during abiotic and bioticresponses (Kotchoni et al., 2006; Missihoun et al., 2014;Sengupta et al., 2015).On the basis of genes annotated for the GO enzyme

substrate glycerol- (Supplemental Table S8), whichwere functionally connected in AraNet (Fig. 6A), aPLS-DA could separate the treatments O from Pand OP, explaining 28% of the variation in gene ex-pression and 60% of the treatment effects (Fig. 6B;Supplemental Table S9). In an orthogonal projectionof latent structures-discriminant analysis (OPLS-DA), the single and sequential O3 treatments (O andOP) could be divided further by the activity of eightgenes (VIP score . 1): LIP1, GPDHc1, SDP6, PLT5,PDAT, DGD1, SQD2, and PAD4 (Fig. 6, C and D; formodel statistics, see Supplemental Table S10). Par-ticularly, the triacylglycerol lipase LIP1 (At2g15230;fatty acid catabolism; El-Kouhen et al., 2005) wasup-regulated in O comparedwith OP (P, 0.05; Fig. 6,E and F), whereas the sulfolipid synthase SQD2(At5g01220; biosynthesis of photosynthetic mem-brane components) was significantly down-regulatedin O (P = 0.02) but not in OP (Fig. 6, E and F; KEGG/KaPPA-View4 paths are shown in Supplemental Figs.S6 and S7). Moreover, the glycerol 3-phosphate (G3P)dehydrogenases GPDHc1 and SDP6 (At2g41540 andAt3g10370; Shen et al., 2006) were both up-regulatedin O and negatively affected by herbivory in OP (P,0.05; Fig. 5E). Pathway visualization in KEGG/KaPPA-View4 (Fig. 6F) showed that GPDHc1 and SDP6(located in the cytosol and on the mitochondrialmembrane, respectively) constitute the G3P shuttlethat is responsible for transporting reducing equiva-lents to the mitochondrial ETC via dihydroxyacetonephosphate recycling (Shen et al., 2003, 2006; Quettieret al., 2008).

Predictive Interactions of GO Functional Networks

During the shift between O3 and herbivore stress,B. nigra actively regulated energy and glycerol meta-bolic processes (Figs. 2, 3, 5, 6), possibly in connectionwith the osmotic stress response and stomatal regula-tion (Fig. 4). To better describe the system transitionsduring the stress treatments (O, P, and OP), we per-formed a comparative correlation analysis (Steuer,2006) between the regulation of glycerol, G3P shuttle(GPDHc1/SDP6), mitochondrial ETC (SDH1), and sto-matal closure (NOGC1/FMO; Fig. 7, A and B), while GOnetwork analyses (AraNet and GeneMANIA) providedbiological insights into potential functional relation-ships between these processes (Fig. 7, C and D). InAraNet, GO functional associations were predictedfor genes involved in glycerol metabolism that moststrongly responded to the dynamics of the sequentialtreatment (Fig. 5). Overall, the network was enriched inprocesses of fatty acid biosynthesis, mitochondrialmetabolism (G3P shuttle), photosynthesis (chlorophyllbiosynthesis, sulfolipids, and photosystem stabiliza-tion), response to phosphate starvation, and stomatalclosure (Table II). New functional links between theseprocesses highlighted the connection between glycerolmetabolism (particularly via the G3P shuttle; GPDHc1and SDP6) and central energy metabolism (e.g. thepentose phosphate shunt, NADPH regeneration, gly-colysis, ATP synthesis, and mitochondrial ETC; Fig.7C). In AraNet, SDP6 was linked to SDH1-1 (Fig. 7C),and GeneMANIA specifically predicted that the corre-sponding proteins interact and are coexpressed (Fig.7D). In keeping with this result, we found that SDH1-1 and SDP6 were strongly coexpressed during O3stress in O (r = 0.97, P = 0.005) and during herbivorestress in P (r = 0.95, P = 0.01; Fig. 7A). Up-regulation of

Table I. Major gene hubs in the B. nigra network for multiple stress response to O3 and herbivory (Fig. 4)

Arabidopsis Genome

Initiative No.Gene Symbol Gene Name GO Processa Cellular Locationb

At4g27740 – Yippee family putative zinc binding Unknown NucleusAt4g16590 CSLA1 Cellulose synthase-like A01 Glycosyl transferase Golgi, cytosolAt5g04440 – Unknown function (DUF1997) Unknown ChloroplastAt1g12390 – Cornichon protein (guard cells) Signal transduction Plasma membraneAt2g29100 GLR2.9 Glutamate receptor2.9 Ca2+ homeostasis Plasma membrane, Golgi,

endoplasmic reticulumAt3g50830 COR413 Cold acclimation WCOR413-like Unknown Plasma membraneAt1g29910 LHCB1.2 Light-harvesting chlorophyll-binding1.2 Photosynthesis ChloroplastAt2g22900 MUCI10 Mucilage-related10 Galactosyltransferase Golgi, trans-GolgiAt2g32990 GH9B8 Glycosyl hydrolase9B8 Cellulose biosynthesis Extracellular (cell wall)At3g09360 – TBP-binding protein RNA polymerase II (TF) NucleusAt1g62580 NOGC1 Nitric oxide-dependent guanylate cyclase1 Stomatal closure Chloroplast, cytosolAt3g50770 CML41 Calmodulin-like41 Signaling, Ca2+ binding ChloroplastAt5g58270 M3 ABC transporter mitochondrion3 Mo-cofactor biosynthesis Mitochondria, chloroplastAt3g24100 – SERF (uncharacterized) Unknown NucleusAt1g07040 – Unknown protein Unknown ChloroplastAt2g02390 GST18 Glutathione S-transferase18 Amino acid biosynthesis Cytoplasm, cytosol

aGO biological processes and/or molecular functions as reported by The Arabidopsis Information Resource database (https://www.arabidopsis.org).bCellular localization confirmed with the ePlant visualization tool (BAR; University of Toronto; http://bar.utoronto.ca/;dev/eplant).









https://www.arabidopsis.org

http://bar.utoronto.ca/&tnqh_x223c;dev/eplant

http://bar.utoronto.ca/&tnqh_x223c;dev/eplant


SDH1-1 persisted in the sequentialOP treatment (Fig. 7B),but its coexpressionwith SDP6was reduced (r = 0.48, P =0.07; Fig. 7A). Glycerol metabolism also was associatedwith osmotic stress, stomatal closure, and the ABA re-sponse via interactions of SDP6 with the flavin mon-oxygenases NOGC1/FMO and SRE1/ABA2 (At1g52340;Nambara et al., 1998; Fig. 7C). NOGC1 and FMO werenegatively correlated with glycerol accumulation duringO3 stress inO (especiallyFMO; r =20.91,P= 0.03) but notin OP (Fig. 7, A and B). Moreover, SDP6 interacted withwater-glycerol protein channels (NIP aquaglyceroporins,vacuole BETA-TIP, and GAMMA-TIP3) and with thephosphate transporterPHT2;1 (viaNOGC1/FMO; Fig. 7D).

Physiological Measurements of Photosynthesis andGas Exchange

Combined omics and network analyses highlighted theconnection of glycerol and energy metabolism with theregulation of the osmotic stress response and stomatalclosure. To assess the actual impact ofO3 andherbivory onB. nigra, we performed a second experiment where wemeasured phenotypic and physiological parameters for

plants exposed to the stress treatments O, P, and OP aswell as a long-term (16-d) O3 stress treatment (OL; Fig.8A). Although few individuals showed visible symptomsof early senescence and chlorosis, the chlorophyll contentof the three youngest fully expanded leaves (L5–L7) de-creased by 9.5% after 5 d of O3 exposure (P, 0.05, n = 20;Fig. 8B). Herbivory alone did not directly affect chloro-phyll levels, but plants exposed previously to O3 in thesequential OP treatment had chlorophyll levels 13.8%lower than plants subjected to the P treatment (P , 0.05,n = 10; Fig. 8B). The deleterious effect of O3 was furtherdemonstrated by the long-term exposure (OL), in whichplants exhibited strong symptoms of senescence andchlorosis, particularly on the central fully expanded leaves(L5–L7; Fig. 8A). Chlorophyll levels in the OL plants wereconsistently around 47.8% lower than those in controlplants of the same age (P , 0.001, n = 10; Fig. 8B).

In B. nigra plants, 5 d of O3 stress (O) also reducedphotosynthetic activity and intracellular CO2 levels andnegatively affected stomatal conductance and leaf tran-spiration (Table III). After 16 d of exposure (OL), stomatalconductance decreased even further, but intracellular CO2levels increased. Twenty-four hours of herbivore stress by

Figure 5. Regulation of energy metabolic networks in B. nigra under multiple O3 and herbivory stress treatments. A, Network model(AraNet) assessing gene connectivity between the 85 genes enriched in energy metabolic processes. Model fitness was calculatedbased on the gene interconnectivity of the receiver operating characteristic (ROC) for the true-positive and false-positive rates betweenthe entry genes (red curve) and a randomly generated gene set (green curve). The network achieved a high area under the curve score(0.89; P = 1.82 E-62). B and C, Multivariate analysis evaluating the change in expression of energy genes in the network during themultiple stress treatments:O3 (O), herbivoryby P. brassicae (P), and sequential treatment (OP). PLS-DA scores and loading plots of firstand second latent variables, LV[1] versus LV[2], are shown. Five LVs cumulatively explained 54% of the variation in X (R2X[c]) and99%of the treatment response (R2Y[c]),with 67% totalmodel predictability (Q2[c]). The importanceofmitochondrial genes SDH1-1 andMSD1 in themodel is highlighted in red (VIP score =1.40 and1.34; Supplemental Tables S4 and S5).Dand E,Multivariate analysis for thegeneral model (970 genes; P# 0.05). PLS-DA scores and loading plots of first and second latent variables, LV[1] versus LV[2], are shown(Supplemental Tables S6 and S7). SDH1-1 (VIP score = 1.57) and ALDH7B4 (VIP score = 1.45), in red, are important for O3 treatments(O andOP). F, Distribution of VIP values for all 970 genes and correlation between the expression of SDH1-1 andALDH7B4 throughoutall treatment conditions (r = 0.65, P = 0.005).


Papazian et al.





P. brassicae (P) did not affect photosynthesis but slightlyreduced stomatal conductance and leaf transpirationcomparedwith untreated controls. However,when plantswere exposed to herbivory after O3 stress in the sequentialOP treatment, photosynthesis, stomatal conductance, andleaf transpiration all were reactivated. The sequentialstress thus reversed the responses induced by either stressalone (Table III). Because stomatal regulation follows thecircadian clock and decreases during the day, we verifiedthese conductance measurements via steady-state porome-try (Fig. 8, C and D). These measurements supported ourinitial findings, showing that conductance decreased in re-sponse to a single stress in theO (P=0.07; Supplemental Fig.S10) and P (P = 0.02) treatments but that the sequential OPtreatment induced stomatal reopening (P = 0.004; Fig. 8F).

DISCUSSION

Multiple stresses to plants may evoke unpredictedmolecular responses with negative, neutral, or positiveconsequences for plant metabolism. We found molecularevidence of changes in the regulation of photosynthesis

and mitochondrial activity during the responses ofB. nigra to sequential O3 stress and herbivore stress byP. brassicae. O3 induced the suppression of photosynthesisand stomatal closure, but this response was redirected tohigher photosynthetic activity after sequential exposureto herbivores. By using omics multivariate and networkanalyses, we identified glycerol metabolism as a centraldriver of this shift. As predicted by our combined omicsmodels, stomatal conductance and gas exchange wereenhanced after the sequential stress treatment, confirmingthe occurrence of a strategic change affecting photosyn-thetic activity and energy metabolism. This response tosequential stresses couldnot have beenpredicted from theindividual stress responses alone.

Effects of O3 Stress on Photosynthesis andStomatal Regulation

High levels of O3 negatively affect photosynthesis inplants (Bagard et al., 2008; Salvatori et al., 2015; Vainonenand Kangasjärvi, 2015). O3 oxidizes thylakoid membranesin the chloroplasts, resulting in chlorophyll degradationand early leaf senescence (Bergmann et al., 1999; Ranieri

Figure 6. Regulation of glycerol metabolic networks in B. nigra under multiple O3 and herbivory stress treatments. A, Networkmodel (AraNet) assessing gene connectivity between the 78 genes enriched in glycerol metabolic processes. The networkinterconnectivity scored high model fitness (area under the curve score = 0.99; P = 9.3 E-57) for true-positive and false-positiverates between the entry genes (red curve) and a randomly generated gene set (green curve). B, Multivariate analysis (PLS-DA)evaluating the change in expression of glycerol genes in the network during the multiple stress treatments: O3 (O), herbivory byP. brassicae (P), and sequential treatment (OP). Five latent variables (LVs) cumulatively explained 74% variation in X (R2X[c]) and100% of the treatment response (R2Y[c]), with 78% total model predictability (Q2[c]; Supplemental Tables S8 and S9). C, OPLS-DA modeling differences for glycerol genes between O and OP (Supplemental Table S10). D, OPLS-DA loading plot related to C(VIP score. 1), colored in blue and green, respectively for their abundance in O andOP. E, Univariate analysis of genes selectedfrom the OPLS-DA model. Asterisks indicate two-tailed Student’s t test significance of P , 0.05 between O and OP. Error barsindicate SE. F, Pathway analysis in KaPPA-View4 for glycerol metabolism highlighting the two G3P dehydrogenases GDPDHc1(cytosolic) and SDP6 (mitochondrial) of the mitochondrial G3P shuttle. The color code indicates up-regulation (red) and down-regulation (green; log expression).









et al., 2001; Goumenaki et al., 2010). However, even beforesymptoms of bleaching and chlorosis appear, as demon-strated in this study, the senescence process may be fullyinitiatedwith the down-regulation of chlorophyll and light-harvesting genes and the induction of the transcriptionfactor involved in leaf senescence, EIN3 (Long and Naidu,2002; Potuschak et al., 2003; Li et al., 2013). In addition tosuppressing photosynthesis, chronic O3 exposure above40nLL21 triggers a reactive oxygen species (ROS) signalingcascade that causes stomatal closure to prevent O3 fromentering the leaf. However, this process also limits CO2

uptake (Bergmann et al., 1999; Ranieri et al., 2001; Bookeret al., 2009;CastagnaandRanieri, 2009;Vahisalu et al., 2010;Settele et al., 2014). Our study confirmed anegative effect ofO3 on B. nigra stomatal conductance and leaf transpiration,which in the short term also led to lower levels of intra-cellular CO2. After long-term exposure (70 nL L21 O3 for16 d), signs of bleaching and chlorosis in leaves becameevident, whereas reduced stomatal conductance and lowphotosynthesis rates were associated with increased levelsof intracellular CO2, a condition that can reinforce stomatalclosure (Paoletti and Grulke, 2005; Singh et al., 2009).

Figure 7. Components of the glycerol metabolic network and mitochondrial ETC under multiple O3 and herbivory stresstreatments. Comparative correlation analysis and predictive GO network interactions (AraNet and GeneMANIA) are shown. A,Comparative correlation analysis between glycerol, components of the G3P shuttle (GPDHC1 and SDP6), mitochondrial ETCcomplex II (SDH1-1), and flavin monoxygenases (NOGC1 and FMO). Treatments are O, P, and OP. Edges indicate r for positive(blue) and negative (red) correlations. All correlation values reported were significant (between P, 0.05 and P, 0.001), exceptthe two correlations in the sequential (OP) treatment of glycerol/FMO (r = 0.85) and SDP6/SDH1 (r = 0.48), which were notsignificant. B, Average relative gene expression (log2) and glycerol abundance for each treatment (O, P, and OP) and corre-sponding Student’s t test significance (*, P, 0.05 and **, P, 0.01) compared with the control conditions. C, Gene interactionspredicted in AraNet for the glycerol metabolic network. Entry genes (in black squares) and emerging new members of a pathwayare colored by GO categories after enrichment analysis (GOlorize/Cytoscape). Predicted interactions with members of the mi-tochondrial ETC complex II, SDH, are highlighted in the box. D, Gene interactions generated in GeneMANIA between querygenes SDH1-1, SDP6, GPDHc1, NOGC1, and FMO. Node functions related the genes to processes of energy and mitchonodriametabolism. Functional links indicated coexpression (purple), colocalization (blue), predicted interaction (orange), and sharedprotein domains (brown) between the network components.


Papazian et al.



O3 stress also induced the expression of MYB44, agene that may play a complex role in the responsesto abiotic and biotic stresses such as drought andwounding (Baldoni et al., 2015).MYB44 overexpressionin Arabidopsis mutants increases drought toleranceby modifying the regulation of stomatal closure (Junget al., 2008). However,MYB44 also negatively regulatesABA responses (Jaradat et al., 2013; Li et al., 2014) in-volved in stomatal closure, leaf senescence, and ROSscavenging (Persak and Pitzschke, 2014). In our study,MYB44 expression correlated positively with glycerollevels and negatively with the expression of NOGC1, anitric oxide-dependent guanylyl cyclase involved instomatal closure via Ca2+ signaling (Mulaudzi et al.,2011; Joudoi et al., 2013). In addition,MYB44 correlatedpositively (r = 0.85, P = 0.001) with the chloroplasticlipid transfer protein LTPc1 (At2g10940), which togetherwith LTPc2 (At2g45180) was up-regulated during O3stress but not herbivore stress (Supplemental Figs. S3 andS4; Supplemental Table S3). LTPs are known to be activein the transfer of glycerolipids between cell membranes(e.g. from chloropasts to the endoplasmic reticulum; Xuet al., 2008). These results suggest a coordinated regula-tion of MYB44 and NOGC1 in response to O3, whichcould serve as a feedback mechanism for ABA signalingand stomatal closure, potentially connecting central en-ergy metabolism and glycerolipid pathways involved inosmotic stress responses.

Importance of Glycerol and Energy Metabolism asSafety Valves

Glycerol metabolism allows plants to adapt to a rangeof environmental stresses. Spinach (Spinacia oleracea)leaves, for example, accumulate triacylglycerol derivedfrom membrane galactolipids in response to O3 fumiga-tion (Sakaki et al., 1990). The regulation of glycerolipidpathways also is induced inArabidopsis, wheat (Triticumaestivum), and saltbush (Atriplex spp.) in response to

temperature stress, which results in diacylglycerol traf-ficking from the endoplasmic reticulum to the chloro-plasts (Li et al., 2015). In Arabidopsis, heat, salt, anddrought also induce triacylglycerol accumulation in thecytosol, an adaptation that enables structural remodelingof membrane lipids (Mueller et al., 2015).

The combined suppression of photosynthesis andreduction in chlorophyll levels observed in our studysuggests that glycerol may originate from the O3 stress-induced degradation of chloroplasts and glycolipidmembranes. Increased glycerol levels also were cor-related with changes in the expression of genes in-volved in stomatal closure and osmotic stress responses(MYB44, NOGC1, and FMO). Glycerol has osmolyticproperties, and its accumulation in mutants that lackglycerol kinase enhances resistance to dehydration stressin Arabidopsis (Eastmond, 2004).

GO network analyses further emphasized a connectionbetween osmotic stress responses, glycerol metabolism,and central energy processes in chloroplasts and mito-chondria. In the absence ofCO2, photons harvested duringphotosynthesis cannot be assimilated, so plants suppressphotosynthesis and activate nonphotochemical quenchingto protect cellular structures against excess excitation en-ergy (Niyogi, 2000; Murata et al., 2007). Similarly, mito-chondria are sinks for excess electrons generated duringoxidative stress and oxidize reducing equivalents via res-piration (Hoefnagel et al., 1998;Niyogi, 2000; Scheibe et al.,2005; Noctor et al., 2007; Nunes-Nesi et al., 2008). Weobserved that O3 exposure increased the activity of themitochondrial manganese superoxide dismutase MSD1(Tsang et al., 1991; Martin et al., 2013; Fig. 5D), suggestingthat mitochondria also play a role in the oxidative stressresponse to O3. This hypothesis was further supportedby the increased activity of the mitochondrial succinatedehydrogenase SDH1-1 and the G3P shuttle (GPDHc1/SDP6). In the mitochondrial ETC, SDH1-1 acts as thebinding site for coenzyme Q in complex II (Huang et al.,2013), while the G3P shuttle is pivotal in supplying the

Table II. Biological processes relative to the B. nigra glycerol metabolic network responsive to O3 and herbivory

Rank GO Identifier Biological Process P a Genes

1 GO:0006636 Unsaturated fatty acid biosynthesis 0.0001632 SQD2, LIP1, PDAT2 GO:0006127 Glycerol phosphate shuttle 0.0002918 GPDHc1, SDP63 GO:0016117 Carotenoid biosynthesis 0.0003421 SQD2, LIP1, PDAT4 GO:0019375 Galactolipid biosynthesis 0.0003564 DGD1, SQD25 GO:0015995 Chlorophyll biosynthesis 0.0004397 DGD1, SQD2, LIP16 GO:0046506 Sulfolipid biosynthesis 0.0005835 DGD1, SQD27 GO:0016036 Phosphate starvation 0.0006321 DGD1, SQD28 GO:0019563 Glycerol catabolism 0.0008752 GPDHc1, SDP6, LIP19 GO:0042550 PSI stabilization 0.0008752 DGD1

10 GO:0019761 Glucosinolate biosynthesis 0.0009609 SQD211 GO:0006072 Glycerol-3-phosphate metabolism 0.001167 GPDHc1, SDP612 GO:0009247 Glycolipid biosynthesis 0.001458 DGD1, SQD213 GO:0019288 Isopentenyl diphosphate biosynthesis

(methylerythritol 4-phosphate pathway)0.001849 DGD1, SQD2

14 GO:0090332 Stomatal closure 0.002915 NOGC115 GO:0006071 Glycerol metabolism 0.00466 GPDHc1, SDP6

aGO term P , 0.01 calculated in AraNet as a hypergeometric test on all 27,416 gene entries in the Arabidopsis database.









ETC with redox energy derived from NADH (McKennaet al., 2006; Shen et al., 2006; Berg et al., 2012; Mrá�ceket al., 2013). Curiously, we found that some phosphateresponse genes (RNS1 and PT2) also were up-regulatedin response to O3 stress, and our GO network analyses

linked glycerolipid metabolism to phosphate starvation(SQD2 and DGD1). Phosphate starvation symptoms aresimilar to those of O3 stress, including down-regulationof photosynthesis, lowCO2 assimilation, photooxidationof membrane lipids (Hernández and Munné-Bosch,

Figure 8. Physiological responses in B. nigra under multiple O3 and herbivory stress treatments. A, Leaf phenotypes of 5-week-oldB. nigra exposed toO3 fumigation at 70 nL L21 for 5 d (O), herbivore feedingwith 30 first instar P. brassicae caterpillars for 24 h(P), sequential stress of O3 followed by herbivore feeding (OP), long-term exposure to O3 at 70 nL L21 for 16 d (OL), or notreatment as controls (C). B, Relative chlorophyll content in leaf tissues determined by optical A653 for the three youngest fullyexpanded leaves (L5–L7). Student’s t test values are shown between C and O (*, P, 0.05; n = 20), Pand OP (*, P, 0.05; n = 10),and C andOL (***, P, 0.001; n = 10). C to E, Stomatal conductance (mmol water m22 s21) of fully expanded leaves (L6) measuredvia steady-state porometry for three experimental setups on separate days (n = 10 per treatment per day). F, Differences in stomatalconductance between the treatments and their controls measured simultaneously. A one-tailed Student’s t test evaluated if themean differences between treatment responses were larger than zero: C versus O (P = 0.07), C versus P (P, 0.05), and O versusOP (**, P , 0.01). Error bars indicate SE.


Papazian et al.



2015), membrane lipid remodeling with down-regulationof SQD2 (Jost et al., 2015), and transfer of digalactosyldia-cylglycerol fromchloroplasts tomitochondria (Jouhet et al.,2004).Overall, the cumulative regulation of glycerol metabo-

lism observed in our study may reflect a flux reconfigu-ration in B. nigra to sustain mitochondrial activity, whichled to the accumulation of glycerol as an intermediatemetabolite in the pathway (Kleijn et al., 2007; Morandini,2013; Gomes de Oliveira Dal’Molin et al., 2015).

Sequential O3 Stress and Herbivory Induce Abiotic andBiotic Cross Talk

Herbivory by P. brassicae induced the down-regulationof photosynthesis and reduced carbon assimilation rates,although herbivore stress was less severe than O3 stress.Chewing herbivores systemically reduce photosyntheticactivity in damaged leaves and in neighboring tissues(Zangerl et al., 2002; Bilgin et al., 2010; Halitschke et al.,2011), whereas JA signaling and LOXs directly affectphotosynthesis and ETC activity in chloroplasts (Nabityet al., 2013; Havko et al., 2016). The underlying mecha-nisms that suppress photosynthesis in favor of induceddefense responses are not fully understood, but a tradeoffbetween growth and defense appears to determine howplants rearrange their metabolism and redirect primaryresources to the production of specialized defensivecompounds (Schwachtje and Baldwin, 2008; Tang et al.,2009; Meldau et al., 2012). Herbivory induced partialstomatal closure in B. nigra, although we did not observeany distinct up-regulation of MYB44, as would be ex-pected after biotic stress and wounding (Jung et al., 2010;Persak and Pitzschke, 2013; Shim et al., 2013). AlthoughO3 stress by itself induced the up-regulation of MYB44and stomatal closure, O3 stress followed by herbivorestress caused the stomata to reopen. Consequently, thephenotype resulting from the sequential stresses resem-bled that observed after herbivore damage alone, with arelatively low stomatal conductance.Sequential herbivory positively induced the expres-

sion of MYC2 (JA signaling) and negatively affectedEIN3 (ET signaling) expression, which was induced byO3.MYC2 and EIN3 are key integrators of plant abioticand biotic stress responses (Abe et al., 2003; Andersonet al., 2004; Fujita et al., 2006; Dombrecht et al., 2007;

Atkinson et al., 2013). In a mutually antagonistic in-teraction, JA-activated MYC2 represses the transcrip-tion of EIN3/EIL1, whereas the induction of EIN3/EIL1reciprocally represses MYC2 and JA responses (Songet al., 2014, 2015; Zhang et al., 2014; Kim et al., 2015).

Similarly, ERF2, which is involved in ET signaling(Fujimoto et al., 2000) and positively regulates JA re-sponses (McGrath et al., 2005; Pré et al., 2008), wasdown-regulated after exposure to O3 stress alone butwas induced during herbivore stress. This asymmetricregulation of the ABA and JA/ET pathways and theobserved relationship with stomatal behavior suggest aspecific cross talk that balances the metabolic responsesto O3 and herbivory in the sequential stress situation.

O3 and Herbivory Affect Central Metabolism inOpposite Ways

In a previous study, we showed that the interactionbetween O3 stress and herbivory in B. nigra inducedchanges in the plant’s distribution of secondary me-tabolites (glucosinolates and phenolics). Moreover,while O3 fumigation at 120 nL L21 promoted feedingdamage by P. brassicae, caterpillars that fed on the fu-migated leaves exhibited reduced performance andfitness, with delayed development and lighter pupae(Khaling et al., 2015).

Here, we show that central (primary) metabolism alsoplays a pivotal role in the way plants respond to con-current abiotic and biotic stressors. Both transcriptomicsand metabolomics indicated that O3 stress inducedchanges in carbon and nitrogenmetabolism. Specifically,pools of central metabolites were expanded, while theactivity of many genes involved in carbohydrate me-tabolism and cell wall biosynthesis was reduced, in-cluding that of F2KP (for Fru-2,6-bisphosphatase/phosphofructokinase), the central regulator of glycolysisand gluconeogenesis (Draborg et al., 2001; Nielsen et al.,2004;McCormick andKruger, 2015; Fig. 5B). In addition,several genes involved in amino acid metabolism wereup-regulated, suggesting an increased flux toward mi-tochondrial activity and nitrogen mobilization, both ofwhich are cellular processes associated with senescence(Bouché et al., 2003; Li et al., 2006; Breeze et al., 2011;Debouba et al., 2013 Watanabe et al., 2013). Similar ef-fects of O3 stress on central metabolism have been

Table III. Photosynthetic and gas-exchange measurements of 5-week-old B. nigra plants under O3 and herbivore stress

Values are means 6 SE (n = 6–10), and significant variation was calculated by Student’s t test as follows: *, P . 0.05; **, P . 0.01; and ***, P .0.001; n.s., not significant. Relative group comparisons are indicated. Treatment abbreviations are as follows: controls (C); exposure to 70 nL L21 O3

for 5 d (O); exposure to 70 nL L21 O3 for 16 d (OL); P. brassicae herbivory for 24 h (P); and sequential 70 nL L21 O3 and herbivory (OP).

ParameterO3 (5 d, 70 nL L21) P. brassicae Herbivory (24 h) O3 (16 d, 70 nL L21)

C O Versus C P Versus C OP Versus O OL Versus C

Photosynthesis (mmol CO2 m22 s21) 11.72 6 0.8 8.63 6 0.63** 11.5 6 0.58 n.s. 11.24 6 0.85* 3.29 6 0.26***

Stomatal conductance (mol water m22 s21) 0.31 6 0.02 0.15 6 0.04** 0.25 6 0.10 n.s. 0.24 6 0.03* 0.12 6 0.01**Leaf transpiration (mol water m22 s21) 3.45 6 0.22 1.84 6 0.22*** 2.97 6 0.90 n.s. 2.66 6 0.26* 1.55 6 0.19***Intracellular CO2 concentration (mL L21) 300 6 3.87 264 6 4.93* 291 6 9.35 n.s. 288 6 9.81 n.s. 333 6 10.19***





reported previously: changes in the activity and ex-pression of genes and pathways involved in detoxificationand redox balance (e.g. the aldehyde dehydrogenaseALDH, chloroplastic superoxide dismutase activity, andNADH regeneration) were linked to stomatal closure,decreased photosynthetic activity, and increased mito-chondrial respiration (Dizengremel et al., 2009; Yendreket al., 2015). However, when herbivory was applied as asecond stress, all these initial effects of O3 stress were re-versed, resulting in decreased levels of sugars and aminoacids and a reconfiguration of gene expression, includingdown-regulation of the amino acid transporter AAP2,which promotes nitrogen accumulation in siliques andseed development in Arabidopsis (Hirner et al., 1998;Ortiz-Lopez et al., 2000).

Interestingly, a common response to both O3 andherbivory was an increased abundance of the centralmetabolite GABA. In mitochondria, the GABA shuntdelivers additional succinate and NADH to the ETCand provides an alternative route to amino acid me-tabolism (Bouché et al., 2003). During abiotic and

biotic stresses, GABA concentrations can spike rap-idly, but its function in plant responses remains un-clear (Bouché and Fromm, 2004; Fait et al., 2008).Studies have shown that GABA is involved in stresssignaling processes, from leaf senescence (Ansari et al.,2005) to plant communication with insects and mi-croorganisms (Shelp et al., 2006; Michaeli and Fromm,2015). Under conditions of oxidative stress andinhibited photosynthesis, GABA participates in ROSscavenging in support of normal growth and stresstolerance (Bouché et al., 2003; Dizengremel et al.,2012). GABA also has been proposed to regulate thecarbon and amino acid metabolism of plants throughinteraction with GABA/Glu receptors (GLR) in con-cert with differential regulation of the ABA/ET sig-naling pathways (Lancien and Roberts, 2006; Fordeand Lea, 2007). Under the conditions examined in thiswork, increased levels of GABA may contribute to themaintenance of central metabolism, oxidative stressresponses, and/or the regulation of stomatal closure,possibly via GLR and Ca2+ signaling.

Figure 9. Summary of omics and physiological responses in B. nigra during sequential O3 and herbivory stress treatments. Themodel links metabolome and transcriptome fluctuations to physiological responses of photosynthesis, CO2 assimilation, andstomatal opening. Stress adaptation mechanisms are proposed (1–5). Blue, O3 fumigation (5 d at 70 nL L21, 16 h per day); red, O3

followed by P. brassicae (24 h, 30 first instar caterpillars). Up arrows indicate up-regulation (genes) or increase (metabolites), anddown arrows indicate the opposite. 1 and 2, O3 induces the abiotic stress responses of senescence (EIN3 and ERF2) and stomatalclosure (MYB44), with feedback onNOGC1 and FMO. ABA and JA/ET cross talk integrates responses betweenO3 and sequentialherbivory (MYC2 and ERF2), with opposite effects on stomatal closure. 3 and 4, Photosystem suppression (LHCs) and non-photochemical quenching (NPQ1) in response to O3 are linked to the regulation of glycerolipid metabolism (LPTs, LIP1, andSQD2). Glycerol derived from degraded chloroplast membranes enters the G3P shuttle (GPDHc1/SDP6) to sustain NAD+

recycling and mitochondrial activity (SDH1-1,MSD1, and GABA) as antioxidative stress mechanisms. A possible role of glyceroland sugars as osmolytes also is suggested. Sequential herbivory restores the glycerolipid pathway for alternative source-sinkpriorities (e.g. JA responses [LOXs/MYC2]). 5, GABA plays multiple roles in plant stress adaptation, regulating Ca2+ homeostasis,carbon-nitrogen metabolism, leaf senescence, ROS scavenging, and signaling of plant-insect interactions.


Papazian et al.



In conclusion, we propose the existence of a tolerancemechanism in B. nigra (Fig. 9) where glycerol and cen-tral energy metabolism play a central role in adaptationto the sequential stresses of O3 exposure and herbivory,enabling B. nigra to prioritize the demands of stomatalosmoregulation and oxidative stress. The suppressionof photosynthesis and the regulation of glycerol andmitochondrial metabolism during O3 stress may dissi-pate excess energy to avoid the formation of oxidativeradicals (Hoefnagel et al., 1998), combining fast NAD+

recycling with the maintenance of central metabolismand growth (Dizengremel et al., 2008, 2012). This hypoth-esis is supported by previous studies on the physiologicalfunctions of glycerol metabolism in NADH/NAD+ ho-meostasis (Shen et al., 2006; Quettier et al., 2008), osmoticstress responses (Biela et al., 1999; Eastmond, 2004; Chenet al., 2011; Geijer et al., 2012), and plant development (Huet al., 2014). These pathways were redirected in the se-quential treatment, during which glycerolipid resourcesmay have been reallocated toward JA signaling and de-fense against herbivores (Turner et al., 2002; Kachroo et al.,2004; Havko et al., 2016). Alternatively, these reversed ef-fects may represent a manipulation of the host’s metabo-lism by the herbivore (Karban and Agrawal, 2002), whichcan interfere with plant defense and water stress responsegenes (Reymond et al., 2000; Consales et al., 2012).

MATERIALS AND METHODS

Plants

Seeds from Brassica nigra plants collected from a natural population growingalong the Rhine River in Wageningen were obtained from the Laboratory ofEntomology ofWageningen University. The seeds were planted individually inplastic pots (93 93 9.5 cm) filled with a 3:1:1 mix of peat, potting compost, andsand. Theywere grown under greenhouse conditions at theUniversity of EasternFinland inKuopio. The plantswerewatered intermittentlywith sprinklers for 5 hevery day without chemical control for pests or diseases. The seedlings werefertilized twice per week with 0.1% 5-Superex (nitrogen:phosphorus:potassium,19:5:20). When the plants were 4 weeks old and had developed approximatelyseven leaves, they were taken to growth chambers (Weiss Bio 1300; WeissUmwelttechnik) and subjected to the different treatments for 6 d.

Treatments

B. nigra plants were subjected to one of four treatments: exposure to ambientair with and without feeding by Pieris brassicae caterpillars (C and P) and O3fumigation with and without subsequent feeding by P. brassicae caterpillars (Oand OP). The experiment was repeated five times (biological replicates), witheach replicate including three plants per treatment. Two extra replicates wereproduced for the evaluation of gene expression after exposure to ambient airand P. brassicae feeding (seven replicates in total). In the follow-up experiment,B. nigra was again treated as in C, O, P, and OP and with additional long-termO3 fumigation (OL).

O3 Fumigation

The plants (12 in total, three per treatment) were moved to plant growthchambers. The chambers had been modified so that each had an independentlycontrolled O3 concentration. High-O3 chambers were set to 70 nL L21 from 4 AM

to 8 PM and maintained at a basal O3 concentration of 30 nL L21 for theremaining hours each day. This treatment was done to imitate natural diurnalvariation in O3 concentration. In ambient chambers, the O3 concentrationfluctuated between 15 and 20 nL L21. Chambers were maintained at a tem-perature of 23°C 6 3°C, relative humidity of 60% during the day and 80% at

night, a photoperiod of 16 h of light/8 h of darkness, with a light intensity of300 mmol m22 s21. The plants were watered daily. The above conditions weremaintained for 5 d, after which the plants subjected to herbivore feeding wereinfested for 24 h. In the follow-up experiment, long-term O3 fumigation wasstarted on 3-week-old plants and continued for 16 d.

Herbivore Feeding

The large cabbage white butterfly, Pieris brassicae (Lepidoptera: Pieridae),was obtained from stocks at the Laboratory of Entomology, WageningenUniversity, and reared on Brussels sprouts plants (Brassica oleracea var gemmifera‘Brilliant’) at the University of Eastern Finland under greenhouse conditions.Before the experiments, P. brassicae adults were presented with B. nigra plantsfor oviposition, and first instar caterpillars were collected soon after hatching ina climate-controlled insect-rearing room with a temperature of 25°C 6 2°C, aphotoperiod of 16 h of light/8 h of darkness, light intensity of 300mmolm22 s21,and relative humidity of approximately 60%. For the herbivore treatment, atotal of 30 first instar P. brassicae caterpillars were mounted on the three highestfully expanded leaves of each plant (10 caterpillars per leaf) and left to feed for24 h. After 24 h (day 6), VOCs were collected and plant samples were harvestedfor metabolomics and transcriptomics analyses.

Physiological Measurements of Photosynthesis andGas Exchange

The relative chlorophyll content of leaf tissueswas determined bymeasuringthe optical A653 (CCM-200 plus; Opti-science) of the three youngest fully ex-panded leaves (L5–L7). Photosynthetic and gas-exchange parameters (i.e. thecarbon assimilation rate [mmol CO2 m

22 s21], intracellular CO2 levels, stomatalconductance, and leaf transpiration [mmol water m22 s21]) were measured forabout 1 h on one fully expanded leaf (L6) per plant (n = 10) using a LI-COR gasanalyzer (LI-6400). The leaf chamber parameters were set to mimic the ambientgrowth conditions, with block temperature at 24°C, relative humidity at 60%,CO2 at 400 mL L21, and saturating light at 1,000 mmol m22 s21. In addition, leafL6 stomatal conductance was determined via steady-state porometry (SC-1;Decagon Devices), which measured the actual water vapor flux (mmol waterm22 s21) from the leaf through the stomata and out to the environment.

Sampling for Metabolomics and Transcriptomics

At the timeof sampling, theplantswere 5weeks old.Counting from the apex,the three youngest fully expanded leaves (L5–L7) of each of the three plants(altogether nine leaves from three plants per treatment) were cut at the petiolewith a sharp knife and pooled together. These leaves were immediatelywrapped in aluminum foil and flash frozen in liquid nitrogen. The leaves werestored at280°C and later ground into a fine powder with liquid nitrogen usinga mortar and pestle. Equal amounts of powder from each sample were expressmailed on dry ice to the laboratories in Umea (SE) and Lausanne (CH) formetabolomics and transcriptomics analyses, respectively. Thus, each laboratoryreceived subsamples of the same material.

Metabolomics Analyses

All metabolomics analyses were performed at the Umeå Plant ScienceCenter-Swedish Metabolomics Center (UPSC-SMC) in Umea. Leaf tissue sam-ples for LC-MS and GC-MS were prepared by extracting 10 to 12 mg of frozensample using 1 mL of cold chloroform:methanol:water (20:60:20) containing7.5 ng mL21 labeled salicylic acid-D4 (m/z [M-H]2 141.046) as an internal stan-dard. A 3-mm tungsten carbide bead was then added to each vial, and thesamples were agitated for 3 min at 30 Hz in an MM 301 Vibration Mill (Retsch).To separate the mixture from tissue debris and avoid contamination, the ex-tracts were centrifuged at 20,800g for 10 min at 4°C, and 200 mL of the su-pernatant was separated and evaporated to dryness using a SpeedVac. ForGC-MS analysis, samples were derivatized using 30 mL of methoxyamine(15 mg mL21 in pyridine) and agitated for 10 min before being left to react for16 h at 25°C. Silylation was achieved using 30 mL of N-Methyl-N-(trimethylsilyl)trifluoroacetamide (MSTFA) for 1 h at 25°C. Finally, samples were diluted with30 mL of heptane containing 15 ng mL21 methyl stearate (internal standard) andinjected into the system. For LC-MS analysis, dried samples were redissolvedin 10 mL of cold methanol and diluted with 10 mL of cold water before beinginjected into the system.





GC-MS

One microliter of the derivatized sample was injected splitless using a CTCCombiPal autosampler (CTCAnalytics) intoanAgilent6890gaschromatographequipped with a 10-m 3 0.18-mm i.d. fused silica capillary column with achemically bonded 0.18-mmDB 5-MS UI stationary phase (J&W Scientific). Theinjector temperaturewas 270°C, the purge flow ratewas set to 20mLmin21, andthe purge was turned on after 60 s. The gas flow rate through the column was1 mLmin21; the column temperature was held at 70°C for 2 min, then increasedby 40°C min21 to 320°C, and held for 2 min. The column effluent was intro-duced into the ion source of a Pegasus III time-of-flight mass spectrometer(Leco). The transfer line and the ion source temperatures were set to 250°C and200°C, respectively. Ions were generated by a 70-eV electron beam at an ioni-zation current of 2 mA, and 30 spectra s21 were recorded in themass range 50 to800 m/z. The acceleration voltage was turned on after a solvent delay of 150 s.The detector voltage was set to 1,700 V.

LC-MS

For the analysis of secondary metabolites, samples were analyzed by ultra-high-performance liquid chromatography-electrospray ionization/time-of-flightmass spectrometry (UHPLC-ESI/TOF-MS; Waters). The Acquity system wasequipped with a 2.1- 3 100-mm, 1.7-mm C18 UPLC column (reverse-phasecolumn/nonpolar stationary phase) held at 40°C. The liquid chromatographysystem was coupled to an LCT Premier time-of-flight mass spectrometer. Twomicroliters of each sample was injected and separated using a mobile phasecontaining a mix of solvents A (water + 0.1% formic acid) and B (acetonitrile +0.1% formic acid). Gradient elution was performed using the following program,gradually increasing the proportion of solvent B over time: 0 to 4 min, 1% to 20%B; 4 to 6min, 20% to 40%B; 6 to 9min, 40% to 95%B; and 9 to 13.5min, 95%B. Thetotal running time for each sample was 19 min, with a flow rate of 500 mLmin21.The source temperature was 120°C, cone gas flow was 10 L h21, desolvationtemperature was 320°C, nebulization gas flow was 600 L h21, and capillary andcone voltages were set at 2.5 kV (negative ionization mode) and 35 V, respec-tively. Datawere acquired in dynamic range enhancementmode every 0.1 s,witha 0.01-s interscan delay. The lock mass compound for accurate mass measure-ments (Leu enkephalin) was infused directly at 400 pg mL21 in 50:50 acetonitrile:water at 20 mL min21. The normal lock mass in dynamic range enhancementmodewas the negative 13C ion of Leu enkephalin (m/z 555.265), and the extendedlockmasswas the normal negative ion (m/z 554.262). Mass spectrawere acquiredin centroid mode with anm/z range of 100 to 1,000, and the data threshold valuewas set to 3.

Orbitrap Tandem Mass Spectrometry

To verify the data acquired by UHPLC-ESI/TOF-MS, samples were rean-alyzed for the determination of selected peaks of interest by ultra-HPLC-tandem mass spectrometry using linear ion traps (LTQ Orbitrap). Separationwas performed on a Thermo Accela liquid chromatography system, equippedwith a column oven (held at 40°C) and a Hypersil C18 GOLD column (2.1 350 mm, 1.9 mm; mobile phase as for UHPLC-ESI/TOF-MS), and analyzed bytandemmass spectrometry using an LTQ Orbitrap mass spectrometer (ThermoFisher Scientific). External mass calibration was performed according to themanufacturer’s guidelines.

Data Processing and Identification

The GC-MS instrument was operated using the LECO ChromaTOF software(optimized for Pegasus HT; Leco). Retention time indexes (RIs) were calculatedrelative to an alkane series (C8–C40). From the raw data, feature extraction andpeak integration were all performed with Matlab, combining target analysis (apredefined list of retention time windows and m/z values) and automated peakdeconvolution. Compounds were identified comparing RIs andmass spectra withthe UPSC-SMC in-house database and with the public Golm Metabolome Data-base of the Max Planck Institute. For comparison with the Golm MetabolomeDatabase, RIs measured on the 5% phenyl-95% dimethylpolysiloxane capillarycolumn VAR5 (Golm Metabolome Database) were transferred to the DB-5 (10m)system of the UPSC-SMC (Strehmel et al., 2008; Hummel et al., 2010). Sampleswere normalized on the unit variance and no centering (UVN) scores of integratedareas for the internal standards (methyl stearate and salicylic acid-D4).

The UHPLC-ESI/TOF-MS instruments were operated with MassLynx ver-sion 4.1 software (Waters). Compounds from LC-MS analysis were compared

with standards of glucosinolates (sinigrin, glucobrassicin, gluconapin, gluco-tropaeolin, gluconasturtiin, and sinalbin [Phytoplan, Diehm & Neuberger]),data from the METLIN mass spectra depository, and additional literature ref-erences for glucosinolates (Clarke, 2010), hydroxycinnamic acid derivatives,and flavonol glucosides (Lin et al., 2011). Tandem mass data analysis from theOrbitrap were used to compare the MSn profiles and further confirm theidentifications. Raw data were processed using Sieve and Matlab software forpeak alignment and integration. The peak areaswere normalized against that ofthe labeled internal standard of salicylic acid-D4 (m/z [M-H]2 141.046).

VOCs Collection and Analyses

All VOCs collection and analyses were performed at the University of EasternFinland. The plants were enclosed in glass jars, and VOCs were collected for60min via dynamic headspace sampling. Filtered air was fed into the glass jars ata rate of 250 mLmin21 and pulled out at a rate of 200 mLmin21 through stainlesssteel tubes filled with Tenax TA and Carbopack B adsorbents (150 mg each; mesh60/80; Markes International). Samples were analyzed by GC-MS (Agilent 7890AGCand 5975CVLMSD). Trapped compoundswere desorbedwith an automatedthermal desorber (TD-100; Markes International) at 250°C for 10min, cryofocusedat210°C, and then injected in split mode onto an HP-5 capillary column (50 m30.2 mm; film thickness, 0.33 mm) with helium as the carrier gas. The oven tem-perature was held at 40°C for 1 min, raised to 210°C at a rate of 5°C min21, andfinally raised to 250°C at a rate of 20°Cmin21. The columnflowwasmaintained ata rate of 1.2 mL min21. The compounds were identified by comparing the massspectrum of an individual compound with the spectra of compounds in an ex-ternal authentic standard and with compounds in the Wiley Library. Relativeemissions were measured by peak integration (absolute rates expressed as nmolm22 h21 are reported in Supplemental Data Set S2).

Transcriptomics Analyses

Transcriptomics analyses were performed at the University of Lausanne.B. nigra leaves (3–6 g) were ground in liquid N2, and total RNA was extracted,reverse transcribed, and processed according to a previously published pro-cedure (Bodenhausen and Reymond, 2007). Labeled probes were hybridizedonto CATMA version 4 microarrays containing 32,998 Arabidopsis (Arabidopsisthaliana) gene-specific tags and gene family tags (Sclep et al., 2007). Hybridi-zation and scanning have been described previously (Reymond et al., 2004).Data analyseswere carried out using an interface developed at the University ofLausanne (Gene Expression Data Analysis Interface; Liechti et al., 2010). Dif-ferentially expressed genes were identified by fitting a linear model for eachgene and evaluating the fold change and moderated t statistic P values (Smyth,2004). To address the issue ofmultiple comparisons, we used the false discoveryratemethod developed by Storey and Tibshirani (2003) and computed a q value.Because we employed Arabidopsis whole-genome microarrays to probe theexpression of B. nigra genes, the number of genes that produced hybridizationsignals was clearly low and overall hybridization signal intensity was weakerthan with Arabidopsis samples. Hence, we noticed that high false discoveryrate values are estimated when the number of induced genes was relativelysmall. However, by comparing gene expression between experiments, geneswith small P values in response to one treatment often had a small P value inanother treatment. Thus, interexperiment comparison adds to data interpreta-tion, and false discovery rate calculations might be too conservative in somecases. Therefore, we used an unadjusted P value of 0.05. False discovery ratevalues are indicated in Supplemental Data Set S1.

Statistical Analyses

Matrices forgene expression andmetaboliteswere created inExcel.All stepsofbasic statistics (i.e. Pearson’s correlation, Student’s t test, ANOVA, and posthocTukey’s test) were performedwith Excel andwithMinitab 17 Statistical Software(Minitab; www.minitab.com). Othermore specific analyseswere performedwiththe open-source software R (https://www.r-project.org/) andRStudio (https://www.rstudio.com/) or with other software as mentioned below.

Multivariate Analyses

Gene expression and metabolite profiles were subjected to multivariateanalysis using the SIMCA 14 software package (Umetrics). Supervised re-gression models such as PLS-DA and OPLS-DA were used to investigate the


Papazian et al.




http://www.minitab.com

https://www.r-project.org/

https://www.rstudio.com/

https://www.rstudio.com/


variation in X variables (gene or metabolites), which were modeled for the Yexplanatory variables corresponding to O3 and herbivore treatments. The cu-mulative (c) variations in X and Y explained by the models are reported as R2X(c) and R2Y(c), respectively. Models were fit to the minimum number of latentvariables corresponding to the highest value of predicted variation: Q2(c). Se-lection of important variables was based on the VIP score, considered signifi-cant if above the threshold of 1.

Gene Expression Data Evaluation and Pathway Analyses

Gene selection for the Student’s t test cutoff ofP# 0.05was executed in Excel.The open-source software R and RStudio were used to create heat maps, hier-archical clustering, and correlation of gene expression profiles using the CRANlibrary packages and functions heatmap (gplots), hclust, and corrgram, re-spectively. More specifically, gene expression values for each sample (up- ordown-regulation) were represented graphically using the heatmap function,while rows (genes) and columns (samples) of the matrix were reordered indendrograms following hierarchical clustering. The default function hclust wasused with its method of complete linkage, which agglomerates clusters com-puting the largest distance between any object in one cluster and the otherobjects in the other clusters. Similarity between each sample group was furthertested using the corrgram function (default method of Pearson’s correlation).Thus, a correlation matrix was produced as a graphical display (i.e. the corre-logram) with cells colored according to the respective correlation coefficient (r)for each paired sample comparison.

GO enrichment analysis of the gene clusters was performed with the Web-based tool Functional Classification SuperViewer of the University of Toronto(http://bar.utoronto.ca/ntools/cgi-bin/ntools_classification_superviewer.cgi), where GO categories for genes in a given cluster are normalized for thefrequency in Arabidopsis, while bootstraps and SD provide confidence intervalsfor the accuracy of the output (Provart and Zhu, 2003). P values of the hyper-geometric distribution were used to select the significant functional classes.Only enrichments with significant P# 0.05were considered and reported in thegraphs. The MapMan tool (Thimm et al., 2004; http://mapman.gabipd.org/web/guest) was used to visualize the gene expression data set in the context ofmetabolic pathways or other processes represented in modules (bins) and tobuild the Venn diagrams for up- and down-regulated genes. In addition, weperformed pathway analysis and interpretation of omics data using the freedatabases KaPPA-View4 (Tokimatsu et al., 2005; Sakurai et al., 2011; http://kpv.kazusa.or.jp/) and the KEGG database (Kanehisa and Goto, 2000; http://www.genome.jp/kegg/). Gene subcellular localizations and expression weresearchedwith the ePlant server (https://bar.utoronto.ca/eplant/; University ofToronto).

Integrative Omics Network Analyses

The open-source platformCytoscape (version 3.2.1)was used to visualize andanalyze the omics networks. The correlation matrix between genes and metab-olites was computed into a similarity network using the ExpressionCorrelationapp (http://apps.cytoscape.org/apps/expressioncorrelation). The network wasbuilt setting the edge (links) parameters to r$ 0.85 for both positive andnegativecorrelations. For network visualization, the graphic layout was set to (yFile) or-ganic. Network topological features were evaluated with Network Analyzer(Max Planck Institut Informatik; Assenov et al., 2008). For comparative correlationanalysis, the network between glycerol and the five gene components GPDHc1(At2g41540), SDP6 (At3g10370), NOGC1 (At1g62580), FMO (At1g12200), andSDH1-1 (At5g66760) was produced with the open-source software R using theCRAN library package qgraph (Epskamp et al., 2012). Correlation significancewastested with Minitab 17.

GO Network Analyses

AraNet (http://www.inetbio.org/aranet) is a free database of cofunctionalgene networks based on The Arabidopsis Information Resource 10 annotations(Lee et al., 2015), which can be used for computational identification of newcandidate genes in functional pathways and for the integration of high-throughput omics data sets (Lee et al., 2010). Closely connected genes arelisted and ranked in a guilt-by-association network on the basis of previousexperimental data sets and GO evidence codes, such as IDA (inferred fromdirect assay), IPI (inferred fromprotein interaction), ISS (inferred from sequenceor structural similarity), and TAS (traceable author statement). Model accuracyand coverage are assessed with the ROC, which calculates the probability rate

of true positives versus false positives between the interconnectivity of the entrygenes and a selection of random genes. The ROC calculates themodel fitness forthe connected genes as true positive and false positive rates between the entrygenes (red curve) and a randomly generated gene set (green curve). The cor-responding score of the area under the ROC curve ranges from approximately0.5 to 1, indicating random and perfect performance. New candidate geneswere searched with the function new members of a pathway. The predictivepower of the new network is automatically estimated from the initial inter-connectivity of its genes on the basis of the connection to the entry genes (rankedby connectivity scores). The AraNet function GeneSet analysis/GO was used toevaluate enrichment of the network. The same GO analysis function also wasused to confirm the network enrichment in glycerol and glycerolipid metabolicprocesses of the gene set selection (by enzyme substrate annotation; 76 genes),which was later used for multivariate effects of the treatments. Results from theAraNet analysis were imported into Cytoscape and colored according to statis-tically overrepresented GO categories using the plugin GOlorize (Garcia et al.,2007; http://apps.cytoscape.org/apps/golorize).

Another public Web server was used for the prediction of biological inter-action, GeneMANIA (http://www.genemania.org/), which also can be used asa Cytoscape plugin (Warde-Farley et al., 2010). The initial entry list consisted offive genes, GPDHc1 (At2g41540), SDP6 (At3g10370), NOGC1 (At1g62580), andFMO (At1g12200), constituting the core region of the glycerol network inanalysis, and the predicted interaction with SDH1-1 (At5g66760). GeneMANIAextended this list to create a network of genes identified as having similarfunctions. The predicted gene interactions and their weights were estimated onthe basis of Arabidopsis knowledge through genomics and proteomics data(e.g. coexpression and protein interaction), which were retrieved from GEO,BioGRID, Pathway Commons, and I2D as well as organism-specific functionalgenomics data sets (Warde-Farley et al., 2010).

Figure Layout and Editing

PhotoshopCS5.1 (Adobe)wasused for editing thefinal graphic layouts of thefigures.

Accession Numbers

Microarray data from the transcriptomics analysis were deposited in theArrayExpress database (www.ebi.ac.uk/arrayexpress) under accession num-ber E-MTAB-5030. Sequence data from this article can be found in the Arabi-dopsis Genome Initiative or GenBank/EMBL databases under the followingaccession numbers: At2g30570 (PSBW ), At3g54890 (LHCA1), At1g19150(LHCA6), At2g34420 (LHB1B2), At2g40100 (LHCB4.3), At4g10340 (LHCB5),At1g15820 (LHCB6), At1g60950 (FED A), At1g08550 (NPQ1), At1g07110(F2KP), At2g21330 (FBA1), At4g38970 (FBA2), At3g54050 (CFBP1/HCEF1),At2g25540 (CESA10), At4g16590 (CSLA01), At2g22900 (MUCI10), At3g46970(PHS2), At5g66760 (SDH1-1), At3g10920 (MSD1), At4g31800 (WRKY18),At1g80840 (WRKY40), At2g46400 (WRKY46), At3g45140 (LOX2), At1g17420(LOX3), At1g32640 (MYC2), At5g67300 (MYB44), At2g37630 (MYB91),At1g72290 (WSCP), At3g45640 (MPK3), At5g45340 (CYP707A3), At5g47220(ERF2), At3g20770 (EIN3), AT2G02990 (RNS1), At2g38940 (PT2), At2g38170(CAX1), At1g68100 (IAR1), At1g62580 (NOGC1), At1g12200 (FMO), At1g29740(LRR-RK ), At1g29310 (SecY protein/sec61), At2g15230 (LIP1), At2g41540(GPDHc1), At3g10370 (SDP6), At3g18830 (PLT5), At5g13640 (PDAT), At3g11670(DGD1),At5g01220 (SQD2),At3g52430 (PAD4), At1g54100 (ALDH7B4), At2g37790(AKR4C10), At2g10940 (LTPc1), and At2g45180 (LTPc2).

Supplemental Data

The following supplemental materials are available.

Supplemental Figure S1. Metabolomics: PLS-DA score plot latent varia-bles.

Supplemental Figure S2. Integrative omics network analysis: topology.

Supplemental Figure S3. Chloroplastic lipid transfer protein (LTPc1;At2g10940).

Supplemental Figure S4. Chloroplastic lipid transfer protein (LTPc2;At2g45180).

Supplemental Figure S5. KEGG pathway analysis for glycerolipid metab-olism (O).




http://bar.utoronto.ca/ntools/cgi-bin/ntools_classification_superviewer.cgi

http://bar.utoronto.ca/ntools/cgi-bin/ntools_classification_superviewer.cgi

http://mapman.gabipd.org/web/guest

http://mapman.gabipd.org/web/guest

http://kpv.kazusa.or.jp/

http://kpv.kazusa.or.jp/

http://www.genome.jp/kegg/

http://www.genome.jp/kegg/

https://bar.utoronto.ca/eplant/

http://apps.cytoscape.org/apps/expressioncorrelation

http://www.inetbio.org/aranet

http://apps.cytoscape.org/apps/golorize

http://www.genemania.org/

http://www.ebi.ac.uk/arrayexpress







Supplemental Figure S6. KEGG pathway analysis for glycerolipid metab-olism (OP).

Supplemental Figure S7. Glucosinolates: correlation analysis.

Supplemental Figure S8. Flavonols: kaempferol glucosides, correlationanalysis.

Supplemental Figure S9. Flavonols: quercetin glucosides, correlation anal-ysis.

Supplemental Figure S10. Stomatal conductance: effect of 5 d of O3 stress.

Supplemental Table S1. Metabolomics: PLS-DA model statistics.

Supplemental Table S2. Metabolomics: metabolite identities and PLS-DAVIP scores.

Supplemental Table S3. Gene correlations with MYB44 in the integrativenetwork.

Supplemental Table S4. Transcriptomics: PLS-DA for the energy meta-bolic network.

Supplemental Table S5. Transcriptomics: gene identities and VIP scoresfor the energy network.

Supplemental Table S6. Transcriptomics: PLS-DA general model(970 genes).

Supplemental Table S7. Transcriptomics: gene identities and PLS-DA VIPscores.

Supplemental Table S8. Transcriptomics: genes in the glycerol metabolicnetwork.

Supplemental Table S9. Transcriptomics: PLS-DA for the glycerol meta-bolic network.

Supplemental Table S10. Transcriptomics: OPLS-DA for the glycerol met-abolic network.

Supplemental Data Set S1. Transcriptomics: microarray and selection (P,0.05).

Supplemental Data Set S2. Metabolomics: identifications and integratedpeak areas.

ACKNOWLEDGMENTS

We thank the University of Lausanne, the University of Eastern Finland, andTimo Oksanen for technical support; the University of Umeå and the UmeåPlant Science Centre for providing a pleasant and supportive working envi-ronment; the staff of the Swedish Metabolomics Centre, particularly KristerLundgren and Ilka Abreu, for technical assistance with the mass spectrometryanalyses; and Rui Climaco Pinto (BILS) for assistance with the bioinformaticanalysis and data evaluation.

Received August 23, 2016; accepted October 3, 2016; published October 6, 2016.

LITERATURE CITED

Abe H, Urao T, Ito T, Seki M, Shinozaki K, Yamaguchi-Shinozaki K(2003) Arabidopsis AtMYC2 (bHLH) and AtMYB2 (MYB) function astranscriptional activators in abscisic acid signaling. Plant Cell 15: 63–78

Anderson JP, Badruzsaufari E, Schenk PM, Manners JM, Desmond OJ, EhlertC, Maclean DJ, Ebert PR, Kazan K (2004) Antagonistic interaction betweenabscisic acid and jasmonate-ethylene signaling pathways modulates defensegene expression and disease resistance in Arabidopsis. Plant Cell 16: 3460–3479

Ansari M, Lee RH, Chen SC (2005) A novel senescence‐associated geneencoding g‐aminobutyric acid (GABA):pyruvate transaminase is upre-gulated during rice leaf senescence. Physiol Plant 123: 1–8

Ashmore MR (2005) Assessing the future global impacts of ozone onvegetation. Plant Cell Environ 28: 949–964

Assenov Y, Ramírez F, Schelhorn SE, Lengauer T, Albrecht M (2008) Computingtopological parameters of biological networks. Bioinformatics 24: 282–284

Atkinson NJ, Lilley CJ, Urwin PE (2013) Identification of genes involved inthe response of Arabidopsis to simultaneous biotic and abiotic stresses.Plant Physiol 162: 2028–2041

Atkinson NJ, Urwin PE (2012) The interaction of plant biotic and abioticstresses: from genes to the field. J Exp Bot 63: 3523–3543

Bagard M, Le Thiec D, Delacote E, Hasenfratz-Sauder MP, Banvoy J,Gérard J, Dizengremel P, Jolivet Y (2008) Ozone-induced changes inphotosynthesis and photorespiration of hybrid poplar in relation to thedevelopmental stage of the leaves. Physiol Plant 134: 559–574

Baldoni E, Genga A, Cominelli E (2015) Plant MYB transcription factors:their role in drought response mechanisms. Int J Mol Sci 16: 15811–15851

Bale JS, Masters GJ, Hodkinson ID, Awmack C, Bezemer TM, Brown VK,Butterfield J, Buse A, Coulson JC, Farrar J, et al (2002) Herbivory inglobal climate change research: direct effects of rising temperature oninsect herbivores. Glob Change Biol 8: 1–16

Barah P, Bones AM (2015) Multidimensional approaches for studying plantdefence against insects: from ecology to omics and synthetic biology. JExp Bot 66: 479–493

Bekaert M, Edger PP, Hudson CM, Pires JC, Conant GC (2012) Metabolicand evolutionary costs of herbivory defense: systems biology of gluco-sinolate synthesis. New Phytol 196: 596–605

Bennett R, Donald A, Dawson G, Hick A, Wallsgrove R (1993) Aldoxime-forming microsomal enzyme systems involved in the biosynthesis ofglucosinolates in oilseed rape (Brassica napus) leaves. Plant Physiol 102:1307–1312

Berg JM, Tymoczko JL, Stryer L (2012) Oxidative phosphorylation. In Bi-ochemistry. WH Freeman, New York

Bergmann E, Bender J, Weigel HJ (1999) Ozone threshold doses and ex-posure–response relationships for the development of ozone injurysymptoms in wild plant species. New Phytol 144: 423–435

Biela A, Grote K, Otto B, Hoth S, Hedrich R, Kaldenhoff R (1999) TheNicotiana tabacum plasma membrane aquaporin NtAQP1 is mercury-insensitive and permeable for glycerol. Plant J 18: 565–570

Bilgin DD, Zavala JA, Zhu J, Clough SJ, Ort DR, DeLucia EH (2010) Bioticstress globally downregulates photosynthesis genes. Plant Cell Environ33: 1597–1613

Bino RJ, Hall RD, Fiehn O, Kopka J, Saito K, Draper J, Nikolau BJ,Mendes P, Roessner-Tunali U, Beale MH, et al (2004) Potential ofmetabolomics as a functional genomics tool. Trends Plant Sci 9: 418–425

Bodenhausen N, Reymond P (2007) Signaling pathways controlling in-duced resistance to insect herbivores in Arabidopsis. Mol Plant MicrobeInteract 20: 1406–1420

Boeckler GA, Gershenzon J, Unsicker SB (2011) Phenolic glycosides of theSalicaceae and their role as anti-herbivore defenses. Phytochemistry 72:1497–1509

Boex-Fontvieille E, Rustgi S, von Wettstein D, Reinbothe S, Reinbothe C(2015) Water-soluble chlorophyll protein is involved in herbivore re-sistance activation during greening of Arabidopsis thaliana. Proc NatlAcad Sci USA 112: 7303–7308

Booker F, Muntifering R, McGrath M, Burkey K, Decoteau D, Fiscus E,Manning W, Krupa S, Chappelka A, Grantz D (2009) The ozonecomponent of global change: potential effects on agricultural and hor-ticultural plant yield, product quality and interactions with invasivespecies. J Integr Plant Biol 51: 337–351

Both M, Csukai M, Stumpf MP, Spanu PD (2005) Gene expression profilesof Blumeria graminis indicate dynamic changes to primary metabolismduring development of an obligate biotrophic pathogen. Plant Cell 17:2107–2122

Bouché N, Fait A, Bouchez D, Møller SG, Fromm H (2003) Mitochondrialsuccinic-semialdehyde dehydrogenase of the g-aminobutyrate shunt isrequired to restrict levels of reactive oxygen intermediates in plants.Proc Natl Acad Sci USA 100: 6843–6848

Bouché N, Fromm H (2004) GABA in plants: just a metabolite? TrendsPlant Sci 9: 110–115

Breeze E, Harrison E, McHattie S, Hughes L, Hickman R, Hill C, KiddleS, Kim YS, Penfold CA, Jenkins D, et al (2011) High-resolution tem-poral profiling of transcripts during Arabidopsis leaf senescence reveals adistinct chronology of processes and regulation. Plant Cell 23: 873–894

Castagna A, Ranieri A (2009) Detoxification and repair process of ozoneinjury: from O3 uptake to gene expression adjustment. Environ Pollut157: 1461–1469

Chen H, Chen SL, Jiang JG (2011) Effect of Ca2+ channel block on glycerolmetabolism in Dunaliella salina under hypoosmotic and hyperosmoticstresses. PLoS ONE 6: e28613

Clarke DB (2010) Glucosinolates, structures and analysis in food. AnalMethods 2: 310–325


Papazian et al.




















Consales F, Schweizer F, Erb M, Gouhier-Darimont C, Bodenhausen N,Bruessow F, Sobhy I, Reymond P (2012) Insect oral secretions suppresswound-induced responses in Arabidopsis. J Exp Bot 63: 727–737

Debouba M, Dguimi HM, Ghorbel M, Gouia H, Suzuki A (2013) Ex-pression pattern of genes encoding nitrate and ammonium assimilatingenzymes in Arabidopsis thaliana exposed to short term NaCl stress. JPlant Physiol 170: 155–160

De Vos M, Van Oosten VR, Van Poecke RMP, Van Pelt JA, Pozo MJ,Mueller MJ, Buchala AJ, Métraux JP, Van Loon LC, Dicke M, et al(2005) Signal signature and transcriptome changes of Arabidopsis dur-ing pathogen and insect attack. Mol Plant Microbe Interact 18: 923–937

Ditchkoff SS, Lewis JS, Lin JC, Muntifering RB, Chappelka AH (2009)Nutritive quality of highbush blackberry (Rubus argutus) exposed totropospheric ozone. Rangeland Ecol Manag 62: 364–370

Dizengremel P, Le Thiec D, Bagard M, Jolivet Y (2008) Ozone risk as-sessment for plants: central role of metabolism-dependent changes inreducing power. Environ Pollut 156: 11–15

Dizengremel P, Le Thiec D, Hasenfratz-Sauder MP, Vaultier MN, BagardM, Jolivet Y (2009) Metabolic-dependent changes in plant cell redoxpower after ozone exposure. Plant Biol (Stuttg) (Suppl 1) 11: 35–42

Dizengremel P, Vaultier MNN, Le Thiec D, Cabané M, Bagard M, GérantD, Gérard J, Dghim AA, Richet N, Afif D, et al (2012) Phosphoenol-pyruvate is at the crossroads of leaf metabolic responses to ozone stress.New Phytol 195: 512–517

Dombrecht B, Xue GP, Sprague SJ, Kirkegaard JA, Ross JJ, Reid JB, FittGP, Sewelam N, Schenk PM, Manners JM, et al (2007) MYC2 differ-entially modulates diverse jasmonate-dependent functions in Arabi-dopsis. Plant Cell 19: 2225–2245

Draborg H, Villadsen D, Nielsen TH (2001) Transgenic Arabidopsis plants withdecreased activity of fructose-6-phosphate,2-kinase/fructose-2,6-bisphosphatasehave altered carbon partitioning. Plant Physiol 126: 750–758

Eastmond PJ (2004) Glycerol-insensitive Arabidopsis mutants: gli1 seed-lings lack glycerol kinase, accumulate glycerol and are more resistant toabiotic stress. Plant J 37: 617–625

El-Kouhen K, Blangy S, Ortiz E, Gardies AM, Ferté N, Arondel V (2005)Identification and characterization of a triacylglycerol lipase in Arabidopsishomologous to mammalian acid lipases. FEBS Lett 579: 6067–6073

Epskamp S, Cramer AOJ, Waldorp LJ, Schmittmann VD, Borsboom D(2012) qgraph: network visualizations of relationships in psychometricdata. J Stat Softw 48: 1–18

Fahey JW, Zalcmann AT, Talalay P (2001) The chemical diversity anddistribution of glucosinolates and isothiocyanates among plants. Phy-tochemistry 56: 5–51

Fait A, Fromm H, Walter D, Galili G, Fernie AR (2008) Highway or byway:the metabolic role of the GABA shunt in plants. Trends Plant Sci 13: 14–19

Fares S, Oksanen E, Lännenpää M, Julkunen-Tiitto R, Loreto F (2010)Volatile emissions and phenolic compound concentrations along a ver-tical profile of Populus nigra leaves exposed to realistic ozone concen-trations. Photosynth Res 104: 61–74

Felton GW, Bi JL, Summers CB, Mueller AJ, Duffey SS (1994) Potentialrole of lipoxygenases in defense against insect herbivory. J Chem Ecol20: 651–666

Fernandez O, Béthencourt L, Quero A, Sangwan RS, Clément C (2010)Trehalose and plant stress responses: friend or foe? Trends Plant Sci 15:409–417

Fiehn O (2002) Metabolomics: the link between genotypes and phenotypes.Plant Mol Biol 48: 155–171

Firn RD, Jones CG (2009) A Darwinian view of metabolism: molecularproperties determine fitness. J Exp Bot 60: 719–726

Forde BG, Lea PJ (2007) Glutamate in plants: metabolism, regulation, andsignalling. J Exp Bot 58: 2339–2358

Fuhrer J (2003) Agroecosystem responses to combinations of elevated CO2,ozone, and global climate change. Agric Ecosyst Environ 97: 1–20

Fujimoto SY, Ohta M, Usui A, Shinshi H, Ohme-Takagi M (2000) Arabidopsisethylene-responsive element binding factors act as transcriptional activatorsor repressors of GCC box-mediated gene expression. Plant Cell 12: 393–404

Fujita M, Fujita Y, Noutoshi Y, Takahashi F, Narusaka Y, Yamaguchi-Shinozaki K, Shinozaki K (2006) Crosstalk between abiotic and bioticstress responses: a current view from the points of convergence in thestress signaling networks. Curr Opin Plant Biol 9: 436–442

Fukushima A, Kusano M, Redestig H, Arita M, Saito K (2009) Integratedomics approaches in plant systems biology. Curr Opin Chem Biol 13:532–538

Garcia O, Saveanu C, ClineM, Fromont-RacineM, Jacquier A, Schwikowski B,Aittokallio T (2007) GOlorize: a Cytoscape plug-in for network visuali-zation with Gene Ontology-based layout and coloring. Bioinformatics 23:394–396

Geijer C, Ahmadpour D, Palmgren M, Filipsson C, Klein DM, Tamás MJ,Hohmann S, Lindkvist-Petersson K (2012) Yeast aquaglyceroporins usethe transmembrane core to restrict glycerol transport. J Biol Chem 287:23562–23570

Geilen K, Böhmer M (2015) Dynamic subnuclear relocalization ofWRKY40, a potential new mechanism of ABA-dependent transcriptionfactor regulation. Plant Signal Behav 10: e1106659

Gomes de Oliveira Dal’Molin C, Quek LE, Saa PA, Nielsen LK (2015) Amulti-tissue genome-scale metabolic modeling framework for the anal-ysis of whole plant systems. Front Plant Sci 6: 4

Goumenaki E, Taybi T, Borland A, Barnes J (2010) Mechanisms under-lying the impacts of ozone on photosynthetic performance. Environ ExpBot 69: 259–266

Halitschke R, Baldwin IT (2003) Antisense LOX expression increasesherbivore performance by decreasing defense responses and inhibitinggrowth-related transcriptional reorganization in Nicotiana attenuata.Plant J 36: 794–807

Halitschke R, Hamilton JG, Kessler A (2011) Herbivore-specific elicitationof photosynthesis by mirid bug salivary secretions in the wild tobaccoNicotiana attenuata. New Phytol 191: 528–535

Halkier BA, Gershenzon J (2006) Biology and biochemistry of glucosino-lates. Annu Rev Plant Biol 57: 303–333

Havko NE, Major IT, Jewell JB, Attaran E, Browse J, Howe GA (2016)Control of carbon assimilation and partitioning by jasmonate: an ac-counting of growth-defense tradeoffs. Plants (Basel) 5: 7

Hernández I, Munné-Bosch S (2015) Linking phosphorus availability withphoto-oxidative stress in plants. J Exp Bot 66: 2889–2900

Hirai MY, Sugiyama K, Sawada Y, Tohge T, Obayashi T, Suzuki A, ArakiR, Sakurai N, Suzuki H, Aoki K, et al (2007) Omics-based identificationof Arabidopsis Myb transcription factors regulating aliphatic glucosi-nolate biosynthesis. Proc Natl Acad Sci USA 104: 6478–6483

Hirner B, Fischer WN, Rentsch D, Kwart M, Frommer WB (1998) Devel-opmental control of H+/amino acid permease gene expression duringseed development of Arabidopsis. Plant J 14: 535–544

Hoefnagel M, Atkin OK, Wiskich JT (1998) Interdependence betweenchloroplasts and mitochondria in the light and the dark. Biochim Bio-phys Acta 1366: 235–255

Hu J, Zhang Y, Wang J, Zhou Y (2014) Glycerol affects root developmentthrough regulation of multiple pathways in Arabidopsis. PLoS ONE 9: e86269

Huang S, Taylor NL, Stroher E, Fenske R, Millar AH (2013) Succinate dehy-drogenase assembly factor 2 is needed for assembly and activity of mito-chondrial complex II and for normal root elongation in Arabidopsis. Plant J73: 429–441

Hui D, Iqbal J, Lehmann K, Gase K, Saluz HP, Baldwin IT (2003) Mo-lecular interactions between the specialist herbivore Manduca sexta(Lepidoptera, Sphingidae) and its natural host Nicotiana attenuata.V. Microarray analysis and further characterization of large-scalechanges in herbivore-induced mRNAs. Plant Physiol 131: 1877–1893

Hummel J, Strehmel N, Selbig J, Walther D, Kopka J (2010) Decision treesupported substructure prediction of metabolites from GC-MS profiles.Metabolomics 6: 322–333

Irmisch S, Clavijo McCormick A, Günther J, Schmidt A, Boeckler GA,Gershenzon J, Unsicker SB, Köllner TG (2014) Herbivore-inducedpoplar cytochrome P450 enzymes of the CYP71 family convert aldox-imes to nitriles which repel a generalist caterpillar. Plant J 80: 1095–1107

Jansen JJ, Allwood JW, Marsden-Edwards E, van der Putten WH,Goodacre R, van Dam NM (2008) Metabolomic analysis of the inter-action between plants and herbivores. Metabolomics 5: 150–161

Jaradat MR, Feurtado JA, Huang D, Lu Y, Cutler AJ (2013) Multiple rolesof the transcription factor AtMYBR1/AtMYB44 in ABA signaling, stressresponses, and leaf senescence. BMC Plant Biol 13: 192

Jost R, Pharmawati M, Lapis-Gaza HR, Rossig C, Berkowitz O, LambersH, Finnegan PM (2015) Differentiating phosphate-dependent andphosphate-independent systemic phosphate-starvation response net-works in Arabidopsis thaliana through the application of phosphite. JExp Bot 66: 2501–2514

Joudoi T, Shichiri Y, Kamizono N, Akaike T, Sawa T, Yoshitake J, Yamada N,Iwai S (2013) Nitrated cyclic GMP modulates guard cell signaling in Ara-bidopsis. Plant Cell 25: 558–571





Jouhet J, Maréchal E, Baldan B, Bligny R, Joyard J, Block MA (2004)Phosphate deprivation induces transfer of DGDG galactolipid fromchloroplast to mitochondria. J Cell Biol 167: 863–874

Jung C, Seo JS, Han SW, Koo YJ, Kim CH, Song SI, Nahm BH, Choi YD,Cheong JJ (2008) Overexpression of AtMYB44 enhances stomatal clo-sure to confer abiotic stress tolerance in transgenic Arabidopsis. PlantPhysiol 146: 623–635

Jung C, Shim JS, Seo JS, Lee HY, Kim CH, Choi YD, Cheong JJJ (2010) Non-specific phytohormonal induction of AtMYB44 and suppression of jasmonate-responsive gene activation in Arabidopsis thaliana. Mol Cells 29: 71–76

Kachroo A, Venugopal SC, Lapchyk L, Falcone D, Hildebrand D, Kachroo P(2004) Oleic acid levels regulated by glycerolipid metabolism modulatedefense gene expression in Arabidopsis. Proc Natl Acad Sci USA 101:5152–5157

Kanehisa M, Goto S (2000) KEGG: Kyoto Encyclopedia of Genes and Genomes.Nucleic Acids Res 28: 27–30

Kangasjärvi J, Jaspers P, Kollist H (2005) Signalling and cell death inozone-exposed plants. Plant Cell Environ 28: 1021–1036

Karban R, Agrawal AA (2002) Herbivore offense. Annu Rev Ecol Syst 33:641–664

Khaling E, Papazian S, Poelman EH, Holopainen JK, Albrectsen BR, BlandeJD (2015) Ozone affects growth and development of Pieris brassicae on thewild host plant Brassica nigra. Environ Pollut 199: 119–129

Kim J, Chang C, Tucker ML (2015) To grow old: regulatory role of ethyleneand jasmonic acid in senescence. Front Plant Sci 6: 20

Kleijn RJ, Geertman JMAM, Nfor BK, Ras C, Schipper D, Pronk JT,Heijnen JJ, van Maris AJ, van Winden WA (2007) Metabolic fluxanalysis of a glycerol-overproducing Saccharomyces cerevisiae strainbased on GC-MS, LC-MS and NMR-derived C-labelling data. FEMSYeast Res 7: 216–231

Koricheva J (2002) Meta-analysis of sources of variation in fitness costs ofplant antiherbivore defenses. Ecology 83: 176–190

Kotchoni SO, Kuhns C, Ditzer A, Kirch HH, Bartels D (2006) Over-expression of different aldehyde dehydrogenase genes in Arabidopsisthaliana confers tolerance to abiotic stress and protects plants againstlipid peroxidation and oxidative stress. Plant Cell Environ 29: 1033–1048

Lancien M, Roberts MR (2006) Regulation of Arabidopsis thaliana 14-3-3gene expression by gamma-aminobutyric acid. Plant Cell Environ 29:1430–1436

Lee I, Ambaru B, Thakkar P, Marcotte EM, Rhee SY (2010) Rational as-sociation of genes with traits using a genome-scale gene network forArabidopsis thaliana. Nat Biotechnol 28: 149–156

Lee T, Yang S, Kim E, Ko Y, Hwang S, Shin J, Shim JE, Shim H, Kim H,Kim C, et al (2015) AraNet v2: an improved database of co-functionalgene networks for the study of Arabidopsis thaliana and 27 other non-model plant species. Nucleic Acids Res 43: D996–D1002

Li D, Li Y, Zhang L, Wang X, Zhao Z, Tao Z, Wang J, Wang J, Lin M, Li X,et al (2014) Arabidopsis ABA receptor RCAR1/PYL9 interacts with anR2R3-type MYB transcription factor, AtMYB44. Int J Mol Sci 15: 8473–8490

Li Q, Zheng Q, Shen W, Cram D, Fowler DB, Wei Y, Zou J (2015) Un-derstanding the biochemical basis of temperature-induced lipid path-way adjustments in plants. Plant Cell 27: 86–103

Li RJ, Hua W, Lu YT (2006) Arabidopsis cytosolic glutamine synthetaseAtGLN1;1 is a potential substrate of AtCRK3 involved in leaf senes-cence. Biochem Biophys Res Commun 342: 119–126

Li Z, Peng J, Wen X, Guo H (2013) Ethylene-insensitive3 is a senescence-associated gene that accelerates age-dependent leaf senescence by directlyrepressing miR164 transcription in Arabidopsis. Plant Cell 25: 3311–3328

Liechti R, Csardi G, Bergmann S, Schutz F, Sengstag T, Boj SF, ServitjaJM, Ferre J, Van Lommel L, Schuit F, et al (2010) EuroDia: a beta-cellgene expression resource. Database (Oxford) 2010: baq024

Lin LZ, Sun J, Chen P, Harnly J (2011) UHPLC-PDA-ESI/HRMS/MS(n)analysis of anthocyanins, flavonol glycosides, and hydroxycinnamicacid derivatives in red mustard greens (Brassica juncea Coss variety). JAgric Food Chem 59: 12059–12072

Lindroth RL (2010) Impacts of elevated atmospheric CO2 and O3 on forests:phytochemistry, trophic interactions, and ecosystem dynamics. J ChemEcol 36: 2–21

Liu ZQ, Yan L, Wu Z, Mei C, Lu K, Yu YT, Liang S, Zhang XF, Wang XF,Zhang DP (2012) Cooperation of three WRKY-domain transcription fac-tors WRKY18, WRKY40, and WRKY60 in repressing two ABA-responsivegenes ABI4 and ABI5 in Arabidopsis. J Exp Bot 63: 6371–6392

Lof ME, De Gee M, Dicke M, Gort G, Hemerik L (2013) Exploitation ofchemical signaling by parasitoids: impact on host population dynamics.J Chem Ecol 39: 752–763

Long SP, Naidu SL (2002) Effects of oxidants at the biochemical, cell, andphysiological levels, with particular reference to ozone. In JNB Bell, MTreshow, eds, Air Pollution and Plant Life. Wiley, London, pp 69–88

Ludwikow A, Sadowski J (2008) Gene networks in plant ozone stress re-sponse and tolerance. J Integr Plant Biol 50: 1256–1267

Martin MVV, Fiol DF, Sundaresan V, Zabaleta EJ, Pagnussat GC (2013) oiwa, afemale gametophyticmutant impaired in amitochondrial manganese-superoxidedismutase, reveals crucial roles for reactive oxygen species during embryo sacdevelopment and fertilization in Arabidopsis. Plant Cell 25: 1573–1591

McCormick AJ, Kruger NJ (2015) Lack of fructose 2,6-bisphosphate com-promises photosynthesis and growth in Arabidopsis in fluctuating en-vironments. Plant J 81: 670–683

McGrath KC, Dombrecht B, Manners JM, Schenk PM, Edgar CI, MacleanDJ, Scheible WR, Udvardi MK, Kazan K (2005) Repressor- and activator-type ethylene response factors functioning in jasmonate signaling and diseaseresistance identified via a genome-wide screen of Arabidopsis transcriptionfactor gene expression. Plant Physiol 139: 949–959

McKenna MC, Waagepetersen HS, Schousboe A, Sonnewald U (2006)Neuronal and astrocytic shuttle mechanisms for cytosolic-mitochondrialtransfer of reducing equivalents: current evidence and pharmacologicaltools. Biochem Pharmacol 71: 399–407

Meldau S, Erb M, Baldwin IT (2012) Defence on demand: mechanismsbehind optimal defence patterns. Ann Bot (Lond) 110: 1503–1514

Michaeli S, Fromm H (2015) Closing the loop on the GABA shunt in plants:are GABA metabolism and signaling entwined? Front Plant Sci 6: 419

Mirabella R, Rauwerda H, Allmann S, Scala A, Spyropoulou EA, de VriesM, Boersma MR, Breit TM, Haring MA, Schuurink RC (2015) WRKY40and WRKY6 act downstream of the green leaf volatile E-2-hexenal inArabidopsis. Plant J 83: 1082–1096

Misra P, Pandey A, Tiwari M, Chandrashekar K, Sidhu OP, Asif MH,Chakrabarty D, Singh PK, Trivedi PK, Nath P, et al (2010) Modulationof transcriptome and metabolome of tobacco by Arabidopsis transcriptionfactor, AtMYB12, leads to insect resistance. Plant Physiol 152: 2258–2268

Missihoun TD, Hou Q, Mertens D, Bartels D (2014) Sequence and func-tional analyses of the aldehyde dehydrogenase 7B4 gene promoter inArabidopsis thaliana and selected Brassicaceae: regulation patterns inresponse to wounding and osmotic stress. Planta 239: 1281–1298

Mittler R (2006) Abiotic stress, the field environment and stress combina-tion. Trends Plant Sci 11: 15–19

Morandini P (2013) Control limits for accumulation of plant metabolites:brute force is no substitute for understanding. Plant Biotechnol J 11: 253–267

Mrá�cek T, Drahota Z, Houšt�ek J (2013) The function and the role of themitochondrial glycerol-3-phosphate dehydrogenase in mammalian tis-sues. Biochim Biophys Acta 1827: 401–410

Mueller SP, Krause DM, Mueller MJ, Fekete A (2015) Accumulation ofextra-chloroplastic triacylglycerols in Arabidopsis seedlings during heatacclimation. J Exp Bot 66: 4517–4526

Mulaudzi T, Ludidi N, Ruzvidzo O, Morse M, Hendricks N, Iwuoha E,Gehring C (2011) Identification of a novel Arabidopsis thaliana nitricoxide-binding molecule with guanylate cyclase activity in vitro. FEBSLett 585: 2693–2697

Murata N, Takahashi S, Nishiyama Y, Allakhverdiev SI (2007) Photo-inhibition of photosystem II under environmental stress. Biochim Bio-phys Acta 1767: 414–421

Nabity PD, Zavala JA, DeLucia EH (2013) Herbivore induction of jasmonicacid and chemical defences reduce photosynthesis in Nicotiana attenu-ata. J Exp Bot 64: 685–694

Nakabayashi R, Saito K (2015) Integrated metabolomics for abiotic stressresponses in plants. Curr Opin Plant Biol 24: 10–16

Nambara E, Kawaide H, Kamiya Y, Naito S (1998) Characterization of anArabidopsis thaliana mutant that has a defect in ABA accumulation:ABA-dependent and ABA-independent accumulation of free aminoacids during dehydration. Plant Cell Physiol 39: 853–858

Nielsen TH, Rung JH, Villadsen D (2004) Fructose-2,6-bisphosphate: atraffic signal in plant metabolism. Trends Plant Sci 9: 556–563

Niyogi KK (2000) Safety valves for photosynthesis. Curr Opin Plant Biol 3:455–460

Noctor G, De Paepe R, Foyer CH (2007) Mitochondrial redox biology andhomeostasis in plants. Trends Plant Sci 12: 125–134


Papazian et al.



Nunes-Nesi A, Sulpice R, Gibon Y, Fernie AR (2008) The enigmatic con-tribution of mitochondrial function in photosynthesis. J Exp Bot 59:1675–1684

Onkokesung N, Reichelt M, van Doorn A, Schuurink RC, van Loon JJ,Dicke M (2014) Modulation of flavonoid metabolites in Arabidopsisthaliana through overexpression of the MYB75 transcription factor: roleof kaempferol-3,7-dirhamnoside in resistance to the specialist insectherbivore Pieris brassicae. J Exp Bot 65: 2203–2217

Ortiz-Lopez A, Chang HC, Bush DR (2000) Amino acid transporters inplants. Biochim Biophys Acta 1465: 275–280

Pandey P, Ramegowda V, Senthil-Kumar M (2015) Shared and uniqueresponses of plants to multiple individual stresses and stress combina-tions: physiological and molecular mechanisms. Front Plant Sci 6: 723

Paoletti E, Grulke NE (2005) Does living in elevated CO2 ameliorate tree responseto ozone? A review on stomatal responses. Environ Pollut 137: 483–493

Persak H, Pitzschke A (2013) Tight interconnection and multi-level controlof Arabidopsis MYB44 in MAPK cascade signalling. PLoS ONE 8: e57547

Persak H, Pitzschke A (2014) Dominant repression by Arabidopsis tran-scription factor MYB44 causes oxidative damage and hypersensitivity toabiotic stress. Int J Mol Sci 15: 2517–2537

Petersen BL, Chen S, Hansen CH, Olsen CE, Halkier BA (2002) Compo-sition and content of glucosinolates in developing Arabidopsis thaliana.Planta 214: 562–571

Pieterse CMJ, Van der Does D, Zamioudis C, Leon-Reyes A, Van WeesSCM (2012) Hormonal modulation of plant immunity. Annu Rev CellDev Biol 28: 489–521

Pinto DM, Blande JD, Souza SR, Nerg AM, Holopainen JK (2010) Plantvolatile organic compounds (VOCs) in ozone (O3) polluted atmospheres:the ecological effects. J Chem Ecol 36: 22–34

Pitzschke A, Hirt H (2009) Disentangling the complexity of mitogen-activated protein kinases and reactive oxygen species signaling. PlantPhysiol 149: 606–615

Poelman EH, Loon JJ, Dam NM, Vet LE, Dicke M (2010) Herbivore‐in-duced plant responses in Brassica oleracea prevail over effects of con-stitutive resistance and result in enhanced herbivore attack. EcolEntomol 35: 240–247

Potuschak T, Lechner E, Parmentier Y, Yanagisawa S, Grava S, Koncz C,Genschik P (2003) EIN3-dependent regulation of plant ethylene hor-mone signaling by two Arabidopsis F box proteins: EBF1 and EBF2. Cell115: 679–689

Prasch CM, Sonnewald U (2013) Simultaneous application of heat,drought, and virus to Arabidopsis plants reveals significant shifts insignaling networks. Plant Physiol 162: 1849–1866

Pré M, Atallah M, Champion A, De Vos M, Pieterse CM, Memelink J(2008) The AP2/ERF domain transcription factor ORA59 integrates jasmonicacid and ethylene signals in plant defense. Plant Physiol 147: 1347–1357

Provart N, Zhu T (2003) A browser-based functional classification super-viewer for Arabidopsis genomics. Currents in Computational MolecularBiology 271–272

Quettier ALL, Shaw E, Eastmond PJ (2008) SUGAR-DEPENDENT6 en-codes a mitochondrial flavin adenine dinucleotide-dependent glycerol-3-P dehydrogenase, which is required for glycerol catabolism and postgerminative seedling growth in Arabidopsis. Plant Physiol 148: 519–528

Ralph SG, Yueh H, Friedmann M, Aeschliman D, Zeznik JA, Nelson CC,Butterfield YS, Kirkpatrick R, Liu J, Jones SJ, et al (2006) Conifer de-fence against insects: microarray gene expression profiling of Sitkaspruce (Picea sitchensis) induced by mechanical wounding or feeding byspruce budworms (Choristoneura occidentalis) or white pine weevils(Pissodes strobi) reveals large-scale changes of the host transcriptome.Plant Cell Environ 29: 1545–1570

Ranieri A, Giuntini D, Ferraro F, Nali C, Baldan B, Lorenzini G, SoldatiniGF, Soldatini F (2001) Chronic ozone fumigation induces alterations inthylakoid functionality and composition in two poplar clones. PlantPhysiol Biochem 39: 999–1008

Rejeb IB, Pastor V, Mauch-Mani B (2014) Plant responses to simultaneousbiotic and abiotic stress: molecular mechanisms. Plants (Basel) 3: 458–475

Renaut J, Bohler S, Hausman JF, Hoffmann L, Sergeant K, Ahsan N,Jolivet Y, Dizengremel P (2009) The impact of atmospheric compo-sition on plants: a case study of ozone and poplar. Mass Spectrom Rev28: 495–516

Reymond P, Bodenhausen N, Van Poecke RMP, Krishnamurthy V, DickeM, Farmer EE (2004) A conserved transcript pattern in response to aspecialist and a generalist herbivore. Plant Cell 16: 3132–3147

Reymond P, Weber H, Damond M, Farmer EE (2000) Differential geneexpression in response to mechanical wounding and insect feeding inArabidopsis. Plant Cell 12: 707–720

Rizhsky L, Liang H, Mittler R (2002) The combined effect of drought stress andheat shock on gene expression in tobacco. Plant Physiol 130: 1143–1151

Rizhsky L, Liang H, Shuman J, Shulaev V, Davletova S, Mittler R (2004)When defense pathways collide: the response of Arabidopsis to acombination of drought and heat stress. Plant Physiol 134: 1683–1696

Rolland F, Baena-Gonzalez E, Sheen J (2006) Sugar sensing and signaling inplants: conserved and novel mechanisms. Annu Rev Plant Biol 57: 675–709

Saito K, Hirai MY, Yonekura-Sakakibara K (2008) Decoding genes withcoexpression networks and metabolomics: ‘majority report by precogs’.Trends Plant Sci 13: 36–43

Saito K, Matsuda F (2010) Metabolomics for functional genomics, systemsbiology, and biotechnology. Annu Rev Plant Biol 61: 463–489

Saito S, Hirai N, Matsumoto C, Ohigashi H, Ohta D, Sakata K, MizutaniM (2004) Arabidopsis CYP707As encode (+)-abscisic acid 89-hydroxy-lase, a key enzyme in the oxidative catabolism of abscisic acid. PlantPhysiol 134: 1439–1449

Sakaki T, Saito K, Kawaguchi A, Kondo N, Yamada M (1990) Conversionof monogalactosyldiacylglycerols to triacylglycerols in ozone-fumigatedspinach leaves. Plant Physiol 94: 766–772

Sakurai N, Ara T, Ogata Y, Sano R, Ohno T, Sugiyama K, Hiruta A, YamazakiK, Yano K, Aoki K, et al (2011) KaPPA-View4: a metabolic pathway data-base for representation and analysis of correlation networks of geneco-expression and metabolite co-accumulation and omics data. Nucleic AcidsRes 39: D677–D684

Salvatori E, Fusaro L, Strasser RJ, Bussotti F, Manes F (2015) Effects ofacute O3 stress on PSII and PSI photochemistry of sensitive and resistantsnap bean genotypes (Phaseolus vulgaris L.), probed by prompt chlo-rophyll “a” fluorescence and 820 nm modulated reflectance. PlantPhysiol Biochem 97: 368–377

Sardans J, Peñuelas J, Rivas-Ubach A (2011) Ecological metabolomics:overview of current developments and future challenges. Chemo-ecology 21: 191–225

Sawada H, Kohno Y (2009) Differential ozone sensitivity of rice cultivars asindicated by visible injury and grain yield. Plant Biol (Stuttg) (Suppl 1)11: 70–75

Scheibe R, Backhausen JE, Emmerlich V, Holtgrefe S (2005) Strategies tomaintain redox homeostasis during photosynthesis under changingconditions. J Exp Bot 56: 1481–1489

Schön M, Töller A, Diezel C, Roth C, Westphal L, Wiermer M, SomssichIE (2013) Analyses of wrky18 wrky40 plants reveal critical roles of SA/EDS1 signaling and indole-glucosinolate biosynthesis for Golovino-myces orontii resistance and a loss-of resistance towards Pseudomonassyringae pv. tomato AvrRPS4. Mol Plant Microbe Interact 26: 758–767

Schwachtje J, Baldwin IT (2008) Why does herbivore attack reconfigureprimary metabolism? Plant Physiol 146: 845–851

Sclep G, Allemeersch J, Liechti R, De Meyer B, Beynon J, Bhalerao R,Moreau Y, Nietfeld W, Renou JP, Reymond P, et al (2007) CATMA, acomprehensive genome-scale resource for silencing and transcript pro-filing of Arabidopsis genes. BMC Bioinformatics 8: 400

Sengupta D, Naik D, Reddy AR (2015) Plant aldo-keto reductases (AKRs)as multi-tasking soldiers involved in diverse plant metabolic processesand stress defense: a structure-function update. J Plant Physiol 179: 40–55

Settele J, Scholes R, Betts S, Bunn S, Leadley P, Nepstad D, Overpeck JT,Taboada MA (2014) Terrestrial and inland water systems. In ClimateChange 2014: Impacts, Adaptations and Vulnerability. Part A: Globaland Sectoral Aspects. Contribution of Working Group II to the FifthAssessment Report of the Intergovernmental Panel on Climate Change.Cambridge University Press, Cambridge, UK, pp 271–359

Shao Y, Jiang J, Ran L, Lu C, Wei C, Wang Y (2014) Analysis of flavonoids andhydroxycinnamic acid derivatives in rapeseeds (Brassica napus L. var. na-pus) by HPLC-PDA-ESI(-)-MS(n)/HRMS. J Agric Food Chem 62: 2935–2945

Sharma YK, Léon J, Raskin I, Davis KR (1996) Ozone-induced responsesin Arabidopsis thaliana: the role of salicylic acid in the accumulation ofdefense-related transcripts and induced resistance. Proc Natl Acad SciUSA 93: 5099–5104

Shelp BJ, Bown AW, Faure D (2006) Extracellular g-aminobutyrate medi-ates communication between plants and other organisms. Plant Physiol142: 1350–1352

Shen W, Wei Y, Dauk M, Tan Y, Taylor DC, Selvaraj G, Zou J (2006)Involvement of a glycerol-3-phosphate dehydrogenase in modulating





the NADH/NAD+ ratio provides evidence of a mitochondrial glycerol-3-phosphate shuttle in Arabidopsis. Plant Cell 18: 422–441

Shen W, Wei Y, Dauk M, Zheng Z, Zou J (2003) Identification of a mito-chondrial glycerol-3-phosphate dehydrogenase from Arabidopsis thali-ana: evidence for a mitochondrial glycerol-3-phosphate shuttle in plants.FEBS Lett 536: 92–96

Shim JS, Jung C, Lee S, Min K, Lee YWW, Choi Y, Lee JS, Song JT, KimJKK, Choi YD (2013) AtMYB44 regulates WRKY70 expression andmodulates antagonistic interaction between salicylic acid and jasmonicacid signaling. Plant J 73: 483–495

Shulaev V, Cortes D, Miller G, Mittler R (2008) Metabolomics for plantstress response. Physiol Plant 132: 199–208

Simmonds MS (2003) Flavonoid-insect interactions: recent advances in ourknowledge. Phytochemistry 64: 21–30

Singh E, Tiwari S, Agrawal M (2009) Effects of elevated ozone on photo-synthesis and stomatal conductance of two soybean varieties: a casestudy to assess impacts of one component of predicted global climatechange. Plant Biol (Stuttg) (Suppl 1) 11: 101–108

Smallegange RC, van Loon JJA, Blatt SE, Harvey JA, Agerbirk N, DickeM (2007) Flower vs. leaf feeding by Pieris brassicae: glucosinolate-richflower tissues are preferred and sustain higher growth rate. J Chem Ecol33: 1831–1844

Smyth GK (2004) Linear models and empirical Bayes methods for assessing dif-ferential expression in microarray experiments. Stat Appl Genet Mol Biol 3: e3

Sønderby IE, Hansen BG, Bjarnholt N, Ticconi C, Halkier BA, Klieben-stein DJ (2007) A systems biology approach identifies a R2R3 MYB genesubfamily with distinct and overlapping functions in regulation of ali-phatic glucosinolates. PLoS ONE 2: e1322

Song J, Zhu C, Zhang X, Wen X, Liu L, Peng J, Guo H, Yi C (2015) Bio-chemical and structural insights into the mechanism of DNA recognitionby Arabidopsis ETHYLENE INSENSITIVE3. PLoS ONE 10: e0137439

Song S, Huang H, GaoH,Wang J, WuD, Liu X, Yang S, Zhai Q, Li C, Qi T, et al(2014) Interaction between MYC2 and ETHYLENE INSENSITIVE3 modulatesantagonism between jasmonate and ethylene signaling in Arabidopsis. PlantCell 26: 263–279

Stam JM, Kroes A, Li Y, Gols R, van Loon JJ, Poelman EH, Dicke M (2014)Plant interactions with multiple insect herbivores: from community togenes. Annu Rev Plant Biol 65: 689–713

Steuer R (2006) Review: on the analysis and interpretation of correlations inmetabolomic data. Brief Bioinform 7: 151–158

Storey JD, Tibshirani R (2003) Statistical significance for genomewidestudies. Proc Natl Acad Sci USA 100: 9440–9445

Strehmel N, Hummel J, Erban A, Strassburg K, Kopka J (2008) Retentionindex thresholds for compound matching in GC-MS metabolite profil-ing. J Chromatogr B Analyt Technol Biomed Life Sci 871: 182–190

Suzuki N, Rivero RM, Shulaev V, Blumwald E, Mittler R (2014) Abioticand biotic stress combinations. New Phytol 203: 32–43

Tang J, Zielinski R, Aldea M, DeLucia E (2009) Spatial association ofphotosynthesis and chemical defense in Arabidopsis thaliana followingherbivory by Trichoplusia ni. Physiol Plant 137: 115–124

Tang JY, Zielinski RE, Zangerl AR, Crofts AR, Berenbaum MR, DeluciaEH (2006) The differential effects of herbivory by first and fourth instarsof Trichoplusia ni (Lepidoptera: Noctuidae) on photosynthesis in Ara-bidopsis thaliana. J Exp Bot 57: 527–536

Textor S, Gershenzon J (2008) Herbivore induction of the glucosinolate-myrosinase defense system: major trends, biochemical bases and eco-logical significance. Phytochem Rev 8: 149–170

Thimm O, Blasing O, Gibon Y, Nagel A, Meyer S, Kruger P, Selbig J,Muller LA, Rhee SY, Stitt M (2004) MAPMAN: a user-driven tool todisplay genomics data sets onto diagrams of metabolic pathways andother biological processes. Plant J 37: 914–939

Tokimatsu T, Sakurai N, Suzuki H, Ohta H, Nishitani K, Koyama T,Umezawa T, Misawa N, Saito K, Shibata D (2005) KaPPA-View: a web-based analysis tool for integration of transcript and metabolite data onplant metabolic pathway maps. Plant Physiol 138: 1289–1300

Tsang EW, Bowler C, Hérouart D, Van Camp W, Villarroel R, Genetello C,Van Montagu M, Inzé D (1991) Differential regulation of superoxide dis-mutases in plants exposed to environmental stress. Plant Cell 3: 783–792

Turner JG, Ellis C, Devoto A (2002) The jasmonate signal pathway. PlantCell (Suppl) 14: S153–S164

Vahisalu T, Puzõrjova I, Brosché M, Valk E, Lepiku M, Moldau H,Pechter P, Wang YS, Lindgren O, Salojärvi J, et al (2010) Ozone-triggered rapid stomatal response involves the production of reactiveoxygen species, and is controlled by SLAC1 and OST1. Plant J 62: 442–453

Vainonen JP, Kangasjärvi J (2015) Plant signalling in acute ozone expo-sure. Plant Cell Environ 38: 240–252

van Dam NM, Witjes L, Svatoš A (2004) Interactions between above-ground and belowground induction of glucosinolates in two wildBrassica species. New Phytol 161: 801–810

Van Dingenen R, Dentener FJ, Raes F, Krol MC, Emberson L, Cofala J(2009) The global impact of ozone on agricultural crop yields undercurrent and future air quality legislation. Atmos Environ 43: 604–618

Warde-Farley D, Donaldson SL, Comes O, Zuberi K, Badrawi R, Chao P,Franz M, Grouios C, Kazi F, Lopes CT, et al (2010) The GeneMANIAprediction server: biological network integration for gene prioritizationand predicting gene function. Nucleic Acids Res 38: W214–W220

Watanabe M, Balazadeh S, Tohge T, Erban A, Giavalisco P, Kopka J,Mueller-Roeber B, Fernie AR, Hoefgen R (2013) Comprehensive dis-section of spatiotemporal metabolic shifts in primary, secondary, andlipid metabolism during developmental senescence in Arabidopsis.Plant Physiol 162: 1290–1310

Weckwerth W (2011) Green systems biology: from single genomes, pro-teomes and metabolomes to ecosystems research and biotechnology. JProteomics 75: 284–305

Xiong L, Yang Y (2003) Disease resistance and abiotic stress tolerance inrice are inversely modulated by an abscisic acid-inducible mitogen-activated protein kinase. Plant Cell 15: 745–759

Xu C, Fan J, Cornish AJ, Benning C (2008) Lipid trafficking between theendoplasmic reticulum and the plastid in Arabidopsis requires the ex-traplastidic TGD4 protein. Plant Cell 20: 2190–2204

Yendrek CR, Koester RP, Ainsworth EA (2015) A comparative analysis oftranscriptomic, biochemical, and physiological responses to elevatedozone identifies species-specific mechanisms of resilience in legumecrops. J Exp Bot 66: 7101–7112

Zangerl AR, Hamilton JG, Miller TJ, Crofts AR, Oxborough K, BerenbaumMR, de Lucia EH (2002) Impact of folivory on photosynthesis is greater thanthe sum of its holes. Proc Natl Acad Sci USA 99: 1088–1091

Zavala JA, Baldwin IT (2004) Fitness benefits of trypsin proteinase inhib-itor expression in Nicotiana attenuata are greater than their costs whenplants are attacked. BMC Ecol 4: 11

Zhang X, Zhu Z, An F, Hao D, Li P, Song J, Yi C, Guo H (2014) Jasmonate-activated MYC2 represses ETHYLENE INSENSITIVE3 activity to an-tagonize ethylene-promoted apical hook formation in Arabidopsis. PlantCell 26: 1105–1117


Papazian et al.



central metabolic responses to ozone and herbivory affect photosynthesis and stomatal ... ·...

Documents