From climate change to molecular response: redox proteomics of ozone-induced responses in soybean

Authors

  • Ashley Galant,

    1. Department of Biology, Washington University, One Brookings Drive, Campus Box 1137, St. Louis, MO 63130, USA
    Search for more papers by this author
  • Robert P. Koester,

    1. Department of Plant Biology, 1201 West Gregory Drive, MC-051, University of Illinois, Urbana-Champaign, Urbana, IL 61801, USA
    Search for more papers by this author
  • Elizabeth A. Ainsworth,

    1. Department of Plant Biology, 1201 West Gregory Drive, MC-051, University of Illinois, Urbana-Champaign, Urbana, IL 61801, USA
    2. USDA-ARS Global Change and Photosynthesis Research Unit, 1201 West Gregory Drive, MC-051, Urbana, IL 61801, USA
    Search for more papers by this author
  • Leslie M. Hicks,

    1. Donald Danforth Plant Science Center, 975 North Warson Road, St. Louis, MO 63132, USA
    Search for more papers by this author
  • Joseph M. Jez

    1. Department of Biology, Washington University, One Brookings Drive, Campus Box 1137, St. Louis, MO 63130, USA
    Search for more papers by this author

Author for correspondence:
Joseph M. Jez
Tel: +1 314 935 3376
Email: jjez@biology2.wustl.edu

Summary

  • Ozone (O3) causes significant agricultural losses, with soybean (Glycine max) being highly sensitive to this oxidant. Here we assess the effect of elevated seasonal O3 exposure on the total and redox proteomes of soybean.
  • To understand the molecular responses to O3 exposure, soybean grown at the Soybean Free Air Concentration Enrichment facility under ambient (37 ppb), moderate (58 ppb), and high (116 ppb) O3 concentrations was examined by redox-sensitive thiol labeling, mass spectrometry, and targeted enzyme assays.
  • Proteomic analysis of soybean leaf tissue exposed to high O3 concentrations reveals widespread changes. In the high-O3 treatment leaf, 35 proteins increased up to fivefold in abundance, 22 proteins showed up to fivefold higher oxidation, and 22 proteins increased in both abundance and oxidation. These changes occurred in carbon metabolism, photosynthesis, amino acid synthesis, flavonoid and isoprenoid biosynthesis, signaling and homeostasis, and antioxidant pathways.
  • This study shows that seasonal O3 exposure in soybean alters the abundance and oxidation state of redox-sensitive multiple proteins and that these changes reflect a combination of damage effects and adaptive responses that influence a wide range of metabolic processes, which in some cases may help mitigate oxidative stress.

Introduction

Global climate change and air pollution pose significant challenges to agriculture and food production worldwide (FAO, 2005). In the northern hemisphere, tropospheric ozone (O3) is a major pollutant that affects agriculture yields of multiple crops (Heagle, 1989; Fiscus et al., 2005; Ainsworth et al., 2008). Since the 19th century, ground-level O3 concentrations have doubled, with concentrations in industrialized nations rising 0.5–2.5% yr–1 and major crop growing regions of the US, India, and China facing more changes of up to 10% yr–1 (Staehelin et al., 1994; Vingarzan, 2004; Ashmore et al., 2006). The current daytime summer O3 concentration in the northern hemisphere and higher localized concentrations are already above the established 40 ppb threshold for crop losses (Heagle, 1989; Fuhrer et al., 1997). Climate models predict that mean surface O3 concentrations may rise 20–25% globally by 2050, with concentrations in India and south Asia reaching comparable values by 2020 (IPCC 2001; Dentener et al., 2005; Van Dingenen et al., 2009). Understanding how crops respond to increasing O3 pollution is essential for meeting the growing demands for sustainable food systems as the world faces increasing population, urbanization, and climate change.

The negative effects of O3 on crop yield are well documented from experiments conducted in open-top growth chambers, free air concentration enrichment (FACE) systems, and other facilities (Heagle, 1989; Morgan et al., 2004, 2006; Fiscus et al., 2005; Ainsworth et al., 2008; Chen et al., 2009; Van Dingenen et al., 2009; Betzelberger et al., 2010). Among major food crops, soybean (Glycine max) is among the most sensitive to atmospheric O3 concentrations (Mills et al., 2007). In FACE trials with soybean, a 13 ppb increase in O3 from 56 to 69 ppb resulted in a 20% decrease in crop yield (Morgan et al., 2004, 2006). Comparable reductions in yield occur across multiple soybean varieties, suggesting that breeding for O3 tolerance may be difficult (Heagle, 1989; Burkey & Carter, 2009; Betzelberger et al., 2010). Economically, annual crop losses due to O3 damage at current tropospheric values are estimated at $2–4 billion in the US and $3–5.5 billion in China, and will likely increase in the future (Van Dingenen et al., 2009).

As an environmental stress, O3 is an oxidant and causes visible symptoms in leaves, including chlorophyll loss, leaf bronzing, and even development of necrotic spots. Elevated O3 decreases photosynthetic rates and both seed mass and number (Chernikova et al., 2000; Robinson & Britz, 2000; Krupa et al., 2001; Baier et al., 2005; Fiscus et al., 2005; Morgan et al., 2006; Chen et al., 2009; Emberson et al., 2009; Betzelberger et al., 2010). At the molecular level, proteomic studies of rice, wheat, soybean, and poplar exposed to continuous short-term (2–5 d) high (> 200 ppb) O3 stress in growth chambers generally reveal drastic reductions in the major leaf photosynthetic and carbon metabolism proteins and induction of defense/stress-related proteins (Agrawal et al., 2002; Bohler et al., 2007; Cho et al., 2008; Feng et al., 2008; Renaut et al., 2009; Ahsan et al., 2010; Sarkar et al., 2010); however, assessments of proteome changes have not examined plants grown in the field under long-term elevated daytime O3 concentrations. Moreover, published studies do not probe the effect of O3 on the plant thiol-redox proteome, as these changes are not observable by standard proteomics methods (Buchanan & Balmer, 2005; Alvarez et al., 2011). This is important for understanding the effect of oxidative stresses on proteins because changes in thiol-redox state can modulate activity by altering active site residues required for catalysis and/or ligand binding, oligomerization, and cellular localization (Paget & Buttner, 2003; Barford, 2004).

Recently, we used differential labeling, two-dimensional gel electrophoresis (2-DE) and mass spectrometry (Fig. 1) to identify plant proteins that respond to changes in redox environment resulting from exogenous and endogenous oxidative stresses caused by application of hydrogen peroxide and buthionine sulfoximine, respectively (Alvarez et al., 2011). Here we employ this method to examine the total and thiol-redox proteomes of leaf and root tissues from soybean grown under ambient (37 ppb), moderate (58 ppb), and high (116 ppb) O3 concentrations at SoyFACE (http://www.soyface.illinois.edu). This analysis shows that long-term exposure to high O3 concentrations in the field results in increased abundance of multiple proteins, in contrast to exposure to moderate O3 concentration, and that many of those proteins also displayed increased oxidation. In comparison to short-term studies in growth chambers, the changes in the soybean total and redox proteomes are more widespread across metabolism. We also provide the first demonstration that O3 exposure in leaf tissue alters the oxidation states of sulfhydryl groups of multiple proteins in soybean. These biochemical changes may play a role in metabolic acclimation to long-term field growth under chronic O3 exposure.

Figure 1.

Redox proteome labeling approach. Proteins (ovals) with free thiols (-SH), disulfide bonds (-S-S-), or modified cysteines (-S-mod) are incubated with N-ethylmaleimide (NEM) to block free sulfhydryl groups. Oxidized thiols are reduced with dithiothreitol (DTT) treatment. The resulting free thiols are labeled with 5-iodoacetamidofluorescein (IAF), and then the proteins are separated by two-dimensional gel electrophoresis and identified by LC-MS/MS.

Materials and Methods

Plant growth, sampling, and SoyFACE O3 treatment

Soybeans (G. max (L.) Merr cv Pioneer 93B15) were planted on 9 June 2009 at SoyFACE and exposed to target O3 concentrations of 40–200 ppb for 9 h d–1, avoiding periods when leaves were wet (Betzelberger et al., 2010). Samples were taken from ambient, 85  and 200 ppb target plots. Across the season, the 1 min average O3 concentrations were within 10% of the target for > 80% of the time; however, due to rainy days and short periods without fumigation, the seasonal 9 h daytime average concentrations were 58 and 116 ppb in the moderate (85 ppb target) and high (200 ppb target) plots, respectively. The corresponding 8 h average O3 concentrations were 61.4 and 126.1 ppb for the 85 and 200 ppb target plots, respectively. The cumulative seasonal AOT40 (i.e. accumulated ozone exposure over a threshold of 40 ppb) in the ambient-, moderate- and high-O3 treatments reached 3.4, 20.5 and 73.6 ppm h, respectively, by the end of the season. The growing season meteorological conditions were not unusual during 2009, with June, July and August precipitation totaling 347 mm, and daytime maximum temperature averaging 26°C. For the proteomic experiments, plants were harvested during reproductive growth on 5 August 2009. Fully expanded leaves at the top of the canopy were cut with a razor and immediately frozen in liquid N2. After cutting the leaves, entire plants were pulled from the soil and roots from those plants were immediately washed in water, clipped from the plant and frozen in liquid N2. Total amounts of protein in tissues were analyzed by Bradford assays. Photosynthetic carbon uptake rates of mature leaves from the upper canopy were determined in the field on 8 August 2009 with an infrared gas analyzer (Li-6400, Li-Cor, Lincoln, NE, USA). Gas exchange measurements were made at midday at ambient [CO2], ambient temperature and photosynthetic photon flux density (PPFD) on three leaves per O3 concentration. Photosynthetic carbon uptake was calculated using the equations of von Caemmerer & Farquhar (1981).

Protein extraction

Frozen root (800 mg) or leaf (400 mg) tissues were ground to a fine powder in liquid N2. Tissue was suspended in extraction buffer (100 mM Tris, pH 8.0; 100 mM N-ethylmaleimide (NEM); 1% CHAPS; 1% plant protease inhibitors (Sigma-Aldrich)) to 200 mg ml−1. Samples were sonicated (3 × 15 s), and then centrifuged (16 000 g; 15 min). Supernatant was mixed with methanol, incubated (4°C; 30 min), and centrifuged (16 000 g; 5 min). The pellet was washed with methanol and resuspended in reduction buffer (50 mM Tris, pH 8.0, 7 M urea, 2 M thiourea, 50 mM dithiothreitol (DTT)). The suspension was incubated (25°C; 15 min), and then reprecipitated with methanol. The pellet was dried and resuspended in labeling buffer (40 mM HEPES, pH 7.5; 50 mM NaCl; 200 mM 5-iodoacetamidofluorescein (IAF; Sigma-Aldrich); 10 min; 25°C). After methanol precipitation, the pellet was dried, resuspended in DeStreak buffer (GE Life Sciences, Piscataway, NJ, USA), and protein concentration determined by CB-X assay (G-Biosciences, Maryland Heights, MO, USA).

Labeling, 2-DE, and proteomic analysis

Protein (200 μg) in DeStreak buffer was absorbed into a pH 4–7 isoelectric focusing (IEF) gel strip for separation in a Protean IEF cell (Alvarez et al., 2011). Gels were imaged to detect IAF-labeled proteins (λex = 488 nm and λem = 520 nm) using a Typhoon 9410 (GE Healthcare). Next, gels were fixed, washed, and stained with Sypro Ruby (λex = 457 nm and λem = 610 nm) to detect total protein. Gel images were aligned using Progenesis SameSpots (Nonlinear Dynamics, Durham, NC, USA). Alignment of replicate control and treatment images was carried out for each pairwise comparison. To quantify abundance and oxidation differences between samples, spot volume was calculated and normalized. Spots differing significantly (ANOVA, < 0.05) between the averaged control and O3 treatment gels were selected for identification. Significant spots were excised using a Gelpix System (Genetix, New Milton, UK) for protein digestion and nanoLC-MS/MS analysis (Tables S1–S4) (Alvarez et al., 2009, 2011). LC-MS/MS data (Supporting Information Tables S1–S4) were processed using Analyst QS v1.1 (AB Sciex, Framingham, MA, USA) and searched against the NCBInr database (July 2010) using an in-house version of Mascot v2.20 (Matrix Science, Boston MA, USA) with the following parameters: tryptic peptides with ≤ one missed cleavage site; precursor and MS/MS fragment ion mass tolerances of 0.8 and 0.8 Da, respectively; variable carbamidomethylation and fluoresceination of cysteine; and variable oxidation of methionine. False-positive rates (average = 1.2%) were estimated during the Mascot analysis by searching against a randomized decoy database. Data were filtered using Scaffold 3 (Proteome Software, Portland, OR, USA). Positive identification criteria included ≥ two peptide sequences, protein probability of 99.9%, and peptide probability of 80%.

Enzyme assays

Spectrophotometric assays were used to determine phosphoglycerate kinase (PGK) (De & Kirtley, 1977), malate dehydrogenase (MDH) (Zeikus et al., 1977), glutamine synthetase (GS) (Kingdon et al., 1968), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Bergmeyer et al., 1974), fructose 1,6-bisphosphate aldolase (FBA) (Robertson & Fridovich, 1980), and ribulose 1,5-bisphosphate carboxylase oxygenase (RuBisCO) (Jordan & Ogren, 1981) activity. To measure chitinase activity, a fluorescence-based assay (Sigma; CS1030) that monitors the hydrolysis reaction was performed according to manufacturer’s instructions, in which 4-methylumbelliferyl N,N′-diacetyl-β-D-chitobioside, 4-methylumbelliferyl N-acetyl-β-D-glucosaminide, and 4-methylumbelliferyl β-D-N,N′,N′′-triacetylchitotriose were used as substrates for chitobiosidase, exochitinase, and endochitinase activities, respectively. Equal amounts of protein from control and treated tissue extracts were added to tubes containing assay-appropriate buffer plus 1% plant protease inhibitors (Sigma-Aldrich).

Immunoblot analysis of RuBisCO large subunit

Total protein extracts from soybean leaf tissue exposed to ambient (37 ppb) and high (116 ppb) O3 were fractionated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Proteins from the gels were transferred on to a nitrocellulose membrane in electrode buffer (25 mM Tris, pH 8.3, 192 mM glycine, and 20% (v/v) methanol). The membrane was washed with Tris-buffered saline (TBS) (80 mM Tris, pH 7.5, 200 mM NaCl) and incubated with TBS containing 5% (w/v) non-fat milk. Following this, membranes were incubated in 1:1000 dilution of anti-RuBisCO large subunit antibody (Agrisera, Vannas, Sweden). Reactive proteins were identified using a 1 : 3000 dilution of goat anti-rabbit IgG-alkaline phosphatase conjugate antibody (Bio-Rad) with color detection by enhanced chemiluminescent substrate (Pierce, Rockford, IL, USA). Comparison of immunoreactive protein amount was evaluated by densiometry (Quantity One software; Bio-Rad).

Results

Two-dimensional gel electrophoresis analysis of spots in O3-treated soybean

To identify O3-responsive proteins in soybean, soluble protein extracts of leaf and root tissue from plants grown at the SoyFACE facility under 37 (ambient), 58 (moderate), and 116 (high) ppb O3 were prepared. A total of nine plants were harvested at the R3 stage to provide three plants for each O3 treatment. From each plant, leaf and root tissue was collected. Total protein content in these plants showed no significant difference. Photosynthetic carbon uptake rates of upper canopy leaves showed a significant linear decline from 32.3 ± 2.8 μmol m−2 s−1 at ambient O3 to 27.1 ± 2.2 μmol m−2 s−1 at 58 ppb O3 to 17.0 ± 2.4 μmol m−2 s−1 at 116 ppb O3.

Extracted proteins were first incubated with NEM, which reacts with free thiols and not oxidized sulfhydryls, then reduced with DTT, and reacted with IAF to label previously oxidized thiols (Fig. 1) (Alvarez et al., 2011). This process allows for differential IAF modification of oxidized sulfhydryl groups in proteins from tissue extracts (Baty et al., 2005; Alvarez et al., 2011).

After 2-DE, gels were imaged for IAF-labeled proteins, then stained with Sypro Ruby and imaged for total protein (Fig. S1). Gels for each of the three biological replicates from each tissue type and O3 treatment combination were compared against the average ambient O3 gel images, and spots that significantly changed in signal intensity were identified. If an aligned spot was identified in each of three replicates, it was marked for further analysis. Across the various combinations of samples (O3 treatment, soybean tissue, and proteome), a total of 1455 spots were detected, of which 277 spots were differentially expressed and/or oxidized (Table 1).

Table 1.   Total number of spots detected and identified as either differentially expressed or oxidized across all experimental conditions relative to ambient controls
O3 treatmentTissueSignalSpots detectedSpots differentially expressed/oxidizedSpots with one proteinSpots with multiple proteinsTotal spots identified
  1. The number of spots detected only includes spots in common to all three gels for each treatment, soybean tissue, and signal combination. Differentially expressed/oxidized spots are further broken down into those with either single or multiple proteins. Sypro, Sypro Ruby; IAF, 5-iodoacetamidofluorescein.

116 ppbLeafSypro1542991726
116 ppbLeafIAF1714791827
116 ppbRootSypro2083641014
116 ppbRootIAF196295611
58 ppbLeafSypro196447411
58 ppbLeafIAF195437310
58 ppbRootSypro1582171017
58 ppbRootIAF1772891524
Total  14552775783140

Spots that differed in abundance and/or oxidation were excised, trypsin-digested, and analyzed by nano-LC/MS/MS. The resulting spectra were searched against the NCBInr database, with 57 spots containing a single protein and 83 spots having two or more proteins (Tables S1–S4). Subsequent protein identification and analysis did not distinguish between single and multiple protein spots. Across the various treatments and tissue combinations, 159 proteins changed in abundance and/or oxidation state (Fig. 2a). Of those proteins, 55, 27, 9, and 30 were unique to the 116 ppb O3 leaf, 116 ppb O3 root, 58 ppb O3 leaf, and 58 ppb O3 root samples, respectively. A further 38 proteins were found to differ in multiple tissue–O3 combinations. Within each of the four tissue–O3 combinations, various proteins changed in abundance and/or oxidation state (Fig. 2b–e). For example, the 116 ppb O3 leaf sample contained 79 proteins, of which 35 changed in abundance, 22 became more oxidized, and 22 changed in both abundance and oxidation.

Figure 2.

Differences in proteins in soybean leaf and root tissues under moderate (58 ppb) and high (116 ppb) O3 compared with ambient (58 ppb) O3. (a) Distribution of proteins across combinations of tissue and O3 concentration. Numbers in overlapping regions of the lobes indicate proteins found in more than one set of conditions. (b–e) Numbers of differentially oxidized (5-iodoacetamidofluorescein (IAF)) and/or abundant (Sypro Ruby) proteins between treated samples and controls. In each panel, the diagram in the top left corner indicates which lobe from panel (a) is analyzed.

High O3-induced changes in the total and redox proteomes of soybean leaf

Comparison between the tissue–O3 treatment combinations revealed increased abundance and greater oxidation of the largest number of proteins in the high-O3 leaf sample (Figs 3, 4). Although analysis of the 116 ppb O3 root and 58 ppb O3 leaf and root samples showed multiple proteins either increasing or decreasing in abundance, these generally changed < twofold (Fig. 3). Moreover, both the number of proteins and the fold changes in these samples were less than those observed in the high-O3 leaf sample. For example, in the 116 ppb O3 root sample, the abundance of six proteins increased and that of 10 proteins decreased within a twofold range and 11 proteins became more oxidized and eight proteins more reduced (Fig. 3a). Overall, the changes in protein abundance observed in the high-O3 root and the moderate-O3 leaf and root samples were comparable to those described in previous studies of plant proteomes following short-term O3 exposure in growth chambers (Agrawal et al., 2002; Bohler et al., 2007; Cho et al., 2008; Feng et al., 2008; Renaut et al., 2009; Ahsan et al., 2010; Sarkar et al., 2010).

Figure 3.

Summary of fold-changes in total and redox proteomes of soybean root tissue exposed to high (116 ppb) O3 (a); and leaf (b) and root tissues (c) exposed to moderate (58 ppb) O3. Panels show the fold changes in oxidation (5-iodoacetamidofluorescein (IAF)) and expression level (Sypro) relative to ambient controls for proteins identified by nano-LC/MS/MS. Information about identified proteins is provided in Supporting Information Tables S2–S4.

Figure 4.

Summary of fold changes in total and redox proteomes of soybean leaf tissue exposed to high (116 ppb) O3. Fold changes, relative to ambient (37 ppb) O3 sample, in oxidation state (5-iodoacetamidofluorescein (IAF)) and abundance (Sypro) for identified proteins are plotted. Names of representative proteins are shown with highly (> threefold) oxidized (orange box) and oxidized (> threefold)/expressed (> 1.5-fold) proteins (pink boxes) indicated. Details of identified proteins are in Supporting Information Table S1. Protein abbreviations are as in the text.

By contrast, the 116 ppb O3 leaf sample contained 35 proteins with up to fivefold greater abundance compared with ambient samples, and two proteins with c. 1.5-fold decreased expression (Fig. 4). More striking was the shift in proteins with increased oxidation in the high-O3 leaf sample. Twenty-two proteins increased in total abundance and became two- to ninefold more oxidized. In addition, 22 other proteins became up to fivefold more oxidized without significant changes in abundance (Fig. 4). For the first time, this observation directly links O3 stress to altered thiol-redox state in plant proteins.

Changes in the total and thiol-redox proteomes of leaf tissue exposed to high O3 occurred across a range of biochemical pathways (Fig. 5), including redox systems, carbon metabolism (glycolysis, Calvin cycle, and citric acid cycle), photosynthesis, signaling and homeostasis systems, amino acid metabolism, specialized metabolism of flavonoids and isoprenoids, and are discussed in more detail later. Between the moderate- and high-O3 concentrations, there is a clear shift toward increased expression and protein sulfhydryl oxidation across metabolism in the leaf tissue of soybean.

Figure 5.

Overview of total and redox proteome changes in soybean leaf exposed to high O3 (116 ppb). A schematic view of different metabolic pathways in which proteins were identified is shown. Certain proteins grouped by metabolic/cellular function are highlighted with colored backgrounds. Proteins that change in oxidation state (italic) or abundance (bold), or both (bold and italic), are shown. Details of identified proteins are given in Supporting Information Table S1. Protein abbreviations are as in the text.

Analysis of proteins in leaf tissue exposed to high-O3 treatment

To examine possible enzyme activity changes in the high-O3 leaf sample, assays of PGK, FBA, GAPDH, MDH, and GS were performed. These enzymes were targeted for activity analysis because all of these enzymes showed increased abundance and oxidation in the 116 ppb O3 leaf sample and are linked to carbon/energy metabolism. Except for PGK, each enzyme exhibited increased activity in the high-O3 leaf sample compared with the ambient control with the fold change correlated to increased protein abundance (Table 2).

Table 2.   Comparison of enzyme activities in soybean leaf tissues exposed to ambient (37 ppb) and high (116 ppb) O3
Enzyme activityAmbient ozone activity (mmol min−1 g−1 FW)High ozone activity (mmol min−1 g−1 FW)Activity fold changeTotal protein (oxidation) fold change
  1. Values are averages ± SD for n = 4–8. ND, no detected change.

  2. Enzymes: PGK, phosphoglycerate kinase; FBA, fructose 1,6-bisphosphate aldolase; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; MDH, malate dehydrogenase; GS, glutamine synthetase; RuBisCO, ribulose 1,5-bisphosphate carboxylase oxygenase.

PGK715 ± 82629 ± 680.91.8 (5.7)
FBA31.6 ± 6.572.1 ± 9.92.31.8 (4.8)
GAPDH183 ± 75693 ± 1013.81.7 (4.4)
MDH72.0 ± 3.393.9 ± 10.41.32.0 (5.4)
GS106 ± 9246 ± 252.31.9 (4.5)
RuBisCO11.1 ± 2.610.0 ± 4.10.9ND (ND)
Exochitinase10.5 ± 0.817.2 ± 0.31.64.0 (9.4)
Endochitinase0.34 ± 0.021.40 ± 0.084.14.0 (9.4)
Chitobiosidase0.012 ± 0.0460.133 ± 0.02311.14.0 (9.4)

In addition to these enzymes, the activity and expression level of RuBisCO and phosphoenolpyruvate carboxylase (PEPC) were examined because O3 exposure can alter amounts of these proteins (Heagle, 1989; Chernikova et al., 2000; Robinson & Britz, 2000; Krupa et al., 2001; Morgan et al., 2004, 2006; Baier et al., 2005; Fiscus et al., 2005; Chen et al., 2009; Betzelberger et al., 2010). For RuBisCO, activity assays (Table 2) and immunoblot analysis of the large subunit (Fig. 6) showed a slight decrease in both activity (10% less) and protein (13% less) between the high O3 and ambient leaf samples. Expression of PEPC in the ambient and high O3 leaf samples, as determined by Western blot, was not altered (not shown).

Figure 6.

Immunoblot analysis of ribulose 1,5-bisphosphate carboxylase oxygenase (RuBisCO) large subunit expression. Protein extracts from soybean leaf tissue exposed to ambient (37 ppb) and high (116 ppb) O3 were probed using anti-RuBisCO large subunit antibody. Lanes 1 and 3 contain 10 μg of total protein extract and lanes 2 and 4 contain 5 μg of total protein extract from the indicated samples.

In the high O3 leaf sample, a glycosyl hydrolase/chitinase increased four- and 9.4-fold in expression and oxidation, respectively. Glycosyl hydrolases are involved in the degradation of various sugars, but are also mechanistically related to chitinases, which cleave glycosidic bonds and are typically involved in responses to herbivory. Because many glycosyl hydrolases display varied activities, a fluorescence-based assay was used to evaluate exochitinase, endochitinase, and chitobiosidase activities in control and high-O3 leaf tissues, which increased 1.6-, 4.1-, and 11.1-fold, respectively (Table 2). It is unclear if these changes result from the identified protein or changes in multiple isoforms.

Discussion

Understanding the molecular mechanisms and metabolic consequences of how climate changes, such as elevated tropospheric O3 concentrations, impact crop plants is essential for efforts to maintain their performance under increasing environmental stresses. Although earlier growth-chamber studies describe the effects of short-term continuous O3 exposure on the total proteomes of soybean, rice, wheat, and poplar (Agrawal et al., 2002; Bohler et al., 2007; Cho et al., 2008; Feng et al., 2008; Renaut et al., 2009; Ahsan et al., 2010; Sarkar et al., 2010), these investigations neither reported on the consequences of seasonal elevated O3 exposure under field conditions nor examined its effect on the protein thiol-redox state.

Tropospheric O3 at concentrations > c. 40 ppb negatively affect soybean growth and yield with decreased shoot and pod biomass, fewer pods produced, and premature leaf senescence (Heagle, 1989; Krupa et al., 2001; Morgan et al., 2004, 2006; Fiscus et al., 2005; Ainsworth et al., 2008; Chen et al., 2009; Emberson et al., 2009; Van Dingenen et al., 2009; Betzelberger et al., 2010). O3 enters leaves at the stomata and produces reactive oxygen species (ROS), which diminishes photosynthesis, changes stomatal conductance by altering calcium influxes, increases photorespiration, and induces premature leaf senescence (Heagle, 1989; Morgan et al., 2004, 2006; Baier et al., 2005; Fiscus et al., 2005; Cho et al., 2008; Chen et al., 2009; Betzelberger et al., 2010). Long-term O3 stress leads to reduced photosynthesis and mobilization of energy reserves that convert leaf starch to sugar (Ahsan et al., 2010). Accordingly, previous studies identified several sugar metabolism proteins, including MDH and phosphoglycerate mutase, as highly expressed under acute O3 stress (Bohler et al., 2007). Likewise, decreased photosynthetic efficiency, reduced RuBisCO activity, and elevated PEPC activity are classic markers for O3 damage and senescence; however, these effects can vary with the length, concentration, and type of exposure (Heagle, 1989; Chernikova et al., 2000; Robinson & Britz, 2000; Krupa et al., 2001; Morgan et al., 2004, 2006; Baier et al., 2005; Fiscus et al., 2005; Chen et al., 2009; Betzelberger et al., 2010).

Previous studies on soybean, rice, and wheat exposed to constant 120–200 ppb O3 for 3–5 d in growth chambers typically led to the identification of 20–50 proteins that changed in abundance either up or down (Agrawal et al., 2002; Bohler et al., 2007; Cho et al., 2008; Feng et al., 2008; Renaut et al., 2009; Ahsan et al., 2010; Sarkar et al., 2010). The actual proteins identified in these short-term O3 exposure experiments vary widely, but decreased abundance of the RuBisCO large and small subunits and increased ascorbate peroxidase concentrations are generally observed.

Analysis of the total and redox proteomes of leaf and root tissues from soybean grown at SoyFACE under moderate (58 ppb) chronic daytime O3 showed changes in the number of proteins and their abundance comparable to short-term growth-chamber experiments (Figs 2, 3). Strikingly, all but two proteins identified in the high-O3 soybean leaf sample increased in abundance and/or oxidation (Figs 4 & 5). This suggests a threshold between 58 and 116 ppb O3 at which the expression and oxidation of multiple proteins in leaf tissue drastically increase.

The data presented here indicate that the changes in the soybean leaf total and redox proteomes resulting from seasonal exposure to high O3 (Fig. 5) differ from those observed at moderate O3, are more widespread across metabolism than previously reported, and differ from those resulting from constant short-term O3 exposure. The increased expression of chlorophyll α/β-binding protein, ferredoxin reductase, and a chlorophyllase-like protein observed in the high-O3 leaf sample (Fig. 5) may help maintain photosynthesis at this growth stage before O3-induced senescence occurs.

Similarly, RuBisCO large and small subunits, RuBisCO activase, RuBisCO-associated protein, and RuBisCO-binding protein showed increased expression and/or oxidation in this sample. RuBisCO was not considered a priori as a target of oxidation, although previous proteomic studies of other plants grown under short-term ozone treatment in growth chambers identified it as changing in abundance (Agrawal et al., 2002; Bohler et al., 2007; Cho et al., 2008; Feng et al., 2008; Renaut et al., 2009; Ahsan et al., 2010; Sarkar et al., 2010). Oxidation of RuBisCO can reduce the catalytic activity of the enzyme (Marcus et al., 2003); however, alterations in proteins associated with RuBisCO (i.e. the activase) may compensate for possible oxidative changes. Likewise, proteins related to iron homeostasis (ferredoxin reductase and ferritin) show increased abundance in soybean leaf under high O3. Because the spots containing RuBisCO included multiple proteins, activity assays and immunoblot analysis were used for further examination and showed a slight 10% decrease at this stage of soybean growth (Table 2, Fig. 6). Moreover, immunoblot analysis of PEPC in leaf tissue showed no significant change in the high-O3 leaf sample compared with control. Consistent with the known effect of O3 on photosynthesis, the PPFD values of leaves from soybean grown under ambient, moderate, and high O3 decreased with increasing O3 exposure. The observed changes in the total and thiol-redox proteomes of soybean described here reinforce physiological data indicating that responses to constant short-term vs seasonal daytime O3 exposure and to exposure at moderate and high O3 differ (Chen et al., 2009).

Multiple proteins (i.e. PGK, GAPDH, FBA, ribose-5-phosphate isomerase, phosphoribulokinase, triosephosphate isomerase, MDH, and isocitrate dehydrogenase) in the reduction and regeneration phases of the Calvin cycle, glycolysis, and the citric acid cycle increase in expression and/or oxidation state in high-O3 leaf samples (Fig. 5). In addition, the activity of FBA, GAPDH, and MDH increased in this sample (Table 2). While consistent with earlier proteomic studies (Agrawal et al., 2002; Bohler et al., 2007; Cho et al., 2008; Feng et al., 2008; Renaut et al., 2009; Ahsan et al., 2010; Sarkar et al., 2010), this analysis indicates a wider range of protein changes across more metabolic pathways than reported before and for the first time identifies thiol-redox alterations resulting from an environmental oxidative stress. All of the proteins above interact with thioredoxin, which is essential for maintaining protein redox-state in plants (Buchanan & Balmer, 2005; Meyer et al., 2009; Montrichard et al., 2009). Moreover, phosphoribulokinase and GAPDH interact via thioredoxin-mediated redox changes in response to light intensity (Howard et al., 2008), which may be affected by O3-related changes in oxidation state. Similarly, MDH is a critical regulatory point in the citric acid cycle; however, the cytosolic form of the enzyme is redox-regulated and inactivated under oxidizing conditions (Hara et al., 2006). In the high-O3 leaf sample, MDH had two and fivefold higher expression and oxidation compared with controls with a modest 1.3-fold increase in total activity (Table 2). It is possible that these changes reflect the need to maintain MDH in the leaf to supply metabolites to the citric acid cycle.

In addition to changes in core carbon metabolism, the starch/sugar mobilization pathways (phosphohexomutase, glucanase, and glycosyl hydrolase/acid chitinase) showed increased abundance and/or oxidation in the high-O3 leaf sample (Fig. 5). The changes in enzymes related to the conversion of starch to sugar are consistent with a shift in energy demand under O3 stress (Bohler et al., 2007). Of the 79 proteins identified in the high-O3 leaf sample, a putative glycosyl hydrolase showed the greatest changes in both abundance and oxidation (Fig. 4). Although it is unclear if this protein functions in cell wall degradation, pathogen response, or sugar mobilization, various glycosyl hydrolase/chitinase activities increased up to 11-fold in the high-O3 leaf sample (Table 2), and may be connected to the mobilization of starch for energy production (Bohler et al., 2007).

Proteins in specialized metabolic pathways related to O3 stress were also identified in the high-O3 leaf sample. In the isoprenoid synthesis pathway, deoxyxylulose phosphate (DXP) oxidoreductase and isopentenyl diphosphate (IPP) isomerase were oxidized (Fig. 5); however, the effect of oxidation on their activities is unclear. Interestingly, volatile isoprenoid emissions may act as an O3 protection mechanism in plants (Loreto & Fares, 2007). Moreover, changes in carotenoid and (iso)flavonoid pathways (carotenoid-associated protein, chalcone isomerase, isoflavone reductase, and caffeoyl-CoA methyltransferase) suggest alterations in the synthesis of these compounds, which act as photoprotective compounds and antioxidants (Middleton & Teramura, 1993).

In leaves exposed to high O3, up-regulation and/or oxidation of proteins in amino acid biosynthesis and nitrogen homeostasis were observed (Fig. 5). The cytosolic form of GS, a central player in nitrogen sensing, increased in abundance, oxidation, and activity in the 116 ppb O3 leaf sample (Table 2). Elevated expression of cytosolic GS is associated with leaf senescence and the recycling of ammonia during stress conditions (Brugière et al., 2000). Aspartate-semialdehyde dehydrogenase is a control point in isoleucine, methionine, lysine, and threonine synthesis (Alvarez et al., 2004; Schroeder et al., 2010). Although redox control has not been described for the plant enzyme, oxidation of a catalytic cysteine in the bacterial homolog alters activity (Alvarez et al., 2004). Likewise, carbamoyl phosphate synthetase, which becomes more oxidized following high-O3 exposure, is sensitive to changes in redox environment (Hart & Powers-Lee, 2009). Also related to nutrient metabolism, the observed expression and oxidation changes in 14-3-3 proteins may further modify the activities of enzymes across the carbon, nitrogen, and sulfur nutrient assimilation pathways and/or signal transduction systems linked to stress responses (Oecking & Jaspert, 2009; Shin et al., 2011).

The proteomic analysis here supports studies demonstrating that redox-protection mechanisms play a critical role in plant responses to O3 exposure (Gillespie et al., 2011), and for the first time directly shows that O3 exposure changes the thiol oxidation state of proteins in various pathways. Chronic exposure to high O3 leads to activation of redox protection mechanisms in plants and increased expression and/or oxidation of proteins in those systems, including ascorbate peroxidase, methionine sulfoxide reductase, and glutathione-S-transferases (Figs 4, 5) (Gillespie et al., 2011). In plants, the ascorbate-glutathione system maintains redox homeostasis (Ishikawa & Shigeoka, 2008). The increased abundance of ascorbate peroxidase, which is critical in this system, is linked to responses that attenuate oxidative stress and to mobilization of sugars, which further enhance ascorbate synthesis (Nishikawa et al., 2005). The threefold increase in oxidation of methionine sulfoxide reductase, which repairs oxidized methionines, suggests a potential role in responding to O3 stress. The plant methionine sulfoxide reductases contain two redox-active cysteines and require regeneration by thioredoxin for catalysis (Rouhier et al., 2006; Tarrago et al., 2009).

Although the role of O3 as an environmental stress is well established (Heagle, 1989; Fiscus et al., 2005; Ainsworth et al., 2008), the extent of redox-linked changes in crops exposed to seasonal O3 concentrations has not been examined before. The study presented here demonstrates that high-O3 exposure alters the thiol-redox state of multiple proteins across metabolic pathways, but these changes in the redox proteome of soybean leaf are likely only a small fraction of the total number of proteins that change in oxidation state. Importantly, changes in cellular redox state provide a mechanism for rapidly modulating the activities of multiple proteins across biochemical pathways by altering active site residues, oligomerization, and/or cellular localization (Paget & Buttner, 2003; Barford, 2004; Jez et al., 2004; Hicks et al., 2007; Yi et al., 2010). The work presented here provides a first insight into the total and thiol-redox proteome changes in soybean grown under seasonal O3 stress, but questions remain.

Future work using these methods could examine both the temporal and tissue-specific changes caused by O3 exposure from the initial short-term response to long-term metabolic compensation to early senescence. Efforts using more sensitive isolation/detection strategies will likely reveal broader redox-linked changes resulting from O3 stress. Under moderate O3 (58 ppb) concentration, plant antioxidant response systems may be able to abrogate some damage, especially considering that seasonal O3 exposure is not constant with decreases at night and in wet weather that provide recovery time for the plant. The increased protein abundance and oxidation in the high-O3 leaf tissue suggests that in addition to antioxidant responses other metabolic changes may help to maintain plant metabolism through the stress. It is possible that oxidative damage alters the activity of key metabolic proteins in soybean grown in high O3, which then requires increased expression to maintain sufficient metabolic activity and leads to acclimation under a stress during the growing season; however, by the end of a growing season, soybean succumbs to senescence with significant reductions in seed yield (Morgan et al., 2004, 2006). A more complete time-dependent and tissue-specific analysis of proteomic changes (total and redox) remains to be performed and could help in providing molecular insights into the changes that occur between initial oxidative shock responses and long-term growth under the stress. Ultimately, understanding how environmental O3 affects redox-sensitive pathways in different plants may help in the development of crops better adapted to global climate change and provide information about how to target the engineering of O3 protection systems.

Acknowledgements

We thank Amy Betzelberger for generously providing photosynthetic data and Sophie Alvarez for help with proteomic analysis. This research was funded by the National Science Foundation (MCB-0904215 to J.M.J. and L.M.H.). A.G. was supported by an American Society of Plant Biologists Pioneer Hi-Bred Graduate Research Fellowship. SoyFACE is supported by the USDA ARS.

Ancillary