Hydrogen Peroxide Response in Leaves of Poplar (Populus simonii × Populus nigra) Revealed from Physiological and Proteomic Analyses

Hydrogen peroxide (H2O2) is one of the most abundant reactive oxygen species (ROS), which plays dual roles as a toxic byproduct of cell metabolism and a regulatory signal molecule in plant development and stress response. Populus simonii × Populus nigra is an important cultivated forest species with resistance to cold, drought, insect and disease, and also a key model plant for forest genetic engineering. In this study, H2O2 response in P. simonii × P. nigra leaves was investigated using physiological and proteomics approaches. The seedlings of 50-day-old P. simonii × P. nigra under H2O2 stress exhibited stressful phenotypes, such as increase of in vivo H2O2 content, decrease of photosynthetic rate, elevated osmolytes, antioxidant accumulation, as well as increased activities of several ROS scavenging enzymes. Besides, 81 H2O2-responsive proteins were identified in the poplar leaves. The diverse abundant patterns of these proteins highlight the H2O2-responsive pathways in leaves, including 14-3-3 protein and nucleoside diphosphate kinase (NDPK)-mediated signaling, modulation of thylakoid membrane structure, enhancement of various ROS scavenging pathways, decrease of photosynthesis, dynamics of proteins conformation, and changes in carbohydrate and other metabolisms. This study provides valuable information for understanding H2O2-responsive mechanisms in leaves of P. simonii × P. nigra.


Introduction
Various environmental stresses usually affect reactive oxygen species (ROS) homeostasis in plants, leading to the generation of excess ROS, such as singlet oxygen ( 1 O 2 ), superoxide anion radicals (O 2 •− ), hydrogen peroxide (H 2 O 2 ), and hydroxyl radicals (HO • ). Among them, H 2 O 2 is the most abundant ROS in plant cells during photosynthesis, photorespiration, and respiration processes [1]. The relatively stable non-radical H 2 O 2 can easily penetrate membrane through water channels, functioning as a likely long-distance signaling molecule in plant growth and stress perception [2]. Moreover, H 2 O 2 has been proven to be a regulator of many physiological processes, such as cell wall modulation, senescence,

Membrane Integrity and Osmolyte Accumulation in Leaves
To evaluate the effects of H2O2 on membrane stability, the malondialdehyde (MDA) content and relative electrolyte leakage (REL) in leaves were determined. MDA contents and RELs were not changed under 12 mM H2O2, but increased under 24 and 36 mM H2O2 ( Figure 3A,B). MDA contents were increased from 17.6 nmol•g −1 fresh weight (FW) in control to about 21.4 nmol•g −1 FW under 24 and 36 mM H2O2 ( Figure 3A). For REL, a 1.3-fold increase under 24 mM H2O2 and a 1.9-fold under 36 mM H2O2 were observed when compared with control ( Figure 3B). The leaves of P. simonii × P. nigra were immersed in 0, 12, 24 and 36 mM H2O2 solutions for 6 h, respectively ( Figure 1). The photosynthetic parameters were measured to evaluate the photosynthetic changes in response to the H2O2 stress. The net photosynthetic rate (Pn) decreased from 4 μmol CO2•m −2 •s −1 in control to about 3.7 μmol CO2•m −2 •s −1 under each H2O2 treatment (Figure 2A). The stomatal conductances (Gs) under 12, 24, and 36 mM H2O2 were also reduced 1.2-, 1.7-, and 2.4-fold, respectively, when compared with control ( Figure 2B). In addition, the intercellular CO2 concentration (Ci) increased slightly from 473.2 μmol CO2•mol −1 in control to 496.3 μmol CO2•mol −1 under 36 mM H2O2, but the transpiration rate (Tr) was not significantly altered under the H2O2 stress ( Figure 2C,D).   (Tr). The values were determined after plants were treated with 0, 12, 24 and 36 mM H2O2, and were presented as means ± standard deviation (SD) (n = 3). The different small letters indicate significant difference (p < 0.05) among different treatments.

Membrane Integrity and Osmolyte Accumulation in Leaves
To evaluate the effects of H2O2 on membrane stability, the malondialdehyde (MDA) content and relative electrolyte leakage (REL) in leaves were determined. MDA contents and RELs were not changed under 12 mM H2O2, but increased under 24 and 36 mM H2O2 ( Figure 3A,B). MDA contents were increased from 17.6 nmol•g −1 fresh weight (FW) in control to about 21.4 nmol•g −1 FW under 24 and 36 mM H2O2 ( Figure 3A). For REL, a 1.3-fold increase under 24 mM H2O2 and a 1.9-fold under 36 mM H2O2 were observed when compared with control ( Figure 3B).

Membrane Integrity and Osmolyte Accumulation in Leaves
To evaluate the effects of H 2 O 2 on membrane stability, the malondialdehyde (MDA) content and relative electrolyte leakage (REL) in leaves were determined. MDA contents and RELs were not changed under 12 mM H 2 O 2 , but increased under 24 and 36 mM H 2 O 2 ( Figure 3A Figure 3B).
In addition, the contents of proline and glycine betaine were increased gradually and significantly with the increasing concentration of H2O2. The proline contents under H2O2 treatment of 12, 24 and 36 mM were increased 1.3-, 1.7-, and 2.8-fold, respectively ( Figure 3C). The contents of glycine betaine under three H2O2 treatments were increased 1.1-, 1.2-, 1.3-fold, respectively ( Figure  3E). In addition, the contents of soluble sugar were increased 1.3-fold under 24 mM and 1.5-fold under 36 mM H2O2 ( Figure 3D).

ROS and Antioxidant Substances Content, and Antioxidant Enzyme Activities
To evaluate the ROS homeostasis in the H2O2-treated leaves, the O2 •− generation rate, H2O2 content, and the activities of several ROS scavenging enzymes were analyzed. The O2 •− generation rate remained constant under 12 and 36 mM H2O2 treatments, but was increased 1.2-fold under 24 mM H2O2 treatment ( Figure 4A). H2O2 content in leaves was increased 1.4-fold under 24 mM H2O2 treatment ( Figure 4A). Superoxide dismutase (SOD) activity was increased about 1.2-fold under 24 and 36 mM H2O2 treatments ( Figure 4B). Besides, the activities of several ROS scavenging enzymes were altered in leaves under certain H2O2 concentrations. The CAT activities were decreased 1.2-fold under 24 mM and 1.5-fold under 36 mM H2O2 treatments ( Figure 4B). However, the activities of ascorbate peroxidase (APX) were increased about 1.2-fold under 12 and 24 mM H2O2, and peroxidase (POD) activity was increased 1.7-fold under 24 mM H2O2 ( Figure 4C). The glutathione peroxidase (GPX) activity was increased 1.1-fold under 36 mM H2O2 treatment ( Figure 4D). Moreover, the activities of three enzymes involved in the regeneration of the reduced antioxidants were all altered under the H2O2 stress. The activity of monodehydroascorbate reductase (MDHAR) was inhibited, while the activities of dehydroascorbate reductase (DHAR) and glutathione reductase (GR) were significantly increased under H2O2 treatments ( Figure 4E,F). The activities of MDHAR under three H2O2 treatments were reduced 1.1-, 1.2-, and 1.3-fold, respectively. DHAR activities were increased about 1.2-fold under 24 and 36 mM H2O2, and GR activities were increased 1.3-fold under 24 mM H2O2 and 1.9-fold under 36 mM H2O2 ( Figure 4E,F). In addition, the glutathione S-transferase (GST) activity was induced 1.2-fold under 36 mM H2O2 treatment ( Figure 4F).
In addition, ascorbate (AsA), dehydroascorbate (DHA), reduced glutathione (GSH), and oxidized glutathione (GSSG) were detected in leaves in response to the H2O2 treatment. The contents of AsA were decreased 1.3-fold under 24 mM H2O2, but increased 1.2-fold under 36 mM H2O2 treatment ( Figure 4G). The contents of DHA and GSH were all reduced under H2O2 treatment. DHA In addition, the contents of proline and glycine betaine were increased gradually and significantly with the increasing concentration of H 2 O 2 . The proline contents under H 2 O 2 treatment of 12, 24 and 36 mM were increased 1.3-, 1.7-, and 2.8-fold, respectively ( Figure 3C). The contents of glycine betaine under three H 2 O 2 treatments were increased 1.1-, 1.2-, 1.3-fold, respectively ( Figure 3E). In addition, the contents of soluble sugar were increased 1.3-fold under 24 mM and 1.5-fold under 36 mM H 2 O 2 ( Figure 3D).  Proteins were separated on 24 cm linear gradient immobilized pH gradient (IPG) strips (pH 4-7) using isoelectric focusing (IEF) in the first dimension, followed by 12.5% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gels in the second dimension. The 2DE gel was stained with Coomassie Brilliant Blue. Molecular weight (MW) in kilodaltons (KDa) and isoelectric point (pI) of proteins are indicated on the left and top of the gel, respectively. Eighty-one H 2 O 2 -responsive proteins identified by matrix-assisted laser desorption/ ionization (MALDI) tandem time of flight (TOF-TOF) mass spectrometry were marked with numbers on the gel, and the detailed information can be found in Table 1, Figure S1 and Table S1.                                      Glyceraldehyde -  Glyceraldehyde -  Fructose                                                                             a Assigned spot number as indicated in Figure 5. b The name of the proteins identified by MALDI TOF/TOF MS. Protein names marked with an asterisk (*) have been edited based on BLAST against NCBI non-redundant protein database. The detailed information of the NCBI BLAST can be found in

Annotation and Functional Categorization of the H2O2-Responsive Proteins
Among the 81 H2O2-responsive proteins, 34 proteins were originally annotated as unknown, hypothetical proteins, or without annotation. They were all re-annotated according to the Basic Local Alignment Search Tool (BLAST) analysis (Tables 1 and S3). Based on BLAST alignments, Gene Ontology, subcellular localization prediction, and information from literature, the 81 proteins were classified into ten functional categories including photosynthetic electron transfer chain, Calvin cycle, carbohydrate and energy metabolism, other metabolism, protein synthesis, protein folding and unfolding, redox homeostasis and stress defense, signaling, cell structure, and miscellaneous or function unknown (Table 1 and Figure 6A). Interestingly, proteins involved in photosynthetic electron transfer and Calvin cycle accounted for the largest group (46% of H2O2-responsive proteins). Besides, carbohydrate and energy metabolism (15%), other metabolism (12%), as well as protein folding and unfolding (10%) were also over-represented.  a Assigned spot number as indicated in Figure 5. b The name of the proteins identified by MALDI TOF/TOF MS. Protein names marked with an asterisk (*) have been edited based on BLAST against NCBI non-redundant protein database. The detailed information of the NCBI BLAST can be found in

Annotation and Functional Categorization of the H2O2-Responsive Proteins
Among the 81 H2O2-responsive proteins, 34 proteins were originally annotated as unknown, hypothetical proteins, or without annotation. They were all re-annotated according to the Basic Local Alignment Search Tool (BLAST) analysis (Tables 1 and S3). Based on BLAST alignments, Gene Ontology, subcellular localization prediction, and information from literature, the 81 proteins were classified into ten functional categories including photosynthetic electron transfer chain, Calvin cycle, carbohydrate and energy metabolism, other metabolism, protein synthesis, protein folding and unfolding, redox homeostasis and stress defense, signaling, cell structure, and miscellaneous or function unknown (Table 1 and Figure 6A). Interestingly, proteins involved in photosynthetic electron transfer and Calvin cycle accounted for the largest group (46% of H2O2-responsive proteins). Besides, carbohydrate and energy metabolism (15%), other metabolism (12%), as well as protein folding and unfolding (10%) were also over-represented.  a Assigned spot number as indicated in Figure 5. b The name of the proteins identified by MALDI TOF/TOF MS. Protein names marked with an asterisk (*) have been edited based on BLAST against NCBI non-redundant protein database. The detailed information of the NCBI BLAST can be found in

Annotation and Functional Categorization of the H2O2-Responsive Proteins
Among the 81 H2O2-responsive proteins, 34 proteins were originally annotated as unknown, hypothetical proteins, or without annotation. They were all re-annotated according to the Basic Local Alignment Search Tool (BLAST) analysis (Tables 1 and S3). Based on BLAST alignments, Gene Ontology, subcellular localization prediction, and information from literature, the 81 proteins were classified into ten functional categories including photosynthetic electron transfer chain, Calvin cycle, carbohydrate and energy metabolism, other metabolism, protein synthesis, protein folding and unfolding, redox homeostasis and stress defense, signaling, cell structure, and miscellaneous or function unknown (Table 1 and Figure 6A). Interestingly, proteins involved in photosynthetic electron transfer and Calvin cycle accounted for the largest group (46% of H2O2-responsive proteins). Besides, carbohydrate and energy metabolism (15%), other metabolism (12%), as well as protein folding and unfolding (10%) were also over-represented.  a Assigned spot number as indicated in Figure 5. b The name of the proteins identified by MALDI TOF/TOF MS. Protein names marked with an asterisk (*) have been edited based on BLAST against NCBI non-redundant protein database. The detailed information of the NCBI BLAST can be found in

Annotation and Functional Categorization of the H2O2-Responsive Proteins
Among the 81 H2O2-responsive proteins, 34 proteins were originally annotated as unknown, hypothetical proteins, or without annotation. They were all re-annotated according to the Basic Local Alignment Search Tool (BLAST) analysis (Tables 1 and S3). Based on BLAST alignments, Gene Ontology, subcellular localization prediction, and information from literature, the 81 proteins were classified into ten functional categories including photosynthetic electron transfer chain, Calvin cycle, carbohydrate and energy metabolism, other metabolism, protein synthesis, protein folding and unfolding, redox homeostasis and stress defense, signaling, cell structure, and miscellaneous or function unknown (Table 1 and Figure 6A). Interestingly, proteins involved in photosynthetic electron transfer and Calvin cycle accounted for the largest group (46% of H2O2-responsive proteins). Besides, carbohydrate and energy metabolism (15%), other metabolism (12%), as well as protein folding and unfolding (10%) were also over-represented. a Assigned spot number as indicated in Figure 5. b The name of the proteins identified by MALDI TOF/TOF MS. Protein names marked with an asterisk (*) have been edited based on BLAST against NCBI non-redundant protein database. The detailed information of the NCBI BLAST can be found in Table S2. c Protein subcellular localization predicted by software YLoc, LocTree3, Plant-mPLoc, ngLOC, and TargetP. Chl, chloroplast; Cyt, cytoplasm; Mit, mitochondrion. d Database accession numbers from NCBInr. e The Mascot score obtained after searching against the NCBInr database. f The number of unique peptides identified for each protein. g The mean values of protein spot volumes relative to total volume of all the spots. The different small letters on the columns indicate significant differences (p < 0.05) among the four samples as determined by one-way ANOVA. Error bars indicate ± SD.

Annotation and Functional Categorization of the H 2 O 2 -Responsive Proteins
Among the 81 H 2 O 2 -responsive proteins, 34 proteins were originally annotated as unknown, hypothetical proteins, or without annotation. They were all re-annotated according to the Basic Local Alignment Search Tool (BLAST) analysis (Table 1 and Table S3). Based on BLAST alignments, Gene Ontology, subcellular localization prediction, and information from literature, the 81 proteins were classified into ten functional categories including photosynthetic electron transfer chain, Calvin cycle, carbohydrate and energy metabolism, other metabolism, protein synthesis, protein folding and unfolding, redox homeostasis and stress defense, signaling, cell structure, and miscellaneous or function unknown (Table 1 and Figure 6A). Interestingly, proteins involved in photosynthetic electron transfer and Calvin cycle accounted for the largest group (46% of H 2 O 2 -responsive proteins). Besides, carbohydrate and energy metabolism (15%), other metabolism (12%), as well as protein folding and unfolding (10%) were also over-represented.
unfolding, redox homeostasis and stress defense, signaling, cell structure, and miscellaneous or function unknown (Table 1 and Figure 6A). Interestingly, proteins involved in photosynthetic electron transfer and Calvin cycle accounted for the largest group (46% of H2O2-responsive proteins). Besides, carbohydrate and energy metabolism (15%), other metabolism (12%), as well as protein folding and unfolding (10%) were also over-represented.

Subcellular Localization and Hierarchical Clustering of H 2 O 2 -Responsive Proteins
The subcellular localization of the 81 proteins was predicted using five different tools (i.e., YLoc, LocTree3, Plant-mPLoc, ngLOC, and TargetP) ( Figure 6B, Table 1 and Table S4). In total, 51 proteins (63%) were predicted to be localized in chloroplasts, 20 in cytoplasm, four in mitochondria, one secreted, and five uncertain. This implied that most chloroplast proteins were obviously affected by H 2 O 2.

Hierarchical Clustering and Analysis of H 2 O 2 -Responsive Proteins
To better understand the abundance patterns of the coordinately regulated proteins, hierarchical clustering analysis of the 81 H 2 O 2 -responsive proteins were performed, which revealed four main clusters (Figure 7). Cluster I included a total of 42 proteins, as the most group cluster, which included the significantly decreased proteins under H 2 O 2 treatment. Cluster II included the proteins decreased under 12 mM, but increased under 24 or 36 mM H 2 O 2 treatment. Cluster III contained the proteins unchanged or increased under relative lower concentration of H 2 O 2 stress, but decreased under relative higher concentration, especially 36 mM H 2 O 2 treatment. Several proteins involved in carbohydrate and energy metabolism were grouped into this subcluster. Cluster IV contained six proteins whose abundances were increased under H 2 O 2 treatment. Notably, several heat shock proteins (HSPs) were categorized into Cluster IV.

Protein-Protein Interaction (PPI) among H 2 O 2 -Responsive Proteins
To discover the relationship of the 81 H 2 O 2 -responsive proteins, the PPI networks were generated using the web-tool STRING 10 [30]. Among the 81 H 2 O 2 -responsive proteins, 59 unique homologous proteins were found in Arabidopsis (Table S5) [31]. Out of the 59 proteins, 40 proteins were depicted in the STRING database ( Figure 8). Six modules forming tightly connected clusters were illuminated, and stronger associations were represented by thicker lines in the networks (Figure 8). Twelve and nine proteins were connected in Module 1 and Module 2, respectively. Most of them were involved in photosynthesis or carbohydrate metabolism. Module 3 contained five proteins mainly involved in energy metabolism, and Module 4 contained six proteins mainly involved in protein folding. Three proteins involved in amino acid metabolism were assigned in Module 5. unchanged or increased under relative lower concentration of H2O2 stress, but decreased under relative higher concentration, especially 36 mM H2O2 treatment. Several proteins involved in carbohydrate and energy metabolism were grouped into this subcluster. Cluster IV contained six proteins whose abundances were increased under H2O2 treatment. Notably, several heat shock proteins (HSPs) were categorized into Cluster IV. The rows represent individual proteins. The increased or decreased proteins are indicated in red or green, respectively. The color intensity increases with increasing abundant differences, as shown in the scale bar. The scale bar indicates log (base2) transformed protein abundance ratios ranging from −1.8 to 1.8. Functional categories indicated by capital letters, spot numbers, and protein names are listed on the right side. A, photosynthetic electron transfer chain; B, Calvin cycle; C, carbohydrate and energy metabolism; D, other metabolisms; E, protein synthesis; F, protein folding and unfolding; G, redox homeostasis and stress defense; H, signaling; I, cell structure; J, miscellaneous or function unknown. The abbreviations refer to Table 1. The rows represent individual proteins. The increased or decreased proteins are indicated in red or green, respectively. The color intensity increases with increasing abundant differences, as shown in the scale bar. The scale bar indicates log (base2) transformed protein abundance ratios ranging from −1.8 to 1.8. Functional categories indicated by capital letters, spot numbers, and protein names are listed on the right side. A, photosynthetic electron transfer chain; B, Calvin cycle; C, carbohydrate and energy metabolism; D, other metabolisms; E, protein synthesis; F, protein folding and unfolding; G, redox homeostasis and stress defense; H, signaling; I, cell structure; J, miscellaneous or function unknown. The abbreviations refer to Table 1. homologous proteins were found in Arabidopsis (Table S5) [31]. Out of the 59 proteins, 40 proteins were depicted in the STRING database ( Figure 8). Six modules forming tightly connected clusters were illuminated, and stronger associations were represented by thicker lines in the networks ( Figure  8). Twelve and nine proteins were connected in Module 1 and Module 2, respectively. Most of them were involved in photosynthesis or carbohydrate metabolism. Module 3 contained five proteins mainly involved in energy metabolism, and Module 4 contained six proteins mainly involved in protein folding. Three proteins involved in amino acid metabolism were assigned in Module 5.
Excessive H 2 O 2 can damage protein and lipid structure, leading to the destruction of cell membrane stability. In this study, the increases of MDA content and REL implied that H 2 O 2 induced plasma membrane lipid peroxidation and plasma membrane permeability in leaves of poplar ( Figure 3A,B). Importantly, our proteomics results indicated that plastid lipid associated proteins (PAPs) and tubulin were 36 mM H 2 O 2 -reduced in leaves ( Figure 9B). PAPs, also termed as fibrillin/CDSP34 proteins, are involved in the structural stabilization of thylakoid membrane upon environmental constraints. The gene expression and protein accumulation of PAPs were induced in tomato (Lycopersicon esculentum) and potato (Solanum tuberosum) under osmotic and oxidative stresses [36,37]. Besides, the expression of PAPs in Brassica napus and Arabidopsis were differentially regulated by various abiotic stresses, such as drought, ozone, cold, NaCl, light, and mechanical wounding [38,39]. The α/β-tubulin heterodimer is the building block of microtubules, which regulates cell division and expansion, as well as organelle movement. Microtubule organization and dynamics quickly responds to various external stress signals, such as low temperature [40], cold acclimation [41], as well as osmotic and salt stresses [42,43]. All these imply that cell structure modulation is critical for the tolerance of various exogenous stresses-induced intracellular oxidative stress in poplar.

H 2 O 2 -Induced Alteration of ROS Scavenging Pathways in Poplar
Generally, when leaves are exposed to H 2 O 2 treatment, exogenous H 2 O 2 can permeate through cell membrane into cells, leading to the increase of intracellular H 2 O 2 level [17]. In this study, the H 2 O 2 concentrations in leaves were increased under 12 and 24 mM H 2 O 2 ( Figure 4A). Interesting, the intracellular H 2 O 2 levels were increased in leaves of grass pea (Lathyrus sativu L.) under 5 and 10 mM H 2 O 2 for 24 h, but returned to the normal level under 20 mM H 2 O 2 , which were revealed from H 2 O 2 content detection and the histochemical localization using DAB staining [44]. These implied that special pathways were employed for intracellular H 2 O 2 scavenging when plants were exposed to relatively higher concentration of H 2 O 2 .
In poplar leaves, we found diverse antioxidative enzymes and antioxidants were involved in intracellular H 2 O 2 scavenging to cope with exogenous H 2 O 2 stress ( Figure 9C) [44]. Importantly, in this study, the induced GPX activity would especially facilitate to reduce relative higher intracellular H 2 O 2 levels ( Figure 4D) Figure 4B) [1]. The activities of several antioxidative enzymes (e.g., MDHAR, DHAR, and GR) were H 2 O 2 -modulated for regeneration of reduced AsA and GSH, such as reduced MDHAR activity, and increased activities of DHAR and GR in poplar ( Figure 4E,F). MDHAR and DHAR catalyze the regeneration of AsA, using nicotinamide adenine dinucleotide phosphate (NADPH) and GSH as electron source/donor, respectively, while GR maintains the cellular reduced GSH pool through converting GSSG to GSH with NADPH. The stress-induced activities of DHAR and GR also have been reported in several plants (e.g., maize and pepper) in response to abiotic stresses, such as salinity, drought, low temperature, and heavy metal [46]. This suggests that the H 2 O 2 -induced activities of DHAR and GR would maintain the reduced AsA and GSH pools for antioxidative processes in leaves. In addition, the altered H 2 O 2 -reponsive contents of antioxidants AsA/DHA and GSH/GSSG indicated that they contributed to H 2 O 2 scavenging, which would function as the substances of aforementioned antioxidative enzyme (i.e., APX, POD, GPX, and DHAR) systems, but also could directly reduce H 2 O 2 as reductants in poplar leaves. Additionally, the exogenous H 2 O 2 -induced osmolytes (i.e., proline, soluble sugar, and glycine betaine) were suggested to protect cellular components from degeneration by scavenging ROS in poplar leaves ( Figure 3C-E). Similarly, the accumulation of proline and soluble sugars was found in wheat leaves under H 2 O 2 treatment [20].

H 2 O 2 -Altered Redox Homeostasis in Poplar Leaves
In addition to the ROS scavenging pathways, glutathione-S-transferase (GST), glyoxylase (Gly) and aldo/keto reductase (AKR) were altered in regulating secondary release of metabolite signals in poplar leaves to cope with H 2 O 2 stress. GSTs are mostly known as detoxifiers of electrophilic compounds by covalently linking glutathione to hydrophobic substrates for sequestration or removal, which plays an important role in improving plant stress tolerance [47]. In this study, GST activity was increased significantly under 36 mM H 2 O 2 , but the abundances of GST U30 and GST F1 were decreased ( Figures 4F and 9C, and Table 1). It can be explained that the enzyme activity is determined by not only protein abundance, but also its changes in conformation and post-translational modification (PTM). Moreover, plant GST group is a large protein family containing at least eight classes, and each family member has different role in response to various stress conditions. For example, the abundances of B. distachyon GST1-like [19] and citrus GST [18] were H 2 O 2 -decreased in leaves, but the abundances of wheat GST 19E50 [20] and rice GST F11 [17] were H 2 O 2 -increased in leaves. Therefore, in this study, the reduced abundances of two GST members would not account for the overall GST activity in polar leaves to cope with H 2 O 2 . The PTM mechanisms of GSTs are valuable to be further investigated.
Gly system comprising of Gly I and Gly II is the primary route for detoxification of methylglyoxal that is a toxic byproduct inhibiting cell proliferation, protein degradation, and antioxidant defense system [48]. Gly I was accumulated in rice leaves under 0.6 and 3 mM H 2 O 2 stresses [17], but decreased in citrus leaves under 10 mM for 8 h [18]. In addition, AKRs are involved in detoxifying lipid peroxidation derived reactive aldehydes, leading to enhance the tolerance against abiotic stress-induced oxidative stress [49]. In this study, all the abundances of Gly I and AKR maintained at normal levels under 12 and 24 mM H 2 O 2 , but decreased significantly under 36 mM H 2 O 2 (Table 1 and Figure 9C). This implied that Gly and AKR systems were probably employed for detoxification under relative lower concentration of H 2 O 2 stress.

H 2 O 2 -Reduced Photosynthesis in Poplar
Photosynthesis is sensitive to ROS accumulation resulted from diverse stresses, because most photosynthetic enzymes are the preferential targets for the oxidation. In this study, the H 2 O 2 immersion resulted in the decreases of net photosynthesis and stomatal conductance of P. simonii × P. nigra seedlings (Figure 2A,B). Importantly, we found 37 H 2 O 2 -responsive proteins were involved in photosynthesis, which accounted for 46% of all the H 2 O 2 -responsive proteins in poplar leaves (Table 1, and Figures 6A and 9D,E). This is similar to what happened in leaves from rice under 0.6 and 15 mM H 2 O 2 for 6 h and citrus under 10 mM H 2 O 2 for 8 h, respectively [17,18]. In these two studies, 32% and 28% H 2 O 2 -responsive proteins in rice and citrus were identified using 2DE-based proteomics approaches, respectively [17,18]. These proteins are involved in light harvest, oxygen evolving, electron transfer, ATP synthesis, and Calvin cycle. Most of them were decreased under certain H 2 O 2 concentration, leading to photosynthesis decline in poplar (Figures 2A and 9D,E, and Table 1). Interestingly, most of these proteins, except for PnsL5, were changed in H 2 O 2 -treated leaves from other trees (e.g., citrus [18]) and gramineous plants (e.g., rice [17], wheat [20], and B. distachyon [19]).
In this study, our 2DE-based proteomics investigation revealed that ten of the H 2 O 2 -resposive photosynthetic proteins had multi-proteoforms in poplar leaves in response to H 2 O 2 (Table 1 and Figure 9D,E). These proteoforms were mainly resulted from various H 2 O 2 -induced PTMs, including oxidative modification. All these proteins have been found to be oxidized in Arabidopsis in response to H 2 O 2 treatments using redox proteomics approaches [50][51][52][53]. This is consistent with the findings from citrus and Arabidopsis that most photosynthesis-related proteins were easily carbonylated [18] and/or oxidized [51] under H 2 O 2 exposure. Therefore, it is necessary to investigate the protein redox PTMs using redox proteomics technologies.

H 2 O 2 -Responsive Carbohydrate and Other Metabolisms in Poplar
Carbohydrate and energy supply are essential for plants in response to oxidative stress [20]. Our proteomics results revealed that phosphoglucomutase (PGM), cytosolic GAPDH, and enolase were altered in poplar leaves under H 2 O 2 stress ( Figure 9F). The increase of cytosolic PGM would enhance reversible interconversion of glucose-1-phosphate and glucose-6-phosphate, providing substrates for glycolysis and synthesis of a variety of cellular constituents. In roots of two black poplar (P. nigra) clones, the PGM gene and soluble sugar level were all induced under drought stress [54]. This implied that the mobilization of stored starch would be triggered in poplar when carbon assimilation was inhibited due to oxidization stress-reduced photosynthesis [54]. Besides, the H 2 O 2 -decreased abundances of cytosolic GAPDH and enolase implied that glycolysis was inhibited in poplar under H 2 O 2 stress. In addition, the decreased abundances of alcohol dehydrogenase and cytosolic NAD-dependent malate dehydrogenase would reduce the regeneration of reducing power nicotinamide adenine dinucleotide (NADH) in H 2 O 2 -stressed poplar leaves [55,56], while the abundance-altered pyruvate dehydrogenase E1 and mitochondria ATP synthase would contribute for modulation of tricarboxylic acid cycle and energy supply in H 2 O 2 -stressed poplar leaves.
Glutamate and cysteine are central metabolites that serve as donors for the synthesis of other amino acids, vitamins, coenzymes, GSHs, and proteins, which play critical roles in plant stress responses. The syntheses of glutamine/glutamate and cysteine are depended on glutamine synthetase (GS)-involved pathway [57] and O-acetylserine (thiol) lyase (OAS-TL)-involved pathway [58], respectively. In this study, two proteoforms of cytosolic GS and OAS-TL were decreased in poplar leaves under H 2 O 2 stress (Table 1, Figure 9G). The oxidative-decreased GS was also found in wheat leaves under 15 mM H 2 O 2 for five days [20]. In addition, two other amino acid metabolism-related enzymes (i.e., 3-isopropyl malate dehydrogenase and alanine aminotransferase family protein) were decreased in poplar leaves under 36 mM H 2 O 2 stress (Table 1 and Figure 9G). These proteomics results implied that amino acid metabolism was reduced in poplar leaves under H 2 O 2 stress.
In addition, two enzymes in plant tetrapyrrole biosynthetic pathway, delta-aminolevulinic acid dehydratase 1 (ALAD1) and uroporphyrinogen decarboxylase (UROD), were all decreased under 36 mM H 2 O 2 stress (Table 1, Figure 9H). ALAD catalyzes the asymmetric condensation of two molecules of δ-aminolevulinic acid to porphobilinogen, and UROD catalyzes the formation of coproporphyrinogen from uroporphyrinogen. Altered ALAD activity concomitant with reduced chlorophyll content have been reported in many terrestrial plants exposed to various metal (e.g., aluminum, cadmium, and lead) stresses [59][60][61]. Interestingly, transgenic tobacco plants with reduced activity of UROD was characterized by the accumulation of photosensitizing tetrapyrrole intermediates, which would induce the enzymatic detoxifying defense system, and especially resemble the hypersensitive reaction observed after pathogen attack [62]. Thus, the reduction of ALAD1 and UROD might result in the accumulation of photosensitizing tetrapyrrole intermediates, which probably play roles in the response of H 2 O 2 .

H 2 O 2 -Responsive Proteins Conformation in Poplar
Maintaining proteins in their functional conformations and preventing the aggregation of non-native proteins are important for plant survival under oxidative conditions. HSPs and other chaperones are responsible for protein folding, assembly, translocation, and degradation in response to stress conditions [63]. In this study, a RuBisCO large subunit-binding protein α subunit (RBP-α) and two proteoforms of chaperone DnaK (DnaK) were increased remarkably in poplar under 36 mM H 2 O 2 stress (Table 1 and Figure 9I). RBP is considered as chloroplast chaperonin 60 (Cpn60), which is most likely involved in mature protein folding/assembly in plants, and facilitates the translocated protein to fold into native conformation. Previous studies showed that RBPs were induced in wheat leaves under drought [64] and salt stress [65], as well as in rice leaves by H 2 O 2 stress [17]. Moreover, Cpn60 subunit β can protect RuBisCO activase from thermal denaturation and function in acclimating photosynthesis to heat stress [66]. In this study, the two proteoforms of DnaK in poplar have relatively high sequence similarity to chloroplast HSPs [67]. The chloroplast HSPs carry out pivotal function in processes related to growth and development and in response to diverse environmental stresses, such as heat, light, and pathological stress [67]. For example, the expression of the chloroplast-localized Hsp70B is induced in Chlamydomonas under heat shock, high light and oxidative stresses [68]. A wheat chloroplast TaHsp70 plays a critical role in defense response elicited by infection of stripe rust fungus [69]. Therefore, the increases of RBP-α and DnaKs in poplar under 36 mM H 2 O 2 treatment suggest they play an important role in protection against the high dose oxidative stress.
In addition, heat shock cognate protein 80 (HSC80) and two proteoforms of heat shock protein 90 (HSP90) were increased in poplar under H 2 O 2 stress (Table 1, Figure 9I). HSC80 was found to be increased 10-fold in tomato cell culture upon heat shock [70]. HSP90 is distinct from many other molecular chaperones in that most of its known substrates to date are signal transduction-related proteins such as steroid hormone receptors and signaling kinases [63]. Recent studies revealed that plant HSP90s were important in plant development, environmental stress response, as well as disease and pest resistance [71]. Therefore, the induced HSC80 and HSP90s might prevent the aggregation of non-native proteins and reestablish normal protein conformation in H 2 O 2 -stressed poplar leaves.

Plant Cultivation and Treatment
The terminal buds or lateral buds excised from P. simonii × P. nigra plantlets were transferred to a culture flask containing 80 mL 1/2 MS solid medium, containing 2% (w/v) sugar and 0.54% (w/v) agar. The explants were cultured in a phytotron at 26 • C/22 • C (day/night), 16 h photoperiod and 200 µmol·m −2 ·s −1 light intensity for 50 days. The shoots of regenerated plantlets were immersed in 0, 12, 24 and 36 mM H 2 O 2 for 6 h, respectively. After the treatments, leaves were harvested and blotted dry on filter paper immediately. For each treatment, at least three biological replicates were performed. For each replicate, more than three whole leaves with similar size from at least three separate poplar seedlings were collected and pooled. The fresh weight was 0.2 g. The samples were either used fresh or stored at −80 • C for future analysis.

Photosynthesis Measurement
Pn, Gs, Ci, and Tr were measured in fully expanded leaves of each plant using a portable photosynthesis system LICOR 6400 XT (LI-COR Inc., Lincoln, NE, USA) [72]. The measurements were done at 10:00 a.m. At least nine leaves for each sample were measured.

Determination of MDA Content, REL, Total Soluble Sugar, Proline, and Glycine Betaine Contents
The MDA content and REL were determined using previous methods described by Wang et al. [73]. For the MDA content assay, 0.2 g fresh leaves were ground in 5 mL 10% trichloroacetic acid, and centrifuged at 10,000× g at 4 • C for 10 min, then the supernatant was collected. Two milliliters of 0.6% (w/v) thiobarbituric acid solution was added to 2 mL of the supernatant. The reaction solution was incubated for 15 min at 100 • C followed by cooling down to ambient temperature. The absorbance was detected under 450, 532 and 600 nm using an Ultrospec 2100 pro UV/Visible spectrophotometer (GE Healthcare Life Science, Uppsala, Sweden). The MDA content was calculated according to Li et al. [74].
For the REL determination, 0.2 g fresh leaves were cut and completely immersed in 20 mL deionized water, then degassed for 10 min. The electrical conductivity of the solution (E1) was measured using a conductivity instrument (DDS-11A) after 20 min. Subsequently, the solution was incubated at 100 • C for 15 min and cooled to room temperature, and then the electrical conductivity of the solution (E2) was determined. In addition, the electrical conductivity of deionized water (E0) was also detected. The REL was calculated according to the equation: REL (%) = (E1 − E0)/(E2 − E0) × 100%.
Contents of proline and total soluble sugar were determined using ninhydrin reaction and an anthrone reagent method developed by Li et al. [74]. For the proline determination, 0.2 g fresh leaves were ground in 3 mL 3% (w/v) sulfosalicylic acid. After incubation at 100 • C for an hour and followed by cooling down to ambient temperature, the homogenate was centrifuged at 15,000× g for 10 min at 25 • C. Then, 1 mL supernatant, 2 mL glacial acetic acid, and 2 mL ninhydrin were incubated at 100 • C for an hour. After cooled to room temperature, the incubated solution with addition of 8 mL methylbenzene was allowed to stand for an hour before the detection at 520 nm using a spectrophotometer.
For the total soluble sugar assay, 0.2 g fresh leaves were ground in 5 mL deionized water. The homogenate was incubated at 100 • C for 30 min followed by cooling down to ambient temperature. After centrifugation at 15,000× g for 10 min at 25 • C, the supernatant was collected and diluted with deionized water to 50 mL. Then, 1 mL the extracting solution, 0.5 mL 2% (w/v) ethyl acetate solution of anthrone, and 5 mL concentrated sulfuric acid were incubated at 100 • C for 2 min. The absorbance was detected under 630 nm using a spectrophotometer after cooled down to room temperature.
Glycine betaine content was measured using reinecke salt method as described by Zhao et al. [75]. Fresh leaves (0.2 g) were ground in liquid nitrogen, and then the powder was incubated for 24 h in 6 mL 0.375% (w/v) reinecke salt solution. The homogenate was centrifuged at 10,000× g for 15 min at 20 • C. The supernatant was collected and filtered through a 0.45-µm-pore-size cellulose acetate filter. The filtrate was dried at 70 • C and then resuspended in 5 mL deionized water. After 5 mL reinecke salt solution was added, the reaction solution was incubated at 4 • C for 2 h, and then centrifuged at 4000× g for 15 min at 4 • C. The precipitate was collected, and then redissolved in 15 mL aether. Then, the solution was centrifuged at 4000× g for 15 min at 4 • C. The supernatant was harvested, dried, and redissolved in 70% acetone. The absorbance was determined at 525 nm using a spectrophotometer. The contents of proline, total soluble sugar, and glycine betaine content were calculated from the standard curve.

Determination of ROS and Antioxidant Substances Contents, and Antioxidant Enzyme Activity Assay
The content of H 2 O 2 was determined using potassium iodide reaction as described in Suo et al. [76]. Generation rate of O 2 •− was obtained according to a method of Zhao et al. [75].
For antioxidant enzyme activity assay, 0.2 g leaves were homogenized on ice in 3 mL 50 mM phosphate buffer (pH 7.8, and containing 2% PVP-40 and 2 mM ascorbate). After centrifugation at 15,000× g for 20 min at 4 • C, the supernatant was collected for enzyme activity assays, including SOD, CAT, APX, GPX, POD, MDHAR, DHAR, GR, and GST. The activities of SOD, CAT, APX, POD, GR, and GST were assayed according to our previous methods [72]. The activities of GPX, MDHAR, and DHAR were measured according to our previous methods described by Suo et al. [76]. For all the enzyme activity assays, protein content was determined using the Bradford method [77]. In addition, contents of AsA, DHA, GSH, and GSSG were measured according to methods of Wei et al. [78].

Protein Sample Preparation, 2DE, and Protein Abundance Analysis
The proteins from leaves under different concentrations of H 2 O 2 treatments were extracted using a phenol extraction method according to Wang et al. [73]. About 1.6 mg protein was loaded on per gel, separated on linear gradient IPG strips (24 cm, pH 4-7) through isoelectric focusing (IEF) in the first dimension, followed by 12.5% SDS-PAGE gels in the second dimension, and stained by CBB staining. Gel image acquisition and analysis were conducted as described in detail in Wang et al. [73]. For quantitative analysis, the volume of each spot was normalized against the total valid spots. The protein spots displaying consistent abundance changes from three biological replicates with greater than 1.5-fold changes and a p value smaller than 0.05 were considered to be DAP [72].

Protein Identification by MALDI-TOF/TOF MS and Database Searching
The DAP spots were excised from the gels and digested with trypsin as previously described [73]. MS/MS spectra were obtained using an ABI 5800 MALDI TOF/TOF MS (AB Sciex, Foster City, CA, USA). The mass error was below 30 ppm at both MS and MS/MS mode, and the resolution was 10,000. The MS/MS spectra were subjected to the online Mascot program [79] to search against all green plants (Viridiplantae) in NCBInr protein databases [80]. The searching parameters were set according to Wang et al. [73], the mass accuracy was 0.3 Da, and the maximum number of missed cleavages was set to one. To obtain high confident identification, proteins had to meet the following criteria: (1) the top hits on the database searching report; (2) a probability-based MOWSE score greater than 52 (p < 0.05); and (3) more than two peptides matched with nearly complete y-ion series and complementary b-ion series.

Protein Classification, Subcellular Localization, Hierarchical Cluster Analysis, and Protein-Protein Interaction Prediction
The identified proteins were searched against the NCBI database [80] and UniProt database [81] to determine if their functions were known. Combined with knowledge from BLAST alignments and literature, proteins were classified into different categories.
The subcellular localization of the identified proteins were predicted using five Internet tools according to Suo et al. [76]: YLoc [82], LocTree3 [83], Plant-mPLoc [84], ngLOC [85], and TargetP [86]. Only the consistent predictions meeting the high criteria from at least two tools were accepted as a confident result.
Self-organizing tree algorithm hierarchical clustering of the protein abundance profiles was obtained from log (base 2) transformed fold change values of protein spots using Cluster software (version 3.0).
The protein-protein interactions were predicted using the web tool STRING 10 [30]. The homologs of the DAPs in Arabidopsis were obtained by sequence BLAST in TAIR database [31], and then the homologs were subjected to the web tool of STRING 10 for creating functional protein association networks, based on published literature, genome analysis of domain fusion, gene neighborhood, phylogenetic profiling/homology, co-expression, co-occurrence, and other experimental evidence [76].

Statistical Analysis
All the physiological results were presented as means ± standard deviation (SD) of at least three biological replicates. Data were analyzed by one-way ANOVA followed by Duncan's test using the statistical software SPSS 17.0 (SPSS Inc., Chicago, IL, USA). A p value smaller than 0.05 was considered to be statistically significant.

Conclusion
In the course of poplar development and stress-response in poplars, the molecular regulation of H 2 O 2 homeostasis in the cells is complicated and fine-tuned. In the present study, we present a primary H 2 O 2 -responsive network in leaves of poplar (P. simonii × P. nigra) using integrative analysis of physiological and proteomic approaches. The molecular network includes 14-3-3 protein-/NDPK-mediated signaling pathway, dynamics of thylakoid membrane structure, enhancement of diverse antioxidative defense system, alteration of photosynthesis, adjustment of carbohydrate and other basic metabolisms, as well as modulation of protein synthesis and conformation ( Figure 9). This study provides new information and insights into underlying H 2 O 2 -responsive mechanisms in poplar plants. Although our proteomics results highlighted some critical candidate proteins/genes in H 2 O 2 -responsive signaling and metabolic pathways, their biological functions for H 2 O 2 response in poplar still need future characterization using molecular genetics and PTM analysis tools.