Evaluation of protein pattern changes in roots and leaves of Zea maysplants in response to nitrate availability by two-dimensional gel electrophoresis analysis

Experimental design and biochemical parameters

The aim of this work was to apply a proteomic approach to study the changes in protein patterns of root and leaf organs of maize plants in the first phase of exposure to high availability of nitrate, comparable to agricultural conditions, after a growth period under nitrogen starvation. This is a typical condition in which the addition of nitrate induces an increase in uptake and assimilation of this nutrient [5, 28].

The need for a simultaneous analysis of the root and the leaf organs of starved plants, with completely developed but not stressed leaf apparatus, led to the definition of the experimental design showed in Figure 1. Briefly, seedlings were transferred into a hydroponic system after 3 days of germination and grown for further 14 days in a solution deprived of nitrogen. After that, at the beginning of the light period (T₀), some plants were maintained in the same nutritional condition (control, C) whereas others were transferred in a nutrient solution containing 10 mM NO₃^- (N). In order to define the sampling time for proteomic analysis, the changes of biochemical parameters in response to NO₃^- were firstly evaluated. Roots and leaves were collected at T₀ time and after 6, 30 and 54 h of nitrate exposure.

At these sampling times, the plants achieved the developmental stage corresponding to the complete expansion of the third leaf (pictures of harvested plants are showed in Additional file 1). The qualitative comparison between the C and N plants revealed some morphological differences. In particular, while the plants appeared very similar at the T₀ sampling time, after 30 h the expansion of the fourth leaf was slightly more evident in N plants with respect to the C ones. This trend was more pronounced at 54 h and, only in C plants, was accompanied by the comparison of faint yellow areas in the leaf blades. In the tested conditions, no significant differences were observed in root system.

In order to characterize the physiological status of the plants, the changes in nitrate content and NR activity (Figure 2) as well as the levels of proteins, amino acids, reducing sugars, sucrose and chlorophyll were evaluated (Figure 3).

In roots and leaves of starved plants, both nitrate and NR activity were undetectable. After the addition of the nutrient to hydroponic solution the levels of nitrate progressively increased in plant tissues, reaching a level of 32.6 and 10.3 μmol of NO₃^- g^-1 FW after 54 h in roots and leaves respectively (Figure 2A). A parallel dramatic increase of NR activity was measured until the 30^th h of NO₃^- exposure, while at the longest time considered (54 h) a decreased activity was observed (Figure 2B). This trend was more evident in the roots in which a more rapid and large availability of nitrate took place. The total protein levels did not change significantly in all the conditions tested (Figure 3A and 3B), while a sharp increase in free amino acids was detected in both organs after nitrate addition (Figure 3C and 3D). Moreover, the levels of amino acids were higher in the leaves than in the roots. Although many factors are involved in the overall amino acid levels, these results may suggest a contribution of translocation of nitrogen compounds between the two organs. Nitrate exposure also induced a decrease in reducing sugars in both organs (Figure 3E and Figure 3F), while only in the roots of the plants exposed for 54 h to 10 mM NO₃^- a drop of sucrose took place (Figure 3G).

Taken together, these results well describe the induction trend of NO₃^- assimilation pathway, as suggested by the increase of NR activity and amino acids accompanied by the consequent decrease of reducing sugars, the main source of carbon skeletons [41]. In roots, where photosynthesis cannot satisfy this request and/or the demand of carbon skeleton is high, sucrose pool was also affected. The changes in carbohydrate availability and the increase of amino acid levels also explain the decrease in NR activity observed in roots at the 54^th h. In fact, these data are in agreement with the inhibitory effect on NR evocated by an increase of some amino acids, mainly asparagine and glutamine [5, 42]. Moreover, it is know that NR activity increases after sucrose addition whilst the low sugar content, condition that we observed in the roots of N plants, affects the nitrate reduction system [5, 42, 43]. The results suggested that this feedback mechanism was activated in roots of the plants exposed for 54 h to 10 mM NO₃^-. Finally, only at the 54^th h, a significant decrease in chlorophyll content (Figure 3I) was measured in the leaves of starved plants, thus suggesting that the first symptoms of stress were appearing.

2-DE analysis and protein identification

The biochemical and physiological data showed that the plants incubated for the last 30 h in the presence of 10 mM NO₃^- were in a condition in which nitrogen metabolism is completely activated in both root and leaf organs and that, at the same time, no stress symptoms were detectable in the control plants. Starting from these results, the proteomic study was conducted by analyzing the soluble protein fractions extracted from roots and leaves of plants incubated for the last 30 h in the absence or in the presence of 10 mM NO₃^-.

The 2-DE representative gels of the soluble fractions of root and leaf samples are shown in Figure 4. The electrophoretic analyses detected about 1100 and 1300 spots in roots and leaves gels, respectively. To ascertain the quantitative changes in the proteomic maps, the relative spot volumes (%Vol) were evaluated by software-assisted analysis. The Student's t-test (p < 0.05), coupled with a threshold of two-fold change in the amount, revealed that 20 spots in roots and 18 spots in leaves were affected by nitrogen availability.

Functional role and quantitative change of the proteins identified in roots

Many of the spots identified in roots were enzymes involved in nitrogen and carbon metabolisms (Table 2). According to the induction of the NO₃^- assimilation pathway, in the roots of the plants incubated for the last 30 h in the presence of the nutrient, we observed an increase in the accumulation of nitrite reductase (spot 268, NiR) and of glutamine synthetase plastidial isoform (spot 483, GS2).

Moreover, in response to the demand of carbon skeletons and NADPH, which is used in non-green tissues for ferredoxin reduction [44], an increase in the levels of phosphoglycerate mutase (spot 216, PGAM-1), glucose-6-phosphate dehydrogenase (spot 1162, G6PD) and 6-phospho-gluconate dehydrogenase (spot 392, 6PGD) took place. These results well agree with previous array data that describe the responses to nitrate exposure in Arabidopsis and tomato [7, 29, 45].

An increase in accumulation of the cytosolic isoform of glutamine synthetase (spot 538, GS1-1) was also detected in roots of N plants. On the basis of identified peptides by MS analysis it was possible to discriminate among the 5 GS1 isoforms known in Zea mays (SwissProt reviewed database) and to restrict the possible identification to 2 of them (GS1-1 [Swiss-Prot:P38559] and GS1-5 [Swiss-Prot:P38563] [46]). The fact that Li and co-workers [46], through a Northern blot hybridization analysis, found that the transcript of GS1-1 gene was the only one expressed in roots, conducted to the specific identification of GS1-1 protein. Moreover, Sakakibara and co-workers [47] showed that GS1-1 transcript was the only induced by NO₃^-. The proteomic approach used in the present work allows to confirm these results at the translational level, demonstrating that in maize roots a cytosolic ammonia assimilation pathway can be activated also in response to nitrate.

Other spots that were found to increase their relative volumes in response to nitrate were a non-symbiotic hemoglobin and a monodehydroascorbate reductase (spot 960, Hb2 and spot 390, MDHAR). In a previous work on Arabidopsis, it was found that NO₃^- induced AtHB1 and AtHB2, two genes that encode for non-symbiotic hemoglobins [7, 29]. Scheible and co-workers [7] suggested that these proteins could change their abundance in relation to the redox status, whereas Wang and co-workers [29] speculated on the possibility that the induction of hemoglobin could aim at reducing oxygen concentration during NR synthesis, since molybdenium can be sensitive to oxygen. Besides, hemoglobin and MDHAR are known to be involved in the scavenging of NO that can be produced by cytosolic and/or plasmamembrane nitrate reductase when nitrite is used as substrate [48, 49]. NO is a signaling molecule which is involved in many biochemical and physiological processes [50]. It has been reported that in plant roots, NO plays a role in growth, development and in some responses to environmental conditions, such as hypoxia [51]. Recently, a possible involvement of NO in the mediation of nitrate-dependent root growth in maize has been suggested [52]. According to this work, that describes a reduction of endogenous NO at high external NO₃^- concentration, the observed concomitant up-accumulation of Hb2 and MDHAR in our experimental condition supports the hypothesis that they might contribute in controlling NO levels in root tissues after exposure to NO₃^- [48, 49, 52].

The last protein found to be present in higher amount in N plants was a pyruvate decarboxylase (spot 231, PDC). This enzyme catalyzes the decarboxylation of pyruvic acid into acetaldehyde, the first step of the alcoholic fermentation. In particular, we identified the PDC isoenzyme 3 that has been previously found to be induced in hypoxia condition [53]. Although further studies are required to understand why PDC is induced by NO₃^-, we can observe that fermentation pathways are induced in response to redox status changes and that this condition could be also linked to the activation of the Hb/NO cycle (see above) [49, 54].

Among the spots identified in roots, six showed a down-accumulation in N plants (Table 2). Three of them were identified as phenylalanine ammonia-lyase (spots 171, 172 and 1160, PAL-a, PAL-b and PAL-c). The MS analysis indicated for all three spots the same protein [GenBank:AAL40137] while the electrophoretic data showed some differences in M_r and pI, suggesting that post-translational modification events may have occurred. It has been shown as low nitrogen availability induces transcripts encoding enzymes of phenylpropanoid and flavonoid metabolism, such as PAL, chalcone synthase and 4-coumarate:coenzyme A ligase, whilst after nitrogen repletion these activities are down-regulated [7, 55]. Our proteomic data appear to be in agreement with these studies.

Previously, it was found that under low nitrogen availability four proteases (e.g. serine, aspartate/metalloproteases and two cysteine proteases) increased their activity to degrade non-essential proteins in order to remobilize this nutrient [56]. In this work, we found an aspartic protease belonging to the A1 family (spot 707, AP) that was down-regulated after NO₃^- exposure. Moreover, the experimental M_r appeared lower with respect to that expected for this protein, thus suggesting that this spot is referable to the active form of the enzyme [57]. These data support a new possible role for A1 protease family [57, 58].

Phosphoenolpyruvate carboxylase activity is known to increase during nitrate assimilation, having a role in cell pH homeostasis and an anaplerotic function [14, 19, 59–61]. In addition, the monoubiquitination of this enzyme was recently well described in germinating castor oil seeds by Uhrig and co-workers [62]. It was found that this event is non-destructive and that this reversible post-translational modification of the enzyme reduces its affinity for PEP and its sensitivity to allosteric activators and inhibitors. The MS analysis of spot 53 (for sequence details see Additional file 4) identified 8 peptides, 7 of which matched with a PEPCase [DDBJ:BAA28170] (theoretical M_r/pI equal to 109.4/5.7), while the last peptide belonged to an ubiquitin (UB) [Swiss-Prot:P69319] (theoretical M_r/pI equal to 8.5/6.6). The experimental M_r and pI of spot 53, that were 115.4 and 5.7 respectively, were in agreement with the monoubiquitination of the PEPCase (PEPCase-UB, theoretical M_r/pI equal to 117.9/5.8 respectively). Moreover, the domain responsible to bind ubiquitin previously identified in PEPCase of other vascular plants is present in this maize PEPCase [62]. These results suggest that in maize roots the modulation of PEPCase activity in response to nitrogen availability could occur also through reversible monoubiquitination.

The last spot identified in roots that was down-regulated by NO₃^- was the 14-3-3-like protein GF14-12 (spot 774, GF14-12). Previously, it was found that this protein is localized in the nucleus where it binds the DNA at the G-box regions in association with transcription factors and that it is involved in the regulation of gene expression [63, 64]. More recently, it was described an interaction of 14-3-3 proteins with some transcription factors such as VP1, EmBP1, TBP and TFIIB [65]. Further studies are required to clarify the effective role of GF14-12, for which the functional information are still lacking.

Functional role and quantitative change of the proteins identified in leaves

Many of the spots identified in leaves by LC-ESI-MS/MS analysis were proteins linked to the NO₃^- assimilation as well as to the photosynthetic activity (Table 3).

The activity of NR can be modulated also at post-translational level through a phosphorylation event followed by binding of inhibitory 14-3-3 protein [66, 67]. One of the spots analyzed in the leaves was identified as TaWIN2 (Table 3, spot 1094, TaWIN2), that was previously described to be involved in the NR inactivation [67]. We found that the level of this protein decreased in leaves of N plants, where NR activity was induced (Table 3).

According to the well known relationships existing between nitrogen and carbon metabolism, the changes in accumulation of some spots after NO₃^- addition are consistent with an increase of photosynthesis rate. Two spots that raise after NO₃^- addition were identified as CPN-60α and CPN-60β (spot 467 and 462, CPN-60α and CPN-60β, respectively), that are chaperonin proteins involved in folding of ribulose-1,5-bisphosphate carboxylase [68]. Moreover, a chloroplastic triosephosphate isomerase was up-regulated by NO₃^- (spot 1065, TIM), while a cytosolic 6-phosphogluconate dehydrogenase (spot 650, 6PGD) was down-regulated, as expected when the request of reducing power could be satisfied by the increase in photosynthetic activity [69].

Spot 500 was identified as the α subunit of the chloroplastic ATP synthase (ATPsyn α), but unexpectedly it was more abundant in leaves of C plants. Although only a speculative interpretation of this result can be made, we could hypothesize that in leaves of the N plants ATP synthase should be activated and this process requires the reconstitution of the enzymatic complex in the thylakoid membranes [70]. Hence, to clarify this point, it should be necessary to investigate if the decrease of ATPsyn α observed in the soluble fraction of N plants is effectively accompanied by an increase of this protein in the membrane fraction.

Thiamine (i.e. vitamin B₁) is required in many pathways, such as the Calvin cycle, the branched-chain amino acid pathway and pigment biosynthesis [71]. Along with higher request of this vitamin in leaves of N plants, where the activation of these pathways could take place, we identified, among the spots up-regulated by N, the thiazole biosynthetic enzyme (spot 999, TH1-1) that is known to be involved in thiamine biosynthesis [71].

Two spots were identified as PEPCase (spot 134 and 138, PEPCase-a and PEPCase-b respectively). In C₄ plants such as maize, this enzyme plays a central role in photosynthesis, because it catalyses the primary fixation of atmospheric CO₂ [72]. The catalytic activity and sensitivity of this enzyme are mediated by a reversible phosphorylation [73]. The experimental pIs of the spots 134 and 138 were 5.8 and 5.7, respectively. Moreover, these two PEPCase forms showed opposite changes in abundance in the leaves of plants grown in the last 30 h in the presence of NO₃^- with respect to the controls. The results obtained in our work suggest that the two spots of PEPCase are referable to the phosphorylated (spot 138) and to the unphosphorylated (spot 134) form with a predicted pI of 5.7 and 5.8, respectively, that are known to correspond to the more and less active states of this enzyme [74]. Interestingly, despite the fact that data suggest an increase in the photosynthetic activity, the phosphorylated form was more abundant in the proteomic map of C plants. These results support the immunological observation by Ueno and co-workers [73] that the diurnal regulation of phosphorylation state of PEPCase appears delayed in nitrogen-limited conditions, suggesting that the circadian control of PEPcase is affected by nitrogen starvation.

Two of the spots down-regulated in leaves of N plants were a phenylalanine ammonia-lyase (spot 313, PAL) and a methionine synthase (spot 254, MetS). The decrease of PAL, observed also in root tissue (see above), is a further evidence that phenylpropanoid and flavonoid metabolisms are affected by nitrogen availability [7, 55]. On the other hand, the change in accumulation of MetS is contrasting with a recent proteomic study performed on wheat by Bahrman and co-workers [39]. These authors found that the induction of this enzyme was positively related to nitrogen availability. This discrepancy could be associated to different genetic traits of the two species, as well as it could be linked to different experimental approaches adopted in the two studies. Nevertheless, it should be observed that in both these works a single spot referable to MetS was detected, while further information on total level and/or on activity of this enzyme is necessary to clarify this point.

The spot 219, which considerably increased in N plants, was a lipoxygenase (LOX). In particular, the analysis of the MS spectra identified the LOX codified by ZmLOX10 gene, which was found to be a plastidic type 2 linoleate 13-LOX [75]. The expression analysis of this gene revealed that its transcript was abundant in leaves and was regulated by a circadian rhythm with a trend strictly linked to the photosynthetic activity. Moreover, it has been proposed that ZmLOX10 is involved in the hydroperoxide lyase-mediated production of C₆-aldehydes and alcohols and not in the biosynthesis of JA [75]. Although some evidences suggest a role of ZmLOX10 in the responses to (a)biotic stresses, its involvement in the diurnal lipid metabolism was also proposed [75, 76].

At the same time, we identified two proteins as an oxygen-evolving enhancer protein 2 (spot 1244, OEE2) and a 23 kDa polypeptide of photosystem II (spot 1612, 23pPSII), which were down-accumulated in leaves of N plants (Table 3). Both have been classified as members of PsbP family that is one of the three extrinsic protein families composing the oxygen-evolving complex (OEC) of photosystem II in higher plants [77–79]. In addition, it was recently demonstrated that PsbP proteins are essential for the normal function of PSII and play a crucial role in stabilizing the Mn cluster in vivo [80]. Moreover, the stability of this class of protein seems related to the lipid composition of chloroplastic membranes that is also affected by nitrogen availability [81, 82].

In order to elucidate the physiological meaning of these variations and to verify if they could be related to a stress status or to an alteration in photosynthetic performance, changes of both maximum quantum yield of photosystem II (F_V/F_M; dark adapted plants) and effective quantum yield of photosystem II (Φ_II; light adapted plants), dry weight and MDA levels of shoot were measured (Figure 5). Although the F_V/F_M parameter, measured on over-night dark adapted plants at time points 0, 24 and 48 hours, resulted in very similar values between C and N plants (about 0.80; see also Figure 5A), the Φ_II values showed a very slight decrease in C plants during the second period of illumination (C plants Φ_II, 0.71 versus N plants, 0.73) and the difference became more marked between 48 and 54 hours of nitrogen starvation. Similar data could be obtained by monitoring biomass production at the different time points (Figure 5B), indicating that photosynthetic performances are highly impaired in C plants after 48–54 hours of treatment. Nevertheless, no changes in MDA were detected in all the conditions tested (Figure 5C).

Taken together these results indicate that at the 30^th h, the time point chosen for proteomic analysis, plants start feeling the different nitrogen content in the growth media without developing major stress symptoms and the associated pleiotropic effects.

These data sustain the hypothesis that ZmLOX10 could be involved in lipid metabolism of the chloroplast that is strictly depending on photosynthetic activity [75, 76]. Further analyses are needed to unravel this possible intriguing role of ZmLOX10.

Considering the PsbP proteins, the change in accumulation of OEE2 and 23pPSII could indicate that OEC stability is affected by the N availability. Through time-course experiments, it will be possible to better correlate the relationship among N nutritional status, lipid metabolism, PsbP protein levels and PSII functionality.

Plant material and growth conditions

Maize (Zea mays L.) seeds of T250 inbred line, kindly provided by Prof. Zeno Varanini of Udine University – Italy, were germinated in the dark at 26°C on blotting paper saturated with deionized water. After 72 h, seedlings were transferred to a hydroponic system placed in a growth chamber with a day/night regime of 16/8 h and a PPFD of 200 μmol m^-2 s^-1 at plant level, with a temperature of 22°C in the dark and 26°C in the light and with a relative humidity of 70%. Seedlings were grown using of the following solutions: (i) 4 mM CaSO₄ for the first 48 h; (ii) 0.4 mM CaSO₄, 0.2 mM K₂SO₄, 0.175 mM KH₂PO₄, 0.1 mM MgSO₄, 5 μM KCl, 20 μM Fe-EDTA, 2.5 μM H₃BO₃, 0.2 μM MnSO₄, 0.05 μM CuSO₄, 0.2 μM ZnSO₄, 0.05 μM Na₂MoO₄ (growing solution) for the following 12 days. After this 17 days-long period of N starvation (T₀), plants were transferred in a fresh growing solution added (N) or not (C) with 10 mM KNO₃. The pH of all the growth solutions was adjusted to 6.1 and the solutions were changed every three days. All hydroponic solutions were continuously aerated by an electric pump.

At T₀ stage and after a period of 6, 30 and 54 h plants were harvested, washed with distilled water and then blotted with paper towels. Finally, roots and leaves were separated and the samples were frozen in liquid N₂ and stored at -80°C. The roots used for determining nitrate content were rinsed twice in ice-cold 0.4 mM CaSO₄ solution for 15 min for removing the anion present in the apoplast before sampling.

Levels of nitrate

Nitrate was extracted from the tissues by homogenizing the samples previously boiled in 4 volumes of distilled water for 15 min. The homogenate was centrifuged at 12,000 g for 20 min to obtain a clarified supernatant. Nitrate content was measured by adding 0.8 ml of 5% (w/v) salicylic acid in concentrated sulfuric acid solution to 0.2 ml of the supernatant. The mixture was stirred vigorously and allowed to react over 20 min, afterwards 19 ml of 2 N NaOH were slowly added and the resulting colour was read at 410 nm [83].

Nitrate reductase activity

NR was extracted by using 4 volumes of ice-cold 50 mM pH 7.8 MOPS-KOH buffer containing 5 mM EDTA, 5 mM NaF, 2 mM MSH, 1 mM PMSF, 10 μM FAD, 1 μM leupeptin and 10 μM chymostatin. The homogenates were centrifuged at 13,000 g for 15 min at 4°C. NR activity was measured as described by Ferrario-Méry et al. [84] using a reaction mixture consisting of 50 mM pH 7.5 MOPS-KOH buffer, 1 mM NaF, 10 mM KNO₃, 0.17 mM NADH, 10 mM MgCl₂ and 5 mM EDTA. The reaction was blocked after 10 or 20 min by adding an equal volume of sulphanilamide (1%, w/v in 3 M HCl) followed by n-naphtylethylethylenediamine dihydrochloride (0.02%, w/v). 30 min later, the concentration of NO₂^- was determined spectrophotometrically at 540 nm. The protein concentration was determined by 2-D Quant Kit (GE Healthcare).

Determination of reducing sugars, sucrose, amino acids, total proteins and chlorophyll

Reducing sugars, sucrose and amino acids were extracted by homogenizing frozen tissues in 4 volumes of ice-cold 0.5 M perchloric acid (PCA). The homogenate was centrifuged for 10 min at 13,000 g at 4°C and the resulting pellet was washed with the same volume of PCA and then centrifuged again in the same conditions. KOH was added to the collected supernatant (to pH 7.6) to remove excess PCA. Reducing sugars were measured according to the colorimetric method by Nelson [85]. Total soluble sugars were determined by the same method boiling an aliquot of PCA extract for 1 h before neutralization. Sucrose was estimated from the difference between total soluble and reducing sugars. Total amino acids were measured by the ninhydrin method [86].

Total proteins were extracted as previously described by Martínez and co-workers [87] by homogenizing the samples, previously powdered in liquid nitrogen, in 4 volumes of a 125 mM pH 8.8 Tris-HCl buffer containing 1% (w/v) SDS, 10% (w/v) glycerol, 50 mM Na₂S₂O₅. The homogenate was centrifuged at 13,000 g for 20 min to obtain a clarified supernatant. The protein content was measured by using 2-D Quant Kit (GE Healthcare).

Chlorophyll was extracted by homogenizing the leaves, previously powdered in liquid nitrogen, in 4 volumes of 80% pre-cooled acetone (v/v). The homogenate was centrifuged at 13,000 g for 20 min at 4°C to obtain a clarified supernatant. Chlorophyll concentration was measured according to Lichtenthaler [88].

Determination of malondialdehyde and chlorophyll fluorescence of the leaves

Malondialdehyde (MDA) was assayed by the method of Heath & Packer [89]. Frozen samples were homogenized with 4 volumes of ice-cold 0.1% (w/v) trichloroacetic acid (TCA) and centrifuged at 13,000 g for 20 min at 4°C. An equal volume of 20% (w/v) TCA plus 0.5% (w/v) thiobarbituric acid was added to the supernatants, which were subsequently heated at 95°C for 30 min. The extracts were then clarified by centrifugation at 13,000 g for 10 min, and the difference between the absorbance at 532 and 600 nm was measured. The MDA equivalent was calculated from the resulting difference using the extinction coefficient of 155 mM^-1 cm^-1.

In order to determine the photosynthetic performance, the chlorophyll fluorescence was measured by using a portable continuous-excitation type fluorometer (Handy-PEA, Hansatech Instrument). The maximum quantum efficiency of photosystem II (F_V/F_M) was calculated on over-night dark adapted plants, according to the equation (F_M-F₀)/F_M, where F₀ and F_M are the fluorescence levels when plastoquinone electron acceptor pool (Qa) is fully oxidized and transiently fully reduced, respectively [90]. The photosynthetic performance of light adapted plants was evaluated by monitoring the effective quantum yield of photosystem II (Φ_II) defined as (F_M'-F₀')/F_M' [91], where F_M'and F₀' represent the maximal and minimal fluorescence emission of photosystem II under light conditions.

Statistical analyses of biochemical and physiological measurements

For all the biochemical and physiological measurements, the experimental design consisted in three independent biological samples each analyzed in triplicate (n = 9).

One-way analysis of variance (ANOVA) followed by the post hoc Tukey test (p < 0.01) was used to verify the significance of the variations measured among all the tested parameters. This statistical analysis was performed using the software STATISTICA 7.

Extraction of protein samples for 2-DE analysis

Three independent biological replicates were extracted for each condition. Frozen samples, each composed by leaves or roots of 6 plants, were finely powdered in liquid nitrogen using a pestle and mortar, added with PVPP [0.5% and 1% (w/w) for roots and leaves samples, respectively], homogenized in 4 volumes of extraction buffer [0.5 M Tris-HCl pH 8, 0.7 M sucrose, 10 mM EDTA, 1 mM PMSF, 1 μM leupeptin, 0.1 mg mL^-1 Pefabloc (Fluka), 0.2% (v/v) MSH] and centrifuged at 13,000 g at 4°C for 20 min. The resultant supernatant was centrifuged at 100,000 g at 4°C for 38 min to obtain the soluble fraction. Proteins were then purified using the method previously described by Hurkman and Tanaka [92] by adding an equal volume of ice-cold Tris buffered phenol (pH 8) to the supernatant. Samples were shaken for 30 min at 4°C, incubated for 2 h at 4°C and finally centrifuged at 5,000 g for 20 min at 4°C to separate the phases. Proteins, grouped in the upper phenol phase, were precipitated by the addition of five volumes of -20°C pre-cooled 0.1 M ammonium acetate in methanol and the incubation at -20°C overnight. Precipitated proteins were recovered by centrifuging at 13,000 g at 4°C for 30 min and then washed again with cold methanolic ammonium acetate and three times with cold 80% (v/v) acetone. The final pellet was dried under vacuum and dissolved in IEF buffer [7 M urea, 2 M thiourea, 3% (w/v) CHAPS, 1% (v/v) octylphenoxy polyethoxy ethanol (NP-40), 50 mg mL^-1 DTT and 2% (v/v) IPG Buffer pH 3–10 (GE Healthcare)] by vortexing and incubating for 1 h at room temperature. Samples were centrifuged at 10,000 g for 10 min and the supernatants stored at -80°C until further use. Protein concentration was determined by 2-D Quant Kit (GE Healthcare).

2-DE analysis

Protein samples (400 μg) were loaded on pH 3–10, 24 cm IPG strips passively rehydrated overnight in 7 M urea, 2 M thiourea, 3% (w/v) CHAPS, 1% (v/v) NP-40, 10 mg mL^-1 DTT and 0.5% (v/v) IPG Buffer pH 3–10. IEF was performed at 20°C with current limit of 50 μA/strip for about 90 kVh in an Ettan IPGphor (GE Healthcare). After IEF, strips were equilibrated by gentle stirring for 15 min in an equilibration buffer [100 mM Tris-HCl pH 6.8, 7 M urea, 2 M thiourea, 30% (w/v) glycerol, 2% (w/v) SDS] added with 0.5% (w/v) DTT for disulfide bridges reduction and for an additional 15 min in the same equilibration buffer to which 0.002% (w/v) bromophenol blue and 4.5% w/v iodoacetamide for cysteine alkylation were added. Second-dimensional SDS-PAGE [93] was run in 12.5% acrylamide gels using the ETTAN DALT six apparatus (GE Healthcare). Running was first conducted at 5 W/gel for 30 min followed by 15 W/gel until the bromophenol blue line ran off. Two replicates were produced for each biological replicate, thus obtaining six gels per condition (n = 6).

Proteins were stained using the colloidal Coomassie Brilliant Blue G-250 (cCBB) procedure, as previously described by Neuhoff and co-workers [94]. The gels were scanned in an Epson Expression 1680 Pro Scanner and analyzed with ImageMaster 2-D Platinum Software (GE Healthcare). Automatic matching was complemented by manual matching. The molecular weights of the spots were deduced on the basis of the migration of SigmaMarkers™ wide range (MW 6.500 – 205.000), while pIs were determined according to the strip manufacturer's instructions (GE Healthcare) reporting on the reference gel of the software-assisted analysis the values of pI predicted for any given length of the strip. Both M_r and pI of the spots of interest were then determined by using software-automated algorithm.

Relative spot volumes (%Vol) of the six replicate gels per condition were compared and were analyzed according to the Student's t-test to verify whether the changes were statistically significant (p < 0.05). This analysis was performed by using SigmaStat software. Only spots showing at least a two-fold change in their relative volumes were considered for successive analyses.

Protein in-gel digestion and LC-ESI-MS/MS analysis

Spots excised from gels stained with cCBB were digested as described by Magni and co-workers [95] with some refinements. In detail, after the destaining procedure, spots were dried under vacuum on a centrifugal evaporator and incubated in 10 mM DTT, 100 mM NH₄HCO₃ for 45 min at 56°C. The solution was replaced with 55 mM iodoacetamide, 100 mM NH₄HCO₃ and the spots were incubated for 30 min in the dark at room temperature. After that, spots were briefly washed with 100 mM NH₄HCO₃ and again incubated for 15 min in 50% (v/v) acetonitrile (ACN), for 3 min in 100% ACN, for 3 min in 100 mM NH₄HCO₃, for 15 min in 50 mM NH₄HCO₃ in 50% (v/v) ACN and finally dried under vacuum. The following phases consisting in the protein digestion with trypsin [Sequencing grade modified Trypsin V5111, Promega, Madison] and in the recovery of peptides were carried out as described in the article above cited.

The LC-ESI-MS/MS experiments were conducted using a Surveyor (MS pump Plus) HPLC system directly connected to the ESI source of a Finnigan LCQ DECA XP MAX ion trap mass spectrometer (ThermoFisher Scientific Inc., Waltham, USA). Chromatography separations were obtained on a BioBasic C18 column (180 μm I.D × 150 mm length, 5 μm particle size), using a linear gradient from 5% to 80% solvent B [solvent A: 0.1% (v/v) formic acid; solvent B: ACN containing 0.1% (v/v) formic acid] with a flow of 2.5 μl/min. ESI was performed in positive ionization mode with spray voltage and capillary temperature set at 3 kV and at 220°C, respectively. Data were collected in the full-scan and data dependent MS/MS mode with collision energy of 35% and a dynamic exclusion window of 3 min.

Spectra were searched by TurboSEQUEST^® incorporated in BioworksBrowser 3.2 software (ThermoFisher Scientific Inc., Waltham, USA) against the Zea mays protein subset, Zea mays EST subset and against the protein NCBI-nr database, all downloaded from the National Center for Biotechnology Information [96]. The searches were carried out assuming parent ion and fragment ion mass tolerance of ± 2 Da and ± 1 Da, respectively, two possible missed cleavages per peptide, fixed carboxyamidomethylation of cysteine and variable methionine oxidation. Positive hits were filtered on the basis of peptides scores [Xcorr ≥ 1.5 (+1 charge state), ≥ 2.0 (+2 charge state), ≥ 2.5 (≥ 3 charge state), ΔCn ≥ 0.1, peptide probability < 1 × 10^-3 and Sf ≥ 0.70] [97]. If needed, identified peptides were used in protein similarity search performed by alignment analyses against the NCBI-nr database using the FASTS algorithm [98]. Physical properties of the characterized proteins were predicted by in silico tools at ExPASy [99].