Arbuscular mycorrhizal symbiosis affects the grain proteome of Zea mays: a field study

Maize is one of the most important crops worldwide and is strongly dependent on arbuscular mycorrhiza (AM) fungi, organisms that form a mutualistic association with land plants. In maize, AM symbiosis enhances spike dry weight, spike length, spike circumference, and the dry weight and dimensions of the grain. Notwithstanding its ubiquitous nature, the detailed relationship between AM fungal colonization and plant development is not completely understood. To facilitate a better understanding of the effects of AM fungi on plants, the work reported here assessed the effects of a consortium of AM fungi on the kernel proteome of maize, cultivated in open-field conditions. To our knowledge, this is the first report of the modulation of a plant seed proteome following AM fungal inoculation in the field. Here, it was found that AM fungi modify the maize seed proteome by up-regulating enzymes involved in energetic metabolism, embryo development, nucleotide metabolism, seed storage and stress responses.

. Relative amounts (%) of maize seed proteins. Seed proteins were classified in groups according to their solubility in a series of solvents including albumin (soluble in water), globulins (soluble in dilute aqueous salt solution), prolamins (soluble in alcohol solution) and glutelins (soluble in dilute alkali). Data are expressed as means ± standard error (Three biological samples were analysed twice in duplicate). ANOVA followed by Fisher's probable least-squares difference test used a cut-off significance at p = 0.05. Different letters indicate significantly different values based on one-way ANOVA (P < 0.05). Small letters indicate comparison between treatments (CTRL vs MIC) at the same time (20 or 60 DAF); capital letters indicate comparison between different times (20 DAF vs 60 DAF) in the same treatment (CTRL or MIC).
The 2D maps of seed proteins, stained with Colloidal Coomassie, showed a mean of 750 reproducible spots (Figs 1(a,b) and 2(a,b)). Significant variations were detected for 141 spots, of which 131 were MS/MS identified (93%). Table 2 lists the information regarding modulated proteins: spot number, number of identified peptides, sequence coverage, optical density variation using colour code, ANOVA P-value, protein name and Blast results when present, theorical molecular weight and pI, accession number and reference organism, and the biological process in which the identified protein is believed to be involved. Supplementary Tables S1, S2 and S3 list optical density raw data (as well as the statistical differences and P values), MS/MS results and BLAST results, respectively.

Table 2. Information regarding modulated proteins spot number, number of identified peptides, sequence coverage, optical density variation using colour code, ANOVA P-value, protein name and Blast results when present, theoretical molecular weight and pI (experimental pI data were not reported because the isoelectrofocusing was performed on non-linear IPG strips and image analysis software was not able to precisely calculate the pI), accession number and reference organism, biological process in which the identified protein was involved.
Despite the fact that the root is the organ that is colonised by AM fungi, the physiology of the entire plant is affected by the symbiosis, with interaction with the fungus having been reported to modulate photosynthesis, leaf hydration, reproduction and fruit quality in both maize and other plant species 6,26,36,37 . This modulation of the plant physiology changes according to the different stages of a plant's development and which plant organs are being analyzed. As shown in Table 2, the maize seed proteome of mycorrhizal plants differed either at the beginning of seed development (20 DAF) or at the end of maturation (60 DAF).
Effects during seed development. A detailed examination of the results revealed that at 20 DAF the AM symbiosis induced the up-regulation of enzymes involved in energetic metabolism, the latter stages of embryo development, nucleotide metabolism, seed storage and stress responses. AM fungi enhances primary metabolism by up-regulating ATP synthase (spot 219); this protein is a key enzyme whose expression is linked to respiratory and photosynthetic phosphorylation, both of which are major processes in the energetic metabolism of above-ground plant tissues. The up-regulation of the cytosolic ascorbate peroxidase (spot 387), a major enzyme involved in detoxification of hydrogen peroxide, was also induced by AM fungi; its expression may be linked with embryo development. Thus, Méchin et al. 24 reported that this enzyme is modulated in maize seeds 14 days after pollination 24 , whereas Finnie et al. 38 showed that a cytosolic form of this enzyme was only detectable in an early developmental stage of barley seeds.
The overexpression of nucleoside diphosphate kinase 1 (spot 111) may lead to reduced constitutive reactive oxygen species (ROS) levels and enhaced tolerance to multiple environmental stress 39 . The expression of nucleoside diphosphate kinase has been reported to increase in response to drought and salinity, thus it is expected to accumulate in the late phases of embryogenesis. This enzyme also plays significant roles in hormone responses, heat stress and, in general, growth and development 39 . An increase of ROS could induce the observed increase of heat shock 70 kDa protein (spot 55) expression.
The accumulation of Legumin 1 (spot 179), a storage protein found in maize seeds, is a confirmation of the data of globulin quantification and is linked with the seed storage process. The AM symbiosis induced down regulation of three starch granule-associated proteins, namely phosphoglucomutase 2 (spot 245), phosphoglycerate mutase (spot 136), and a pyruvate ortophosphate dikinase (spot 74) as well as seven proteins involved in cellular metabolic processes, an elongation factor 2 (spot 108), a translational initiation factor eIF-4A (spot 360), an ATP-dependent Clp protease ATP-binding subunit (spot 86), a ketol-acid reductoisomerase (spot 292), a stress responsive protein (spot 59) and succinate dehydrogenase flavoprotein subunit (spot 394). In order to use their Scientific RepoRts | 6:26439 | DOI: 10.1038/srep26439 stored carbon reserves, plants must be able to degrade their starch granules to oligosaccharides and monosaccharides. In particular, as previously reported, phosphoglucomutase 2 converts glucose 1-phosphate to glucose 6-phosphate facilitating the use of this compound in glycolysis 40 . The orthophosphate dikinase partly controls the composition of the storage protein fractions and the starch-protein balance 24 . The classical role of orthophosphate dikinase in both C3 and C4 plants involves catalyzing the reversible reaction of pyruvate, ATP and phosphate to phosphoenol-pyruvate, AMP and diphosphate. In rice, the expression of ortophosphate dikinase was found to be highest at 5-15 days after pollination; after that time this enzyme was likely rapidly degraded or inactivated through phosphorylation 41 . This pool of inactivated orthophosphate dikinase was also present in mature seeds, suggesting a role in developmental processes during seed germination 39,42 . Effects at seed maturation. At 60 DAF, the presence of AM fungi induced the modulation of 33 maize seed proteins, 4 up-regulated and 29 down-regulated. The degradation of the reserves (starch and storage proteins) and of some functional proteins could provide enough energy and amino acids for seed germination and for embryo development 43  Maize seeds acquire the ability to germinate during the stage of maturation drying 44 . The decreased water content plays an important role for the seeds to acquire the ability to germinate and for protection against fungal infection. Germination is a potentially stressful process and the reactivation of metabolism may provide an important source of ROS 44 . This can explain the increase in the abundance of proteins linked to the ROS response and AM symbiosis an overexpression of the same proteins such as salt tolerance protein (395) and down-regulation of APx-1 cytosolic ascorbate peroxidase (387), splicing factor (9), two spots belonging to nucleoside diphosphate kinase 1 (spots 111 and 319) that can lead to decreased constitutive reactive oxygen species (ROS) levels and enhanced tolerance to multiple environmental stress 39 , a peroxiredoxin (335), superoxide dis-mutase3 (273), a 22 kDa heat shock protein (4), the activator of a 90 kDa heat shock protein ATPase (49), and a stress responsive protein (59). The down regulation, in AM-treated seeds, of different isoforms of storage proteins such as Zein-alpha 19D1 precursor (21), vicilin-like embryo storage protein (spots 26, 51 and 242), z1B alpha zein protein (64) and lactoylglutathione lyase (213) could be linked with the seed protein turnover induced by embryo maturation.

Ripening effect on maize seed proteome.
Maize is an excellent model for research on cereal seed development because of the relatively large size of both its embryo and endosperm. Despite the importance of seed maturation information for agricultural purpose, there is scant data available in literature regarding the effects of root AM fungal inoculation on seed maturation. Kegg maps (Fig. 3 and Table S4) summarizes the main biochemical pathways involved in the maize seed proteome modification during ripening, i.e. carbon fixation; starch and sucrose metabolism; the pentose phosphate pathway; the citrate cycle; glycolysis/gluconeogenesis; valine, leucine and isoleucine biosynthesis; alanine, aspartate and glutamate metabolism; glyoxylate and dicarboxylate metabolism; pyruvate metabolism; purine metabolism; and cysteine and methionine metabolism. Both in control and in mycorrhizal plants, seed maturation induced the same proteome evolution with the exception of: malate dehydrogenase, succinate dehydrogenase, adenosine kinase, adenylate kinase, acetolactate synthase, ketol-acid reductoisomerase, homocysteine S-methyltransferase and methionine synthase (down-regulated in CTRL and not modified in MIC); alanine transaminase, alanine-glyoxylate transaminase, polygalacturonase, 6-phosphofructokinase, glycine transaminase, 1-aminocyclopropane-1-carboxylate synthase (up-regulated in CTRL and not modified in MIC); nucleoside diphosphate kinase (up-regulated in CTRL and down-regulated in MIC); and zein-protein precursor (up-regulated in MIC and not modified in CTRL). These changes in protein abundance could be linked with the higher content of starch in the seeds of plants treated with mycorrhizal fungi.
Our results are in agreement with those of Huang et al. 45 , who reported, on the basis of the metabolic and functional features of maize embryos, the identification of proteins classified into 7 major categories belonging to 3 functional groups: protein metabolism (26%), stress response (21%) and carbohydrate and energy metabolism (17%). At maturity, the maize seed accumulates large amounts of starch and storage proteins 45 . However, proteins involved in stress response (24%) were often up-regulated during seed maturation 45 .
A large body of literature describes the effects of AM fungi on the physiology of whole plants 6,7,16,17 , with a particular focus on fruit composition. The work presented here is a first step in filling the gap in the knowledge of the effect of AM fungi on seed composition. In the work described here, it was demonstrated that AM fungi strongly modify the seed proteome, particularly up-regulating enzymes involved in energy metabolism, embryo development, nucleotide metabolism, seed storage and stress responses.
Finally, this work underlines the importance of using soil microorganisms as inocula in field production to sustainably improve crop quality.

Materials and Methods
Experimental Field, Plant Growth And Seed Sampling. The experiment was conducted as described in Berta et al. 26 . In accordance with standard agricultural practices, field soil was fertilized with potassium sulfate (400 Kg/ha) and 18 Five, randomly selected, roots per treatment were used to evaluate frequency (F%), mycorrhizal degree (M%) and arbuscule abundance (A%) 46 . Selective extraction of different protein classes. Ten grams of seeds were ground in a mortar using liquid nitrogen and extracted twice with milliQ water containing a protease inhibitor cocktail (Sigma-Aldrich), in the ratio 1:10 (p/v), at 4 °C for 2 hours. The slurries were centrifuged at 10,000 × g for 15 min. The two supernatants (albumin fraction) were pooled and stored at − 20 °C, whereas the pellet was extracted twice with Tris-HCl 50 mM, pH 8.0, containing 0.3 M NaCl. The slurries were centrifuged at 10,000 × g for 15 min. The supernatants (globulin fraction) were pooled and stored at − 20 °C. The insoluble pellet was extracted twice with 70% ethanol containing 0.2% 2-mercaptoethanol. After stirring for 3 hours at 4 °C, the suspension was centrifuged at 10,000 × g for 30 min at 4 °C. The supernatants (prolamin fraction), were pooled and dried with a Rotavapor device. The insoluble pellet was then resuspended in 0.1 M NaOH to extract the glutelin fraction at 4 °C for 2 hours.
Three biological replicates were analysed in triplicate. Protein concentrations were determined according to Bradford 47 .
Seed water content was determined by placing one gram of ground seeds at 110 °C and then in a jar containing silica gel. Samples were analyzed twice in duplicate.
Gels were stained according to Candiano et al. 49   us to perform an analysis of variance to detect statistical differences between the quantitation of the same spot in all replicates.

Protein identification by nano-LC-Q-TOF MS/MS. For MS analysis, spots of interest were cut from
the gel and destained overnight with a solution of 25 mM ammonium bicarbonate and 50% acetonitrile. The proteins were digested with trypsin (Roche, Segrate, Milano, Italy) in-gel digested as described by Hellmann et al. 50 . All nano-HPLC-MS/MS experiments were performed on a Q-TOF mass spectrometer Q-Star XL (AB Sciex, Concord, Ontario, Canada) controlled by the Analyst QS 1.1 software (AB Sciex) connected to an Ultimate 3000 nano-HPLC system. The peptide pellets were resuspended in 10 μ l of solvent A (95% v/v water, 5% v/v acetonitrile, 0.1%v/v formic acid). Five microliters of each sample were loaded onto the precolumn, 300 μ m i.d. × 5 mm, C18 PepMap, 5 μ m beads, 100 Å, (LC-Packings) and washed for 5 min using a flow rate of 40 μ l min −1 solvent A. The peptides were subsequently eluted at 300 nl min −1 from the precolumn over an analytical column, 15 cm × 75 μ m, C18 PepMap100, 3 μ m beads, 100 Å (LCPackings) using a 35 min gradient from 5 to 60% solvent B (5% v/v water, 95% v/v acetonitrile, 0.1% v/v formic acid) delivered at 300 μ l min −1 . The analytical column was connected with a 15 μ m inner diameter Silica Tip (Pico Tip) nanospray emitter (New Objective, Woburn, MA). The spray voltage (set between 1800 and 2000 V) was applied to the emitter through a stainless steel union and tuned to get the best signal intensity using a standard BSA tryptic digest before every sample's batch submission. The QStar-XL was operated in information-dependent acquisition (IDA) mode. Mass spectra were acquired from 400 to 1800 m/z. The two most intense ions with charge states between 1 and 4 in each survey scan were selected for the MS/MS experiment. MS/MS data were acquired from 60 to 1800 m/z. Each acquisition cycle was comprised of a 1 s MS and a 3 s MS/MS. The MS to MS/MS switch threshold was set to 15 counts per second (c.p.s.). All precursor ions subjected to MS/MS in the previous cycle were automatically excluded for 60 s using a 3 amu.
Homology-driven proteomics. Mascot Distiller (Matrix Science, London, UK) was used to create peak lists from MS and MS/MS raw data. Mascot Server (Matrix Science) was used for database searching versus NCBInr. The last check for proteins homology assignments was made versus NCBInr 20151214 (78002046 sequences; 28422168805 residues). Carbamidomethylation of cysteine residues, oxidation of methionine, deamidation of asparagine and glutamine were set as possible variable modifications and trypsin was selected as the protease. One missed cleavage site was allowed, and the peptide MS and MS/MS tolerance was set respectively to 100 ppm and 0.2 Da. Positive identifications were assigned with a minimum of two unique peptides with at least one peptide having a significant ion score (underlined in red in Table S2 in the supporting information). Considering the scarce number of corn sequences in the databases, if we obtained an automatic hit without a significant score, sequence tags were manually interpreted from the ESI-MS/MS spectra to confirm the hypothetical assignment. We also accepted hits identified by at least one peptide with a significant ion score according to the MASCOT MS/MS ion search algorithm as being confident assignments. When a protein has only one spectrum with a significant Mascot score, but in the results there are more spectra with lower scores, they were manually inspected and if they had a pattern compatible with the theoretical peptide, they were considered for homology searching. The sequence obtained from the manually reconstructed peptide was submitted to MS homology and if the first positive hit was the same protein or a homologue sequence of the one automatically recognized, the peptide was inserted in the table as assigned to the protein. This approach allow the use of partial "de novo" sequences that can be more fitting to the sequences in the database 51 .
Data analysis was performed with Blast2GO standard parameters. The EC annotations, obtained by mapping from equivalent GO annotations, were visualized reconstructing the structure of the Gene Ontology relationships and ECs on KEGG maps (http://www.genome.jp/kegg). In KEGG maps were displayed the enzymatic functions of sequences in the context of the metabolic pathways in which they participate. Statistical analysis. Data were analyzed by a one-way ANOVA followed by Fisher's test with cut-off significance at p = 0.05 using Stat View 4.5 (Abacus Concepts) software.