Proteome alteration of soybean as a function of pod distortion syndrome

Soybean is one of the main crops throughout the world. It’s one of the richest and cheapest sources of protein as well as oil. The production of soybean corresponding to its area of cultivation has increased steadily over the past years. (Murithi et al., 2016). Different factors can influence soybean production and one of the main factors leading to loss of soybean production is pod distortion syndrome (Hill et al., 2006). Generally, PDS may be distinguishable at reproductive growth stages (R₁-R₆) marked by abnormal flowers (Fig. 1B), pods (Fig. 1 C, D, E) and seed (Fig. 1F) as well as ever green phenotype (Fig. 1A-right section).The PDS in soybean is influenced by several factors such as environmental (Shimada et al., 2005), agronomical (Ahmadi and Zelentsov, 2002), biological (Hill et al., 2006), physiological (Ahmadi and Zelentsov, 2002) and genetic factors (Hill et al., 2006). Despite massive damage caused by PDS, and it’s huge significance in soybean productivity, the molecular mechanisms underlying the syndrome remain undiscovered yet. Therefore, the main objective of the current study was to find out the causes that provide evidences of pod distortion in soybean at molecular scale. So, protein expression profiles of PDS and non-PDS (normal soybean plants) in two cultivars (Katul and Gorgan 3) were compared to unfurl their roles in causing PDS. To our knowledge, this is the first report on proteomics linked to PDS in soybean.
 

Plant Materials
 
Leaf samples of the Soybean cultivars (Katul and Gorgan 3) containing both non- PDS (Fig. 1G) and PDS (Fig. 1H) were collected from fifteen fields during two successive growing seasons (2013-2014) in Gorgan city, Golestan province, Iran. All the PDS and non-PDS plants (Fig. 1A) with three biological replications were gathered using random roving method at R₁ growth stage. The leaf samples were transferred to the laboratory (Tabriz University), and stored at -80ºC in a deep fridge for proteomic study. All management practices followed during the study period were recorded (Table 1).
 

 
Chlorophyll Extraction
 
In order to estimation of chlorophyll contents in the leaves of non-PDS (Fig. 1G) and PDS (Fig. 1H) plants of  both cultivars, one gram of fresh leaves were sampled and ground with 20 ml acetone 80%. Then, the solution centrifuged at 5000-10000 rpm for 5 minutes. The absorbance of the solution red at 645 nm and 663 nm againest the acetone bllank by spectrophotometer (Arnon, 1967). Finally, the concentration of chlorophyll a, chlorophyll b and totall chlorophyll were estimated using the following equation:  
Protein extraction
 
The TCA/acetone extraction method was used to extract leaf proteins (Damerval et al., 1986) and protein concentration was determined using Bradford method (Bradford, 1976).
 
EIF and SDS-PAGE
 
The total protein (400 µg/100 µL) was separated using 2-DE [isoelectric focusing (IEF) tube gel for the first dimension, and SDS-PAGE for the second dimension] (O’Farrell, 1975). The isoelectric focusing (IEF) tube gel was 11 cm in length and 3 mm in diameter, consisting of 0.1 M urea, 145 µL polyacrylamide 30%, 250 µL NP-40, 31.25 µL ampholites (pH 3.5–10.0 and 5.0–8.0), 1.875 µL ammonium persulfate 10%, 1.251 µL N, N, N0, N0-tetramethylethylenediamine and 355 µL ddH₂O. The SDS-PAGE gel consisted of 8.5 mL polyacrylamide, 6.3 mL gel spacer buffer (pH: 8.8), 120 µL ammonium persulfate 10%, 20 µL N, N, N0, N0-tetramethylethylenediamine and 2 mL ddH₂O. After electrophoresis, the gel was stained using Coomassie brilliant blue (CBB) (Pink et al., 2010). The isoelectric point (pI) and molecular weight (MW) of each protein was determined using 2-DE markers (Bio-Rad, Hercules, CA, USA).
 
Image acquisition
 
The 2-DE gels were scanned with GS-800 calibrated densitometer (Bio-Rad) and analyzed with PDQuest ver. 8.0 (Bio-Rad).
 
Cleavage and peptide mapping
 
After 2-DE separation, the gel pieces containing protein spots were excised, sealed and delivered to Alberta Proteomics and Mass Spectrometry Facility (Pharmacy and Health Research Building of Alberta University, Edmonton, Alberta, Canada). Protein samples were subjected to in-gel trypsin digestion. To be concrete, the excised gel bands were destained twice in 100 mM ammonium bicarbonate/acetonitrile (50:50), reduced by 10 mm BME in 100 mm bicarbonate, and then alkylated by 55 mM iodoacetamide in 100 mm bicarbonate. After dehydration, trypsin (6 ng/ul) was added to the gel pieces and incubated overnight (~16 hrs.) at room temperature to allow digestion. The trypsin-treated protein spots were then extracted from the gel using 97% water/2% acetonitrile/1% formic acid followed by a second extraction using 50% of the first extraction buffer and 50% acetonitrile (Cottrell, 2011).
 
Protein identification by Mass Spectrometry
 
Fractions containing tryptic peptides dissolved in aqueous 25% v/v ACN and 1% v/v formic acid were resolved and ionized using Nano flow HPLC (Easy-nLC II, Thermo Scientific) coupled to the LTQ XL-Orbitrap hybrid mass spectrometer (Thermo Scientific). Nanoflow chromatography and electrospray ionization were accomplished using a PicoFrit fused silica capillary column (ProteoPepII, C18) with 100 ìm inner diameter (300Å, 5 ìm, New Objective). Peptide mixtures were injected into the column at a flow rate of 3000 nL/min and resolved at 500 nL/min using 45 min linear gradients from 0 to 45% v/v aqueous ACN in 0.2% v/v formic acid. The mass spectrometer was operated in data-dependent acquisition mode, recording high-accuracy and high-resolution survey Orbitrap spectra using external mass calibration, with a resolution of 60000 and m/z range of 400–2000. The fourteen most intense multiply charged ions were sequentially fragmented using collision induced dissociation and spectra of the fragments recorded in the linear ion trap. After two fragmentations, all precursors selected for dissociation were dynamically excluded for 60 s. Data were processed using Proteome Discoverer 1.4 (Thermo Scientific), and the UniPort soybean database was searched using SEQUEST (Thermo Scientific) (McCormack et al., 1997). Search parameters included a precursor mass tolerance of 10 ppm and a fragment mass tolerance of 0.8 Da. Peptides were searched with carbamidomethyl cysteine as a static modification as well as oxidized methionine, deamidated glutamine and asparagine as dynamic modifications (Cottrell, 2011).
 
Identification of uncharacterized proteins
 
In order to characterize and identify probable unknown proteins a number of bioinformatics tools were exploited, including ClustalW (for multiple alignment and selection of high-score sequences), ProtParam (for determination of physico-chemical properties) and Inter ProScan (for identification of protein families and domains).
 
Statistical analysis
 
The 2-DE gels were analyzed with PDQuest ver. 8.0 (Bio-Rad) and the T-test (SAS Institute, 1997) through repeated measure analysis of variance (ANOVA) was carried out to examine whether there were significant differences between non-PDS and PDS in both the soybean cultivars. Only those proteins whose expression differed significantly (p < 0.01, at least a 2-fold difference in abundance) were considered to be increased or decreased in expression. Also, obtained chlorophylls data were analyzed using t-test to reveal whether there were a signicant differences between two types of growth in both cultivars.

Data analysis showed that the contents of chlorophyll a, b and a+b in PDS plants were higher than non-PDS plants (Pr>0.001) (Table 2). That is, at the same growth stages, the PDS plants retain their chlorophylls content; while, the non-PDS plant destructed their chlorophylls. Generally, senescence begins with proteolyzing of chloroplast polysomes, ribosomes and leaf proteins, ultimately leading in yellowing leaves. In the plant with stay green phenotype deconstruction of photosynthetic machinery during leaf aging process may partially or completely knocked out by different genetic, metabolic and physiologic pathways (Thomas, 2013).
 

       
Protein samples of non-PDS and PDS soybean plants were separated by 2-DE and stained by CBB (Fig. 2). PDQuest revealed 155 reproducible spots in Katul (Fig. 2A and Fig. 2B) and 143 reproducible spots in Gorgan 3 (Fig. 2C and Fig. 2D). Totally, the 5 and 11 proteins in PDS Katul and Gorgan 3 were found to be significantly different (p≥0.1) from those of their non-PDS counterparts, respectively (Fig. 2). The identified proteins were categorized based on their functions (Table 3) (Bevan et al., 1998). In Katul, 2 out of the 5 proteins (40%) were involved in energy generation. One was involved in protein synthesis (20%), and one in signal transduction (20%). The rest (1 protein, 20%) was an unknown protein. In Gorgan 3, proteins were involved in energy generation (37%), metabolism (27%), protein destination and storage (9%), and signal transduction (9%). The rest 18% (2 out of 11 proteins) were unknown proteins.
 

 

       
Ribulose 1, 5 bisphosphate carboxylase/oxygenase (Rubisco) large subunit in Katul (spot 6603-Fig. 2A and Fig. 2B) and Gorgan 3 (spot 5602- Fig. 2C and Fig. 2D) were the only protein whose expression increased, indicating that this enzyme may play a pivotal role in PDS state of growth. The large subunit is responsible for Rubisco catalytic activity (Ellis and Van Der Vies, 1988). Enhanced expression level of this enzyme in PDS plants may provide an evidence that energy consumption is increased in this syndrome. Since PDS soybean plants display a delayed senescence and a prolonged period of growth (even after pod maturation), the energy produced in PDS plants may contribute to more growth rather than plant maturation and grain-filling. Leaves may continue to grow and remain green instead of transferring their nutritional elements to the seeds. It is estimated that rubisco accounts for about half of the soluble proteins of chloroplast. Therefore, the big change in the Rubisco contents would lead to the apparent change of the other proteins. The observed fact that the number of decreased proteins is higher than that of increased proteins. It may be explained by the change in rubisco contents (Table 3).
       
The expression level of Rubisco large subunit binding protein subunit alpha, which is necessary for the assembly of Rubisco subunits, decreased in Gorgan 3 (spot 1804- Fig. 2C and Fig. 2D) (Tables 3 and 4). This subunit improves the folding and further the assembly of the other subunits, resulting  in enhanced solubility and activity of the enzyme. Therefore, decreased expression of this subunit may lead to decreased solubility and activity of the enzyme which may play as a regulatory system of biological activation of Rubisco (Miernyk, 1999; Ellis, 2006).
 

       
We also found that the expression level of Rubisco activase decreased in PDS Gorgan 3 (spot 1502- Fig. 2C and Fig. 2D) (Tables 3 and 4). The main role of Rubisco activase is activation and regulation of Rubisco activity. This enzyme removes sugar phosphates from Rubisco active site and leads to its  activation (Rokka et al., 2001). The results of our study are consistent with previous reports. In tobacco, anti-Rubisco activase transgenic plants have been found to have an increased content of Rubisco. Down-regulation of Rubisco activase gene in the transgenic tobacco decreased Rubisco carbamilation and proteolysis rates (He et al., 1997).
       
Moreover, in our study the expression level of carbonic anhydrase decreased in PDS Gorgan 3 (spot 7104- Fig. 2C and Fig. 2D) (Tables 3 and 4). By converting HCO₃ to CO₂ in an ATP-dependent manner, this enzyme increases the chloroplastic level of CO₂, ultimately resulting in increased Rubisco carbamilation. Thus, decreased expression of carbonic anhydrase may reduce Rubisco enzymatic activity (and therefore photosynthesis) in CO₂ limited conditions and may play an important regulatory mechanism for Rubisco active state (Merewitz et al., 2011).
       
One of proteins whose expression deceased in both PDS Katul (spot 1301- Fig. 2A and Fig. 2B) and Gorgan 3 (spot 0202-Fig. 2C and Fig. 2D) was 14-3-3 like protein (Tables 3 and 4), This protein regulates processes as diverse as guard cell signaling, cascade stress and mitogenic signal transduction, cell cycle and apoptosis, seed filling, ethylene biosynthesis, activity of phosphoprotein II, carbon and nitrogen metabolism, and nodule development in legumes (Fu et al., 2000; Cotelle and Leonhardt, 2016) (Fig. 3). It has already been proved that osmotic and salt stresses influenced 14-3-3 like protein expression in soybean and maize (Zörb et al., 2010; Nouri and Komatsu, 2010). Therefore, 14-3-3 like protein may have an important role in PDS incidence in soybean.
 

       
Oxygen evolving enhancer protein 2 (OEE2) is another energy related protein whose expression decreased in PDS plants of Katul (spot 3301- Fig. 2A and Fig. 2B) and Gorgan 3 (spot 2103- Fig. 2C and Fig. 2D). Reduced expression of OEE2 may lead to decrease of photosystem II activity and maintenance of thylakoid integrity (Bahrman et al., 2004). It can be seen clearly that about half of the chloroplast proteins are estimated as membrane (insoluble) proteins (Table 3), and there is a possibility of big changes in the amount of membranes proteins.  
       
Glutamate dehydrogenase (GDH) was another protein whose expression decreased in PDS soybean plants of Gorgan 3 (spot 7405- Fig. 2C and Fig. 2D). This enzyme exist in all organisms that reversibly incorporate ammonium into 2-oxoglutarate to form glutamate. Decrease of GDH expression is in parallel with decrease of glutamine deamination, resulting in dwindling the Krebs cycle (Fait et al., 2008) which may decrease Rubisco activation by metabolic inhibition. Decrease of GDH expression occurs in heat stress (Hossain et al., 2013) and flooding (Nanjo et al., 2011).
 
Nonetheless, malate dehydrogenase (MDH) was one of proteins whose expression decreased PDS in Gorgan 3 cultivar (spot 6401- Fig. 2C and Fig. 2D), but remained unchanged in PDS Katul plants (Tables 3 and 4). This enzyme is involved in both Krebs cycle and glycolysis pathway that converts oxaloacetate to malate (Wang et al., 2012). Therefore, decreased expression of this enzyme will result in decrease of malate to oxaloacetate conversion and vice versa. Decreased amount of oxaloacetate may decrease the Krebs cycle flow and metabolic inhibition of Rubisco activity. Decrease of MDH expression has been reported in heat shock (in Agrostis stolonifers) (Xu and Huang, 2010).
       
Stem 31 kDa glycoprotein is one of the proteins whose expression decreased in PDS Gorgan 3 (spot 6201- Fig. 2C and Fig. 2D) but not in PDS Katul (Tables 3 and 4). For the first time, this protein was discovered in depodded soybean plants (Wittenbach, 1983) as our samples have not any pods (Fig. 1B). This protein has been reported to play an important role in reserving of carbon and nitrogen, plant nutrition and adaptation under diverse growth conditions during development (Sözen, 2004). It has been reported that abiotic stresses such as salt (Yin et al., 2014), osmotic (Nouri and Komatsu, 2010) and UV-B (Lee et al., 2014) decreased the expression level of this protein. The decreased level of this protein may indicate that the sinks filled completely by source via Rubisco activation and this phenomenon may suppress whose genes which express Stem 31 kDa glycoprotein.
       
However, two spots [one in Katul (spot 1101- Fig. 2A and 2B) and one in Gorgan 3 (spot 2001- Fig. 2C and 2D)], were revealed by nESI-LC-MS/MS to be uncharacterized proteins. Using bioinformatics tools we found that these spots were NACA protein (Tables 3 and 4). NACA is a heterodimeric complex which can reversibly bind to eukaryotic ribosomes with chaperones activity (Breiman et al., 2016). NACA is located in direct proximity to newly synthesized polypeptide chains as they emerge from the ribosome and play an important role in embryonic lethality in metazoans. It also plays a fundamental role as a proteostasis sensor (Rospert et al., 2002). NACA delocalizes from ribosome to protein aggregates when proteostasis is imbalanced, in response to heat stress or aging (Rospert et al., 2002). Postponed senescence, flower and pod abnormalities which we observed in PDS soybean plants may be connected with decreased expression of NACA.
       
Moreover, spot 7301 in Gorgan 3 (Fig. 2C and Fig. 2D) and spot 7401 in Katul (Figs. 2A and Fig. 2B) which were recognized by nESI-LC-MS/MS as unknown proteins, were found by bioinformatics tools to be Methylecgonone reductase and 50S ribosomal protein, respectively (Fig. 2, Tables 3 and 4). Expression of these proteins decreased in PDS plants. Methylecgonone reductase is an NADPH-dependent oxidoreductase involved in cocaine biosynthesis pathway, and 50S ribosomal protein L4 is an rRNA binding protein that binds to 23S rRNA at a site near the 5'-end, which is necessary for protein synthesis.

Comparison of protein expression profiles of PDS and non-PDS soybean plants indicated that particular genes contribute to PDS state of growth. We found that 5 proteins in Katul and 11 proteins in Gorgan- 3 were to be changed (in term of expression level) in PDS. Among the proteins identified to be different in PDS and non-PDS plants, Ribulose 1, 5 bisphosphate carboxylase/oxygenase (Rubisco) large subunit was the only one whose expression increased in PDS plants of both cultivars, indicating that this enzyme plays a critical role in PDS. In addition, 14-3-3 like protein, NACA and OEE2 were the proteins whose expressions decreased significantly in PDS plants of both cultivars, showing that these proteins play important roles in PDS, too. Therefore, we conclude that these four proteins Rubisco large subunit, 14-3-3 like protein, NACA and OEE2 may be key mediators of PDS in soybean plants.

We would like to thank Prof. Jack Moore (University of Alberta, Edmonton, Canada) for analysis of protein samples and for his assistance in interpreting the Mass Spectrometry data.