Plasma membrane proteomic analysis by TMT-PRM provides insight into mechanisms of aluminum resistance in tamba black soybean roots tips

Plant culture and measurements of Al³⁺ treatments

Seeds of TBS were rinsed soaked in double distilled H₂O for 30 min and then transferred to moist towel in the dark at 25 °C. After germination, uniform seedlings were transferred onto a floating mesh in opaque beaker with 1/2 strength hoagland nutrient solution. The nutrient solution was renewed every day. Plants were cultured in a tissue culture room at 27/22 °C day/night temperatures with 14 h of light (200 µmol photons m⁻²·s⁻¹). When the true leaf was fully expanded, the seedlings were pre-treated with 0.5 mM CaCl₂ (pH 4.3) for 24 h. To choose optimum treatment concentration of AlCl₃, the roots were then transferred into 0.5 mM CaCl₂ solution containing 0, 25, 50, 75, 100 µM AlCl₃ at pH 4.3 for 24 h. Furthermore, to choose optimum time, the pre-treated roots were transferred to 0.5 mM CaCl₂ (pH 4.3) containing 50 µM AlCl₃ for 1, 2, 3, 4, 5, 6 d, respectively. After treatments, the relative growth rate (RRG) (Jiang et al., 2018), Evans blue staining (Arroyave et al., 2013) and Chrome Azurol S staining (Jiang et al., 2018) methods were used to measure the effects of Al treatment, root exudates were collected and concentrated for citrate content by enzymic determination (Zhao et al., 2003). Each experiment was repeated for five times. Two-week-old seedlings were transferred into 0.5 mM CaCl₂ solution containing AlCl₃ at 0 (CK), 50 µM (Al treatment) at pH 4.3 for 3 d. After treatment, root apices (0–2 cm) were excised, immediately frozen in liquid nitrogen and then harvested for isolating total RNA and the PM proteins.

Plasma membrane protein enrichment

The plasma membrane proteins were extracted as reported (Voothuluru et al., 2016). Briefly, the frozen root tissues were taken out from liquid nitrogen and quickly ground into fine powder. After adding ice-cold buffer H (330 mM sucrose, 50 mM Na₄P₂O₇, 25 mM NaF, 5% glycerol, 0.5% PVP, 10 mM EDTA, 1 mM Na₂MoO₄, 0.1% protease inhibitor cocktail, 3 mM DTT) (1 mL per g fresh weight of tissue), the samples were centrifuged for 10 min (10,000×g 4 °C). Next, we collected the supernatant and centrifuged for 30 min (100,000×g at 4 °C). The obtained crude microsomal pellets were washed with buffer H without DTT, and incubated in buffer H with 0.02% Brij-58 but without DTT (2 μL/μg of crude microsomal protein on ice) for 45 min to dissolve. The PM-enriched protein fractions were obtained after centrifuged for 30 min (100,000×g at 4 °C). The protein concentration was measured by BCA kit (Beyotime Biotechnology, Beijing, China) according to the manufacturer’s instructions.

TMT proteomics analysis

According to Guo et al. (2019), protein samples were subjected to trypsin digestion and TMT Labeling, then the tryptic peptides were fractionated into fractions by high pH reverse-phase HPLC and subjected to NSI source followed by tandem mass spectrometry (MS/MS) in Q Exactive TM Plus (Thermo, Waltham, MA, USA) coupled online to the UPLC. The resulting MS/MS data were processed using Maxquant search engine (v.1.5.2.8). Spectra search was performed against Phytozome 12 (the plant genomics resource, Glycine max Wm82.a2.v1; https://phytozome.jgi.doe.gov/pz/portal.html#!info?alias=Org_Gmax) concatenated with reverse decoy database. FDR of protein identification and PSM identification was set to 1%. To obtain PM proteins for analysis, subcellular localization was performed by wolfpsort (http://www.genscript.com/psort/wolf_psort.html). For protein abundance ratios measured using TMT data, we considered a 1.3-fold change at p < 0.05 as the thresholds for identifying significant changes.

Bioinformatics analysis

Functional annotation of PM proteins was performed using the GO annotation (http://www.ebi.ac.uk/GOA/) and protein domain annotation (http://www.ebi.ac.uk/interpro/). To classify and group the identified proteins, the KEGG pathway (http://www.genome.jp/kegg/) was analyzed. Among GO annotation, GO annotation proteome was derived from the UniProt-GOA database. Firstly, Converting identified protein ID to UniProt ID and then mapping to GO IDs by protein ID. If some identified proteins were not annotated by UniProt-GOA database, the InterProScan soft would be used to annotated protein’s GO functional based on protein sequence alignment method. Then proteins were divided into biological process and molecular function. To classify and group the identified proteins, KEGG (http://www.genome.jp/kegg/) database was used to annotate protein pathway. Firstly, using KEGG online service tools KAAS to annotated protein’s KEGG database description. Then mapping the annotation result on the KEGG pathway database using KEGG online service tools KEGG mapper. Based on GO, KEGG pathway and protein domain, enrichment analysis of differentially expressed proteins were performed. For each functional enrichment, a two-tailed Fisher’s exact test was used to test the enrichment of differentially expressed proteins relative to all identified PM proteins. GO, KEGG pathway and protein domain with a p < 0.05 were considered significant.

Targeted protein quantification by parallel reaction monitoring

To confirm TMT data, twenty proteins were randomly chosen and quantified by PRM-MS analysis. According to the TMT data, signature peptides for the target proteins were defined and only unique peptide sequences were determined for PRM analysis. As in the TMT experiment, protein preparation and trypsin digestion were performed. After HPLC-MS/MS Analysis, data of each sample was collected using the PRM acquisition method to quantify the targeted proteins abundance. Three biological replicates of each group were performed. The PRM-MS analysis in our study is supported by Jingjie PTM BioLabs (Hangzhou, China).

cDNA synthesis and RT-qPCR analysis

Total RNA was isolated using the RNAiso Plus kit (Takara, Dalian, China) according to the manufacturer’s description. A Nano Drop 2000 spectrophotometer (Thermo, Waltham, MA, USA) was used to measure concentration of total RNA. After DNase treatment, 1 µg total RNA was used as template for each cDNA synthesis by using PrimeScript RT reagent kit (Takara, Dalian, China). Real-time fluorescent quantitative PCR (RT-qPCR) was performed according to our previous work (Jiang et al., 2018). 18s rRNA gene (M16859) was used as endogenous control for RT-qPCR. Primers used were given in Table S1.

Statistical analyses

Statistical analysis was performed using one-way ANOVA or two-sided Student’s t-test. p < 0.05 was considered as of significant difference. Origin Pro (Version 8.5) was used to prepare graphs.

Effect of Al³⁺ treatment on citrate secretion of TBS roots

To optimize Al³⁺ treatment for proteome analysis, we measured citrate secretion at different concentration and duration of Al³⁺ treatment. The results showed that citrate secretion increased with the increase of AlCl₃ concentration and reached the maximum at 50 μM AlCl₃ with no apparent damage to the roots as shown by Evans blue and Chrome Azurol S staining (Fig. 1A). Then, citrate secretion was detected after different time of 50 μM AlCl₃ treatment. Although citrate secretion was increased with the prolonging of treatment time, a sharp enhancement of citrate secretion was detected at the third day with no more than a 25% inhibition of relative root growth (RRG) compared to the control (−Al³⁺) (Figs. 1B and 1C). Moreover, much lower of Evans blue and Chrome Azurol S staining was detected in the group treated with 50 μM AlCl₃ for 3 d than that in the groups with 50 μM AlCl₃ treatment for 4–6 d (Fig. 1D). Therefore, 50 μM AlCl₃ for 3 d was chose for further experiments.

Proteome profiles in TBS roots after Al³⁺ treatment

Proteome profiles were evaluated using TMT labeling and HPLC-MS/MS techniques in the PM of TBS seedlings after treatment with 0 and 50 μM AlCl₃ for 3 d (Fig. S1). The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD017160. A total of 6,877 proteins were obtained, of which 5,524 were available for quantifiable information (Table S2). The error rate of peptides was distributed between −5 and 5 ppm, which suggested that the accuracy of the mass spectrometry data meet the requirement (Fig. 2A). The length of more than 95% peptides was distributed between 7 and 21 amino acids (Fig. 2B), which was in consistence with the properties of tryptic peptides. The percentage of protein with a coverage of (0, 20%) was 69.74% (Fig. 2C). By subcellular localization, 907 PM proteins were identified, of which 758 PM proteins with quantifiable information (Table S2), and a good correlation was observed among the replicates of each treatment group (Fig. S2).

Al³⁺-induced differentially expressed proteins and enrichment analysis

Differentially expressed proteins (DEPs) in PM of TBS roots between Al³⁺-treated and untreated control groups (CK) were examined. Ninety PM proteins were found to be significant difference in expression in the Al³⁺-treated group, which included 46 up-regulated (FC (fold change) > 1.3, p < 0.05) and 44 down-regulated (FC < 0.77, p < 0.05) (Table S3).

The DEPs were annotated using Gene Ontology (GO) terms based on biological process (BP) and molecular function (MF). For BP, enriched GO terms in the up-regulated proteins were drug transport and drug transmembrane transport (Fig. 3A), whereas those in the down-regulated proteins were mainly associated with monovalent inorganic cation transport process (Fig. 3B). For MF, enriched GO terms in the up-regulated proteins included drug transporter activity, drug transporter transmembrane activity, antiporter activity, active transmembrane transporter activity, primary active transmembrane transporter activity, ATPase activity, nucleoside-triphosphatase activity, pyrophosphatase activity, transporter transmembrane activity, transporter activity and hydrolase activity (Fig. 3C), while the most over-represented GO terms in the down-regulated proteins were transporter activity and potassium ion transmembrane transporter activity (Fig. 3D).

For KEGG pathway enrichment analysis, ABC transporters were significantly enriched in both the up-regulated (Fig. 4A) and down-regulated (Fig. 4B) DEPs.

Protein domain enrichment analysis was performed at p < 0.05 using InterPro database. From high to low, the order of enriched domains in the up-regulated proteins included ABC transporter type 1, ABC transporter-like, AAA+ ATPase domain, P-loop containing nucleoside triphosphate hydrolase and Gnk2 homologous domain (Fig. 5A), whereas the down-regulated proteins contained major facilitator superfamily domain, aquaporin-like, nodulin-like and cellulose synthase (Fig. 5B).

To test the correlation of proteins with different FC changes, DEPs were divided into four Q groups: Q1 (0 < FC ≤ 0.67), Q2 (0.67 < FC ≤ 0.77), Q3 (1.3 < FC ≤1.5) and Q4 (FC >1.5). For each Q group, enrichment and cluster analysis were performed based on GO, KEGG pathway and protein domain. As shown in Figs. S3–S5, the results were consistent with GO, KEGG pathway and protein domain enrichment analysis.

Validation of TMT data by PRM and RT-qPCR

To validate the TMT data, 20 and 9 proteins were randomly chosen for PRM and RT-qPCR, respectively. Of the 20 chosen for PRM analysis, 16 proteins had quantitative information (Table S4), and 9 proteins with a good consistency between TMT and PRM results were shown in Figs. 6A–6I, including the protein abundance of four up-regulated (Glyma.09G144200.1.p, Glyma.19G263100.1.p, Glyma.10G047100.1.p and Glyma.19G016400.1.p) and five down-regulated proteins (Glyma.08G015300.1.p, Glyma.07G267800.1.p, Glyma.02G080900.1.p, Glyma.20G179700.1.p and Glyma.20G148200.1.p). RT-qPCR confirmed enhanced transcription of 5 corresponding genes (Glyma.01G210500.1, Glyma.02G181800.1, Glyma.10G047100.1, Glyma.09G144200.1, Glyma.13G203000.1). However, mRNA expression of 4 related genes (Glyma.06G076100.1, Glyma.11G146500.1, Glyma.17G041200.1, Glyma.15G274600.4) showed no significant difference between Al³⁺-treated and untreated groups (Figs. 6J–6R), differing from the TMT data, which may be due to discrepancy in transcription and post-transcriptional and translational regulation.

DEPs involved in membrane trafficking and transporters

Fifty-one DEPs involved in membrane trafficking and transporters were identified, of which 29 proteins were up-regulated and 22 proteins were down-regulated by Al stress (Table S5). These proteins mainly included ABC transporters, aquaporins, NRT1, potassium transporters, MATE efflux family proteins, sugar transporters, calcium transporters and H⁺-ATPase. Noticeably, 13 ABC transporters were found to be up-regulated and only one was down-regulated. However, aquaporins, NRT1, potassium and sulfur transporters were found almost to be down-regulated.

DEPs related to modifying cell wall composition

Five DEPs were related to cell wall synthesis and organization were identified (Table S5), of which cellulose synthases (CESA) (Glyma.04G067900.1.p, Glyma.02G080900.1.p), C2 domain-containing protein (Glyma.11G237800.1.p) and casparian strip membrane protein (CASP) 5 (Glyma.18G290100.1.p) were down-regulated except CASP-like protein (Glyma.20G033800.1.p).

DEPs of PM associated with defense response

Four DEPs associated with defense response were up-regulated and 2 DEPs were down-regulated (Table S5). The up-regulated DEPs included probable methyltransferase PMT16 (Glyma.09G144200.1.p), MLO-like protein 15 (Glyma.03G093600.2.p), Pleiotropic drug resistance protein 1 (Glyma.07G033500.1.p, Glyma.19G169400.1.p). The down-regulated DEPs comprised probable methyltransferase PMT26 (Glyma.04G211300.1.p) and Cytochrome B561 (Glyma.16G024500.2.p).

DEPs of PM belonged to signal transduction

Eighteen DEPs were belonged to signal transduction with 13 belonging to protein kinases (7 up-regulated and 6 down-regulated), 3 up-regulated associated with Ca²⁺ signaling and 2 down-regulated related to transmembrane receptors (Table S5). Among protein kinases, 6 up-regulated DEPs were classified as receptor kinase and determined by qRT-PCR (Figs. 7A–7F).

DEPs for Al stress in TBS roots

It has been showed that Al toxicity influenced the uptake of water and nutrients, thereby resulting in the inhibition of plant growth (Sade et al., 2016; Singh et al., 2017). Aquaporins are membrane channels that facilitate the transport of water and small neutral molecules across membrane and the buildup of turgor pressure driving cell elongation (Li et al., 2017). Previous data revealed that Al inhibits the expression of aquaporin genes in buckwheat (Xu et al., 2017), but aquaporin confers Al resistance by mediating Al uptake and translocation in blue hydrangea and Arabidopsis (Negishi et al., 2012; Wang et al., 2017). The decreased abundance of 6 aquaporin proteins found in this study indicates interfere of Al³⁺ with nutrient uptake in TBS roots (Table S5). This result was further supported by expression changes of proteins linked with nitrogen, potassium, sulfur and others transportation.

We also found Al³⁺ affected proteins associated with modifying cell wall composition. Plant cell wall is mainly consisted of cellulose, generated by integral membrane cellulose synthase (CESA) through depositing cellulose polymers into the developing cell wall (Watanabe et al., 2018). For example, SOS6 encoded a cellulose synthase-like protein in Arabidopsis, and SOS6 mutant results in sensitive to salt-induced osmotic stress due to defect in cell wall formation (Zhu et al., 2010). Indeed, 3 CESAs were markedly reduced under Al³⁺ treatment in TBS membrane (Table S5). Furthermore, 2 Casparian strip membrane proteins (CASPs) were affected by Al³⁺ in this study. CASPs provide the key scaffold for the formation and localization of casparian strips and functions in nutrient uptake and stress resistance (Roppolo et al., 2011), which was supported by the findings that cold and cadmium tolerance in upland cotton and Arabidopsis was associated with increased expression of CASPs (Chen et al., 2019). Thus, Al stress-responsive in TBS roots involves cell wall modification via affecting the expression of CESAs and CASPs in plasma membrane, thereby affecting plant root growth.

DEPs for Al detoxification in TBS roots

The Al-activated release of citrate from root apex, has been characterized as one of the most important Al-resistance mechanisms in soybean (Kochian et al., 2015), and MATE transporters participate in citrate secretion (Li et al., 2017; Liu et al., 2018). Here, the expression of three MATE proteins (GmMATE13, GmMATE75 and GmMATE87) were up-regulated by Al stress (Table S5). In consistence, overexpression of GmMATE75 and GmMATE87 in root apex or base zone resulted in more citrate efflux, thus alleviated root elongation inhibition by Al³⁺ in Arabidopsis (Zhou et al., 2019). Similarly, GmMATE13 has been confirmed to mediate citrate exudation and then to enhance Al resistance in soybean hairy roots (Wang et al., 2019a, 2019c). In addition, up-regulation of H⁺-ATPase was found under Al stress (Table S5). PM H⁺-ATPase extrudes protons from cells to create an electrochemical gradient across the plasma membrane that enables the release of citrate under Al toxicity and phosphorus deficiency (Yu et al., 2016). Our previous data indicated that Al enhanced the expression the PM H⁺-ATPase and interaction between PM H⁺-ATPase and the 14-3-3 protein led to higher activity of the PM H⁺-ATPase and more citrate exudation in broad bean (Chen et al., 2013). Although further investigations are needed, these results lead us to conclude that GmMATE13, GmMATE75, GmMATE87 and H⁺-ATPase collectively function in Al resistance via releasing citrate in TBS roots (Fig. 8).

On the other hand, most ABC transporters are present in cell membranes and are essential for plant nutrition, hormone transport, pathogen responses and lipid deposition (Kochian et al., 2015; Wang et al., 2019b). ABC transporters have been implicated in the tolerance to metal toxicity (Ezaki et al., 2015), such as cadmium in rice (Fu et al., 2019). Accordingly, Al³⁺ induced twelve ABC transporters in the present study, including 2 ALS3 proteins and 5 ABCB transporters (Table S5). ALS3 functions in redistributing Al in plants away from the sensitive root tips, and transport Al into vacuole for detoxification (Larsen et al., 2005; Reuna-Llorens et al., 2014). Furthermore, ALS3 interacts with its functional partner sensitive to Al rhizotoxicity 1 (STAR1), and STAR1-STAR2 complex transport UDP-glucose into the cell wall, thereby masking Al-binding in the cell wall (Kochian et al., 2015). In addition, ABCB transporters participate in the transport of auxin, which enhances Al-induced citrate exudation through up-regulation of GmMATE and activation of the plasma membrane H⁺-ATPase in TBS roots in our previous study (Wang et al., 2016). Thus, the expression patterns of above important transporters indicated that TBS harbors avoidance and tolerance strategies for Al-detoxification.

DEPs for defense responses in TBS roots

Evidence shows methyltransferases decreasing ROS, lipid peroxidation and ion leakage, and heterologous expression of methyltransferase in Escherichia coli confers Al resistance through melatonin production (Luo, He & Han, 2018). Likewise, pleiotropic drug resistance (PDR) proteins confer the plant to have the capacity of stress response and cellular detoxification (Wolfger, Mamnun & Kuchler, 2001). NtPDR1 is involved in the general defense response in tobacco (Sasabe et al., 2002). It showed that the silence of NpPDR1 resulted in an increased susceptibility to sclareol and a reduced resistance to plant pathogen (Stukkens et al., 2005). Additionally, MLO proteins play a broad role in plants, such as stress response processes and cell death protection (Kim et al., 2002; Stein & Somerville, 2002). MLO-mediated defense suppression functions independently of heterotrimeric G-proteins, and functions in Ca²⁺-dependent interaction with calmodulin (Kim et al., 2002). The enhanced abundance of PMT16, PDR1 and MLO15 (Table S5) indicated their involvement in Al³⁺ resistance in TBS.

DEPs for signal transduction

We previously revealed that Al³⁺ signal transmits through heterotrimeric G-proteins, PLC, IP3, DAG, Ca²⁺ and PK, which result in citrate secretion from roots via modifying tricarboxylic cycle (Jiang et al., 2018) (Fig. 8). Recently, Al receptors or sensors have been put forward, and identification of the upstream elements for sensing extracellular Al signal remains to be an open and important question in plant Al stress research. The plasma membrane is the first site of signal perception through receptor proteins and membrane receptors for Al³⁺ perception has been recognized (Yang, Fan & Zheng, 2019). Particularly, it has been established that plant receptor-like kinases (RLKs) act as receptors for extracellular signal perception in abiotic stress responses (Morris & Walker, 2003; Ye et al., 2017). Accordingly, many RLKs have been identified for stress signal detection, such as CPKs for salt sensing (Gao et al., 2018), CRK5 for drought sensing (Burdiak et al., 2015), RLCK for Pseudomonas syringae sensing (Liu et al., 2019). As expected, 6 RLKs were found to be enhanced in expression after Al³⁺ treatment (Table S5). This result provided candidate proteins for further investigation on the upstream signal element of Al signaling in TBS.

Calcium ion (Ca²⁺) serves as an essential second messenger to regulate many cellular activities, thereby achieving a specific physiological response (Yang et al., 2019a, 2019b). Our previous data also established that cytoplasmic Ca²⁺, an essential intracellular second messenger, mediates Al signaling in stylo (Jiang et al., 2018). Calcium channel proteins are critical regulators of calcium homeostasis, and calcium-binding proteins can activate many calcium dependent protein kinases (Chen et al., 2015). Disruption of intracellular Ca²⁺ upon Al³⁺ exposure triggers a significant decrease of citrate release in the root (Chen et al., 2015; Jiang et al., 2018). In this research, abundance of 5 proteins associated calcium signaling were up-regulated under Al stress, including 2 calcium channel proteins, 2 calcium-binding proteins and 1 calcium-dependent lipid-binding family protein (Table S5). These results lend supports to the involvement of Ca²⁺ in Al-induced citrate exudation in TBS.