A comparative proteomic analysis for adventitious root formation in lotus root (Nelumbo nucifera Gaertn)

2.1 Plant materials

Lotus root seeds were harvested from the open cultivated field of YangZhou University, which is located in South-Eastern China. One hundred seeds were placed in a container with water at 10 cm deep for germination. The containers were placed in a light incubator with temperature 26°C/day and 20°C/night (a diurnal cycle of 16 h light/8 h darkness and a light intensity of 200 m⁻²s⁻²) during the whole experimental period. When the plants grew to the 1–2 leaves stage (about 5 days after germination), ARs started to form at the hypocotyls. For the analysis of protein expression, three developmental stages (0 day: germinated, 5 days: AR formed, 10 days: maximum number of AR) were chosen to monitor the changes in expressed proteins.

2.2 Protein extraction and iTRAQ labeling LC-MS/MS

Lotus generally contains low amounts of protein, so we carried out the protein extraction according to the protocol described by Hurkman and Tanaka [26], with some minor modifications. Hypocotyls from three stages (induction stage, initial stage and expression stage) were collected. For extraction of proteins, 2 g of hypocotyls were placed in liquid nitrogen and ground into powder with a pestle and mortar. After homogenizing the powdered hypocotyls, 2 mL of ice-cold phenol was added and the samples were incubated at 4°C for 30 min. Two milliliters of extraction buffer (pH 8.0) [250 mM sucrose, 25 mM Tris–HCl (pH 8.0), 1 mM PMSF, 10 mM EDTA and 1 mM DTT] were then added and the samples were centrifuged at 16,000×g at 4°C for 30 min. The phenol phase was transferred to another tube and the proteins were precipitated by the addition of an equal volume of extraction buffer. The procedure was repeated twice. The final phenol phase was added with 3–4 mL cold solution of 100 mM acetamide in methanol, and the mixtures were placed at −20°C for at least 2 h, and then were centrifuged at 16,000×g at 4°C for 30 min, and the supernatant discarded.

Finally, Bradford protein quantification and SDS-PAGE were used to determine the protein concentrations, taking bovine serum albumin as the standard (Bradford 1976). Then, 100 µg protein from three AR development stages was digested with Trypsin Gold (protein:trypsin=20:1), and then placed at 37°C for 4 h. The above step was repeated with an 8-h digestion, and 0.5 M TEAB was added to reconstitute the digested peptides, according to the manufacturer’s instructions (Applied Biosystems). One unit of iTRAQ reagent was thawed and reconstituted in 24 µL of isopropanol. The digested peptides were labeled using the isobaric tags at room temperature for 2 h, and then dried and pooled after vacuum centrifugation. To fractionate the labeled peptides, strong cationic exchange (SCX) chromatography was applied with a Shimadzu LC-20AB high-pressure liquid chromatography (HPLC) pump system. Digested peptides of each sample were reconstituted in about 4 mL of buffer A [pH 2.7, 25 mM NaH₂PO₄ in 25% acetonitrile (ACN)], and then loaded onto a 4.6 9 250-mm Ultremex SCX column containing 5-µm particles (Phenomenex). A 1 mL/min flowing rate was used to elute peptides using the following processes: buffer A for 10 min, then buffer B (pH 2.7, 25 mM NaH₂PO₄, 1 M KCl in 25% ACN) for 11 min, 35%–80% buffer B for 1 min before equilibrating with buffer A for 10 min before the next injection; the whole system was kept in 80% buffer B for 3 min. Absorbance at 214 nm was used to monitor the eluent; fractions were collected each minute and then pooled as 10 fractions. Samples were dried using a vacuum and desalted using a Strata X C18 column.

2.3 LC–ESI–MS/MS analysis

Each fraction was resuspended in a certain volume of buffer A (2% ACN, 0.1% FA) and centrifuged at 20,000×g for 10 min. In each fraction, the final concentration of peptide was about 0.5 µg/µL on average; 10 µL of supernatant was loaded on a Shimadzu LC-20AD nano-HPLC by the auto sampler onto a 2-cm C18 trap column (inner diameter, 200 μm), and the peptides were eluted onto a revolving 10-cm analytical C18 column (inner diameter, 75 μm) made in-house. The samples were loaded at 15 μL/min for 4 min, then the 44-min gradient was run at 400 nL/min starting from 2% to 35% B (98%ACN, 0.1%FA), followed by a 2-min linear gradient to 80%, and maintenance at 80% B for 4 min, and finally returned to 2% in 1 min. Data acquisition was performed with a Triple TOF 5600 System (AB SCIEX, Concord, ON, Canada) fitted with a Nanospray III source (AB SCIEX) and a pulled quartz tip as the emitter (New Objectives, Woburn, MA, USA). Data were acquired using an ion spray voltage of 2.5 kV, curtain gas of 30 PSI, nebulizer gas of 15 PSI, and interface heater temperature of 150°C. The MS was operated with an RP of greater than or equal to 30 000 FWHM for TOF MS scans. For IDA, survey scans were acquired in 250 ms and as many as 30 product ion scans were collected if they exceeded a threshold of 120 counts per second (counts/s) and with a 2+ to 5+ charge-state. Total cycle time was fixed to 3.3 s. The Q2 transmission window was 100 Da for 100%. Four time bins were summed for each scan at a pulse frequency value of 11 kHz by monitoring the 40 GHz multichannel TDC detector with a four-anode channel detect ion. A sweeping collision energy setting of 35±5 eV adjusted rolling collision was applied to all precursor ions for collision-induced dissociation. Dynamic exclusion was set for 1/2 of the peak width (18 s), and then the precursor was refreshed off the exclusion list.

2.4 Protein identification, quantification and bioinformatics analysis

We used the Mascot protein identification software, which was declared the gold standard for bioinformatics by the Frost/Sullivan research organization. The MGF files were searched using Mascot version 2.3.02 against the NCBInr, SwissProt and UniProt database. The peptide and protein data were extracted using high peptide confidence and top one peptide rank filters. For the protein relative quantity, a more than 2-fold change in abundance and a p-value <0.05 identified a differentially abundant protein. Proteins were confirmed by BLAST against the NCBI, and then gene ontology analysis was performed by Blast2 GO. Using the Kyoto Encyclopedia of Genes and Genomes (KEGG) tool, the differentially expressed proteins were further annotated into biological pathways. The differentially abundant proteins were classified according to their function using Blast2 Go with default parameters (https://blast2go.com).

3.1 Analysis of lotus AR formation under normal growth conditions

ARs are very important for lotus growth and development; therefore, proteomic changes during AR formation in hypocotyls were assessed in a local traditional species (Taikong lotus) using iTRAQ coupled with LC-MS/MS. Initially AR formation was assessed under 26°C/day and 20°C/night conditions. We observed that AR formation could be divided into three phases including phase I (no ARs stage: about 0–5 days), phase II (ARs stage: about 5–10 days) and phase III (maximum number of ARs stage: more than 10 days). During phase I, ARs began to form, which is traditionally referred to as the induction of AR formation. The ARs began to emerge at 5 days, and the number of ARs significantly increased in phase II. The number of ARs did not change after 10 days of growth. Therefore, the three key time points of 0 days (C0 stage), 5 days (C1 stage) and 10 days (C2 stage) were used for further analysis of proteomic changes during AR formation in the hypocotyls (Figure 1).

Figure 1:

Analysis of lotus AR formation. One hundred lotus seeds were selected and placed in a container with water at 10 cm in depth for germination. The seedlings with consistent growth were chosen for further analysis of AR formation.

3.2 Identification and quantification of proteins

Proteins were extracted from the hypocotyls of 0 day, 5 days and 10 days grown lotus seedlings under normal growth condition. After determining the protein concentrations, 100 μg was used for protein separation and identification by iTRAQ and LC-MS/MS. To reduce the probability of false peptide identification, only peptides at the 95% confidence interval, as assessed by a Mascot probability analysis greater than “identity”, were counted as identified, and each confident protein identification involved at least one unique peptide. The error distribution of the theoretical values and true values of the relative molecular weights for all matching peptides are shown in Figure 2A; 323,375 spectra during AR formation were obtained. All the spectra were filtered to eliminate low-scoring spectra, resulting in 66,943 spectra, including 53,106 unique spectra, being identified. These unique spectra were matched to 28,905 peptides, among which 24,992 were unique (Figure 2B). Based on these identified peptides, 6686 proteins were assembled (e.g. see Supplementary Material, File. 1). The molecular mass distribution of the proteins showed that about 14% of the proteins were larger than 100 kDa, and about 6% were 10–20 kDa. Eighty percent of the proteins were between 20 kDa and 100 kDa (Figure 2C).

Figure 2:

Detailed information on protein identification by iTRAQ analysis. (A) analysis of the error distribution of spectra matches quality of lotus during AR formation. (B) identification of the proteome relevant to AR formation. Total spectra, spectra and unique spectra represent the total number of secondary spectra, the spectra number that were matched and unique peptides that were matched, respectively. Peptide, unique peptide and protein represent the number of identified peptides, the number of unique peptides identified and the final number of identified proteins, respectively. (C) Relative molecular mass of identified proteins related to AR formation. The x-axis represents the molecular weights of identified proteins (unit: kDa); the y-axis represents the distribution of these identified proteins according to their molecular mass.

3.3 Protein annotation and classification

Protein functions were annotated by comparison against the existing NCBI database. First, GO functional annotation analysis was performed to study the identified proteins. All proteins were classified into three processes: biological process, cellular component and molecular function. For the biological process, the proteins involved in cellular and metabolic processes were the largest group, with 16.48% and 16.41% of the total identified proteins, respectively. Only 0.04% of proteins were observed to take part in locomotion. For the cellular component category, 23.94% of identified proteins were associated with metabolism of cell parts, and 19.33% were associated with the membrane. We observed that 44.89% of the proteins had catalytic activity, suggesting that the process of AR formation is very complex. For protein orthologous classification, all identified proteins were compared with the database of Cluster of Orthologous Groups (COG) to predict possible protein function (Table 3). The annotated proteins could be classified into 24 clusters. The largest cluster (1124 proteins) of proteins was ‘general function prediction only’, followed by ‘posttranslational modification, protein turnover, chaperones’ and ‘translation, ribosomal structure and biogenesis’. The smallest cluster (three proteins) was relevant to ‘nuclear structure’ (Figure 3).

Figure 3:

Classification of proteins by COG according to protein function in plant metabolism.

3.4 Identification of differentially abundant proteins during AR formation

Based on protein classification and annotation, differentially abundant proteins were identified from the three developmental stages during AR formation to uncover the changes in metabolism at the molecular level. A greater than 2-fold difference in protein abundance and a p-value <0.05 were used as a threshold to judge whether a protein’s change in abundance was significant. In the C0/C1 stages, 239 differentially abundant proteins were identified, among which 66 were upregulated and 173 were downregulated. In the C1/C2 stages, 105 differentially abundant proteins were identified, including 32 upregulated proteins and 73 downregulated proteins. Thus, there were more differentially abundant proteins in the C0/C1 compared with the C1/C2 stages, which suggested that the initial stage of AR formation is more complex than the late stage of AR formation (Figure 4).

Figure 4:

The number of differentially abundant proteins in three different stages.

The differentially abundant proteins were selected to monitor changes in metabolism during AR formation. Proteins in the C0/C1 stages could be mainly classified into 12 categories of energy metabolism, protein and amino metabolism, coenzyme metabolism, lipid metabolism, antioxidant activity, calcium regulation, cell/wall/membrane biogenesis, ion transport, secondary metabolites, plant hormone, hydrolysis and defense mechanisms. Proteins in the C1/C2 stages could be mainly classified into 14 categories, such as energy metabolism, protein and amino acid metabolism, nucleotide metabolism, carbohydrate metabolism, lipid transport and metabolism, transcription, cell wall/membrane biogenesis, posttranslational modification, inorganic ion transport, secondary metabolites, antioxidant activity, plant hormone, defense mechanisms and no function predicted. Many proteins were downregulated in the C0/C1 and C1/C2 stages. Most of the downregulated proteins in the C0/C1 stages are involved in the synthesis of substances, and those in the C1/C2 stages were related to cell growth and differentiation, translation, RNA process, and modification (Table 3).

3.5 Proteins related to AR formation

All the proteins were checked in the C0/C1 and C1/C2 stages during AR formation. The data showed that most of the proteins changed in abundance in above two stages were different, suggesting that the metabolic processes might be different in the C0/C1 and C1/C2 stages. To test whether the expression patterns in this article represented all the well-defined proteins, the data sets were compared with previous reports. We found 17 identified proteins that were relevant to AR formation. These proteins and their biological functions are listed in Table 4. Among them, three proteins, pectin methylesterase, peroxidase 3 and glutathione S-transferase, showed enhanced expression levels in the C0/C1 and C1/C2 stages, and 13 proteins, including S-adenosylmethionine synthase, L-ascorbate peroxidase 2, peroxidase 2, peroxidase 7, cationic peroxidase 1, peroxidase 51, calcium sensing receptor, calcium-binding protein, calcium-dependent protein kinase 2, indole-3-acetic acid synthetase GH3, salicylic acid-binding protein 2, exordium protein and multicopper oxidase LPR1, showed increased abundance only in the C0/C1 stages; no significant changes in their expressions were found in the C1/C2 stages. One protein (Phospholipase D) showed decreased abundance in the C1/C2 stages of AR development.

3.6 Expression analysis of nine genes by qRT-PCR

Nine genes encoding proteins involved in AR formation (pectin methylesterase, peroxidase 3, l-ascorbate peroxidase 2, peroxidase 2, peroxidase 7, calcium sensing receptor, calcium-dependent protein kinase 2, indole-3-acetic acid synthetase GH3, salicylic acid-binding protein 2 and glutathione S-transferase) were analyzed by quantitative RT-PCR. The expressions of eight genes were enhanced in the C0/C1 stage during AR formation. The abundance profiles of pectin methylesterase, calcium-dependent protein kinase 2, peroxidase 7 and salicylic acid-binding protein were very similar. The mRNA level of these genes was improved at 4 days, and then decreased. The peroxidase 3, L-ascorbate peroxidase 2 and indole-3-acetic acid synthetase GH3 genes showed significantly enhanced expression at 10 days. Only one gene (encoding peroxidase 2) showed decreased expression during all AR formation stages (Figure 5). The above results indicated a correspondence of the results from qRT-PCR analysis with the iTRAQ analysis.

Figure 5:

Expression analysis of nine important genes by real-time PCR. The genes encoded pectin methylesterase, peroxidase 3, L-ascorbate peroxidase 2, peroxidase 2, CDPK2, peroxidase 7, Glutathione S-transferase, indole-3-acetic acid synthase and salicylic acid-binding protein 2. Actin was used as an internal standard. Data are from three replicates.

4.1 Protein identification in three stages during AR formation

Recently, the iTRAQ technique coupled with LC–MS/MS has identified regulatory mechanism [25]. There is evidence that the formation of AR in lotus (Nelumbo nucifera) is strictly regulated by many exogenous factors, especially by plant hormones [27].

AR formation is classified into three stages: the induction stage, the initiation stage and the expression stage [28, 29]. In the induction stage, new meristematic cells related to AR are established [9]. Root meristems and primordia are formed, and the root emerges from the stem in the initial and expression stages [29]. In this experiment, lotus root was cultivated in an illuminated incubator under 26°C/day and 22°C/night conditions. Three obvious developmental stages (no AR, occurrence of AR and maximum AR) were observed (Figure 1), and these three stages were used to analyze the changes in the abundance of proteins during AR formation; 323,375 spectra for AR formation were obtained. All the spectra were filtered to eliminate low-scoring spectra, and 66,943 spectra, including 53,106 unique spectra, were identified. These unique spectra were matched to 28,905 peptides, among which 24,992 were unique. Based on these identified peptides, 6686 unique proteins were identified (e.g. see Supplementary Material, File. 1), and these genes’ function were annotated accordingt to GO and COG classification (Figure 3, Table 2). We observed that many proteins showed significant changes in their abundance during AR development. The higher number of proteins in the C0/C1 stages suggested that regulation from the no AR stage (C0) to the occurrence of the AR stage (C1) is more complex compared with that of the other (C1/C2) developmental stages (Figure 4).

4.2 Proteins related to anaerobic adaptation

Adaptation to submergence is very important for the survival of lotus. In this article, we found that several proteins, including ADH and calcium-dependent protein, which are related to submergence, showed increased abundance during AR formation (Table 3). Some species are also sensitive to flooding, and their productivity is seriously affected by exposure to anaerobic stress [30]. However, most aquatic plants (including lotus root) exposed to submergence show high resistance [31]. According to our results, the regulation of the expressions of the genes encoding these submergence adaptation-related proteins is essential for the formation of ARs. ADH (alcohol dehydrogenase) plays an important role in low oxygen tolerance. The expression and activity of ADH are thought to be an indicator of oxygen limitation [32]. Therefore, enhanced expression of ADH during AR development would be essential for lotus roots to adapt to submergence conditions.

Reactive oxygen species (ROS) production increases under abiotic stresses, including anaerobic stress [33]. ROS cause oxidative damage to membrane lipids, proteins and nucleic acids. Thus, ROS detoxification forms an important defense against abiotic stresses. Detoxifying enzymes include superoxide dismutase, catalase and enzymes of the ascorbate-glutathione cycle. We found that many peroxidases showed increased abundance under anaerobic condition (Table 3). Peroxidase is an antioxidant enzyme that is usually upregulated to reduce forms of antioxidants to prevent the formation of ROS [34]. There is evidence that these antioxidant enzymes have critical roles in the adaptation to anaerobic stress in some plants [35, 36]. In this article, we also observed that glutathione S-transferases (GSTs) were upregulated in lotus roots under anaerobic conditions. Some studies have shown that GSTs mainly function by catalyzing the conjugation of GSH to a variety of electrophilic, hydrophobic substrates to produce more water-soluble conjugates [37, 38]. Increased levels of GSTs would improve the low oxygen adaptation ability in the seedlings of lotus root. The plant plasma membrane has been implicated as the primary site of injury, because of a transition in the molecular ordering of membrane lipids. As mentioned above, ROS produced by low oxygen stress also damage the plasma membrane. Therefore, upregulation of peroxidases and GST help to protect plants from ROS damage.

4.3 Plant hormone metabolism during AR formation

AR formation is affected by various environmental and endogenous factors [28, 39]. Plant hormones are widely believed to regulate the growth and developmental processes of plants. Formation of AR is affected by plant hormones, among which auxin has been proven to regulate AR formation [22]. Leyser [40] showed that auxin is involved in shoot and root branching, and in vascular differentiation. Application of exogenous 1-Naphthaleneacetic acid (NAA) results in higher percentages of rooting, larger numbers of ARs and heavier root dry weight [41]. Auxin plays an essential role in the induction stage rather than the initial stage of AR formation, and metabolism relevant to AR formation is very sensitive to auxin in the induction stage [42]. Auxin signal perception and transduction are reduced in Arabidopsis PLD2 mutants [29]. Further analysis revealed that phosphatidic acid accumulation via phospholipase D is required for AR formation, and in cucumber, this process is an early signaling event during AR formation induced by auxin [43]. In the present article, indole-3-acetic acid synthetase and phospholipase D showed enhanced abundance from the induction stage to the initial stage, indicating that IAA and phospholipase D are involved in AR formation, which might be critical for AR induction in lotus (Table 3, Figure 4). However, we observed that low concentrations of IAA promoted AR formation, and AR formation was inhibited by high concentrations of exogenous IAA (data not shown). Therefore, maintaining the balance of IAA metabolism in lotus is important for AR formation.

Oxidative decarboxylation of IAA by plant peroxidases is thought to be a major degradation reaction involved in controlling the in vivo level of IAA [44]. We observed increased abundances of l-ascorbate peroxidase 2, peroxidase 2, cationic peroxidase 1, peroxidase 51 peroxidase P7 and peroxidase 3 in C0/C1 stages, and peroxidase P7 and peroxidase 3 showed increased abundance in stages C1/C2 (Table 3). These peroxidases possibly play roles in the regulation of IAA content through oxidation during AR formation in lotus. Enhancing IAA level in induction stage and decreasing IAA levels in the initial stage and expression stages probably benefited the formation of AR. At the same time, peroxidase activity was remarkably enhanced in IBA-treated tissues, compared with the control, during the formation of AR in soybean hypocotyl [45]. This suggested that a interaction might exist in the regulation between peroxidase and IAA during AR formation.

1-aminocyclopropane-1-carboxylate oxidase is a critical enzyme in ethylene synthesis. In this experiment, we found that the abundance of 1-aminocyclopropane-1-carboxylate oxidase increased in the C1/C2 stages, while no significant change was found in C0/C1 stages (Table 3), suggesting that ethylene might not be necessary for induction of AR formation. Ethylene is considered a critical plant hormone in ethylene-mediated growth promotion. Ethylene not only mediates a range of different biotic and abiotic stress responses [46], but is also involved in plant growth and development. Phatak et al. reported that ethylene has a positive effect on AR formation in tomato [47]. In sunflower, Liu and Reid [48] observed that both endogenous auxin and ethylene promote AR formation. A further study provided evidence that root growth is facilitated by the induced death of epidermal cells of the node external to the tip of the root primordium until root occurrence, and that this process is mediated by ethylene action [49]. The analysis by application of inhibitors of ethylene biosynthesis and perception, as well as of the precursor aminocyclopropane-1-carboxylic acid, showed that ethylene is a stimulator for AR formation in petunia cuttings [42]. However, Coleman et al. [50] believed that ethylene plays a negative role in AR formation.

Recently, ethylene and auxin interaction was observed to be involved in lateral root (including AR) development [51]. Ethylene increases the sensitivity of the root-forming tissues to endogenous indole acetic acid, which directly results in initiating AR formation [17]. We found that protein related to IAA synthesis was enhanced abundant only in C0/C1 stages, and protein involved in ethylene synthesis was improved abundance only in C1/C2 stages. Therefore, we believe that IAA was a promoter of AR induction, and that the synthesis of ethylene is related to AR development (Table 3, Figure 5).

4.4 Ca²⁺ signal transduction

Ca²⁺ is an intracellular second messenger that regulates plant cell physiology and cellular responses to the environment. Cytosolic Ca²⁺ is mobilized from both intracellular and extracellular sources to allow plant cells to adapt to growth stimuli. Recently, many reports have documented the relationship between Ca²⁺and AR formation. Cytosolic Ca²⁺ is a downstream component of the H₂O₂ signaling pathway and depends on the auxin response for AR development [52]. In addition, Ca²⁺ and CaM participate in AR development, which is induced by NO and H₂O₂ in marigold [53]. Ca²⁺-dependent protein kinases are considered as Ca²⁺ sensors and are a family of serine/threonine proteins. Ca²⁺ dependent protein kinase (CDPK) is involved in AR development, where it acts as a downstream messenger in the signaling pathway triggered by auxins to promote AR formation [54]. The expressions of CDPK2, calcium-binding protein and calcium sensing receptor in lotus were studied in detail, and we concluded that these proteins play an important role in AR formation, according to their expression profiles.

In addition, pectin methylesterase, s-adenosylmethionine synthase and glutathione S-transferase increased their abundances in stages C0/C1. Pectin methylesterase belongs to a large multigene family in Arabidopsis, which is involved in various physiological processes, including cell elongation [55], cell diffraction [56] and microsporogenesis [57]. Wu et al. [58] observed that pectin methylesterase activity is required for cell wall remodeling during heat stress in soybean, which maintains plasma membrane integrity and coordinates with heat shock protein to confer thermotolerance. Antisense mRNA expressed in transgenic pea hairy roots prevented the normal separation of root border cells from the root tip into the external environment, which ultimately affects root development [59]. In the pectin methylesterase atpme3-1 mutant, the number of ARs is affected, suggesting that expression of atpme3-1 is highly correlated with AR development. The present study could not confirm whether pectin methylesterase participates in AR formation through Ca²⁺ regulation. According to expression of CDPK2, calcium-binding proteins and calcium sensing receptors, we concluded that Ca²⁺might regulate the activity of pectin methylesterase or other downstream genes to control AR formation (Tables 3 and 4).