Comparative proteomic analysis of pepper (Capsicum annuum L.) seedlings under selenium stress

Plant materials and Se treatments

Pepper seeds (C. annuum 8 #, a cultivar provided by pepper breeding group in Fujian Agriculture and Forestry University) were sterilized with 1% sodium hypochlorite for 30 min and grown in steam-sterilized soil. The seedlings were grown in a greenhouse under the following conditions: 12 h light (150 μm² s⁻¹) at 26 °C, 12 h dark at 23 °C, and relative humidity of 60%. Seedlings were irrigated with half-strength Hoagland solution (pH 5.6). Pepper plants at the four true leaf stages were used for Se treatment. Seedlings were sprayed with half-strength Hoagland solution containing 0 or 100 ppm Na₂SeO₄. After 24 h, the shoots were collected for protein extraction.

Protein extraction and trypsin digestion

Samples were removed from storage at −80 °C, and fixed amounts of tissue samples were ground to powder while liquid nitrogen was added. The samples in each group were treated with four volumes of phenol extraction buffer (containing 10 mM dithiothreitol, 1% protease inhibitor, and two mM EDTA), then sonicated three times on ice with a high intensity ultrasonic processor (Scientz, Ningbo, China). The supernatant was centrifuged for 10 min at 4 °C and 5,500×g with an equal volume of Tris equilibrium phenol. The supernatant was collected and precipitated overnight with five volumes of 0.1M ammonium acetate/methanol. The protein precipitate was washed with methanol and acetone successively. Finally, the precipitate was re-dissolved in 8M urea, and the protein concentration was determined with a BCA kit (code P0010; Beyotime, Beijing, China) according to the manufacturer’s instructions.

For digestion, the final concentration of dithiothreitol in protein solution was five mM, and reduction was performed by incubation at 56 °C for 30 min. The mixture was then alkylated with 11 mM iodoacetamide at room temperature for 15 min. Finally, the urea concentration of the sample was diluted to less than 2M by addition of 100 mM triethyl ammonium bicarbonate. Trypsin was added at a mass ratio of 1:50 (trypsin:protein), and enzymatic hydrolysis was carried out overnight at 37 °C. Trypsin was then added at a mass ratio of 1:100 (trypsin:protein), and enzymatic hydrolysis continued for 4 h.

TMT labeling and HPLC fractionation

The trypsinized peptide segments were desalted with a Strata X C18 column (Phenomenex, Torrance, CA, USA) and then freeze-dried under vacuum. The peptides were dissolved in 0.5M triethyl ammonium bicarbonate and labeled with a TMT kit (ThermoFisher, Shanghai, China) according to the manufacturer’s instructions. The procedure was as follows: the labeled reagent was dissolved in acetonitrile after thawing, incubated at room temperature for 2 h after mixing with the peptide segments, then desalinated after mixing with the labeled peptide segment and freeze-dried in a vacuum.

Peptide segments were classified with high pH reverse-phase high performance liquid chromatography (HPLC) with an Agilent 300 Extend C18 column (five μm diameter, 4.6 mm inner diameter, 250 mm length) (Agilent, Shanghai, China). The gradient of peptide segments was 8–32% acetonitrile (pH 9.0), and more than 60 min was required to separate the peptide segments into 60 components. The peptide segments were then merged into 18 components, which were then vacuum freeze-dried.

LC-MS/MS analysis

The tryptic peptides were dissolved in 0.1% formic acid (solvent A) and directly loaded onto an Acclaim PepMap 100 reverse-phase pre-column (ThermoFisher, Shanghai, China). The gradient comprised an increase from 6% to 23% solvent B (0.1% formic acid in 98% acetonitrile) over 26 min, from 23% to 35% in 8 min, and to 80% in 3 min; then the concentration was held at 80% for the last 3 min. All steps were performed at a constant flow rate of 500 nL/min on an EASY-nLC 1000 ultra-HPLC system.

After equilibration, the peptides were ionized with an NSI ion source and analyzed by MS/MS (Q ExactiveTM mass spectrometer; ThermoFisher, Shanghai, China) coupled online to ultra-HPLC. The ion source voltage was set to 2.0 kV, and the parent ions of peptide segments and their secondary fragments were detected and analyzed with a high resolution Orbitrap. The scanning range of the primary MS was set to 350–1,800 m/z, and the scanning resolution was set to 70,000. The scanning range of the secondary MS was set to 100 m/z, and the scanning resolution of the secondary MS was set to 17,500. The data acquisition mode used a data-dependent scanning program; that is, the first 20 peptide ions with the highest signal intensity were selected to enter the higher collision dissociation collision pool in turn after the first scan, and the fragmentation energy was 28% for fragmentation. Similarly, secondary MS analysis was carried out sequentially. To improve the efficiency of MS, AGC was set to 5E4, the signal threshold was set to 2E4, the maximum injection time was set to 100 ms, and the dynamic elimination time of MS/MS scanning was set to 30 s to avoid repeated scanning of parent ions.

Database search

The resulting MS/MS data were processed and searched against a public proteome C. annuum database (https://www.uniprot.org/uniprot/?query=proteome:UP000222542) with the Maxquant search engine (v.1.5.2.8, https://maxquant.org/) concatenated with a reverse decoy database. The retrieval parameter settings were as follows: trypsin/P was used for digestion; the number of missing digestion sites was set to two; and the minimum length of peptide segments was set to seven amino acids. The maximum number of modifications of the peptide segment was set to five. The mass error tolerance of the primary parent ions of the first search and main search was set to 20 and 5 ppm, respectively. The mass error tolerance was 0.02 Da. Cysteine alkylation was set as a fixed modification. Oxidation on Met was specified as a variable modification. The quantitative method was set to TMT-6plex, and the false discovery rate for protein identification and PSM identification was set to 1%.

Protein annotation

The Gene Ontology (GO) annotation proteome was derived from the UniProt-GOA database (http://www.ebi.ac.uk/GOA/). Subsequently, proteins were classified according to GO annotation on the basis of three categories: biological process, cellular component, and molecular function. The Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used to annotate protein pathways. KEGG online service tools KAAS (http://www.genome.jp/kaas-bin/kaas_main) were used to annotate the KEGG database descriptions of proteins, and then the annotation results were mapped on the KEGG pathway database by using the KEGG online service tool KEGG mapper (http://www.kegg.jp/kegg/mapper.html). For domain annotation, the InterProScan database (http://www.ebi.ac.uk/interpro/) was used to annotate the domain functional descriptions of identified proteins. For subcellular localization, WoLFSPORT (a subcellular localization predication software (http://www.genscript.com/psort/wolf_psort.html) was used. WoLFSPORT is an updated version of PSORT/PSORT II for the prediction of eukaryotic sequences.

Statistical analysis

For GO enrichment and pathway analysis, a two-tailed Fisher’s exact test was used to test the enrichment of the DEPs against all identified proteins. A GO term or KEGG pathway with a P-value < 0.05 was considered significant. The proteins with TMT intensity values were considered quantified, and the minimal PIF was set as 0.75. Statistical analyses were carried out in SPSS ver. 19.0 (SPSS Inc., Chicago, IL, USA). All reported values represent the averages of three replicates with the standard deviation (mean ± SD).

Quantitative real-time PCR validation

Total RNA was extracted using a Ultrapure RNA Kit according to the manufacturer’s protocol (Code: CW0597, CWBIO, Beijing, China). First-strand cDNA synthesis was carried out using a SuperScript™ IV First-Strand Synthesis System according to the manufacturer’s protocol (Code:18091050; ThermoFisher, Waltham, MA, USA). QRT-PCR was performed on ABI 7500 Real-Time PCR System (Roche, Basel, Switzerland) using UltraSYBR Mixture(High ROX)Kit (Code: CW2602, CWBIO, Beijing, China) with the primers listed in Table S1. The CaACTIN (Capana12g001934) was used as an internal standard to calculate relative fold-differences based on comparative cycle threshold (2^−ΔΔCt) values.

Primary MS data and quantitative proteome analysis

A total of 252,213 secondary spectra were obtained by MS analysis of the mock treated and Se treated pepper seedlings. A Pearson correlation coefficient analysis indicated high replicability of the experiment (Fig. 1A). A search of the proteome C. annuum database indicated 47,316 spectra were available, and the utilization rate of the spectra was 18.8%. A total of 24,048 peptide fragments were identified by spectral analysis, including 21,704 unique peptide fragments (Fig. 1B). Most of the peptides contained 7–20 amino acids, results consistent with the trypsin enzymatic hydrolysis and high-energy collision dissociation fragmentation (Fig. 1C). The molecular weights of the proteins were negatively correlated with their coverage (Fig. 1D). The first-order mass error of most spectra was less than 10 ppm, in agreement with the high accuracy of orbital well MS. The results indicated high mass accuracy of the MS data (Fig. 1E). Principal component analysis of the quantitative protein data for all samples is presented in Fig. 1F. Detailed information on the identified peptides, including amino acid sequences, protein descriptions, carried charge of peptide, The maximal posterior error probability for peptides (PEP) is listed in Table S2.

A total of 4,693 proteins were identified, among which 3,938 were quantifiable. To understand the functions and characteristics of the proteins identified and quantified in the data, we performed detailed annotation analysis on the basis of GO, protein domain, KEGG pathway, KOG functional classification (eukaryotic orthologous groups), and subcellular localization (Table S3).

Effects of Se treatment on the global proteome of pepper seedlings

A total of 200 DEPs were identified with a fold-difference expression threshold of 1.5 (Se/Mock ratio ≥ 1.5 or  ≤0.667) and a t-test P-value < 0.05 (Table S4). All identified proteins and DEPs under Se treatment were grouped into three GO categories (biological process, cellular component, and molecular function) (Fig. 2A). In the biological process category, 1,594 identified proteins and 67 DEPs were involved in “metabolic processes,” 1,423 identified proteins and 46 DEPs were involved in “cellular processes,” and 1,107 identified proteins and 46 DEPs were involved in “single-organism processes.” In the molecular function category, 1,959 identified proteins and 74 DEPs had “catalytic activities,” 1,899 identified proteins and 74 DEPs had “binding activities,” and 178 identified proteins and one DEP had “structural activities.” In the cellular components category, 737 identified proteins and 10 DEPs were “cell”-related proteins; 418 identified proteins and three DEPs were “macromolecular complex”-related proteins; and 402 identified proteins and four DEPs were “organelle”-related proteins. The distribution of the GO annotations of the up-regulated and down-regulated DEPs is shown in Fig. S1. We also used WoLFPSORT (http://wolfpsort.org/) software to determine the subcellular location prediction and classification statistics for all identified proteins and DEPs (Fig. 2B), which were grouped according to their subcellular localizations. All identified proteins were classified into 15 subcellular components, including 1,728 chloroplast-localized proteins, 1,283 cytoplasm-localized proteins, and 784 nuclear-localized proteins. For the DEPs, 13 subcellular components were identified, including 72 cytoplasm-localized DEPs, 44 chloroplast-localized DEPs, and 42 nuclear-localized DEPs.

The expression profiles of the DEPs in six samples are presented in a heat map (Fig. 3A). To reveal the changing trends among the six samples, we assigned all DEPs to one of three clusters (I–III) by using MeV (https://sourceforge.net/projects/mev-tm4/) with the K-means method (Fig. 3B). The proteins in clusters I and II showed up-regulation, whereas the proteins in cluster III were down-regulated in the Se stress treatment replicates. Among the DEPs, 172 were up-regulated and 28 were down-regulated (Figs. 3C and 3D). The 16.6 kDa heat shock protein (HSP) (A0A1U8FR15) and two 18.5 kDa class I HSPs (A0A1U8DWR1, A0A1U8E6M6) were up-regulated over ninefold by Se treatment compared with the mock treatment. In addition, a glycine-rich protein (A0A2G2ZTC5) and histone H1 protein were down-regulated more than twofold by Se treatment compared with the mock treatment. A total of 136 DEPs were classified into 20 KOG terms. “Post-translational modification, protein turnover, chaperones” contained the largest DEPs (Fig. 4).

Enrichment analysis of DEPs under Se treatment

To determine whether the DEPs were significantly enriched in some functional types, we performed an enrichment analysis of DEPs  by using GO classification, KEGG pathways, and protein domains. Among the DEPs, most up-regulated proteins were enriched in “sequence-specific DNA binding,” “iron ion binding,” “prephenate dehydrogenase (NAD⁺/NADP⁺) activity,” “nucleic acid binding transcription factor activity,” and “apoplast” (Fig. 5A). For the down-regulated proteins, the top five enriched GO terms were “thiamine-containing compound metabolic process,” “chlorophyll metabolic process,” “porphyrin-containing compound biosynthetic process,” “water-soluble vitamin biosynthetic process,” and “vitamin biosynthetic process” (Fig. 5B).

Under Se treatment, 131 DEPs were grouped into different KEGG pathways, seven of which were enriched (P < 0.05). For the up-regulated proteins, the DEPs were associated with “protein processing in endoplasmic reticulum,” “endocytosis,” “sesquiterpenoid and triterpenoid biosynthesis,” “SNARE interactions in vesicular transport,” and “plant-pathogen interaction” (Fig. 6A). For the down-regulated proteins, the DEPs were associated with “thiamine metabolism” and “porphyrin and chlorophyll metabolism” (Fig. 6B). In the present study, 32 DEPs in the Se treated pepper were identified to be involved in nine metabolic pathways, most of which were significantly up-regulated by Se treatment. However, in the “thiamine metabolism” pathway, four proteins (A0A2G3ALC4, A0A1U8FC73, A0A2G3A131, and A0A2G2Z8I0) were significantly down-regulated by Se treatment (Table 1).

Protein domain enrichment analysis revealed that 25 protein domains were enriched in the DEPs (Fig. 7A). The five most enriched protein domains were “alpha crystallin/Hsp20 domain” (13 proteins), “HSP20-like chaperone” (14 proteins), “target SNARE coiled-coil homology domain” (four proteins), “HSP 70 kDa” (five proteins), and “DnaJ domain” (six proteins). Many HSPs were identified according to the results of the protein domain enrichment analysis. In total, 23 HSPs were significantly up-regulated (Fig. 7B). The expression levels of some HSP genes were basically consistent with the proteomic analyses (Fig. S2).