Phenotypic changes in GE and JS under cold treatment
After 12 days of cold storage, we observed several phenotypic changes in GE, but not in JS. After 24 days, GE fruits were suffering more severe chilling injuries including rotting and bacterial infections, while JS fruits retained a higher quality with fewer injuries (Fig. 1a). Cold stress resulted in increases in relative electrolyte leakage (REL) in both cultivars during 12 days of storage, which was higher in GE than in JS, but the RELs of both cultivars decreased after 24 days of storage (Fig. 1b). Similar tendencies were also identified among the changes in malondialdeobhyde (MDA) and H2O2 levels (Fig. 1b). Overall, our analysis confirmed that cold-sensitive GE suffered a more severe cold damage, with higher levels of electrolyte leakage, lipid peroxidation, and H2O2 than JS during the early phase of cold storage.
Identification and quantitation of DEPs by iTRAQ
We compared protein levels in JS and GE before and after the cold treatment to identify differentially expressed proteins (DEPs). Using iTRAQ labeling LC-MS/MS analysis, 5,450 proteins were specifically identified from 107,101 LC-MS/MS spectra and 30,927 peptides in GE, and 5,291 proteins were identified from 107,048 LC-MS/MS spectra and 29,829 peptides in JS (Additional file 1: Table S1).
We used ratio fold changes of >1.200 or <0.833 in expression (P <0.05) as the cut-off points for upregulated and downregulated proteins, respectively, and found a total of 807 DEPs (12 days: 360; 24 days: 447) in GE, and 722 DEPs (12 days: 391; 24 days: 331) in JS, after cold treatment (Additional file 2: Table S2). After 12 days of cold treatment, we identified 360 DEPs in GE, 142 of which were up-regulated and 218 of which were down-regulated (Fig. 2a). In JS, there were 391 DEPs, of which 251 were up-regulated and 140 were down-regulated (Fig. 2a). After 24 days, there were 447 DEPs in GE, 160 of which were up-regulated and 287 of which were down-regulated (Fig. 2a). Similarly, we found 331 DEPs in JS after 24 days, of which 138 were up-regulated and 193 were down-regulated (Fig. 2a). A higher number of DEPs were identified in JS than in GE at the early phase of cold treatment, but during the later phase of treatment, the number of DEPs in GE increased (Fig. 2a).
According to these observations, we assumed that the cold-tolerant cultivar JS responded faster than GE in terms of expressing cold-responsive proteins. Rapid up-regulation of proteins that regulate the response to the chilling stress and protect the plant cells from damage induced by ROS is important for cold tolerance. Therefore, the delayed cold response in GE may be a critical reason for the severe chilling injury [17]. Meanwhile, among the 807 DEPs in GE, 181 (22.42%) and 268 (33.20%) were specifically identified at the 12 and 24 day time points, respectively. A further 179 (22.18%) DEPs were shared by both time points (Fig. 2b). However, in JS, only 114 out of 722 DEPs (15.79%) were common to both time points, while 277 (38.37%) and 217 (30.06%) were specifically identified after 12 or 24 days, respectively (Fig. 2b). These findings demonstrate that much more different groups of proteins were mobilized in cold-tolerant JS under low temperature treatment [12].
In order to identify the proteins that are most likely to be related to the acquisition of cold tolerance in cantaloupe, a careful analysis of common expressed DEPs was carried out. The selected proteins with differential expression patterns were commonly expressed during the whole treatment period in both cultivars. A hierarchical cluster analysis (HCL) was performed to analyze the correlations of common expressed DEPs in GE and JS after cold treatment. Notably, the changes among 24 common DEPs were statistically significant and their abundance can be illustrated as seven clusters, revealing that two cultivars mobilized numerous proteins and differentially regulated their abundance to cope with cold stress (Fig. 2c; Additional file 2: Table S2). Furthermore, the principal component analysis (PCA) we performed on the expression data above indicated that, in all conditions, the two cultivars presented different protein expression patterns. The changes in protein expression between JS chilled at 12 and 24 days were smaller than those observed in GE, clustering close together with little separation in either axes (Fig. 2d; Additional file 2: Table S2). In contrast, the changes in protein expression in GE between 12 and 24 days clustered away from each other. Consequently, we speculated that, compared with JS, a longer duration of cold stress had a greater impact on the expression of proteins in GE, which may explain a more severe damage to GE during cold treatment [18].
Primary functional classification of DEPs
From the Clusters of Orthologous Groups (COG) database, we found that the largest group of DEPs are involved in ‘posttranslational modification, protein turnover, chaperones’ (119 DEPs), followed by ‘general function prediction only’ (94 DEPs) and ‘translation, ribosomal structure and biogenesis’ (83 DEPs; Fig. 3a; Additional files 3: Table S3). The further analysis will be discussed below.
DEPs in response to the early phase of cold stress
Using gene ontology (GO) analysis, the DEPs were classified into three categories: cellular components (CC), molecular function (MF) and biological processes (BP) [19]. During the early phase of cold stress in both cultivars, the most highly represented categories were ‘cell’, ‘cell part’, ‘intracellular’, ‘intracellular part’ and ‘cytoplasm’ in CC (Fig. 3b); ‘catalytic activity’, ‘binding’ and ‘heterocyclic compound binding’ in MF (Fig. 3b); and ‘metabolic process’, ‘organic substance metabolic process’ and ‘cellular process’ in BP (Fig. 3b). The results indicated that the majority of DEPs were involved in metabolic processes, cellular processes, cell and catalytic activities, suggesting the cold treatment mainly affected physiological metabolism and cell differentiation in cantaloupe (Additional files 4: Table S4). More intriguingly, all three categories of proteins were expressed at higher levels in JS compared with GE, revealing that, the proteome of cold-tolerant JS responds more rapidly to cold stress than that of cold-sensitive GE.
To further identify the roles of the DEPs, we performed KEGG pathway analysis. Only significantly enriched categories with P-values < 0.05 were selected. We found that cold stress affected ribosome, phagosome and phenylpropanoid biosynthesis in both cultivars. Proteins involved with protein processing in the endoplasmic reticulum, plant-pathogen interactions and photosynthesis-antennae were highly enriched in GE compared with JS. On the other hand, proteins involved in photosynthesis, galactose metabolism and cyanoamino acid metabolism, were considerably enriched in JS (Table 1; Fig. 4a, c; Additional files 6: Table S5). More interestingly, there were conspicuous protein-protein interactions among ribosome and other functions (P < 0.05) (Fig. 5a, c). Thus we speculate that ribosome related DEPs may play a significant role in regulating the metabolic mechanisms in cantaloupe at the early phases of cold stress.
DEPs in response to the later phase of cold stress
As above, during the later phase of the cold stress, proteins were characterized by ‘cell’, ‘cell part’, ‘intracellular’, ‘intracellular part’ and ‘cytoplasm’ in CC (Fig. 3c); ‘catalytic activity’, ‘binding’ and ‘heterocyclic compound binding’ in MF (Fig. 3c); and ‘metabolic process’, ‘cellular process’ and ‘organic substance metabolic process’ in BP (Fig. 3c). After 12 days of cold storage, all three categories were dramatically higher in GE compared with JS, indicating that GE experienced greater levels of cold stress as the length of storage increased (Additional files 4: Table S4). The tardiness of the cold response in GE may be a critical reason for its severe cold damage.
KEGG pathway analysis indicated that protein processing in the endoplasmic reticulum and galactose metabolism may be affected in both cultivars after cold stress. Proteins involved in ribosomes, carbon fixation in photosynthetic organisms, plant-pathogen interactions, one carbon pool by folate, and phenylpropanoid biosynthesis were highly enriched in GE while proteins involved in photosynthesis-antennae, phagosomes, amino sugar and nucleotide sugar metabolism, fructose and mannose metabolism, pentose and glucuronate interconversions, and linoleic acid metabolism were enriched in JS (Table 1; Fig. 4b, d; Additional files 5: Table S5). Like the hallmark of the protein-protein interaction at 12 days, there was still a significant interaction among ribosome and other functions in GE, while dramatic changes happened in JS (Fig. 5b, d). This further demonstrates that various proteins were mobilized in JS during the cold treatment and that there was a positive relationship between the diversity of proteins and cold tolerance.
Functional distribution analysis of cold induced proteins in JS
Based on GO analysis, functional distribution analysis were performed and the DEPs identified in cold-tolerant JS after 12 days cold treatment were selected as the cold induced proteins [20] (Additional files 4: Table S4; Additional files 6: Fig. S1). In terms of cellular components, membrane, cell part, cell, and organelle proteins were significantly enriched in JS (P <0.05). In the molecular function category, proteins with catalytic activity and binding were the most positively regulated (P <0.05). In biological processes, proteins involved in cellular processes, metabolic processes and organic substance metabolic processes were the most highly enriched (P <0.05).
Analysis of metabolites in response to cold stress
A limit of a 1.000-fold change coupled with a Student’s t-test (P <0.05) was used to identify the differentially expressed metabolites. Metabolic data indicated that amino acids, such as proline, 3-hydroxy-L-proline 3 and 3-cyanoalanine, accumulated to higher levels during the whole period of cold storage in JS compared with GE (Table 2; Additional files 7: Table S6).
Validation of DEPs by analysis of gene expression
To validate the expression patterns of proteins obtained from the iTRAQ analysis, we randomly selected nine of the corresponding genes for q-PCR analysis using specific primers (Fig. 6; Additional files 8: Table S7). The results indicated that the expression patterns of five of the nine genes (55.56%) in GE and six of the nine genes (66.67%) in JS were consistent with the iTRAQ data, suggesting that these independent evaluations revealed a reliability of the iTRAQ data [21].
Putative candidate proteins for cold tolerance in cantaloupe
In an attempt to identify proteins that may be involved in cold tolerance mechanisms in cantaloupe, we selected 258 and 247 proteins that showed differential expression among two cultivars. The selected candidate proteins were grouped according to the above phenotypic analysis and bioinformatics analysis as follows: carbohydrate and energy metabolism, stress responses, structural proteins, amino acid metabolism and signal transduction (Additional files 9: Table S8). Their expression patterns and possible roles will be discussed below.