Revealing the Specific Regulations of Brassinolide on Tomato Fruit Chilling Injury by Integrated Multi-Omics

Tomato fruit is susceptible to chilling injury (CI) when stored at low temperatures, limiting its storage potential, and resulting in economic loss if inappropriate temperatures are used. Brassinolide (BR) is a plant growth regulator that is known to decrease the susceptibility of fruit to CI. In this study, transcriptome, metabolome, and proteome analysis revealed the regulation mechanism of BR treatment in alleviating tomato fruit CI. The results showed that the differentially expressed metabolites mainly included amino acids, organic acids, carbohydrates, and lipids. Differentially expressed genes (DEGs) were involved in plant cold stress response (HSFA3, SHSP, and TPR), fruit redox process (POD, PAL, and LOX), related to the fruit texture (CESA, β-Gal, and PAE), plant hormone signal transduction (ACS3, ARF, and ERF,), transcription factors (TCP, bHLH, GATA). Moreover, differentially expressed proteins were associated with fruit texture (CESA, PE, PL, and CHI), plant oxidation processes (LOX, GPX, CAT, and POD), plant cold stress response (HSF, HSP20, HSP70, and HSP90B), plant hormone signal transduction (BSK1 and JAR1) and transcription factors (WRKY and MYB). Our study showed that BR alleviates CI symptoms of tomato fruit by regulating LOX in the α-linolenic acid metabolism pathway, enhancing jasmonic acid-CoA (JA-CoA) synthesis, inhibiting cell wall and membrane lipid damage. The results provided a theoretical basis for further study on the CI mechanism of tomato fruit.


INTRODUCTION
The tomato is a highly valuable vegetable that is widely cultivated in China and around the world (1,2). Tomato (Solanum lycopersicum) fruits play a vital role in terms of both their nutritional value and their economic value (3). In order to delay the ripening process and extend the shelf life of harvested fruits, a non-freezing low-temperature treatment called cold storage or low-temperature storage is often used (4). Low-temperature storage is one of the most effective technologies for controlling the quality of fruits and vegetables after harvest (5). However, as a subtropical crop, tomato fruit, like other tropical and subtropical fruits, is sensitive to low temperatures and susceptible to chilling injury (CI) and has been widely used as a model for studying cold-induced responses (6,7).
Chilling injury is a type of physiological defect that frequently occurs in the tissues of susceptible plants when they are exposed to low temperatures but not frozen (8,9). The tomato, as a subtropical fruit, is susceptible to cold stress (7). CI manifests as pitting and lesions in the fruit, as well as irregular ripening (10). The severity of CI in tomato fruit is determined by internal factors, such as the cultivar and growing conditions, and external factors, such as the storage temperature and time (11). However, CI usually occurs when the fruit is stored for 6-8 days at temperatures of 5 • C or less (12). Treatments with jasmonic acid, gibberellin, and oxalic acid, as well as atmosphere and humidity packaging, have been reported to reduce the development in the CI of tomato fruit (13)(14)(15)(16).
Brassinolide (BR), a polyhydroxyl steroid, is a naturally occurring plant-growth regulator (17,18) that affects a range of physiological processes in plants (19)(20)(21). BR is considered to be the most widely used plant hormone in post-harvest storage and the preservation of fruits and vegetables (22,23). Several studies have shown that treatments with BR can delay the development of CI symptoms in cucumbers, green peppers, and bamboo shoots (17,24,25). BR can also accelerate the change in tomato fruit from the mature green to full red stages and can increase the contents of ascorbic acid and lycopene (19).
In this study, the regulatory role of BR treatment in alleviating tomato fruit CI was studied through combined metabolomic, transcriptomic, and proteomic analysis, in order to provide a new perspective on studying the molecular regulatory mechanisms for alleviating tomato fruit CI.

Sample Preparation
Tomato fruit was acquired at the green ripe stage from a glasshouse in Beijing Yanqing Dongxiaoying village (Yanqing District, Beijing, China) and transported directly to a laboratory. Roughly 200 tomato fruits of similar sizes and ripeness, excluding those exhibiting wounds or decay, were separated evenly into two portions. One group was submerged in 0.75 µmol/L BR solution for 10 min, and the other, in pure water for 10 min as a control. After air drying, both the control and treated groups were stored in baskets at 0 ± 1 • C with 85-90% relative humidity. After cold storage, CI symptoms appeared in both groups; however, the CI symptoms were less apparent in the BR-treated group. During storage at 0 ± 1 • C, the pericarp of the tomato fruit was sampled at 0 and 8 days. The samples at 0 d were set as the control group (CK). Three replicate samples of three fruits were collected, with pericarp tissues taken and frozen in liquefied nitrogen. The tissues were kept at −80 • C for subsequent analysis.

CI Index
The CI index was determined as described in a previously published article (16).

RNA Extraction and Detection and Differential Expression Analysis
RNA degradation and contamination were monitored on 1% agarose gels. The RNA purity was checked using a NanoPhotometer R spectrophotometer (IMPLEN, CA, USA). The RNA integrity was assessed using the 2100 Bioanalyzer RNA Nano 6000 Assay Kit (Agilent Technologies, CA, USA). Differential expression analysis was based on former studies (26)(27)(28).

GO and KEGG Enrichment Analysis
Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses of differentially expressed genes were performed using the clusterProfiler (3.4.4) software (29,30).

Extraction and Identification of Metabolites
For each replicate, 100 mg of tissue was powdered in liquid nitrogen and placed in a fresh Eppendorf tube. Next, 500 u-L of 80% methanol (containing 0.1% formic acid), was added to the tube, and the mixture was vortexed, put on ice, and left to stand for 5 min. The tubes were centrifuged at 15,000 rpm for 10 min at 4 • C, and the supernatant liquid was diluted to the final concentration (including 53% methanol) with liquid chromatography-mass spectrometry (LC-MS) grade water and then centrifuged at 15,000 × g at 4 • C for 20 min. The supernatant liquid was used for analysis by LC-MS.
In accordance with the Novogene database, multi-reaction monitoring mode (MRM) was used to analyze the experimental samples. The compounds were quantified, and a qualitative analysis was conducted. SCIEX OS version 1.4 (SCIEX, USA) was used to open the off-machine mass spectrum file, and the chromatographic peaks were integrated and corrected according to a set minimum peak height of 500, a signal-to-noise ratio of 5, smooth points of 1, and other information. The chromatographic peaks were screened. Finally, the integral data of all the chromatographic peaks were derived to obtain the qualitative and quantitative metabolite results.

Extraction, Identification, and Functional Analysis of Protein
The method for the extraction of protein from the tomato peel was based on previous research (31)(32)(33)(34)(35). The Proteome Discoverer 2.4 software (Guangzhou guangdian International Trade Co., LTD, Guangdong, China) was used to search and identify the differential proteins, and a t-test was used to analyze the quantitative protein results (36). Interproscan was used for GO functional annotation, and KEGG was used for the functional protein pathway analysis of the identified proteins (37,38).

Statistical Analysis
A one-way ANOVA was used for data comparison, and the least significant difference (LSD) test at P = 0.05 was used for the comparison of means. The figures were drawn using Origin 8.0 (Microcal, USA).

Effect of BR Treatment on CI Index of Tomato Fruit
After 8 days of storage at 0 • C, tomato fruit in both the BR-treated group and the CK group showed symptoms of CI, while the CI index of the fruits in the BR-treated group was 15.6% lower than that of the CK group. The CI indices of the CK group and BR-treated group were 25 and 9.4%, respectively (Figure 1).

Analysis of Metabolites Differentially Expressed in Response to Chilling in Tomato Fruit
For CK 8 vs. 0 day, a total of 121 differential metabolites were found to be affected by CI, with 63 downregulated and 58 upregulated. A cluster analysis of the differentially expressed metabolites s revealed different expression patterns between CK 8 and 0 day (Figure 2A). These differentially accumulated metabolites represented the main components in the ripening tomato fruit and were divided into 11 categories, the largest being nucleotides, and amino acids and their derivatives (∼22 and 17%, respectively), followed by carbohydrates (16%), organic acid and its derivatives (10%), lipids and lipid-like molecules (9%), flavonoids (6%), alcohols and polyols (4%), phenolamides (4%), phenols (4%), alkaloids (4%), and flavanone (4%) ( Figure 2B). KEGG pathway analysis showed that the main changes in response to CI involved metabolites such as sphingolipid and inositol phosphate metabolism, cutin and suberine wax, fatty acid, monobactam biosynthesis, and vancomycin resistance ( Figure 2C, Appendix 4).
Through detailed analysis, many differential metabolites were found that were regulated to different degrees during CI. For instance, erucic acid, 1-linoleoyl-rac-glycerol, sarsasapogenin, and D-3-phosphoglyceric acid were downregulated, while notoginsenoside R2 was upregulated in lipids and lipidlike molecules. Moreover, 2, 3-dihydroxybenzoic acid, 2hydroxyisocaproic acid, lactic acid, and phosphoric acid were downregulated, while pyrrole-2-carboxylic acid was upregulated, in organic acid and its derivatives. Additionally, inositol and D-tagatose were downregulated, while D-galacturonic acid, beta-D-lactose, N-acetyl-D-glucosamine, and sucrose were upregulated, in carbohydrates. Furthermore, tiliroside, lonicerin, and luteolin 7-O-glucoside were greatly downregulated, while methylophiopogonanone A was upregulated, and in flavonoids. Finally, inosine diphosphate and uridine 5 ′ -diphosphate were downregulated, while uracil, cytidine, adenine, adenosine, and uridine were upregulated in nucleotides and their derivatives ( Table 1, Appendix 1). In order to determine correlations between the differential metabolites, a correlation analysis was performed and a relationship between the different metabolites was found ( Figure 2D).

Effects of BR on the Differential Accumulation of Metabolites in Chilled Tomato Fruit
The joint analysis results for the two comparison groups (CK 8 vs. 0 day and BR 8 vs. CK 8 days) indicated that 43 differential accumulations of metabolites were regulated by BR to different degrees during the process of CI in tomato fruit. These included homocysteine, 1-methylhistidine, D-methionine and N, N-dimethylglycine in amino acid and its derivatives; uridine 5 ′diphosphate, inosine diphosphate, and l-dihydroorotic acid in nucleotides and its derivates; tiliroside and luteolin 7-o-glucoside in flavonoids; D-galacturonic acid in carbohydrates; pyridoxine in vitamins; 2-hydroxyIsocaproic acid, 2,3-dihydroxybenzoic acid, 3-hydroxy-3-methyl butyric acid and lactic acid in organic acid and its derivatives; notoginsenoside R2, erucic acid, 1linoleoyl-rac-glycerol, sarsasapogenin, and aucubin in lipids and lipid-like molecules (Figure 3).

Regulation of DEGs in BR in Tomato Fruit CI
The further screening and analysis of the two comparison groups (CK 8 vs. 0 day and BR 8 days vs. CK 0 day) revealed that numerous important genes related to the process of CI in tomato fruit were affected by BR regulation. The DEGs that were discovered involved genes that plant cold stress, including expansin (EXP), the heat shock protein Dnaj with tetratricopeptide repeat-containing protein (TPR), heat stress transcription factor A3 (HSFA3), MADS-box protein 5, small heat shock protein (SHSP) and zinc finger protein constans-like 5; peroxidase (POD), and phenylalanine ammonia-lyase (PAL), is related to the redox process of tomato fruit CI. Many enzymes were found that were regulated by BR and had effects on the flavor and texture of tomato fruit, including beta-galactosidase (β-GAL), cellulose synthase (CESA), cysteine synthase (CYS), pectin acetylesterase (PAE), sucrose synthase (SUS), xyloglucan endotransglucosylase/hydrolase (XTH), zeaxanthin epoxidase (ZEP), lipoxygenase (LOX), and putative alcohol dehydrogenase (ADH). Some DEGs involved in plant hormone and signal transduction were found to be regulated by BR during the CI process, such as 1-aminocyclopropane-1-carboxylate synthase 3 (ACS3), ABC transporter B family member 1(ABCB1), auxin response factor (ARF), auxin transport protein BIG, auxinregulated protein, EIN3-Binding F-box protein 1(EBF1), ethylene receptor (ETR), and ethylene-responsive factor (ERF). Several transcription factors, including GATA, MADS, TCP, bHLH3, and guided entry of tail-anchored proteins factor 4 (GTE4), were found to be involved in the regulation of BR in the process of CI in tomato fruit, and brassinolide LRR receptor kinase (BRI) was also found to be involved in the regulatory process of CI ( Figure 5). Interestingly, it was discovered that BR-regulated DEGs were involved in plant hormone signal transduction, mitogen-activated protein kinase (MAPK) signaling pathways, the biosynthesis of amino acids, phenylpropanoid biosynthesis, carotenoid biosynthesis, and starch and sucrose metabolism during CI.

Combined Transcriptome and Metabolome Analysis
In the fatty acid biosynthesis pathway, it was found that the expression of palmitoyl-acyl carrier protein thioesterase in tomato fruit affected by CI was upregulated, resulting in the annotation of the metabolite hexadecanoic acid, affecting the accumulation of long chain-Acyl-CoA synthetase 2. In turn, phosphoethanolamine N-methyltransferase in the glycerophospholipid metabolism pathway was regulated to varying degrees, affecting the accumulation of choline phosphate. Additionally, changes in the expression of palmitoylprotein thioesterase in fatty acid elongation led to the annotation of hexadecanoate, which affects the fatty acid degradation pathway. It was also found that short-chain dehydrogenasereductase (SDR) was downregulated in carotenoid biosynthesis, which affected the accumulation of abscisic aldehyde, resulting in abscisic acid 8'-hydroxylase being downregulated. The regulation of PYR/PYL, PP2C, and ABF by abscisic acid in the plant hormone signal transduction pathway was affected. Finally, the interaction between the DNA and stomatal closure seed dormancy was terminated (Figure 6).

Differential Protein Analysis in Tomato Fruit CI
The proteome was quantitatively analyzed to determine protein amounts in fruit with and without CI. A total of 7,865 proteins were identified in the tomato fruit protein samples. Among the 7,865 proteins, 308 were found to be differentially expressed between CI and non-CI, of which 153 were downregulated and 155 were upregulated. The KEGG pathway enrichment analysis showed that the differentially expressed proteins were majorly enriched in four pathways: phagosomes, fatty acid biosynthesis, phenylalanine metabolism, and amino sugar and nucleotide sugar metabolism (Appendix 6). A GO functional enrichment analysis was performed to elucidate the molecular functions and biological processes involved in differentially expressed proteins between chilling and non-chilling tomato fruit (Appendix 6).
According to the enrichment degrees, differentially expressed proteins were mainly involved in six groups of CI biological processes, including the cellular protein modification process (15 proteins), response to stimulus (11 proteins), the carbohydrate metabolic process (10 proteins), the organic acid metabolic process (8 proteins), the cellular amino acid metabolic process (5 proteins), and the alpha-amino acid metabolic process (4 proteins).

Combined Transcriptomic, Metabolites, and Proteomics Analysis
A combination of transcriptomic and proteomic analyses revealed many different substances that were produced by the interaction between genes, metabolites, and proteins. The aldehyde dehydrogenase (ADH) of DEGs and the alcohol dehydrogenase 1/7 of differentially expressed proteins appeared in the fatty acid degradation pathway and the glycolysis/gluconeogenesis pathway, and they acted together to produce acetaldehyde and aldehyde ( Figure 8A). In alanine, aspartate, and the glutamate metabolism pathway, three DEGs were found to be up-regulated [omega amidase, glutamate dehydrogenase (GDH), and glutamine-fructose 6-phosphate transaminase (GFPT)] and glutamate synthase (GS) was downregulated, while differentially expressed proteins glutamine  Figure 8B).

DISCUSSION
Tomato fruit stored at 0 • C is susceptible to CI, which reduces the nutritional quality of the fruit and prevents its normal change of color from affecting the ripening process (39,40). BR applied to the fruit prior to storage can alleviate CI in several species (green bell pepper, bamboo shoot, rice) of fruit during cold storage (17,24,41). In our study, compared with untreated tomato fruit, the BR-treated fruit showed fewer symptoms of the chilling response. These findings suggest that BR is useful for improving the cold tolerance of tomato fruit. A color change is one of the most important indicators for judging the ripening of tomato fruit (42). Lycopene and β-carotene in the pathway of carotenoids are important components of ripe tomato fruit (43). The degradation of chlorophyll, along with the synthesis of lycopene and β-carotene, causes the fruit to change color from green to orange and eventually to red (44). Studies have shown that phytoene synthase (PSY), encoding a key enzyme in the carotenoid biosynthesis pathway, is induced in expression during the fruit-ripening process (45,46). Our results showed that PSY was also upregulated throughout the storage period in tomato fruit, despite CI, which was consistent with previous reports.  Moreover, PSY was upregulated to a greater extent in the BRtreated fruit than in the control fruit, indicating that BR could indeed inhibit the CI process of tomato fruit. However, in our results, no significant changes in the genes encoding chlorophyll, lycopene, or beta-carotene were found. When the fruit was transferred from 0 • C to room temperature, it was not able to change from green to red normally and accelerated decayed. Therefore, we hypothesized that the genes encoding lycopene and β-carotene in the tomato fruit that was subjected to lowtemperature stress were inhibited, resulting in the abnormal color transformation.
β-Galactosidase and pectin methylesterase (PME) in polysaccharides together form the complex structure of plant cell wall and participates in cell wall modification which is the key enzyme in cell wall degradation (47)(48)(49)(50)(51). The synergistic effects of various cell-wall-modification enzymes promote the degradation of cell-wall polysaccharides, among which pectin degradation is the main reason for fruit softening (52). CEL is involved in fruit ripening (53). Endoglucanase (EG) is one of the most important components of CEL due to its ability to degrade the cellulose matrix (50). Studies have found that the activities of PME and β-GAL increase with the ripening of tomato fruit, along with a decrease in fruit hardness (54)(55)(56)(57)(58)(59). However, we found that the genes encoding β-GAL, EG, pectin lyase (PL), and CESA were downregulated in tomato fruit is subjected to low-temperature stress and CI. This was not consistent with previous studies and may be caused by the dysfunction of the internal organism of the tomato fruit, which no longer had the ability to ripen normally. However, the genes encoding β-GAL and PL in BR-treated tomato fruit were upregulated. The results showed that BR can alleviate the symptoms in tomato fruit CI.
During fruit ripening, the fatty-acid pathway and the aminoacid pathway play major roles in the synthesis of the fruit's aromatic volatiles (60). Studies have shown that the expression of lipid compounds depends on the expression of branched-chain amino acid transaminase (61). Amino-acid metabolism is also thought to be closely related to the cold-resistance mechanism of fruit (62,63). It is well-known that LOX, ADH, and propylene oxide cyclase are important enzymes in the fatty-acid metabolism pathway that affect the synthesis of fruit-flavor substances (40). It has been reported that low temperatures can inhibit the activity of ethanol dehydrogenase (EDH) (7,64). Our results were consistent with this; the gene encoding EDH was downregulated during CI. Linolenic acid is processed by 9-lipoxygenase (9-LOX) to form 9-hydrogen peroxide, which, in turn, is processed by ADH to produce aromatic volatile substances (65). When linoleic acid is acted upon by LOX, the phospholipid bilayer is destroyed, reducing the cold resistance of the fruit (66,67). In our study, BR increased the expression of the genes encoding 9-LOX and EDH, so the treatment of the fruit with BR was an effective way to maintain the fatty-acid biosynthesis in the FIGURE 9 | Changes of genes and related enzymes involved in the CI process of tomato fruit under the regulation of BR. Among them, black fonts indicate differentially expressed metabolites, red fonts indicate that DEGs were upregulated, blue fonts indicate that DEGs were downregulated, and green fonts indicate that DEGs are both upregulated and downregulated and black fonts indicate those differential metabolites.
formation of aroma substances and effectively alleviate the CI process. Studies have shown that the accumulation of proline is helpful in alleviating CI in fruits, while proline dehydrogenase (ProDH) negatively affects the accumulation of proline (24,(68)(69)(70). Our results showed that ProDH was inhibited by BR, which increased the accumulation of proline and increased the cold resistance of the fruit, consistent with previous reports. A putative regulatory network model involving CI in response to BR treatment in the α-linolenic acid metabolism pathway was proposed (Figure 9). In this pathway, phosphatidylcholine can first produce α-linolacid through a series of biosyntheses, and, finally, produce methyl jasmonate through the action of certain enzymes and CoA, relieving the symptoms of CI in tomato fruit.

DATA AVAILABILITY STATEMENT
The data presented in the study are deposited in the ProteomeXchange Consortium repository, accession number PXD028500.

AUTHOR CONTRIBUTIONS
CB and DM wrote the manuscript. QW and JZ offers experimental ideas. YZ analyzed the data. ZY revised the manuscript. AF, LM, HG, SY, SZ, and LG sorted out the pictures and attachments in the manuscript. All authors contributed to the article and approved the submitted version.