Identification of Key Genes during Ethylene-Induced Adventitious Root Development in Cucumber (Cucumis sativus L.)

Ethylene (ETH), as a key plant hormone, plays critical roles in various processes of plant growth and development. ETH has been reported to induce adventitious rooting. Moreover, our previous studies have shown that exogenous ETH may induce plant adventitious root development in cucumber (Cucumis sativus L.). However, the key genes involved in this process are still unclear. To explore the key genes in ETH-induced adventitious root development, we employed a transcriptome technique and revealed 1415 differentially expressed genes (DEGs), with 687 DEGs up-regulated and 728 DEGs down-regulated. Using Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, we further identified critical pathways that were involved in ETH-induced adventitious root development, including carbon metabolism (starch and sucrose metabolism, glycolysis/gluconeogenesis, citrate cycle (TCA cycle), oxidative phosphorylation, fatty acid biosynthesis, and fatty acid degradation), secondary metabolism (phenylalanine metabolism and flavonoid biosynthesis) and plant hormone signal transduction. In carbon metabolism, ETH reduced the content of sucrose, glucose, starch, the activity of sucrose synthase (SS), sucrose–phosphate synthase (SPS) and hexokinase (HK), and the expressions of CsHK2, pyruvate kinase 2 (CsPK2), and CsCYP86A1, whereas it enhanced the expressions of β-amylase 1 (CsBAM1) and β-amylase 3 (CsBAM3). In secondary metabolism, the transcript levels of phenylalanine ammonia-lyase (CsPAL) and flavonoid 3′-monooxygenase (CsF3′M) were negatively regulated, and that of primary-amine oxidase (CsPAO) was positively regulated by ETH. Additionally, the indole-3-acetic acid (IAA) content and the expressions of auxin and ETH signaling transduction-related genes (auxin transporter-like protein 5 (CsLAX5), CsGH3.17, CsSUAR50, and CsERS) were suppressed, whereas the abscisic acid (ABA) content and the expressions of ABA and BR signaling transduction-related genes (CsPYL1, CsPYL5, CsPYL8, BRI1-associated kinase 1 (CsBAK1), and CsXTH3) were promoted by ETH. Furthermore, the mRNA levels of these genes were confirmed by real-time PCR (RT-qPCR). These results indicate that genes related to carbon metabolism, secondary metabolite biosynthesis, and plant hormone signaling transduction are involved in ETH-induced adventitious root development. This work identified the key pathways and genes in ETH-induced adventitious rooting in cucumber, which may provide new insights into ETH-induced adventitious root development and will be useful for investigating the molecular roles of key genes in this process in further studies.


Introduction
An adventitious root is a kind of lateral root that often arises from leafstalk, stems, and some non-pericycle tissue [1,2]. The developmental process of an adventitious root is complicated and can be divided into three successive phases: induction, formation, and expression [3]. Adventitious rooting represents the regenerative ability of plants and these roots can obtain water and nutrients from the soil, fix plants on the substrate, and
In order to verify the reliability of the transcriptome data, we measured the sucrose, glucose, and starch contents, relevant enzyme activity, and gene expression during ETHinduced adventitious rooting in cucumber. As shown in Figure 2B, compared with the control, ETH treatment significantly decreased the sucrose, glucose, and starch contents in cucumber explants. Moreover, the activities of SS, SPS, and HK with ETH treatment were significantly lower than those in the control ( Figure 2C). Additionally, the expressions of six DEGs were determined by quantitative real-time PCR (RT-qPCR) assays ( Figure 2D). Among them, except for CsGBE1, the expressions of other genes, including Cseg9, Cseg6, Csβ-G11-like, CsHK2, and CsSS5-like were decreased with ETH treatment.
In the TCA cycle, five DEGs were detected. Among them, one PDG and two DRAT genes appeared in the glycolysis pathway ( Figure 3B and Supplementary Table S6). Exogenous ETH down-regulated the expression of ATP citrate synthase (CS; LOC101209539) for citrate synthesis and isocitrate dehydrogenase NADP + (IDH; LOC101202770).
There were four DEGs in the oxidative phosphorylation pathway ( Figure 3C and Supplementary Table S6). Among them, the transcript levels of NADH dehydrogenase [ubiquinone] 1 alpha subcomplex subunit 2 (NDUFA2; LOC 101208633) and NADH dehydrogenase subunit 5 (nad5; LOC11123919) in complex I were enhanced. In complex V, the expression of uncharacterized gene LOC101205555 was down-regulated, whereas the expression of plasma membrane ATPase 4 (PMA4; LOC101213505) was up-regulated by ETH treatment.
Compared with the control, ETH treatment significantly decreased PK activity ( Figure 3D). Moreover, the relative expression of Csnad5 with ETH treatment was remarkably enhanced in comparison with the control, whereas the relative expressions of CsPFP, CsPK2, CsACS2, and

Fatty Acid Biosynthesis and Fatty Acid Degradation
Totals of four and eight DEGs were mapped to the pathways of fatty acid biosynthesis and fatty acid degradation, respectively ( Figure 4A and Supplementary Table S7). In the fatty acid biosynthesis pathway, ETH increased the transcript levels of FabG (LOC101221009) and long-chain-fatty-acid-CoA ligase (ACSL; LOC101207461 and LOC101212158), whereas it downregulated the expression of oleoyl-[acyl-carrier-protein] hydrolase (OLAH; LOC101210955). In the fatty acid degradation pathway, except for ADH and long-chain fatty acid omegamonooxygenase (CYP86A), the other gene expressions were up-regulated by ETH treatment ( Figure 4A).   According to the RT-qPCR analysis, the relative expression of CsACSL6 with ETH treatment had a significant increase compared with the control, whereas the relative expression of CsCYP86A1 in the pathway of fatty acid degradation significantly declined with ETH treatment in comparison with the control ( Figure 4B).

Phenylalanine Metabolism
Seven DEGs were detected in the phenylalanine metabolism ( Figure 5A and Supplementary  Table S8). Among them, the transcript levels of two amidase genes (LOC101214819 and LOC101204792) were up-regulated and down-regulated, respectively. ETH treatment upregulated the relative expression of aspartate transaminase (AST; LOC101217299) and three primary-amine oxidase (PAO) genes (LOC101210998, LOC101213059, and LOC101212818). Moreover, the expression of phenylalanine ammonia-lyase (PAL; LOC101218856) for phenylalanine synthesis was down-regulated by ETH treatment ( Figure 5A).
Seven DEGs were detected in the phenylalanine metabolism ( Figure 5A and Supplementary Table S8). Among them, the transcript levels of two amidase genes (LOC101214819 and LOC101204792) were up-regulated and down-regulated, respectively. ETH treatment up-regulated the relative expression of aspartate transaminase (AST; LOC101217299) and three primary-amine oxidase (PAO) genes (LOC101210998, LOC101213059, and LOC101212818). Moreover, the expression of phenylalanine ammonia-lyase (PAL; LOC101218856) for phenylalanine synthesis was down-regulated by ETH treatment (Figure 5A).
As shown in Figure 5B, the activity of PAL with ETH treatment was significantly lower than that of the control. The relative expression of CsPAL was decreased, whereas the expression of CsPAO was enhanced by ETH treatment ( Figure 5C).   Supplementary Table S8. Black characters with yellow background are genes, whereas black characters with blue background are proteins. The two squares under the names of genes or proteins indicate the change in abundance of the control and ETH. A red square means up-regulation, whereas a blue square means down-regulation. (B) The activity of PAL and PAO enzyme during adventitious root development in cucumber. (C) Quantitative real-time PCR analysis of the expression profiles of 2 genes related to phenylalanine metabolism during adventitious root development in cucumber. Asterisk (*) and double asterisk (**) indicate the significance at p < 0.05 and p < 0.01, respectively. The bar represents the value of standard error. Values represent the mean ± SE of three replicates. Each experiment was repeated three times independently.
As shown in Figure 5B, the activity of PAL with ETH treatment was significantly lower than that of the control. The relative expression of CsPAL was decreased, whereas the expression of CsPAO was enhanced by ETH treatment ( Figure 5C).

Flavonoid Biosynthesis
In flavonoid biosynthesis, except for shikimate O-hydroxycinnamoyltransferase (HCT; LOC101207100), the transcript levels of chalcone isomerase (CHI; LOC101218716) and flavonoid 3 -monooxygenase (F3 M; LOC101207160) were down-regulated by ETH treatment ( Figure 6A and Supplementary Table S9). The relative expression of CsF3 M was decreased by ETH treatment, whereas the expression of CsHCT was significantly increased ( Figure 6B). low background are genes, whereas black characters with blue background are proteins. The two squares under the names of genes or proteins indicate the change in abundance of the control and ETH. A red square means up-regulation, whereas a blue square means down-regulation. (B) The activity of PAL and PAO enzyme during adventitious root development in cucumber. (C) Quantitative real-time PCR analysis of the expression profiles of 2 genes related to phenylalanine metabolism during adventitious root development in cucumber. Asterisk (*) and double asterisk (**) indicate the significance at p < 0.05 and p < 0.01, respectively. The bar represents the value of standard error. Values represent the mean ± SE of three replicates. Each experiment was repeated three times independently.

Flavonoid Biosynthesis
In flavonoid biosynthesis, except for shikimate O-hydroxycinnamoyltransferase (HCT; LOC101207100), the transcript levels of chalcone isomerase (CHI; LOC101218716) and flavonoid 3′-monooxygenase (F3′M; LOC101207160) were down-regulated by ETH treatment ( Figure 6A and Supplementary Table S9). The relative expression of CsF3′M was decreased by ETH treatment, whereas the expression of CsHCT was significantly increased ( Figure 6B).  Supplementary Table S9. Black characters with yellow background are genes. The two squares under the names of genes or proteins indicate the change in abundance of the control and ETH. A red square means up-regulation, whereas a blue square means  Supplementary Table S9. Black characters with yellow background are genes. The two squares under the names of genes or proteins indicate the change in abundance of the control and ETH. A red square means up-regulation, whereas a blue square means down-regulation. (B) The expression profiles of 2 genes related to flavonoid biosynthesis during adventitious root development in cucumber. Asterisk (*) and double asterisk (**) indicate the significance at p < 0.05 and p < 0.01, respectively. The bar represents the value of standard error. Values represent the mean ± SE of three replicates. Each experiment was repeated three times independently.
As shown in Figure 7B, in comparison with the control, ETH treatment significantly decreased the IAA content, whereas it enhanced the ABA content. In the auxin signaling transduction pathway, the relative expressions of auxin transporter-like protein 5 (CsLAX5), CsGH3.17, and CsSUAR50 were significantly decreased by ETH treatment ( Figure 7C). In ETH signal transduction, the expressions of CsERS and CsERFC3 with ETH treatment were lower than that in the control ( Figure 7C). ETH treatment significantly enhanced the expressions of CsPYL1 and CsBAK1 ( Figure 7C).

Discussion
As an important small gaseous molecule, ETH affects the growth and development of plants, and it is also a major regulator in response to diverse stresses. Many studies have shown that ETH can induce adventitious rooting in cucumber and marigold [1,15,16]. Using the iTRAQ technique and proteome analysis, Lyu et al. [16] revealed that ETH treatment increased photosynthesis and starch degradation to provide energy for adventitious rooting in cucumber. However, some key genes involved in ETH-promoted adventitious root development in cucumber have not been identified. Here, we revealed the mechanism of ETH-induced adventitious root development at the transcriptional level. First, GO enrichment and KEGG pathways in the transcriptome and proteome were analyzed and compared (Figure 1). In the top 20 KEGG pathway terms, "starch and sucrose metabolism", "phenylalanine metabolism", "plant hormone signal transduction", "fatty acid metabolism", "pentose and glucuronate interconversions", "cutin, suberin, and wax biosynthesis", "photosynthesis-antenna proteins", "vitamin B6 metabolism", "purine metabolism", and "alanine, aspartate, and glutamate metabolism" were enriched in transcriptome and proteome (the order was ranked by transcriptome data), which suggests that exogenous ETH induced adventitious rooting via regulation of the decomposition and metabolism of nutrients, the accumulation and degradation of plant hormones, the synthesis of secondary metabolites, and photosynthesis. In order to identify key genes during ethylene-induced adventitious root development in cucumber, several pathways associated with adventitious rooting and enriched in the KEGG pathway of the transcriptome were selected and analyzed further.

Carbon Metabolism in ETH-Induced Adventitious Root Development
Carbon metabolism is an important metabolic pathway during adventitious root development and can provide metabolites for starch synthesis, a carbon skeleton for amino acid synthesis, and intermediates for secondary metabolism [20]. Further, it influences energy generation and respiration through the TCA and photorespiratory cycles [21].
In our study, 17 DEGs were involved in ETH-inducing adventitious rooting in starch and sucrose metabolism (Figure 2). Among them, HK is considered to be a key enzyme involved in glucose sensing, signal transduction, and photosynthesis [22,23]. Andriunas et al. [24] found that glucose was able to act positively through an HK-dependent signaling pathway, suggesting that glucose could regulate the HK signaling pathway. In our study, ETH treatment decreased the content of glucose, the activity of HK and the relative expression of CsHK2 during rooting (Figure 2). In Arabidopsis, the transcript level of ERF1 was inhibited by low glucose signaling, which indicated that the ETH might have had an antagonistic effect on low glucose signaling [25,26]. Hence, exogenous ETH might decrease HK activity and CsHK2 expression during rooting. During this process, energy was obtained by consuming sucrose and glucose. HK enzyme and its related gene in the sucrose metabolism pathway were inhibited by ETH, thus accelerating hypocotyl growth and finally inducing adventitious root development in cucumber. Additionally, starch metabolism may also be involved in adventitious rooting [27]. β-amylase 1 (BAM1) and β-amylase 3 (BAM3) are key genes involved in starch degradation. Starch degradation causes a large amount of starch to be enriched in leaves and delays the growth and development of storage roots [27,28]. Our transcriptome results show that the transcript levels of CsBAM1 and CsBAM3 were down-regulated by ETH treatment (Figure 2). Therefore, ETH-treated explants completed energy conversion by reducing the expression of CsBAM1 and CsBAM3 in the starch signaling pathway and the content of starch, which may be beneficial for adventitious rooting. According to Lyu et al. [16], 48 h after treatment is the root elongation phase. At this stage, starch might be converted to glucose to provide energy for root growth. Therefore, ETH could induce adventitious rooting by decreasing the transcript levels of the starch degradation-related genes CsBAM1 and CsBAM3 and preventing starch degradation.
Further, glycolysis, the TCA cycle, and oxidative phosphorylation are classified as respiratory metabolism and can produce energy in the ATP form [29,30]. In the present study, 12, 5, and 4 DEGs were involved in these pathways by ETH treatment, respectively ( Figure 3). A previous study showed that the changes in glycolysis were closely related to the changes in root respiration in sugarbeet (Beta vulgaris L.) [29]. The change trend of PK activity was similar to that of the respiratory rate, first increasing, then decreasing, and finally becoming stable [29]. Lyu et al. [16] found that ETH enhanced ATP synthase subunit protein expression at 12 and 24 h, and then declined at 48 h, confirming that ETH-induced adventitious rooting might require energy. This agreed with the findings of our metabolomic analysis showing that the activity of PK and the expression of CsPK2 were decreased with ETH at 48 h after treatment (Figure 3), suggesting that at this stage, the respiration rate might decrease, resulting in the decline of ATP generation. Hence, ETH could induce adventitious root development by decreasing the PK activity and its transcript level. At 48 h after ETH treatment, ATP synthetase provided more energy for inducing adventitious rooting.
Fatty acids provide energy for some metabolic processes and act as signal transduction regulators [31]. Twelve DEGs were involved in fatty acid biosynthesis and the fatty acid degradation pathway in ETH-induced adventitious root development (Figure 4). Among them, CYP86A1 was a key gene for root aliphatic suberin biosynthesis in Arabidopsis [32,33]. Root suberin could control the absorption and transportation of water and mineral ions from roots to leaves [34]. Therefore, CYP86A1 also might play an important role in rooting. Our work showed that ETH treatment decreased the expression of CsCYP86A1 during rooting (Figure 4), suggesting that the energy might be transferred to the hypocotyl to induce adventitious root development.

Secondary Metabolite Biosynthesis Pathway in ETH-Induced Adventitious Root Development
Phenylalanine metabolism and flavonoid biosynthesis are secondary metabolite biosyntheses. They benefit plants throughout the whole developmental process. Wang et al. [35] demonstrated that phenylpropanoid and flavonoid biosynthesis pathways were associated with root development. However, there has been no research showing that phenylalanine metabolism could be involved in adventitious root development. In our study, ETH regulated seven DEGs in phenylalanine metabolism during rooting, including CsPAL, CsPAO, and CsAST ( Figure 5). PAL was a critical enzyme that directly catalyzes phenylalanine synthesis and is a precursor for the phenylpropanoid biosynthetic pathway [36]. The activity of PAL varied with the development stage of plants and the differentiation degree of cells and tissues [37,38]. Under excess Fe nutrition, plant growth was suppressed, and the activity of PAL was enhanced [39], indicating that the increase in PAL activity might inhibit plant growth. In our study, the activity and relative expression of PAL decreased with ETH treatment (Figure 5), indicating that the ETH might have induced adventitious rooting by reducing PAL activity and its gene expression.
In addition, flavonoid biosynthesis was reported to have an important role in hairy roots, adventitious roots, and seedling roots [36]. However, there is little information regarding the roles of flavonoid in ETH-induced adventitious rooting. In our transcriptome results, three DEGs, including CsCHI, CsF3 M, and CsHCT were involved in this pathway ( Figure 6). Among them, the expression of CsHCT was significantly increased by ETH treatment. There was a negative correlation between the expression of CsHCT and the content of total flavonoids [40]. When the flavonoid content decreased, the content of lignin increased, and the fibers between the vascular bundles became longer [41]. Hence, ETH treatment might promote adventitious root primordium growth and break through the epidermis by increasing the expression of CsHCT, and the newly formed adventitious roots could be connected to the vascular bundle, leading to fiber growth between vascular bundles.

Plant Hormone Signal Transduction in ETH-Induced Adventitious Root Development
Phytohormone is an essential factor affecting adventitious rooting. We identified 26 DEGs in the plant hormone signal transduction pathway in total (Figure 7). Various studies have shown that auxin could induce adventitious root development [42,43]. In the present study, ETH regulated eight DEGs, including CsAUX1, CsAUX/IAA, CsGH3, and CsSAUR in the auxin signaling transduction pathway (Figure 7). AUX1 and LAX5 are auxin influx carriers of the same family [42]. AUX1 could mediate the auxin influx specifically in the adventitious root-initiating cells, whereas LAX5 was involved later than AUX1 in adventitious rooting [42]. Recently, da Costa et al. [44] reported that LAX1 and LAX3 were used as down-regulation genes during adventitious rooting in Arabidopsis thaliana, which was consistent with our result. We found that ETH decreased the expression of CsAUX1 and CsLAX5 (Figure 7), which proves that the auxin influx genes were negatively regulated in ETH-promoted adventitious root development. Moreover, Harkey et al. [45] found that the ETH precursor ACC reduced the expression of GH3.17 in Arabidopsis roots. The decreased activity of GH3 could lead to a decrease in the IAA level [46]. Our study shows that ETH down-regulated the expression of IAA-amido synthetase CsGH3.17 (Figure 7). To expand on the results of the expression studies, the endogenous IAA content was further determined in control and ethephon-treated cucumber explants. In our results, ETH significantly decreased the endogenous IAA content (Figure 7). Therefore, ETH could induce adventitious rooting by negatively regulating the auxin signal pathway, in which ETH might decrease the IAA content and the expressions of CsGH3.17, CsAUX1, and CsLAX5.
ABA negatively controls root emergence [47]. However, the crosstalk between ABA and ETH in adventitious root development was unclear. We found that ETH up-regulated the expression of six DEGs in the ABA signal transduction pathway, which included CsPYR/PYL, CsSnRK2, and CsABF (Figure 7). PYR1, as a member of the family of PYR/PYL/ RCAR ABA receptors, mediated plant immunity, seed germination, and seedling establishment [48,49]. PYL8 has been reported to induce lateral root growth [50]. Here, we found that ETH up-regulated the expression of the ABA receptors CsPYR1, CsPYL5, and CsPYL8 during adventitious rooting (Figure 7), indicating that CsPYR1, CsPYL5, and CsPYL8 might positively regulate ETH-induced adventitious root development. Moreover, ETH significantly enhanced ABA content (Figure 7). Thus, ETH could induce rooting by increasing ABA signal transduction-related gene expression and ABA content.
Furthermore, ETH decreased the expression of four DGEs, including CsETR, CsEBF1/2, and CsERF1/2 in ETH signal transduction (Figure 7). ERS is an ETH receptor gene, which can negatively regulate ETH response [51]. In the present study, ETH decreased the transcript level of CsERS (Figure 7). Hua et al. [52] also found that ETH treatment reduced the ERS transcript level, indicating that ERS could be negatively mediated by exogenous ETH during rooting.
In Arabidopsis, exogenous BR could promote adventitious root initiation and growth [53,54]. In our study, ETH up-regulated the expressions of CsBAK1 and CsXTH3 in the BR signaling transduction pathway (Figure 7). BAK1 is a BR receptor gene. Xie et al. [55] reported that graphene oxide could increase BAK1 expression to enhance root growth in Brassica napus L., confirming that BAK1 might positively respond to root growth. Similarly, our results show that ETH enhanced the transcriptional level of CsBAK1 (Figure 7), which suggests that ETH induced adventitious rooting via positively regulating CsBAK1 expression in BR signal transduction. Therefore, BR might have a positive role in ETH-induced adventitious root development. Xyloglucan endotransglucosylase/hydrolase (XTH), as a cell wall loosening and extension factor, can induce cell elongation and root growth [56]. However, there has been no direct evidence that XTH positively or negatively regulates adventitious root development. We found that ETH up-regulated the expression of CsXTH3 (Figure 7), which implies that ETH might promote rooting by increasing the transcript level of CsXTH3.

Plant Material and Growth Conditions
Cucumber (C. sativus L. "Xinchunsihao") seeds of the same size and full grains were soaked in distilled water for 5-6 h and then placed on a tray with wet filter paper. After this step, the seeds in the tray were moved to the incubator with illumination with a 14 h photoperiod (the photosynthetically active radiation was 200 µmol m −2 s −1 ) at 25 ± 1 • C for 7 d, and the filter paper was kept wet in the tray during this period. The primary roots of explants were carefully cut off after 7 d. In accordance with Zhao et al. [57], the induction of adventitious roots was completed 48 h after treatment. Hence, the cucumber explants were placed in Petri dishes with different treatments and allowed to grow for 48 h. The temperature and photoperiod conditions were the same as the above conditions.

Chemicals and Treatments
Based on our previous results [1], 0.5 µM of 2-chloroethylphosphonic acid (ethrel, an ETH donor) was the optimal concentration to induce adventitious rooting in cucumber. After cutting the primary roots, every ten explants of the same size were selected and placed in Petri dishes containing 70 mL of with distilled water or 0.5 µM ethrel, and the solutions used in the study were changed every 24 h [1]. Each experiment was conducted three times with 3 replicates. At 48 h after treatment, samples (cucumber explants) were collected and stored at −80 • C until used.

RNA Extraction
The total RNA for RNA-Seq and RT-qPCR analyses was extracted from about 500 mg cucumber explants with Trizol reagent (Sangon, Shanghai, China) in accordance with the manufacturer's instructions [58]. The total RNA was dissolved in 100 µL of RNase-free water and quantified using a NanoDrop spectrophotometer (Thermo Scientific, Waltham, MA, USA) [59]. The RNA quality was evaluated using the 6000 Pico LabChip of the Agilent 2100 Bioanalyzer (Agilent, Shanghai, China) [60]. RNA samples were sent to Novogene Co., Ltd. (Beijing, China) for cDNA library construction and sequencing. The reference genome for transcriptome analysis was acquired from https://www.ncbi.nlm.nih.gov/ genome/1639?genome_assembly_id=749658 (accessed on 18 March 2020). Each treatment had 3 technical replications and 3 biological replications.

cDNA Synthesis and Illumina Sequencing
The mRNA was purified from the total RNA using poly-T oligo-attached magnetic beads [61]. In the NEBNext First Strand Synthesis Reaction Buffer (5×), divalent cations were used for fragmentation at a high temperature. First-strand cDNA was synthesized using random hexamer primer and M-MuLV Reverse Transcriptase (RNase H-). Secondstrand cDNA synthesis was subsequently performed using DNA Polymerase I and RNase H. Remaining overhangs were converted into blunt ends via exonuclease/polymerase activities [62]. After adenylation of 3 ends of the DNA fragments, a NEBNext Adaptor with a hairpin loop structure was ligated to prepare for hybridization. In order to screen 250-300 bp cDNA fragments preferentially, the library fragments were purified using an AMPure XP system (Bethe controlman Coulter, Beverly, CA, USA) [20]. Before PCR, 3 µL of USER Enzyme (NEB, Boston, MA, USA) was used at 37 • C for 15 min followed by 5 min at 95 • C with size-selected, adaptor-ligated cDNA [63]. Then, PCR was performed with Phusion High-Fidelity DNA polymerase, Universal PCR primers, and Index (X) Primer. At last, the PCR products were purified (AMPure XP system; Bethe controlman Coulter, Beverly, CA, USA) and the library quality was assessed on the Agilent Bioanalyzer 2100 system (Agilent, Shanghai, China) [20]. The library products were sequenced with the Illumina Novaseq platform. The RNA-seq data produced in this study were deposited in the NCBI Gene Expression Omnibus (GSE189315).

Bioinformatics Analysis
The raw reads were raw sequencing data, and they could be transferred from the original image data through a base call. The clean reads were a high-quality sequence after clearing data from the original reads. The error rate was the total sequencing error rate of the data. Q20 and Q30 refer to the percentage of bases with Phred values greater than 20 and 30 to the total bases, respectively. GCpct was the percentage of G and C among the four bases in the clean reads. The transcriptome sequencing was of high quality and can be used for further analysis (Supplementary Table S1). In addition, the Pearson correlation analysis showed high consistency of gene expression patterns of different biological replicates of each sample, suggesting the high reproducibility of transcriptome data (Supplementary Figure S1). According to the RPKM (reads per kilobase per million reads) criteria, levels of DEGs were calculated by quantifying reads. Transcriptome libraries should be based on the criteria: |log 2 (FoldChange)| > 1 and padj ≤ 0.05. Analyses of GO functional enrichment and KEGG pathway enrichment of the DEGs were carried out using ClusterProfiler software. In the functional enrichment analysis of GO, the DEGs were mainly enriched in the biological process (BP), molecular function (MF), and cellular component (CC). Using the KAAS server, the metabolic pathway was analyzed according to the KEGG pathway database (http://www.genome.jp/kegg/pathway.html) (accessed on 25 May 2020) [64].
Additionally, because the treatment time and ETH donor ethrel concentration were the same in our transcriptome analysis and in the proteomic analysis of Lyu et al. [16], we compared the DEGs in the transcriptome with differentially expressed proteins (DEPs) in the proteome.

Determination of Sucrose, Glucose, and Starch Content
The determination of sucrose content was performed according to Tauzi and Giardina [65]. On the basis of our experimental requirements, we modified the procedure appropriately. The cucumber explants (1.0 g) were added to 10 mL of 80% ethanol and diluted with distilled water in a 100 mL volumetric flask. The volumetric flask was placed in a water bath at a constant temperature of 80 • C for 45 min, then cooled to room temperature, and filtered with filter paper. This step was repeated twice, and the filtrate was collected as the reaction solution. The filtrate (0.4 mL) was mixed with 0.2 mL of 2 M sodium hydroxide (NaOH), and then bathed at 80 • C for 10 min. After cooling to room temperature, the sucrose content was determined. The absorbance was measured at 540 nm, and the soluble sugar content was calculated according to the standard curve.
The determination of the glucose content was performed according to Smith and Zeeman [66]. Fresh explant (0.5 g) was homogenized with 10 mL distilled water, diluted with distilled water, and mixed in a 50 mL volumetric flask. Then it was placed in a boiling water bath for 10 min. The reaction mixture contained 3,5-dinitrosalicylic acid (DNS) regent (6.3 g), 262 mL 2 M NaOH solution, 185 g potassium sodium tartrate with 500 mL of hot water, 5 g of redistilled phenol, and 5 g of sodium sulfite. After cooling the mixture to room temperature, the volume was accurately adjusted to 1000 mL with distilled water. Then the homogenate was centrifuged at 4000 rpm for 15 min at 4°C. The supernatant (2 mL) and deoxyribonucleic acid (1.5 mL) were mixed together, and the mixture was centrifuged at 4000 rpm for 15 min at 4 • C. Then it was placed in boiling water at 100 • C and bathed for 5 min. After cooling to room temperature, the solution was filtered into a 25 mL volumetric flask. The sample absorbance was measured at 540 nm against the controls.
For the starch content analysis, a 0.5 g sample was homogenized with 2 mL of distilled water and 3.2 mL of 60% HClO 4 . Distilled water was added to the mixture to 100 mL and then centrifuged at 5000 rpm for 5 min. Then, 0.5 mL of the supernatant was pipetted and diluted to 3 mL with distilled water. Then, 2 mL of iodine reagent was added, mixed well, and the mixture was left standing for 5 min. Finally, the solution was added to 10 mL of distilled water. The optical density (OD) was monitored at 660 nm [66]. The enzyme activities of SS, SPS, and HK were determined by enzyme-linked immunosorbent assay (ELISA; Andy Gene Biotechnology Co., Ltd., Beijing, China). A 1.0 g sample was ground with liquid nitrogen and homogenized with 9 mL PBS (pH = 7.4). Then the homogenate was centrifuged at 3000 rpm for 15 min at 4 • C [67]. The subsequent measurement method was carried out on the basis of the manufacturer's instructions.

Determination of Phenylalanine Ammonia-Lyase (PAL) Activity
The PAL activity was assessed according to González-Mendoza et al. [68]. A fresh sample (0.5 g) was mixed into 5 mL of Tris-HCl buffer (0.1 M, pH 7.5), 1 mM 1% polyvinylpolypyrrolidone and ethylenediaminetetraacetic acid (EDTA). Then the mixture was centrifuged at 12,000 rpm, at 4 • C for 10 min. The reaction mixture, including 100 µL of enzyme extract and 200 µL of Tris-HCl (50 mM, pH 8.8) containing 20 mM L-phenylalanine was incubated at 40 • C for 30 min. Then 100 µL of 1 M hydrochloric acid (HCl) was added to stop the reaction. The absorbance value was measured at 290 nm. The amount of enzyme that increased the OD 290 by 0.01 per minute was defined as an enzyme activity unit IU.

Determination of IAA and ABA Contents
The concentrations of the IAA and ABA contents were determined by Quaternary gradient ultra-fast liquid chromatograph using a Waters Acquity ARC 600-2998 equipped with the Symmetry-C18 column (4.6 × 250 mm, 5 µm). The frozen sample (1.0 g) was taken and ground with liquid nitrogen quickly and then transferred to a 10 mL centrifuge tube. Then, 5 mL of 80% cold pure methanol was added and the sample was stored in the dark in the refrigerator at 4 • C for 12 h. After 12 h, the sample was centrifuged at a speed of 4000 rpm for 15 min at 4 • C. Then the supernatant was pipetted into another 10 mL centrifuge tube. The residues were suspended in 2.5 mL of extraction solution and stored at 4 • C for 1 h. Then the mixture was centrifuged at a speed of 4000 rpm for 15 min, and the filtrate was collected and combined in the tube. This step was repeated twice. The filtrate was diluted to 10 mL with 80% methanol. Two mL of the supernatant was evaporated in a rotary vacuum at 38 • C for 4 h and then dissolved with 2 mL of 80% cold pure methanol. The extract was filtered with a 0.22 µm filter for liquid chromatograph detection. The mobile phase was chromatographic methanol (A) + 0.1% phosphoric acid (B). The flow rate was 1.0 mL min −1 . The wavelength was 254 nm, and the column temperature was 30 • C [69].

RT-qPCR Assays
In accordance with the manufacturer's instructions, the RNA (400 ng) was reversetranscribed with 5 × Evo M-MLVRT Master Mix (AG, Changsha, China). SYBR Green Pro Taq HS Premix (AG, Changsha, China) was used to amplify cDNA. The gene expression data were normalized using Csactin as an internal control [70]. The primers were designed by primerbank (https://pga.mgh.harvard.edu/primerbank/) (accessed on 1 May 2021). The RT-qPCR reactions containing 7.2 µL of ddH 2 O, 0.4 µL of forward primer, 0.4 µL of reverse primer, 2 µL of cDNA (800 ng), and 10 µL of Taq enzyme were performed on the ABI StepOne plus system (Foster, CA, USA) with the following program: 95 • C for 30 s, 40 cycles of 95 • C for 5 s and 60 • C for 30 s. In order to assess possible contaminations and primer dimer formations, no-template controls (NTCs) were used as a negative control [71]. The RT-qPCR efficiencies of amplicons were calculated by performing a standard curve. Moreover, melting curve analysis was conducted to ensure the amplification of specific amplicons. StepOne TM Software v2.3 (Life Technologies, Boston, MA, USA) was used to analyze the ct values, and 2 −∆∆ct was used to calculate the normalized gene expression levels [72]. The experiment was conducted with three bio-replicates. The primers are listed in Supplementary Table S2.

Statistical Analysis
The statistical analysis was conducted using the Statistical Package for Social Sciences 22.0 (SPSS 22.0, Inc., Chicago, IL, USA) software. All results were expressed as the mean values ± standard error (SE) from at least three independent duplications. The data from substance contents, enzyme activities, plant hormone contents and RT-qPCR assays were analyzed by t-test at p < 0.01 and p < 0.05, respectively.