RhWRKY33 Positively Regulates Onset of Floral Senescence by Responding to Wounding- and Ethylene-Signaling in Rose Plants

Rose plants are one of the most important horticultural crops, whose commercial value mainly depends on long-distance transportation, and wounding and ethylene are the main factors leading to their quality decline and accelerated senescence in the process. However, underlying molecular mechanisms of crosstalk between wounding and ethylene in the regulation of flower senescence remain poorly understood. In relation to this, transcriptome analysis was performed on rose flowers subjected to various treatments, including control, wounding, ethylene, and wounding- and ethylene- (EW) dual treatment. A large number of differentially expressed genes (DEGs) were identified, ranging from 2,442 between the ethylene- and control-treated groups to 4,055 between the EW- and control-treated groups. Using weighted gene co-expression network analysis (WGCNA), we identified a hub gene RhWRKY33 (rchiobhmchr5g0071811), accumulated in the nucleus, where it may function as a transcription factor. Moreover, quantitative reverse transcription PCR (RT-qPCR) results showed that the expression of RhWRKY33 was higher in the wounding-, ethylene, and EW-treated petals than in the control-treated petals. We also functionally characterized the RhWRKY33 gene through virus-induced gene silencing (VIGS). The silencing of RhWRKY33 significantly delayed the senescence process in the different treatments (control, wounding, ethylene, and EW). Meanwhile, we found that the effect of RhWRKY33-silenced petals under ethylene and EW dual-treatment were stronger than those under wounding treatment in delaying the petal senescence process, implying that RhWRKY33 is closely involved with ethylene and wounding mediated petal senescence. Overall, the results indicate that RhWRKY33 positively regulates the onset of floral senescence mediated by both ethylene and wounding signaling, but relies heavily on ethylene signaling.


INTRODUCTION
Rose plants are one of the most important horticultural crops worldwide; their ornamental and economic value mainly depends on the full expansion and slow senescence of their petals (Lu et al., 2014). Most cut flowers are produced in developing countries, whereas their sales are mainly concentrated in developed countries. Therefore, long-distance transportation from the production region to the point of sale is the mainstream channel for cut rose trade. However, long-distance transportation, usually in darkness under, low temperature, and without water, can easily trigger a massive burst of exogenous ethylene production (Faragher et al., 1987;Mor et al., 1989;Reid et al., 1989;Noodén and Schneider, 2004). Furthermore, stressinduced exogenous ethylene accelerates flower opening and senescence and decreases the vase life of cut rose by coupling with the mechanical damage during postharvest treatment (Müller et al., 2000). Therefore, avoiding and reducing the negative effects of long-distance transportation is crucial in maintaining the postharvest quality of cut roses.
As sessile organisms, plants have developed intricate mechanisms to defend themselves against external adversities, including wounding and hormones, and other biotic, and abiotic stresses. Plants are also able to sense injured tissues and protect themselves by regulating defense system-related genes. The endogenous signals released from wounded tissues may act to activate the plant defense system, and wounding-induced responses are distributed in both the site of damage (local response) and systemically (systemic response), mediated by hormone signaling, including jasmonic acid (JA), ethylene (Eth), brassinolide (BRs), salicylic acid (SA), and abscisic acid (ABA).
Previous studies revealed that mechanical damage results in a large increase in the content (∼25-fold) of endogenous JA and JA-isoleucine (JA-Ile) within 5 min of tissue injury (Chung et al., 2008;Glauser et al., 2008;Koo et al., 2009). The phytohormone, JA, is a well-characterized signaling molecule for its role in protection against wounding damage (Leon et al., 2001); it also plays a central role in activating the systemic defense in many vital processes, including flower development and senescence (Wasternack, 2007). The transcript levels of several JA biosynthesis-related genes, including LIPOXYGENASE 2 (LOX2), LIPOXYGENASE 3 (LOX3), and ALLENE OXIDE SYNTHASE (AOS), are upregulated together with the JA acceleration (Bell et al., 1995). Arabidopsis genes such as LOX2, AOS, and VEGETATIVE STORAGE PROTEIN (VSP) were found to be induced by IAA (Indoleacetic Acid), and endogenous auxin content also showed a decline after wounding, showing a significant cross-talk between the two hormone-dependent signaling pathways (Devoto and Turner, 2003). The rea1 (RAP2.6 expresser in shoot apex) mutants show a constitutive cell death phenotype, and early wounding response requires an intact JA signaling pathway (Zou et al., 2016); however, the application of methyl jasmonate (MeJA) has been shown to provide significant protection against cell death (Overmyer et al., 2000).
Ethylene is another essential systemic signal of protection of plants against wounding damage. Ethylene released from wounded plant tissues and its participation in wounding response signaling has been extensively studied for decades. Upon wounding at harvest, ethylene was rapidly synthesized in the damaged stem tissue as a result of the marked increase in 1aminocyclopropane-1-carboxylic acid (ACC) synthase activity and the increased abundance of its transcripts. In melons, the ACC oxidase gene (CmACO1), the encoding enzyme controlling the last step of ethylene biosynthesis in plants, was rapidly induced (within 10 min) by ethylene treatment or upon wounding in leaves. 1-methylcyclopropene (1-MCP), a powerful inhibitor of ethylene action, significantly inhibited the accumulation of ethylene-induced CmACO1 messenger RNA (mRNA) transcripts, while the findings did not apply to wounding-induced CmACO1 expression (Bouquin et al., 1997). Previous pieces of evidence indicated that the transcripts of several genes encoding ACC oxidase (ACO) and ACC synthase (ACS) were rapidly and transiently accumulated after mechanical injury (Kato et al., 2002). In some cases, the enrichment of ACS transcripts was detected within a few hours, even as quickly as 10 min after touching or mechanical wounding in mung bean or soybean plants, respectively (Liu et al., 1993;Botella et al., 1995). Similarly, 1-aminocyclopropane-1-carboxylic acid oxidase (ACO) mRNA levels were also induced within several hours, while some were generally induced at a shorter period (20 or 30 min) (Balague et al., 1993;Kim and Yang, 1994;Jones and Woodson, 1999).
An increasing number of studies have shown that many members of the WRKY family are involved in multiple plant processes, such as stress defense responses, plant senescence, wounding, seed dormancy, and seed germination (Rushton et al., 2010;Birkenbihl et al., 2012;Rinerson et al., 2015;Zhao M. et al., 2020). Moreover, functional studies have shown that WRKY transcription factors (TFs) are involved in regulating senescence by interacting with JA, ABA, auxin, gibberellins (GAs), and SAmediated signaling. In wheat, TaWRKY40-D positively regulated leaf senescence following JA and ABA treatment (Zhao L. et al., 2020). Another WRKY TF, TaWRKY42-B, interacted with and promoted the JA biosynthesis gene (AtLOX3, and its ortholog-TaLOX3), resulting in an increased JA content and an initiation of leaf senescence in Arabidopsis and wheat (Zhao M. et al., 2020). In Arabidopsis, AtWRKY54 and AtWRKY70 functioned redundantly in regulating leaf senescence, as revealed by single and double mutant studies (Besseau et al., 2012). Concerning the diverse roles of WRKY TFs, AtWRKY57 functioned as a repressor in the JA-induced senescence, which was downregulated by JA and upregulated by auxin . AtWRKY45 played a positive role in the regulation of leaf senescence via GA-mediated signaling . Genetic analysis revealed that the crosstalk of AtWRKY75-SA/ROS functions correlatively in agedependent leaf senescence (Guo et al., 2017). Finally, eight SlER-WRKYs-SlWRKY16, 17, 22, 25, 31, 33, 53, and 54 were reported to be responsive to ethylene during tomato fruit ripening (Wang et al., 2017). Taken together, WRKY proteins play critical roles in regulating plant senescence.
Ethylene is required for the signal transduction pathway in injury. However, the components involved in the initial wounding signal perception and transmission leading to ethylene synthesis are not clear. In addition, the detailed molecular mechanisms of the action between ethylene and wounding to regulate the onset of floral senescence remain poorly understood. In relation to this, identifying the initial signal of wounding and signal crosstalk between wounding and the ethylene pathway are crucial steps in understanding the early onset of floral senescence. Here, we monitored the status of the flower opening and senescence after the wounding, ethylene gas, or dual treatment daily. Furthermore, we employed RNA-sequencing (RNA-seq) and weighted gene co-expression network analysis (WGCNA) assays to investigate the signal crosstalk involvement of wounding and the ethylene pathway in the senescence process of rose flowers, and screened a WRKY transcription factor, RhWRKY33, which closely responds to wounding-and ethylene signaling. A biological analysis revealed that RhWRKY33 plays a positive role in regulating petal senescence. In summary, our findings highlight that RhWRKY33 mediates petal senescence by integrating wounding and ethylene signaling (a leading role).

Plant Materials
Cut roses (Rosa hybrida cv. Samantha) were grown in a greenhouse at the China Agricultural University (Beijing, China) and harvested at the second stage of a flower opening. The pretreatment methods and user definitions of flower opening stages were performed based on a previous description (Ma et al., 2006).
For the ethylene treatment, the cut roses and petal discs were exposed to gaseous ethylene (10 ppm) in an airtight growth container for 24 h (Ma et al., 2006). An aqueous solution of sodium hydroxide (NaOH) (1 M) was also placed into the sealed container to absorb the accumulated carbon dioxide (CO 2 ) as previously described (Ma et al., 2005;Wu et al., 2017). For the wounding treatment, the outermost layer petals of the cut rose were wounded with a dissecting needle across its surface, and the petal discs were wounded and slightly squeezed. Then, the gastreated and wounded cut rose or discs were incubated under the following conditions: 22 ± 1 • C, 16 h/8 h photoperiod, and ∼60% relative humidity. Three whole outermost layer petals of the cut rose and petal discs were harvested for RNA extraction.

Total RNA Extraction and RNA-Seq Library Preparation
Total RNA was extracted from the rose petals as previously described (Tian et al., 2014;Wu et al., 2017), and removed from genomic DNA contamination using RNase-free DNase I (Promega); the quality and quantity were evaluated using a NanoPhotometer R spectrophotometer (Implen, California, USA) and an Agilent 2100 Bioanalyzer (Agilent Technologies, California, USA). The RNA integrity number (RIN) of all the samples (>8.0) was used for RNA-seq. The RNA-seq was performed by Beijing Novogene Bioinformatics Technology Co., Ltd. (Beijing, China). Three micrograms of the total RNA per sample were used as the input material for cDNA library construction and Illumina sequencing, and the RNA-seq data were processed, assembled, and annotated as previously described (Yang et al., 2015). Briefly, the clean RNA-seq reads were aligned to the reference genome (Rosa chinensis cv. Old Blush, GenBank ID 8255808) and analysis was performed using HISAT2 (version 2.0.4) (Acosta et al., 2005) for the reads derived from mRNA fragments using default parameters. The mapped reads of each sample were assembled using Stringtie (v2.1.5) (Pertea et al., 2015), and functional annotations were accomplished by searching against the National Center for Biotechnology Information (NCBI) non-redundant protein (Nr) database using the Basic Local Alignment Search Tool (BLAST)x with an E-value threshold of ≤10 −5 .

Differential Expression Analysis
Differentially expressed genes between mock and other treatment conditions were performed using the DESeq R package (version 1.10.1) (Anders and Huber, 2010). The DESeq provides statistical routines for determining differential expressions in digital gene expression data using a model based on a negative binomial distribution. The resulting p values were adjusted using Benjamini and Hochberg's approach to control the false discovery rate. The differentially expressed genes (DEGs) were identified by DESeq, with fold changes ≥ 2 and adjusted p-value ≤0.05. All raw data were deposited in the NCBI BioProject database under the accession number PRJNA550484 (https://www.ncbi.nlm.nih. gov/bioproject/?term=PRJNA550484).

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) Enrichment Analysis of DEGs
The Blast2GOsoftware package was used to identify the GO enriched terms (Conesa et al., 2005). The GO terms with a corrected FDR < 0.05 were considered significantly enriched in DEGs (Conesa et al., 2005). The KEGG Orthology-Based Annotation System (KOBAS) software was used to test the statistical pathway enrichment of DEGs among the KEGG pathways (http://www.genome.jp/kegg/) (Mao et al., 2005).

MapMan Visualization
To identify the gene functions that may play essential roles in regulating the onset of floral senescence, a MapMan visualization was performed as described previously (Thimm et al., 2004). Based on the Mercator with a blast cutoff of 50, contigs were classified into a set of hierarchical functional categories (BINs) (Lohse et al., 2014). Facing classification problems with one unigene that might have multiple contigs, the functional term of unigene was determined by the highest bit score (Nham et al., 2015). Enrichment analysis was performed using Fisher's test and Mefisto (http://www.usadellab.org/cms/index.php?page$= $mefisto) with Bonferroni correction. Gene expression changes were assessed using MapMan 3.5.1 R2.

Ethylene Production Measurements
The petals of each flower were collected and placed in an airtight container (13 ml). The containers were capped and incubated for 1 h at 25 • C. Then, a gas sample of 1 ml was withdrawn, using a gas-tight hypodermic syringe, and injected into a gas chromatograph (GC 17A, Shimadzu, Kyoto, Japan) for the ethylene concentration measurement (Ma et al., 2006). All measurements were performed with three replicates.

Co-expression Analysis of Protein-Coding Genes With RNA-Seq Data
The WGCNA R software package (Langfelder and Horvath, 2008) was used to test whether the expression of protein-coding genes correlated with the senescence degree of rose flowers under the different treatments.

Virus-Induced Gene Silencing (VIGS)
The virus-induced gene silencing of RhWRKY33 in rose petal discs was performed as previously described (Zhang et al., 2019a). A specific fragment of RhWRKY33 (417 bp in length) was cloned and inserted into the pTRV2 vector using XbaI and KpnI sites. The recombined-TRV2 vector, pTRV2, and pTRV1 were transformed into Agrobacterium tumefaciens strain GV3101, and the transformed A. tumefaciens cells were cultured in an LB medium (supplemented with 50 µg/ml kanamycin and 50 µg/ml rifampicin) on a rocking platform (200 rpm) at 28 • C for 18 h. Cells were harvested and resuspended in an infiltration buffer (10mM MgCl 2 , 200 mM acetosyringone, 10 mM MES, pH 5.6) to a final OD 600 of ∼1.0. The culture mixtures contained pTRV1and pTRV2-RhWRKY33 at a ratio of 1:1 (v/v) or pTRV1 and pTRV2 (negative control). The culture mixtures were incubated in the dark at room temperature for 3-4 h before infiltration. Discs (1 cm diameter) were submerged in the infiltration buffer and were exposed to a vacuum (−25 kPa) two times (each for 5 min), washed with deionized water, and then incubated in the dark at 8 • C for 3 days. The phenotypes of the discs were observed and counted.

RT-qPCR
The genomic DNA of the total RNA samples was removed using RNase-free DNase I (Promega). Reverse transcription was performed with 1 µg of the total RNA using HiScript II Q Select RT SuperMix for qPCR (+gDNA wiper) (Vazyme, Nanjing, China). Together with 2 µl [20ng] cDNA used as a template, RT-qPCR was performed using a StepOne Real-Time PCR System (Applied Biosystems, California, USA) with a Kapa SYBR Fast Universal qPCR Kit (Kapa Biosystems, Boston, Massachusetts, USA). Rose RhUBI2 and RhActin5 were used as reference genes (Bustin et al., 2009;Meng et al., 2013;Pei et al., 2013;Gao et al., 2016;Wu et al., 2017;Zhang et al., 2019bZhang et al., , 2021, and the mean of reference genes was used in the following analysis to avoid discrepancies; Relative quantification of the transcript abundance of each gene was performed using the 2 − Ct method (Livak and Schmittgen, 2001). All primers are listed in Supplementary Table 6.

Statistical Analyses
GraphPad Prism 8.0.2 software (http://www.graphpad.com/) was used to analyze the statistical differences in the data. Two groups of data were compared by using two-sided Student's t-tests ( * p <0.05, * * p <0.01, * * * p <0.001, * * * * p <0.0001). The means of the multiple groups of data were compared via one-way ANOVA and Bonferroni's post-hoc test, with p <0.05 considered significant.

Wounding Accelerates Rose Floral Senescence Dependent on Ethylene Signaling
The role of ethylene in the acceleration of floral senescence has been reported (Ma et al., 2018), but limited research has been conducted on the function of wounding in petal senescence. To determine the effects and signaling crosstalk between wounding and ethylene during floral senescence, we studied the lifespan of cut rose flowers under different treatments. Results showed that the wounded cut roses displayed a slight wilting until 7 days after treatment (DATs); their lifespan was 6.6 ± 0.6 days and there existed significant differences compared with the controls (7.4 ± 0.7 days) (Figures 1A,B). The rose flower that underwent ethylene treatment showed severe wilting after 6 days, and after the dual treatment with wounding and ethylene, the wilting started at 4 DAT, and nearly all flowers displayed complete wilting after 6 DAT. The life cycle of the cut rose that underwent the dual-treatment by wounding and ethylene was 4.7 ± 0.7 days, which showed more severe wilting than those that dealt with single treatments (Figures 1A,B). Furthermore, the expression level of senescence-associated gene RhSAG12, a molecular marker of rose floral senescence progression, was significantly higher in the dual-treated flowers compared with the control-treated flowers at 1, 3, and 6 DAT ( Figure 1C). Moreover, we measured the ethylene production under different treatments, and we found that the ethylene production in the wound-treated petals (4.2 ± 0.7 C 2 H 4 g −1 FW h −1 ) was higher than that in the control (2.2 ± 0.6 C 2 H 4 g −1 FW h −1 ) but lower than that in the ethylene (4.5 ± 0.8 C 2 H 4 g −1 FW h −1 ) or wounding-and ethylene-(EW) dual treatment (6.9 ± 0.9 C 2 H 4 g −1 FW h −1 ) at 6 DAT (Supplementary Figure 1). Briefly, these results indicate that the single wounding or ethylene treatment could significantly accelerate the senescence process of rose flowers to some content; however, the role of the wounding was of a senescence-facilitating factor in ethylene signaling.

Global Analysis of Transcriptome Data
To determine the alteration in gene expression during the onset of flower senescence under the dual treatment with wounding and ethylene, we extracted RNA samples from the four different treatments with three biological replicates. The RNA sequencing produced 48,033,008; 55,887,823; 57,773,772; and 58,588,791 clean reads from the four different treatments, respectively. Sequences were mapped to the reference genome (Rosa chinensis cv. Old Blush, GenBank ID 8255808) for annotation, and the mapping rate reached over 83% for the samples from each treatment (Supplementary Table 1). Differentially expressed genes were screened by comparing three different treatments with the control treatment, and 6,771 unigenes that were differentially expressed across these treatments were generated (fold changes ≥ 2, p ≤0.05). The number of DEGs was counted as 274 from Group 1 (wounding vs. control), 2,442 from group 2 (ethylene vs. control), and 4,055 from group 3 (EW vs. control) (Supplementary Figure 2A). Values are means ± SD, RhUBI2 and RhActin5 were used as reference genes, and asterisks indicate statistically significant differences (*p < 0.05, **p < 0.01, ***p < 0.001, Student's t-test).
Results showed that the three groups that shared 96 DEGs attributed to the wounding, drawing few gene expression changes, which was consistent with the single wounding not accelerating the floral senescence process. In addition, Groups 2 and 3 shared 2,093 DEGs, and Group 3 specifically induced 1,903 DEGs (Supplementary Figure 2B), both of which had enriched functional categories related to protein degradation, hormone metabolism, and transcription regulation.

Complex Hormone Biosynthesis and Signaling Pathways Involved in Wounding and Ethylene Signaling
To further understand the detailed regulatory network during the onset of rose floral senescence under wounding and ethylene treatments, all DEGs related to hormone biosynthesis and signaling pathways across the three groups were analyzed using the MapMan software. The largest numbers of DEGs were related to biosynthesis and signal transduction of auxin and ethylene, followed by DEGs in the cytokinin pathway (Figure 2;  Supplementary Table 2).
Differentially expressed genes related to ABA, BRs, JA, SA, and GA were also identified. For the ABA pathway, a total of 11 genes, including the biosynthesis genes AAO3 (rchiobhmchr5g0049511), NCED (rchiobhmchr4g0397001 and rchiobhmchr5g0027901), and signal transduction genes PP2C (rchiobhmchr2g0094051), were regulated by the ethylene or EW treatments. A total of 27 DEGs were found in Group 3, referred to as brassinosteroid biosynthetic and signaling pathways, these included biosynthetic genes DEF4 (rchiobhmchr2g0101791), CYP90D1 (rchiobhmchr2g0142901), CAS1 (novel00019), and signaling transduction genes BEH4 (rchiobhmchr5g0011971) and BRH1 (rchiobhmchr7g0197001). Among the JA-related genes, the upregulated DEGs, including AOS2 (rchiobhmchr6g0261031, rchiobhmchr6g0261011, and rchiobhmchr6g0300221), OPR2 (rchiobhmchr2g0167371), and LOX1 (rchiobhmchr3g0455111), were related to biosynthesis and detected in Group 3. In the SA pathway, the expression of UDP-glucosyltransferase gene UGT74E2 (rchiobhmchr6g0248381, rchiobhmchr6g0248411, rchiobhmchr7g0179611, and rchiobhmchr6g0248381) was enhanced, whereas SABATH methyltransferase expression (novel01210 and rchiobhmchr6g0265391) was decreased in Group 3. The expression of eight GA biosynthetic and four signal transduction genes showed changes in Groups 2 and 3. Among those DEGs, the biosynthetic genes were GA2 (rchiobhmchr5g0023471), GA20ox1 FIGURE 2 | Display of differentially expressed genes (DEGs) involved in hormone biosynthesis and signaling transduction pathway. Significant DEGs (fold changes ≥ 2 and adjusted p-value ≤0.05) were organized into functional categories (BINs) (Thimm et al., 2004) and visualized using the MapMan software (version 1.0). The color set scale was placed on the top right, and the gene expression increased is indicated in red and decreased in blue. Detailed information on gene expression is listed in Supplementary Table 2.

Screening of Candidate Genes Based on Co-expression Network
To further identify the major groups, we performed a WGCNA and obtained 13 distinct modules ( Figure 4A). Focused on the rose senescence degree, we found that the "Yellow" module was the most significantly associated with rose senescence (Figure 4B). Moreover, we analyzed the expression profile of hub genes in the "Yellow" module and found that RchiOBHmChr5g0071811 was the most significantly upregulated gene when compared with control ( Figure 4C).
Phylogenetic and protein sequence analyses were performed (Supplementary Figures 3, 4), and RchiOBHmChr5g0071811 was closely related to AtWRKY33, which contains two WRKY domains. Furthermore, it was located in the nucleus by subcellular localization in Nicotiana benthamiana (Supplementary Figure 4B).
Based on these results, RchiOBHmChr5g0071811 was named RhWRKY33. RhWRKY33 accumulates in the nucleus, where it may function as a transcription factor.
To better understand its expression characteristics, the transcription level of RhWRKY33 was measured in different organs and petal stages through RT-qPCR. As shown in Supplementary Figure 4C, the expression level of RhWRKY33 was higher in the sepals, petals, and stamens, and lower in young leaves, receptacles, and pistils, whereas it significantly increased from stage 1 to stage 5 (Supplementary Figure 4D), implying that RhWRKY33 may function in rose petal senescence.

RhWRKY33 Accelerates the Petal Senescence Depended on Wounding and Ethylene Signaling
To evaluate whether RhWRKY33 was involved in regulating the senescence of rose petals, we silenced the RhWRKY33 transcripts in petal discs by VIGS and found that the color FIGURE 3 | Display of differentially expressed genes related to transcription factors. Significant DEGs (fold changes ≥ 2 and adjusted p-value ≤ 0.05) were organized into functional categories (BINs) (Thimm et al., 2004) and visualized by MapMan software (version 1.0). The color set scale was placed on the top right, and the gene expression increased is indicated in red and decreased in blue. Detailed information on gene expression is listed in Supplementary Table 3. Frontiers in Plant Science | www.frontiersin.org fading of TRV discs began after 6 days, whereas the TRV-WRKY33 discs, with WRKY33 silenced, did not display any significant color fading ( Figure 5A); in contrast, the degree of the TRV disc fading was more serious after 9 and 12 days compared with the TRV-WRKY33 discs ( Figure 5D). The expression of RhWRKY33 was significantly lower in the TRV-WRKY33 discs than in the TRV control discs ( Figure 5B); Concurrently, we measured the expression of RhSAG12 and ion leakage in TRV and TRV-WRKY33 discs at 0, 6, 9, 12 days. Consistent with this, the WRKY33-silenced discs presented lower expression of RhSAG12 and less ion leakage than TRV control (Figures 5C,E). These data provide evidence that RhWRKY33 plays a positive regulatory role during rose petal senescence. further investigate whether RhWRKY33 was involved in EW signaling, we also performed ethylene, wounding, and EW treatments through the VIGS assay. As shown in Figures 6A-G, the fading color and ion leakage were significantly increased in the ethyne and EW treatment compared with wounding-treatment at 6, 9, and 12 DAT, either in TRV or in TRV-RhWRKY33, suggesting the stronger role of ethylene than wounding on petal senescence. Irrespective of the wounding, ethylene, or EW treatment, all RhWRKY33-silenced discs exhibited longer floral longevity, lower color fading rate, and ion leakage when compared with the TRV. In addition, the expression of RhSAG12 was significantly lower in RhWRKY33-silenced discs at 6 DAT than in TRV discs under the above treatments The error bars represent the means of three biological replicates ± SDs, RhUBI2, and RhActin5 were used as reference genes, and the asterisks indicate significant differences based on Student's t-test (*p < 0.05; **p < 0.01). RhWRKY33-silenced and TRV control petal discs under different treatments. The error bars represent the means of three biological replicates ± SDs, RhUBI2, and RhActin5 were used as reference genes, and the asterisks indicate significant differences based on Student's t-test (*p < 0.05; **p < 0.01; ***p < 0.001).
( Figure 6). These results indicate that RhWRKY33 can accelerate the senescence process of rose petals depending on the wounding and ethylene signaling, but mainly on ethylene signaling.
We previously found that JA and ethylene could promote senescence in rose flowers (Zhang et al., 2019b), and JA levels increased significantly in plants subjected to wounding (Chung et al., 2008;Glauser et al., 2008;Koo et al., 2009); whether RhWRKY33 was induced by JA remains unknown. To examine the relationship between JA and RhWRKY33, we measured the expression of RhWRKY33 in JA-and ethylene-induced petals and found that RhWRKY33 was upregulated by JA and ethylene treatment (Supplementary Figure 5). These results further indicate that RhWRKY33 accelerates the petal senescence process mediated by both ethylene and wounding signaling, but relies heavily on ethylene signaling.

Ethylene Is a Senescence-Promoter Hormone in Rose Petals
Phytohormones possess pivotal roles in initiating and modulating floral senescence programs and finally influence the quality and long vase life of fully open petals. Among those hormones, ethylene and ABA function as the senescenceaccelerating factors, while auxin, CTKs, SA, and GA act the inhibited function in floral senescence (Lu et al., 2014;Trivellini et al., 2015;Wu et al., 2017;Shabanian et al., 2018). Particularly, ethylene has been considered as a major endogenous regulator of fruit ripening, leaf-and petalsenescence, and floral organ abscission Han et al., 2016;Shabanian et al., 2018). Once the senescence process of leaves and flowers was initiated, the expression of ethylene biosynthesis genes ACS1/2 was strongly increased, accompanied by a massive burst of ethylene production (Ma et al., 2005). In this study, two ethylene biosynthesis genes, ACS (rchiobhmchr6g0310191) and ACO (rchiobhmchr3g0474401, rchiobhmchr5g0001831), had higher levels of expression in all the ethylene treatment flowers (group 2 and 3) (Figure 2; Supplementary Table 2). Meanwhile, we also found that the expression of eight ERF genes (rchiobhmchr4g0392451, rchiobhmchr4g0392501, rchiobhmchr5g0008991, rchiobhmchr 5g0009711, rchiobhmchr5g0032721, rchiobhmchr6g0288231, rchiobhmchr6g0299771, and rchiobhmchr7g0204641) were increased, and three ERF genes (rchiobhmchr2g0135921, rchiobhmchr6g0298011, and rchiobhmchr7g0230931) were decreased in response to the ethylene signal, and RhERF1 (rchiobhmchr2g0135921) and RhERF4 (rchiobhmchr6g0298011) have been proved to function in petal abscission mediated by ethylene pathway (Gao et al., 2019).
The exogenous application of ethylene also triggers the gene expression change related to the biosynthesis and signaling pathway of other hormones. During the leaf senescence and abscission process, the massive production of ethylene could inhibit auxin synthesis and transport, enhance auxin degradation, and lower diffusible auxin levels (Burg, 1968).
Here, up to 55 DEGs associated with the auxin pathway were induced and 76.4% of those genes (42/55) were downregulated by exogenous ethylene applying, including auxin polar transport PIN family genes (PIN5, rchiobhmchr3g0448071; PIN8, rchiobhmchr2g0169101), auxin-responsive SAUR, and the genes of GH3 family (Figure 2; Supplementary Table 2). We speculate that ethylene may directly or indirectly regulate the expression of auxin-related genes to affect flower senescence. Meanwhile, the exogenous application of ethylene further enhanced senescence-promoting hormone responses, such as ABA, and while declined the genes expression associated with synthesis and signaling transduction of cytokinin, which functions in delaying senescence.

Wounding Function in Acceleration Floral Senescence Depend on Ethylene Signaling
Numerous studies have revealed that wounding is implicated in multiple signaling pathways, such as JA, auxin, and ethylene. Wounding is considered one of the early triggers of plant regeneration , and the content of JA and auxin accumulates within seconds to minutes after wounding, making them candidates for early wounding signal. Further analysis indicated that cell damage leads to JA accumulation, and tissue integrity loss leads to local auxin accumulation by impeding polar auxin transport (Glauser et al., 2009;Larrieu et al., 2015;Zhou et al., 2019). Endogenous JA levels acceleration in response to mechanical wounding and herbivore damage have been extensively reported (Creelman and Mullet, 1997;Rao et al., 2000). JA signal marker genes, namely, LOX2, AOS, and VSP, were found to be induced, whereas endogenous levels of auxins were also shown to decline after wounding, showing a significant cross-talk between the hormone and damage signaling pathways (Devoto and Turner, 2003;Zou et al., 2016). Previous studies confirmed that the damage signal is one of the primary factors in the induction of the senescence phenotype in detached leaves or petals, whereas ethylene is implicated as a regulator of the rate of the process (Philosophhadas et al., 1991). Additional evidence has revealed that ethylene content is generally induced by various environmental stresses, such as damage signaling, exposure to low temperatures, and water stress, in plants (Hyodo et al., 1991).
Our previous study revealed that RhMYB108 could accelerate rose petal senescence by mediating JA and ethylene signaling (Zhang et al., 2019b;Fan et al., 2020). In this study, we found that the rate of petal senescence was not significantly accelerated or delayed in the single wounding treatment. However, the senescence rate of EW-treated petals was faster than those treated by wounding or ethylene, which suggested that wounding could strengthen ethylene signaling, which was consistent with those of a study by Hyodo (Hyodo et al., 1991). In conclusion, further research is needed to reveal the mechanisms underlying the wound-ethylene crosstalk in petal senescence. In addition, our study suggested that RhWRKY33 could accelerate the floral senescence process of cut roses by mediating EW-signaling (Figures 1, 6).
Up to date, increasing studies have shown that many WRKY family members are involved in stress-defense responses, plant senescence, wounding, seed dormancy, and seed germination (Rushton et al., 2010;Rinerson et al., 2015;Zhao M. et al., 2020). In addition, a complex regulatory network consisting of WRKY TFs and plant hormone functions is involved in plant senescence. In our study, the expression of RhWRKY33 was significantly upregulated more than 9 and 6 times, when induced by JA and ethylene, respectively. However, it was significantly repressed by 1-MCP (Supplementary Figure 5). We speculate that RhWRKY33 may be induced by wounds by increasing the JA content. However, in this study, we discovered that the expression of RhWRKY33 was upregulated more than two times under wounding treatment, and less than nine times under JA treatment. Whether there are other factors involved in the regulation of rose petal senescence requires further study.
Overall, using RNA-seq analysis and WGCNA, we identified the RhWRKY33, which is closely related to AtWRKY33 from Arabidopsis thaliana and is located in the nucleus. The silencing of RhWRKY33 (TRV-RhWRKY33) significantly delayed the rose petal senescence process either on the control or on other treatments (ethylene, wounding, and EW). Ethylene plays a major role in petal senescence. From the above results, we conclude that RhWRKY33 is involved in regulating rose petal senescence by mediating woundand ethylene-signaling.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are publicly available. This data can be found here: National Center for Biotechnology Information (NCBI) BioProject database under accession number PRJNA550484 (https://www.ncbi.nlm.nih.gov/bioproject/? term=PRJNA550484).

AUTHOR CONTRIBUTIONS
WJ and QZ performed the experiments and analyzed the data. YLi and YLiu designed the study. SZ, DZ, JX, HZ, and FW provided technical support and conceptual advice. WJ, QZ, SZ, and YLi wrote the manuscript. All authors read and approved the final manuscript.

FUNDING
This study was supported by the General Project of Shenzhen Science and Technology and Innovation Commission (21K270360620) and the National Natural Science Foundation of China (Grant No. 31902054).

SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2021. 726797/full#supplementary-material Supplementary Figure 1 | The ethylene production of different treatments was determined at 0, 1, 3, and 6 days. Values are means ± SD, and asterisks indicate statistically significant differences ( * p < 0.05, * * p < 0.01, Student's t-test).  Supplementary Table 4 | Detailed information of DEGs relating to transcription factors, protein modification, protein degradation (A), and hormone signaling (B) was co-regulated between Groups 2 and 3.
Supplementary Table 5 | Detailed information of DEGs relating to transcription factors, protein modification, protein degradation (A), and hormone signaling (B) was specifically regulated by group 3.