Melatonin Relieves Ozone Stress in Grape Leaves by Inhibiting Ethylene Biosynthesis

Ozone (O3) stress severely affects the normal growth of grape (Vitis vinifera L.) leaves. Melatonin (MT) plays a significant role in plant response to various abiotic stresses, but its role in O3 stress and related mechanisms are poorly understood. In order to understand the mechanism of MT in alleviate O3 stress in grape leaves, we perform a transcriptome analyses of grapes leaves under O3 stress with or without MT treatment. Transcriptome analysis showed that the processes of ethylene biosynthesis and signaling were clearly changed in “Cabernet Sauvignon” grapes under O3 and MT treatment. O3 stress induced the expression of genes related to ethylene biosynthesis and signal transduction, while MT treatment significantly inhibited the ethylene response mediated by O3 stress. Further experiments showed that both MT and aminoethoxyvinylglycine (AVG, an inhibitor of ethylene biosynthesis) enhanced the photosynthetic and antioxidant capacities of grape leaves under O3 stress, while ethephon inhibited those capacities. The combined treatment effect of MT and ethylene inhibitor was similar to that of MT alone. Exogenous MT reduced ethylene production in grape leaves under O3 stress, while ethephon and ethylene inhibitors had little effect on the MT content of grape leaves after O3 stress. However, overexpression of VvACO2 (1-aminocyclopropane-1-carboxylate oxidase2) in grape leaves endogenously induced ethylene accumulation and aggravated O3 stress. Overexpression of the MT synthesis gene VvASMT1 (acetylserotonin methyltransferase1) in tobacco (Nicotiana tabacum L.) alleviated O3 stress and reduced ethylene biosynthesis after O3 stress. In summary, MT can alleviate O3 stress in grape leaves by inhibiting ethylene biosynthesis.


INTRODUCTION
Ozone (O 3 ) in the troposphere is a highly oxidizing atmospheric pollutant. Elevated O 3 concentration severely affects the growth and development of plants (Serengil et al., 2011), as well as human health (Karnosky et al., 2007;Borowiak, 2013). At present, the near-surface O 3 concentration is increasing at an annual rate of 0.5-∼2.0% (Vingarzan, 2004) and is projected to increase by 40-60% at the end of the 21st century, when the tropospheric O 3 concentration will reach 80 nL L −1 (Fiscus et al., 2005). O 3 stress induces the release of large amounts of ethylene from leaf stomata, the damage of plant leaves caused by O 3 is correlated with the release of ethylene (Tingey et al., 1976;Mehlhorn and Wellburn, 1987). As an important signal molecule, ethylene plays an important role in plant response to abiotic stress (Zhang M. et al., 2016). Ethylene biosynthesis begins with the formation of S-adenosyl-Lmethionine (SAM) from methionine by SAM synthetase. Then, 1-aminocyclopropane-1-carboxylic acid (ACC) synthase (ACS) catalyzes SAM to produce ACC, and ACC oxidase (ACO) oxidizes ACC to ethylene (Najeeb et al., 2018).
Grapes (Vitis vinifera L.) are an important fruit crop grown worldwide. Previous investigation suggested that O 3 stress adversely affects the photosynthetic system of grape (cultivar Cabernet Sauvignon) leaves, and exogenous MT treatment can alleviate O 3 stress (Geng et al., 2016;Chen et al., 2020). Grapes and other plants have evolved various strategies to withstand abiotic stresses, such as regulating interactive hormone networks, including MT and ethylene (Arnao and Hernández-Ruiz, 2015;Ryu and Cho, 2015;Nguyen et al., 2016).
Melatonin is an indole derivative of tryptophan that is ubiquitous in plants and animals and has high efficiency, conservation, and strong antioxidant effects (Galano et al., 2011;Yang et al., 2018). MT synthesis in plants requires the participation of many enzymes, and the last step of the reaction is catalyzed by acetylserotonin o-methyltransferase (ASMT) (Kang et al., 2011). MT is an essential plant growth regulator, and both external application and endogenous induction can improve plant tolerance to drought, salinity, and other abiotic stresses (Zuo et al., 2014;Arnao and Hernández-Ruiz, 2015). For example, exogenous MT treatment can enhance the antioxidant capacity of cabbage (Brassica oleracea L.)  and tea tree (Camellia sinensis L.) (Li et al., 2021) by increasing their anthocyanin content. In apple (Malus domestica), MT can regulate reactive oxygen species (ROS) signaling and activate the CBL1-CIPK23 (calcineurin B-like 1-interacting protein kinases23) pathway to regulate the expression of potassium channel protein genes, thereby improving salt tolerance . Furthermore, exogenous MT can reduce the ion poisoning of mushrooms (Agaricus campestris)  and wheat (Triticum aestivum L.) (Al-Huqail et al., 2020). MT can also interact with other plant hormones [abscisic acid (ABA), jasmonic acid, salicylic acid, ethylene, etc.] to form a significant component of the plant immune system (Arnao and Hernández-Ruiz, 2018). For example, MT can increase GA (gibberellin) and reduce ABA content by regulating the expression of GA and ABA synthesis-related genes in cucumber seedlings and alleviating the inhibitory effect of a high salt environment on seedlings (Zhang et al., 2014). In Arabidopsis thaliana, MT reduces root meristem size by inhibiting auxin synthesis and polar transport (Wang Q. et al., 2016) and interacts with ethylene signaling pathways to improve disease resistance (Lee et al., 2014).
So far, the relationship between MT and other signaling molecules under abiotic stress is obscure, especially under O 3 stress. Thus, this experiment has explored the key metabolic changes caused by increasing the MT content in grape leaves under O 3 stress and its possible action mechanism. This research will promote the application of MT in improving the O 3 tolerance of grapes and reveal the potential molecular mechanism of MT in regulating other signal molecules under O 3 stress.

O 3 Fumigation System
Two O 3 fumigation systems were set up in the school vineyard (36 • 11 N, 117 • 06 E), which were divided into four parts, including open-top air chambers (OTCs), gas supply systems, O 3 generation system, and O 3 concentration monitoring system (Geng et al., 2017). The mainframe of the OTC is composed of galvanized steel pipes with a 3 cm diameter and is divided into two parts: the lower part is a regular octagonal prism with length and height of 1.1 and 2.2 m. The upper part is a regular octagonal pedestal; the area of the upper base of the pedestal is one in third of the area of the lower base, and the angle between the side and the vertical is 45 • . The top is open to the atmosphere, and the sides are covered with particular polyethylene plastic film for the greenhouse; the outside is covered with a sunshade net. The installation height of the LED light source (LED cold light source plant light, SP501-N, 405 W, Shanghai Sanhao Electromechanical Co., Ltd.) is 1.5 m. To ensure the stability of the gas concentration in the OTC, the gap between the exhaust ports is gradually reduced from the center of the OTC to the four sides. The O 3 generating system (WJ-HY5, Jinan Sankang) is a high-frequency O 3 generator. The oxygen intake of the O 3 generator can be modulated by adjusting the rotor flowmeter to control the O 3 concentration. The O 3 concentration monitor (DR70C-O 3 type) in the OTC was used to measure the O 3 concentration in real-time and transmits the data to the computer for observation and storage.

Plant Materials, Growth Conditions, and Experimental Treatments
Two-year-old potted seedlings of grapevine cultivar "Cabernet Sauvignon" were used to explore the effects of exogenous MT and ethylene on grape leaves under O 3 stress. Cuttings were planted in cylindrical pots with a diameter of 25 cm and a height of 35 cm (substrate:sand:soil = 2:1:1). The potted seedlings were cultivated in a greenhouse. When the new shoot leaves grew to 10-12 pieces, the plants with the same growth potential were selected and treated with water, 50 µM MT (Xu et al., 2019), 250 mg L −1 ethephon (Ma et al., 2021), or 2 µM aminoethoxyvinylglycine (AVG) (Xu et al., 2019) every 2 days at 6 p.m. (three times in total), and each treatment (1.5 L) was replicated in five plants. After that, the plants were exposed to 110 nL L −1 O 3 for 3 h at 800 µmol m −2 s −1 light intensity at 8 a.m. (Geng et al., 2016(Geng et al., , 2017. All treatments were as follows: The roots were irrigated with clean water and the leaves sprayed with clean water without O 3 treatment (control); the roots were irrigated with clean water, the leaves sprayed with clean water, and then plants were exposed to O 3 (O 3 ); The roots were irrigated with 50 µM MT, the leaves sprayed with water, and then plants were exposed to O 3 (MT + O 3 ); the leaves were sprayed with 250 mg L −1 ethephon, the roots irrigated with clear water, and then plants were exposed to O 3 (Ethephon + O 3 ); The leaves were sprayed with 2 µM AVG, the roots irrigated with clean water, and then plants were exposed to O 3 (AVG + O 3 ); the roots were irrigated with 50 µM MT, the leaves sprayed with 2 µM AVG, and then plants were exposed to O 3 (MT + AVG + O 3 ). After the treatment, leaves with similar nodes and sizes were selected for RNA-Seq analysis and determination of physiological indexes.
"Cabernet Sauvignon" tissue culture seedlings were used to evaluate the effect of increasing endogenous ethylene content on grape leaves under O 3 stress. Healthy apical growth tips of "Cabernet Sauvignon" vines were removed in early summer to establish grapevine in vitro shoot cultures. The plant materials were sterilized (75% alcohol for 2 min, 4% sodium hypochlorite for 15 min) and cultured on MS medium supplemented with 30 g L −1 sucrose, 7.5 g L −1 agar powder, and 0.2 mg L −1 indole-3-butyric acid (IBA). The plants were kept in a growth chamber maintained at 25/20 • C, with a photoperiod of 16 h light/8 h dark, and branches with at least one bud and leaf were used for subculture every month. Healthy 2-month-old seedlings with consistent growth were selected for infection treatment. The tissue culture bottle caps were opened one week before treatment adapt the seedlings to the external environment gradually.

Chlorophyll Fluorescence Imaging and Determination of Related Enzyme Activities and Physiological Indexes
Rapid chlorophyll fluorescence imaging of grape leaves was performed using a fluorescence imaging system (PSI, Czechia). Hydrogen peroxide (H 2 O 2 ) contents in leaves were estimated using the trichloroacetic acid (TCA) method at 390 nm (Velikova et al., 2000). Superoxide radical (O 2 − ) was measured as described by Elstner and Heupel (1976) by monitoring the nitrite formation from hydroxylamine in the presence of·O 2 − . The tissue staining methods of H 2 O 2 and O 2 − were according to Thordal-Christensen et al. (1997) and Orozco-Cardenas and Ryan (1999), respectively. Reduced ascorbic acid (AsA) content was measured by bipyridine colorimetry, while reduced glutathione (GSH) was determined using 5,5 -Dithiobis (2-nitrobenzoic acid) (DTNB) (Zhou et al., 2005). The total glutathione content was determined by the method of previous descripted (Kosugi and Kikugawa, 1985). Oxidized glutathione (GSSG) content was calculated by the difference between total glutathione content and GSH content, and then GSH/GSSG value was obtained. Determination of superoxide dismutase (SOD) was by photochemical reduction of nitroblue tetrazolium (NBT) (Giannopolitis and Ries, 1977). The SOD activity unit U was 50% inhibition of NBT photochemical reduction. Peroxidase (POD) activity was determined by monitoring the increase in absorbance at 470 nm, caused by guaiacol oxidation (Scebba et al., 2001). One unit of POD activity was defined as the change of A470 by 0.01 per min. The activity of catalase (CAT) was determined according to the method of Cakmak and Marschner (1992). CAT can decompose H 2 O 2 , and H 2 O 2 has a strong absorption peak at 240 nm wavelength, reducing A 240 by 0.1 per minute to a unit (U) of CAT enzyme activity. The chlorophyll content was measured by UV (ultraviolet) spectrophotometry (Yang et al., 2009), while the activity of ascorbate peroxidase (APX) was measured as previously reported (Nakano and Asada, 1981). One unit of activity for APX was defined as the amount of enzyme that degraded 1 µmol of AsA per min. The activities of glutathione reductase (GR), monodehydroascorbate reductase (MDHAR), and dehydroascorbate reductase (DHAR) were measured with a kit (Keming Biotechnology Co., Ltd., Suzhou, China). One unit of activity for GR was defined as catalyzing the oxidation of 1 nmol NADPH (nicotinamide adenine dinucleotide phosphate) per gram of sample per min. One unit of MDHAR activity was defined as 1 nmol NADH (nicotinamide adenine dinucleotide) oxidized per min per gram of sample. One unit of DHAR activity was defined as 1 nmol AsA produced per min per gram of sample.

Determination of MT, Ethylene, and ACC Contents
The MT content was determined in reference to the previously reported method (Qianqian et al., 2015), with some modifications. The sample weight was 3.0 g, and the MT was extracted with analytical grade methanol; the final extract was purified by the C 18 solid-phase extraction cartridge (ProElut TM , DIKMA, China) with the help of a vacuum pump, and then the volume was adjusted to 1 mL. The ethylene production rate was measured by gas chromatography (Shimadzu GC-16A, Japan) and repeated three times (Farmer et al., 1986). In order to avoid the wounding effect on ethylene production, the wound was wrapped in cotton with water after sampling and then sealed with a sealing film. The leaves were immediately put into a container and sealed, maintained in a light incubator at 25 • C for 24 h, and then the gas was extracted into a 1 mL syringe for determination. The extraction and determination of ACC were according to Tucker et al. (2010). After O 3 treatment, the samples were stored in liquid nitrogen.

RNA-Seq and Quantitative Real-Time PCR
Transcriptome sequencing was conducted by OE Biotech Co., Ltd. (Shanghai, China). Total RNAs were extracted using TRIzol Reagent (Invitrogen, Carlsbad, CA, United States), and the mRNA was enriched using magnetic beads containing Oligo (dT). The quality of the constructed gene library was checked by the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, United States). After passing the quality test, the HiSeq X Ten sequencer of Illumina Company was used for sequencing, and the double-terminal data of 150 bp was produced. Raw data (raw reads) were processed using Trimmomatic (Bolger et al., 2014). The reads containing ploy-N and the low quality reads were removed to obtain the clean reads. Then the clean reads were mapped to reference genome 1 using hisat2 (Kim et al., 2015). The reads were reassembled using StringTie (Pertea et al., 2015). The protein-coding gene expression was calculated in FPKM (Fragments Per kb Per Million Reads). The default screening difference condition was P < 0.05 and log 2 (fold change) >1. FDR (false discovery rate) error control method was used for P-value multiple hypothesis testing and correction. GO (Gene Ontology) enrichment and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway enrichment analysis of differentially expressed genes (DEGs) were respectively performed using R based on the hypergeometric distribution. The quantitative real-time PCR (qRT-PCR) analysis was carried out with the UltraSYBR Mixture kit (CWBIO, Beijing, China) on a Bio-Rad iQ5 (Hercules, CA, United States) instrument. The reaction mixture was 20 µL: double distilled water (ddH 2 O) 7.0 µL, forward primer (10 µmol L −1 ) 1.0 µL, reverse primer (10 µmol L −1 ) 1.0 µL, 2 × UltraSYBR Mixture 10.0 µL, cDNA 1.0 µL. The Vvactin gene was used as an internal reference. The relative quantitative gene expression values were calculated using the 2 − CT method from three replicates. The primer sequences used are shown in Supplementary Table 1.

Genetic Transformation of ACO2 and ASMT1 in Grape and Tobacco
The open reading frames (ORFs) of ACO2 and ASMT1 from "Cabernet Sauvignon" leaves were cloned and then respectively ligated to the pRI101-AN expression vector driven by the 35S promoter. Then, the plasmid was transformed into Agrobacterium tumefaciens strain GV3101 by the heat shock method. The "Cabernet Sauvignon" tissue culture seedlings were immersed in an Agrobacterium suspension adjusted to OD 600 = 0.6, placed in a closed container, and then the bacteria solution was completely immersed in grape leaves (with obvious water stains) by vacuum extraction. The bacterial suspension on the leaf surfaces was dried, and the seedlings were cultured in bottles with medium. After 2 days, a qRT-PCR analysis was done to detect the expression level, and the overexpression strain was used for the O 3 treatment experiment. The plants infected with an empty carrier were used as a control, and each line was set with three replicates. The tobacco was infected by the leaf disc method (Wang F. et al., 2016), and the T0 tobacco plants overexpressing VvASMT1 were obtained after screening in selection medium. The transgenic lines were further identified by PCR and confirmed by qRT-PCR, after which T2 transgenic lines were obtained for experimental treatment.

Statistical Analysis
All statistical analyses were performed by SPSS 24.0 software. A one-way analysis of variance (ANOVA) followed by Duncan's multiple range test was employed, standard deviation (SD) was calculated from three replicates. The differences between individual means were deemed to be significant at P < 0.05.

Exogenous MT Inhibits the Ethylene Biosynthesis and Signaling Caused by O 3 Stress
To explore the mechanism by which MT alleviates O 3 stress in grape leaves, RNA-Seq analysis was performed on "Cabernet Sauvignon" grape leaves treated with control, O 3 , and MT + O 3 . DEGs were represented by a Venn diagram (Figure 1A). Compared with the control, O 3 significantly (P-value < 0.05) up-regulated and down-regulated 5121 and 2935 genes in grape  GO enrichment analysis showed that the DEGs were primarily associated with biological regulation, cellular process, metabolic process, signaling, cell, cell part, membrane, membrane part, binding, catalytic activity, nucleic acid binding transcription factor activity, and transporter activity (Figures 1B,C). KEGG enrichment analysis indicated that all of the annotated DEGs were primarily related to signal transduction, amino acid metabolism, biosynthesis of secondary metabolites, carbohydrate metabolism, and environmental adaptation (Figure 1D). In the classification of signal transduction pathways, the most apparent change in the number of DEGs occurred in the ethylene signal pathway (Supplementary Tables 5, 6). Compared with the control, O 3 resulted in significant changes in the expression levels of 71 genes related to ethylene biosynthesis and signaling pathways. Except for the obvious down-regulation of three genes, all the others were up-regulated; the expression of ACO2 was up-regulated by 5.2-fold (Supplementary Table 5). Compared with O 3 treatment, MT + O 3 caused significant changes in the expression of 38 ethylene biosynthesis and signaling pathway-related genes. Among these 38 genes, 22 were significantly down-regulated, and ACO2 was down-regulated by 1.03-folds (Supplementary Table 6). When the expression levels of these 22 genes in O 3 treatment and control were compared, it was found that 19 had higher expression levels than the O 3 treatment (Supplementary Table 7). qRT-PCR analysis was done to further determine changes in the expression of ethylene-related genes under O 3 and MT + O 3 treatments. The results showed that the expression levels of 11 genes increased significantly with the extension of O 3 treatment time, among which the expressions of ACO2 and ERF16 (ethylene-responsive transcription factor16) were up-regulated by 12.7 and 12.9-folds, respectively ( Figure 1E). During stress, the relative expression levels of all detected genes under the MT + O 3 treatment were significantly lower than that under the O 3 treatment (Figure 1E), which was similar to the transcriptome analysis results.

MT Relieves O 3 Stress by Regulating the Ethylene Pathway
To determine the effects of MT and ethylene on grape leaves under O 3 stress, "Cabernet Sauvignon" grapes were treated differently. O 3 stress caused more yellowing spots on grape leaves, and ethephon aggravated leaf damage symptoms under O 3 stress, causing obvious chlorosis. In addition, AVG, MT, and a combination of the two treatments significantly reduced the leaf injury symptoms after O 3 stress, and the yellowing area was smaller (Figure 2A). To further explore the mutual influence of MT and ethylene under O 3 stress, the contents of MT and ethylene under different treatments were determined (Figures 2B-D). The results showed that compared with the control, O 3 stress increased the ethylene release rate and the ACC content of "Cabernet Sauvignon" leaves by 100.8 and 82.19%, respectively. Meanwhile, the ethylene production after MT + O 3 treatment was significantly less than in the O 3 treatment (Figures 2B,C). Compared with the control, the MT content after O 3 stress was significantly reduced by 59.25% (Figure 2D). After watering, the MT content in grape leaves increased by 97.64% relative to the control (Figure 2D). Compared with O 3 treatment, the MT content after MT + O 3 and MT + AVG + O 3 treatment increased by 41.48 and 35.7%, respectively; however, there was no significant difference between ethephon + O 3 , AVG + O 3 , and the O 3 treatment ( Figure 2D).

Effects of MT and Ethylene on F v /F m and Reactive Oxygen Species in Grape Leaves After O 3 Stress
Compared with the control, the F v /F m of grape leaves after O 3 stress decreased by 28.01% (Figure 3A)

Effects of MT and Ethylene on Antioxidant System in Grape Leaves After O 3 Stress
Ascorbic acid and GSH are antioxidants involved in scavenging of active oxygen free radicals under stress conditions. Compared with the control, O 3 stress significantly reduced GSH, AsA contents, and GSH/GSSG in grape leaves, while increased GSSG content (Figures 4A-D). Compared with O 3 treatment, the GSH, AsA content, and GSH/GSSG were increased after MT + O 3 , AVG + O 3 , and AVG + MT + O 3 treatments. However, ethephon + O 3 treatment reduced GSH, AsA content, and GSH/GSSG, but increased the content of GSSG (Figures 4A-D).
Compared with the control, the GR, CAT, SOD, and POD activities in grape leaves increased significantly after O 3 stress. In contrast, the activities of MDHAR, DHAR, and APX were significantly inhibited (Figures 4E-K). Compared with the O 3 treatment, MT + O 3 , AVG + O 3 , and AVG + MT + O 3 treatments increased the GR, MDHAR, DHAR, APX, CAT, SOD, and POD activities of grape leaves, but the difference between AVG + MT + O 3 and MT + O 3 treatment was not significant (Figures 4E-K). Compared with the O 3 treatment, the SOD and POD activities of grape leaves after treatment with ethephon + O 3 were significantly increased, the activities of MDHAR and APX were inhibited. In contrast, the activities of GR, CAT, and DHAR did not change (Figures 4E-K). The above results indicate that both MT and AVG can alleviate the damage caused by O 3 stress on grape leaves by regulating the antioxidant system.

Overexpression of VvACO2 Intensifies O 3 Stress in Grape Leaves
To further verify that ethylene can exacerbate the stress effect of O 3 on grape leaves, the VvACO2 gene was transiently overexpressed in grape leaves to promote ethylene biosynthesis.  The expression levels of the two plants infected with 35S: VvACO2 carrier were 7.98 and 10.83-folds that of plants infected with the empty carrier, respectively ( Figure 5A). These results confirm the successful expression of VvACO2 in grape leaves. The ethylene release rate and ACC content were also significantly higher in the leaves of plants overexpressing VvACO2 than those of the control group (Figures 5B,C). Under normal conditions, the growth of the control and overexpression plants was the same. After 110 nL L −1 O 3 treatment for 3 h, the yellowing degree (Figure 5D), H 2 O 2 content (Figure 5E), and O 2 − production rate (Figure 5F) in the leaves of overexpressed plants were significantly higher than those of the control group. Meanwhile, the chlorophyll content ( Figure 5G) of overexpressed plant leaves was significantly lower than that of , and peroxidase (POD, K) activities in "Cabernet Sauvignon" grape leaves. Values represent the mean of three replicates ± SD. For the values labeled with the same lowercase letter, the difference was not significant, according to Duncan's multiple range test at a 5% significance level. the control group. These results show that the endogenous induction of ethylene biosynthesis in grape leaves aggravated O 3 damage to the leaves.

Overexpression of VvASMT1 Enhances Tobacco Tolerance to O 3
To analyze the effect of endogenous MT on plant O 3 tolerance, the overexpression vector 35S: VvASMT1 was transformed into tobacco to increase the endogenous MT content and test its O 3 resistance. The upstream primer of the 35S promoter and the downstream primer for amplifying the ASMT gene were used for PCR amplification. The results showed no specific band in the control group, while specific bands appeared in the five tobacco lines. The band sizes were the sum of the 35S promoter fragment length and the ASMT ORF sequence length ( Figure 6A). The results indicated that VvASMT1 was successfully transferred into tobacco. Two lines with moderate expression and high expression, line 1 and line 3, respectively, were selected for functional analysis (Figure 6B). The results showed that the MT content in the leaves of the wild-type and the two transgenic tobacco lines were 0.2299, 0.4957, and 0.5676 ng g −1 , respectively ( Figure 6C). After treatment with 110 nL L −1 O 3 for 3 h, the tobacco showed yellowing and wilting symptoms, and the stress degree of transgenic tobacco was significantly lower than that of wildtype tobacco (Figure 6D). The leaf color in the transgenic lines after H 2 O 2 and O 2 − staining was significantly lighter than that of the wild-type (Figures 6E,F), indicating that the ROS content was significantly lower than that of the wild-type. Compared with the wild-type, overexpression of VvASMT1 significantly increased GSH, AsA content, and related antioxidant enzymes (GR, SOD, POD, CAT, MDHAR, DHAR, and APX) activities after stress (Supplementary Figure 1). After O 3 treatment, the ethylene release rate and ACC content of transgenic tobacco leaves were significantly lower than the wild-type (Figures 6G,H). The above results indicate that overexpression of the VvASMT1 gene in tobacco increased the MT content, alleviated the O 3 stress, and reduced the ethylene content after stress.

DISCUSSION
At present, only a few studies have examined the molecular mechanism of O 3 stress on fruit trees, especially grapes (Valletta et al., 2016). O 3 enters plant leaves mainly through the stomata and increases ROS production (Langebartels et al., 2002). Thus, ROS level can be an essential indicator to determine the degree of cellular oxidative stress (Collén et al., 2003). When large amounts of ROS are produced under stress, its clearance system gets damaged, resulting in membrane lipid peroxidation, massive chlorophyll degradation, hindrance of photosynthetic electron transfer, inactivation of antioxidant enzymes, and inhibiting plant growth development (Degl'Innocenti et al., 2007;Iriti and Faoro, 2008;Bose et al., 2014). PSII is the most sensitive component of the photosynthetic electron transport chain under O 3 stress (Tran et al., 2013). Similarly, our research showed that O 3 stress could damage grape PSII and significantly reduce the chlorophyll content in grape leaves. Stressed leaves also produced large amounts of ROS and changed the activity of antioxidant enzymes, resulting in yellowing and wilting of grape leaves. Further, O 3 stress significantly induced the upregulation of genes related to ethylene biosynthesis and some ethylene responsive transcription factors, as well as the increase of ethylene release rate and ACC content. In our study, increase in exogenous and/or endogenous ethylene content aggravated O 3 stress. Therefore, it was speculated that O 3 destroys the photosynthetic and antioxidant grapes systems by increasing the ethylene content.
Many studies have shown that MT can alleviate various abiotic stresses; plants with higher MT content are more tolerant to O 3 (Dubbels et al., 1995). The present study also proved that MT could relieve O 3 stress in grape leaves and improve various physiological indexes. Exogenous MT application significantly increased the F v /F m of "Cabernet Sauvignon" grape leaves under O 3 stress and alleviated leaf chlorosis. Similarly, MT can reduce  the damage of barley leaf photosystem II (PSII) under stress and maintain the chlorophyll content (Arnao and Hernández-Ruiz, 2009). These results may be related to the protection of chloroplast structure by MT (Zhao et al., 2016). Additionally, this experiment found that watering the roots of "Cabernet Sauvignon" with MT could increase the leaf MT content. Similarly, treating grapes with MT at the rhizosphere increased the MT levels of the roots, but also in the leaves, thus enhancing salt tolerance in "Crimson seedless" grapevines (Xu et al., 2019). This result suggests that the MT provided through external sources is absorbed in plants (Tan et al., 2007) and can accumulate in distant organs through long-distance transportation to tolerate abiotic stress. In addition, our experiment for the first time verified that overexpression of VvASMT in tobacco could alleviate O 3 stress by increasing the MT content. This result is similar to that obtained following overexpression of the key enzyme gene caffeic acid-O-methyltransferase (COMT) for MT synthesis, which increased the endogenous MT content of tomatoes and enhanced salt resistance .
Plants can resist ROS damage through a defense system composed of enzymatic and non-enzymatic ROS scavenging systems (Apel and Hirt, 2004). Antioxidant enzyme scavenging systems mainly include SOD, POD, CAT, etc., while nonenzymatic systems include the AsA-GSH cycle. GSH can catalyze the degradation of excess H 2 O 2 and activate various defense mechanisms by participating in redox signal transduction (Szalai et al., 2009). The increase in GR activity reduces the cellular glutathione pool, providing sufficient GSH for DHAR to reduce dehydroascorbate (DHA) to AsA (Noctor and Foyer, 1998). MT can enhance the scavenging ability of the ROS system to improve stress tolerance. For example, the application of 0.5 µmol L −1 MT can improve salt tolerance in tomatoes by increasing the antioxidant enzyme activity and the accumulation of AsA and GSH (Liu et al., 2015). Moreover, the leaves of O 3 -tolerant soybean varieties maintain higher AsA levels than susceptible varieties (Chutteang et al., 2015). Thus, varieties with high antioxidant contents in the leaves are more resistant to the O 3 damage. This experiment also found that MT could protect antioxidant enzymes and the AsA-GSH signaling system under O 3 stress, while ethylene had the opposite effect. MT may act as an antioxidant to antagonize ethylene and remove excessive ROS, thereby sharing the pressure of other antioxidants.
Melatonin has opposite regulatory effects on ethylene in different crop species. For example, MT can promote the ripening of grape berries by increasing the ethylene content (Xu et al., 2018) and enhancing the salt tolerance of grapes by promoting VviMYB108A-mediated ethylene biosynthesis (Xu et al., 2019). On the contrary, MT treatment reduces ethylene release and improves fruit quality by inhibiting the expression of genes related to ethylene biosynthesis during apple storage (Onik et al., 2020). In our experiment, MT treatment of O 3 stressed plants down-regulated the expression of ethylene biosynthesis and signal transduction genes in grape leaves. MT treatment also reduced the ethylene release rate and ACC content in O 3 stressed leaves, thus alleviating O 3 stress. The increase of endogenous MT in tobacco also reduced ethylene biosynthesis after stress and alleviated O 3 stress. However, MT treatment of 'Crimson seedless' grapevines roots increased the ethylene release rate of leaves (Xu et al., 2019). These contrasting effects of MT may be due to the multi-pathway characteristics of MT synthesis, making it functionally specific in the developmental stage, tissues, and organs. In addition, the physiological effects of MT are pleiotropic Zhang et al., 2014), and its regulation of ethylene could be indirect. The different inducing effects could also change the mutual regulation with other hormones; thus, the specific mechanism needs to be further explored. Although both MT and ethylene inhibitor significantly alleviated O 3 stress in grapes, the effect of MT treatment was better than treatment with ethylene inhibitor treatment. In addition, the combined treatment effect of MT and ethylene inhibitor was similar to that of MT alone, and the treatment of ethylene and ethylene inhibitor did not affect the MT content under O 3 stress. It can be seen that MT and ethylene may play upstream and downstream roles, respectively, in the signal pathway under O 3 stress.
Finally, as depicted in Figure 7, O 3 stress increased ROS content and decreased the photosynthetic and antioxidant capacities of grape leaves by inducing ethylene biosynthesis. Melatonin pretreatment or overexpression of ASMT1 can enhance the in vivo melatonin level, reduce ethylene production in grape leaves under O 3 stress and increase plant O 3 tolerance. In addition, overexpression of ACO2 in grape leaves decreased O 3 tolerance by increasing endogenous ethylene content. Taken together, MT can alleviate O 3 damage to grape leaves by inhibiting ethylene biosynthesis.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are publicly available. This data can be found here: NCBI repository, accession number: PRJNA733572.

AUTHOR CONTRIBUTIONS
CL performed most parts of the experiment, analyzed the data, and wrote the manuscript. HK and YW participated in performing the experiments. ZG participated in the manuscript writing and revison. YD and YY designed the research. All authors have read and approved the final manuscript.