Based on NCBI (http://blast.ncbi.nlm.nih.gov/Blast.cgi) database for the reported OPRs in Arabidopsis (Biesgen and Weiler, 1999; Schaller and Weiler, 1997), the BLASTP and TBLASTP programs were used to search the Grape Genome Browser (http://www.genoscope.cns.fr/externe/GenomeBrowser/Vitis/) and Database (http://genomes.cribi.unipd.it/grape/), TAIR (http://www.arabidopsis.org/) and PlantTFDB (http://rice.plantbiology.msu.edu/) and a local protein database of homologous OPR genes in plants was constructed and OPR family genes in grapevine were obtained. A pairwise comparison of nucleotide and amino acid sequence similarity (%) between VvOPR family members was developed. The physical and chemical properties of the VvOPR family members, including the number of amino acids, molecular weight (MW), and isoelectric point (theoretical pI) were calculated using the online ExPaSy tool (http://www.ExPASy.org). The subcellular localization of the VvOPR proteins was predicted via the TargetP (https://services.healthtech.dtu.dk/services/TargetP-2.0/) and WoLF PSORT (https://wolfpsort.hgc.jp/) online servers.

A paired comparison analysis between VvOPRs was developed via ClustalX software (2.1) (Larkin et al., 2007). OPRs from grapevine, rice and Arabidopsis were used to construct a phylogenetic tree via the neighbor-joining (NJ) method using MEGA software 6.0 (Tamura et al., 2013). The MEME program (http://meme-suite.org/tools/meme) was used to identify the conserved motifs of the VvOPR family members (Bailey et al., 2009). The gene structure display server program (GSDS 2.0) was used to illustrate the exon/intron organization of the VvOPR genes (Guo et al., 2007). The 2 kb upstream sequences of all of the VvOPR genes were considered as the promoters, and these sequences were extracted for the prediction of cis-regulatory elements via PlantCARE (Mou et al., 2019) Each position of the VvOPR gene on the grapevine chromosome was mapped via the software MapInspect.

The plants (grapevine ecotype Maincure Finger, Arabidopsis thaliana ecotype Columbia L. and rice ecotype Wuyun Jing) were stored in our laboratory. The seeds of Arabidopsis and rice were surface-sterilized with 75% ethanol for 1.5 min, followed by 0.5% calcium hypochlorite for 20 min, and then rinsed at least three times with sterile distilled water. The sterilized seeds were placed on Murashige and Skoog (MS) medium with 1% agar and stratified in the dark at 4°C for 48 h. Then these plates were transferred to a controlled environmental chamber at a light intensity of ~120 μmol photons m^-2s^-1 at 22°C, maintained on a 16/8 h day/night cycle, with 70% relative humidity, and grown horizontally and/or vertically. After the rice seeds germinated, the seedlings were transferred into Hoagland’s liquid medium and/or pots filled with a 9:3:1 mixture of vermiculite/peat moss/perlite under the above conditions.

Cu/Zn treatments (Mikkelsen et al., 2012; Moravcová et al., 2018) were performed on grapevine, VvOPR1-overexpressing Arabidopsis and rice seedlings. Six-month-old grapevine seedlings were grown in Hoagland’s liquid medium supplemented with 10 μM CuSO₄·5H₂O or 0.1 mM ZnSO₄·7H₂O for 0, 3, 6, 9, 12 or 24 h. The whole grapevine plants from each treatment and/or the different organs (roots, stems and leaves) were sampled, frozen in liquid N₂, and stored at -70 °C. The germination rate was determined when VvOPR1-overexpressing Arabidopsis seedlings were grown horizontally in MS medium supplemented with Cu (150 or 200 μM) or Zn (1.25 or 1.5 mM) for 2 weeks, respectively. The germination rate was calculated as the proportion of seedlings for which the radicle had begun to emerge from the seed coat. The morphological characteristics (root length and fresh weight) were determined at various Cu (50, 100 μM)/Zn (0.75, 0.87 mM) concentrations after VvOPR1-overexpressing Arabidopsis seedlings were oriented vertically in Petri dishes containing MS medium incubated for 14 days. A Cu/Zn solution of deionized water was added to Hoagland’s liquid medium to obtain concentrations of Cu 0, 200, 600 and 900 μM/Zn 0, 2.5, 5.0 and 7.5 mM, and two-week-old transgenic rice seedlings were used to detect Cu/Zn tolerance. Then, 4-week-old transgenic rice seedlings were subjected to Cu (750 μM), Zn (7.5 mM), Cu+Zn (750 μM+7.5 mM) treatment for 5 d, and immediately transferred to liquid nitrogen and then a freezer (-80 °C) until further physical and chemical analyses.

Three seedlings were selected randomly from each treatment, and frozen grapevine and rice tissues (0.2 g) were ground thoroughly in liquid nitrogen using a mortar and pestle, respectively. Grapevine RNA extraction was performed via the modified CTAB method (Wang et al., 2011), and total RNA from VvOPR1-overexpressing rice was extracted from the disrupted tissues using the extraction medium (Qiagen) according to the manufacturer’s instructions. The RNA concentration was measured using a NanoDrop 2000 (Thermo Fisher Scientific, Waltham, MA, USA) and RNA integrity was assessed using an RNA Nano 6000 Assay Kit in conjunction with an Agilent Bioanalyzer 2100 system (Agilent Technologies, Santa Clara, CA, USA). The genomic DNA was removed by incubating the RNA sample in gDNA wipeout buffer (Qiagen) at 42 °C for 2 min before complementary DNA (cDNA) synthesis. First-strand cDNA was synthesized from 5 μg of total RNA with the QuantiTect Reverse Transcription Kit (Qiagen) in a 20 μL reaction volume according to the manufacturer’s instructions and it was stored at -20 °C until it was used in real-time quantitative PCR assays.

Quantitative real-time PCR was performed with a 7500 Fast Real-Time PCR system (Applied Biosystems), using SYBR Green chemistry. PCRs were conducted in a final volume of 20 μL containing 2 μL of a 1:10 dilution of cDNA sample, 10 μL of iQ SYBR Green Supermix, and 0.5 μmol/L of each primer. The thermal cycling conditions included initial denaturation at 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s, and 65 °C for 15 s for annealing and extension. PCR efficiency was checked for all primers used for gene expression analyses ( Supplementary Table S1 ). Only primers with higher amplification efficiency (>90%) were used in this experiment. The VvActin and OsUBQ5 genes were used as references in grapevine and rice seedlings, respectively. The relative expression levels of the target genes (antioxidative enzyme-encoding genes and ABA- and JA-related genes) were determined via the standard 2^−ΔΔCT method of Livak and Schmittgen (2001). All samples were tested in duplicate for the reference genes as well as for the genes of interest.

The full-length of VvOPR1 (XP_002281119) gene was artificially biosynthesized via PTDS method (Xiong et al., 2004). The synthesized VvOPR1 was cloned into the TA cloning vector Simple pMD-18 (pMD18-T) and the high integrity of the construct was confirmed by sequencing. The coding sequence was subsequently excised from the pMD18-T vector and inserted into the pCAMBIA1301 vector under the control of the CaMV 35S promoter. The constructed plant expression vectors were transformed into Agrobacterium (strain EHA105) via the freeze-thaw procedure, and strains containing this vector were transformed into the rice cultivar Wuyun Jing according to the methods of Tang (Tang et al., 2017). VvOPR1-overexpressing Arabidopsis was constructed via the floral dip method (Zhang et al., 2006). Positive transformants were selected via hygromycin (50 μg/ml) resistance and confirmed by PCR using the specific primers. In addition, VvOPR1 was transformed into the Arabidopsis NCED1 mutant which is defective in ABA synthesis to construct VvOPR1/NCED1 lines.

Three VvOPR1-overexpressing A. thaliana seedlings were selected randomly from each treatment, and their morphological characteristics (germination rate, root length and fresh weight) were measured and compared. Four-week-old OT and WT rice seedlings grown in Hoagland’s liquid medium supplemented with Cu (750 μM), Zn (7.5 mM), Cu+Zn (750 μM+7.5 mM) for five days were sampled for physical and chemical analyses. The Chl and Car contents were estimated via the Lichtenthaler (1987) and Wellburn (1994) methods. The root activity and MDA was determined as described by Wang et al. (2016b). The content of H₂O₂ was determined via the method of Zhou et al. (2018) with slight modifications. The free proline content was determined according to the methods of Bates et al. (1973) and Song et al. (2011). The antioxidants and antioxidant enzymes in the AsA-GSH cycle were detected refer to previous studies (Zhou et al., 2018; Geng et al., 2018; Thounaojam et al., 2012).

Rice leaves were extracted for ABA (Lou and Baldwin, 2003) and JA (Hung and Kao, 2003) analysis following previous studies with slight modifications. The crude extract was centrifuged and passed through polyvinylpyrrolidone column and C18 cartridges to remove plant pigments and other nonpolar compounds. JA extraction was analyzed by GC-MS with a doubly labeled internal standard ([1,2-¹³C]-JA). The ABA eluates were then concentrated to dryness by vacuum evaporation and resuspended in Tris-buffered saline. Afterwards, ABA was subsequently determined spectrophotometrically at 405 nm with an ABA immunoassay detection kit (model PGR-1; Sigma-Aldrich, St. Louis MO, USA).

For Cu²⁺/Zn²⁺ determination, both the aerial and root parts of the Cu/Zn-treated OT and WT seedlings were harvested, carefully rinsed with distilled water, oven dried at 80 °C for two days and ground (Li et al., 2017). The dried plant parts were digested with HNO₃ (11N) at 200°C for 10 h, and the digested samples were then diluted with HNO₃ (0.1 N) until a clear liquid formed. The filtrate was analyzed via ICP-OES (GBC INTEGRAXL, Australia) and standard solutions from MERCK and Analytika Praha were used to determine the metal content. The Cu²⁺ content was determined at a wavelength of 324.754 nm, and the Zn²⁺ content was measured at 213.856 nm (Tanhan et al., 2007). The TF, TI and AR for Cu²⁺/Zn²⁺ were detected using the following formulas developed by previous studies (Shu et al., 2002; Khan et al., 2015) based on the dry weight of the seedlings ( Supplementary Tables S1, S2 ).

The treatments of the experiments were arranged in a completely randomized block design, and three replicates were performed for each treatment. The data analysis was performed statistically using analysis of variance (ANOVA) via SPSS 17.0 and the data are presented as the means ± SEs (n=3). A statistically significant difference (p<0.05) was calculated by the least significant difference (LSD) test and is shown in the figures.

Based on NCBI database and literature, 121 homologous OPR genes were identified from 12 plant species representing six major plant lineages, including algae, moss, pteridophyta, gymnosperm, monocot and dicot ( Table 1 ). Nine VvOPRs (OPRs from Vitis vinifera) were identified from grapevine and named VvOPR1~9. The putative VvOPR genes were predicted to encode proteins ranging from 340 to 398 amino acids in length, the molecular weight (Mw) and isoelectric point (pI) of the VvOPR proteins ranged from 37.63 to 43.71 KDa and from 5.45 to 8.23, respectively ( Table 2 ). In terms of amino acid composition, positive amino acids accounted for the greatest percentage, followed by aliphatic amino acids and aromatic acids, and negative amino acids accounted for the lowest percentage ( Table 2 ). Almost all of the VvOPR proteins were predicted to accumulate in the cytoplasm (Cyto). The chromosomal positions of the VvOPR genes revealed that OPR genes were distributed in clusters, the VvOPR genes were distributed across two chromosomes with different densities ( Figure 1A ), the putative VvOPR3 genes were mapped to chromosome 11, and the other eight VvOPRs were mapped to chromosome 18. VvOPR1~2 mapped to the top of chromosome 18, and VvOPR4~9 mapped to the bottom of chromosome 18.

To evaluate the evolutionary relationships within the OPR gene family in grapevine, a rooted maximum-likelihood (ML) phylogenetic tree with 25 OPR genes from monocots (rice) and dicots (Arabidopsis and grapevine) ( Figure 1B ) was established. Phylogenetic analysis revealed that the OPR gene family can be subdivided into five well-conserved subfamilies, and these subfamilies are numbered sub.I to V. All OPR genes from grapevine, Arabidopsis and several OPRs from rice were grouped into sub.I and II, indicating that all OPR genes from angiosperms shared a common ancestor before the divergence between dicots and monocots. Several OPRs from rice were grouped into sub.III to V, showing that lineage-specific expansion and divergence events occurred in monocots after divergence from dicots and sub.III to V were generated exclusively in monocots. To explore different selective constraints on duplicated VvOPR genes, the Ka (synonymous nucleotide substitution), Ks (non-synonymous nucleotide substitution) and Ka/Ks ratios for each pair of duplicated VvOPR genes were calculated, and seven duplicated pairs of paralogous genes were identified in the nine VvOPRs. The results showed that the ratios of Ka/Ks for two duplicated pairs were <1, ranging from 0.1986 to 0.6999 ( Table 3 ). These results revealed that the gene duplication events occurred in these genes and that the functions of the duplicated genes did not diverge during genome evolution after the duplication events.

To further detect the origin and evolution of the VvOPR family genes, the nucleotide and amino acid sequences of the nine VvOPRs were blasted via pairwise alignment, respectively ( Table 4 ). The results revealed that the level of nucleotide and amino acid sequence identity among the VvOPRs varied from 28.3% to 98.1% and 44.4% to 97.4%, respectively. Five pairs of VvOPRs showed extremely high homology (the level of sequence identity between VvOPRs was above 90%), as follows: VvOPR4/VvOPR6 > VvOPR1/VvOPR2 > VvOPR7/VvOPR8 > VvOPR8/VvOPR9 > VvOPR7/VvOPR9. The level of nucleotide/amino acid identity between VvOPR3 and the other eight VvOPRs varied from 28.3% to 36.7%/44.4% to 51.9%. These results indicated that VvOPR3 may have undergone rapid evolution and functioned differently from the other eight VvOPR genes.

An exon-intron analysis was performed by comparing the predicted coding sequence (CDS) with the sequence of the VvOPR genes. As shown in Figure 1C , the VvOPR family members presented a highly conserved exon-intron structure, which contained 4~6 exons and 3~5 introns. VvOPR4 shared the longest intron sequence, an untranslated region of 53 kb. In addition, the nine VvOPR proteins shared at least four conserved motifs (motifs 6, 4, 2 and 3), which showed the same alignment. VvOPR7, VvOPR8 and VvOPR9 were completely consistent and motif 9 is characteristic of the three VvOPR proteins. Motif 7 was only deleted in VvOPR3. Phylogenetic analysis revealed that VvOPR members in the same branch were structurally conserved, indicating that motif 7 may play an important role in the evolution and function of VvOPRs. Taken together, the fact that the VvOPR proteins not only share very similar intron/exon structures but also contain common motifs supports their close evolutionary relationships and membership in the same subfamily. All of these results suggested that the strategies of VvOPR genes classification were relevant and reliable.

To further understand the biological regulation of the VvOPR genes, the cis-regulatory elements (CREs) in the upstream of these genes were identified via the PlantCare database. The cis-elements included plant growth and development elements, hormone responsive elements, and environmental stress responsive elements and so on ( Figure 1D ). The frequency of these elements in the regulatory region of each corresponding gene, as well as their overall frequency in family members, is very diverse. Most CREs involved in plant growth and development, such as Box 4 (52 times), GT1-motif (13 times) and I-box (8 times) are related to light-responsive elements. Hormone-responsive elements include those involved in ethylene response (ERE) and methyl jasmonate (MeJA). Classic plant hormones such as SA, ETH and MeJA are involved in the regulation and integration of plant immune responses against pests and pathogens.

Different tissues (roots, stems and leaves) of six-month-old grapevine (Maincure Finger) seedlings were sampled, and real-time quantitative PCR (qRT-PCR) was used to examine the VvOPR1 expression profile. The results revealed that various expression levels of VvOPR1 could be detected in all of the tissues tested, and the expression level of VvOPR1 was the highest in the roots and the lowest in the leaves ( Figure 2A ). As shown in Figure 2B , under Cu, Zn stress, VvOPR1 expression rapidly decreased throughout the 3 h period, and the expression of VvOPR1 was higher under Zn stress than under Cu stress. Then, VvOPR1 expression increased and peaked at 6 h, and the expression of VvOPR1 level was higher under Cu stress than under Zn stress from 6 h to 12 h, the plateau lasted from 9 h to 12 h testing period. Exposure to excess Cu, Zn stress could stimulate the overproduction of reactive oxygen species (ROS) and abscisic acid (ABA), the former impairs antioxidant defense systems and the latter triggers downstream stress-responsive pathways (Mittler et al., 2011). Here, the expression level of VvOPR1 almost mirrored the same trend after exposure to 10 mM H₂O₂ and 200 μM ABA, the expression level of VvOPR1 increased gradually from 0 h to 12 h under the treatment conditions ( Figures 2C, D ). All these results suggest that the expression pattern of VvOPR1 is tissue-specific and that its expression can be induced by various stresses.

VvOPR1 was synthesized via the PTDS method and transferred into both Arabidopsis and rice ( Figure 3A ). Several independent lines of transgenic seedlings (T1 generation) were subsequently generated. After selecting for hygromycin resistance, PCR was performed (T3 generation) ( Figure 3B ), the putative transgenic plants were confirmed, and several transgenic T3 plant lines (OT-1, OT-2 and OT-3) were chosen for further experiments. The phenotypes of the transgenic plants (T3 generation) were similar to those of the wild-type (WT) plants both on MS plates and in soil pots. These findings indicated that overexpression of VvOPR1 in Arabidopsis and rice caused no visible morphological changes.

The performance of WT and OT Arabidopsis seedlings under both normal and Cu and Zn stress conditions was compared to investigate whether the overexpression of VvOPR1 could potentiate the tolerance of OTs to Cu, Zn stress ( Figures 4A-E ). No significant difference in growth performance was detected between the two types of seedlings in the absence of stress, but all seedlings from each type were obviously suppressed when they were grown for three weeks in the medium containing Cu, Zn, and the suppression was more severe in the WT seedlings than in the OT seedlings ( Figures 4A-E ). The germination rate of OT Arabidopsis seeds was nearly twofold higher compared with that of WT seeds on medium supplemented with 150, 200 μM Cu. A similar pattern was observed for OT seeds under 1.2 mM Zn stress condition ( Figures 4A, B ). Exposure to Zn stress, compared with that of WT seedlings, the root length of OT seedlings increased by 32% to 46%, whereas the fresh weight increased by 56% to 80% ( Figures 4C, D, E ). Under Cu stress, the root length increased by 30% in the 50 μM Cu treatment, however, the root length increased from 50% to 78% in the OT lines compared with that in the WT plants under the 150 μM Cu treatment ( Figures 4C, D ), whereas the fresh weight increased by from 20% to 45% ( Figures 4C, E ).

The effects of Cu, Zn stress on OT rice seedlings grown in MS medium and Hoagland’s liquid medium were detected ( Figures 4F–H ). When treated with Cu, WT rice showed neither any roots nor any cotyledons, while several cotyledons were detected in OT grown in MS medium ( Figure 4F ). When Zn stress was imposed on the rice seedlings, all of the WT presented no any roots and much fewer cotyledons than the OT did, whereas all of the Zn-treated OT seedlings showed several leaves and roots that became shorter with increasing Zn concentration in the MS medium ( Figure 4F ). After exposure to Hoagland’s liquid medium containing Cu, Zn for 5 days, several morphological disturbances were detected in the OT and WT. Visible toxicity symptoms such as chlorosis, necrosis and rolling of the leaves were more obvious upon Cu and Cu+Zn impositions ( Figure 4G ), and the roots of the Cu- and Cu+Zn-stressed seedlings became reddish, and their health deteriorated compared with that of the control seedlings ( Figure 4G ). Both the OT and WT seedlings exhibited a decrease in shoot length, root length and fresh weight significantly when they were exposed to 750 μM Cu solution ( Figure 4G ); to some degree, the greatest reduction in shoot length, root length and fresh weight occurred in the seedlings subjected to Cu+Zn stress ( Figures 4G, H ). Compared with those grown under Cu, Cu+Zn conditions, the Zn-stressed seedlings presented fewer toxic effects and retained a similar leaf color ( Figures 4G, H ). Together, compared with the WT seedlings, the OT presented significantly higher shoot length, root length and fresh weight under stress conditions, indicating that the tolerance of the OT seedlings to Cu, Zn stress was greater than that of the WT seedlings.

Chlorophyll fluorescence has been routinely used to monitor the photosynthetic performance of plants noninvasively and to screen plants for tolerance to environmental stresses (Baker and Rosenqvist, 2004). To determine whether the chlorophyll content of OT is affected by Cu, Zn stress, the chlorophyll (Chla) and carotenoid (Car) contents were detected. As shown in Table 5 , there was no significant difference in the Chla, Chlb, Chla/Chlb or Car contents between WT and OT under normal conditions. In contrast, following the Cu, Zn and Cu+Zn treatments, the Chla, Chlb, Chla/Chlb and Car contents decreased in both the WT and OT seedlings compared with those under the control conditions, but the Chla, Chlb and Chla/Chlb values in the leaves of the OT were significantly higher than those in the WT, whereas the Car content in the OT was lower than that in the WT. The Chla, Chlb and Car contents in the leaves of the rice seedlings treated with 7.5 mM Zn were greater than those in the rice lines treated with 0.75 mM Cu and 0.75 mM Cu+7.5 mM Zn. Cu, Zn damage to the OT plants was evaluated both as photosystem II (PSII) stability and as injury to the whole plant. Fv/Fm (variable fluorescence/maximal fluorescence) was used to estimate the quantum yield of PSII photochemistry. The F₀ (minimal fluorescence), Fm and Fv of WT and OT plants were measured, and no significant difference was detected between WT and OT plants under control condition. The values of OT were higher than those of WT seedlings after treatment with 0.75 mM Cu, 7.5 mM Zn and 0.75 mM Cu+7.5 mM Zn, respectively ( Table 6 ). In particular, the combined Cu, Zn treatment markedly inhibited PSII, as indicated by a decrease in the Fv/Fm value. Together, these results are consistent with the hypothesis that plants overexpressing VvOPR1 are more tolerant of photoinhibition than are WT seedlings.

The roots were affected the most under heavy metal stress ( Figure 4 ), and the OT and WT seedlings experienced different degrees of damage when exposed to Cu, Zn stress ( Figure 5 ). Compared with that of the control group, the root activity of the OT-treated group decreased by 14.5%/30.8% when it was subjected to Zn/Cu stress, and there were significant differences between the two genotypes. However, the root activity reduced to the maximum level when these seedlings were subjected to Cu+Zn stress, and only the activity of the OT-2 line was significantly different from that of the WT line ( Figure 5 ). Malondialdehyde (MDA) and H₂O₂ were induced by Cu, Zn stress ( Figure 5 ), and the MDA content was remarkably reduced by 21.2%, 18.9% and 12.13%, respectively, compared with that of the WT with Zn, Cu and Cu+Zn treatment ( Figure 5 ), whereas the H₂O₂ content increased by 69%, 134% and 173% respectively, under Zn, Cu and Cu+Zn stress compared with that the CK ( Figure 5 ). In addition, the proline content was significantly higher in the OT (increased by 24.14%, 23.05% and 26.58%) than that in the WT under the stressed conditions (Cu, Zn and Cu+Zn stress, respectively) ( Figure 5 ).

The response of the ascorbate-glutathione (AsA-GSH) cycle to Cu, Zn stress in the OT is characterized in Figure 5 . Compared with that under the control conditions, the AsA content of the OT/WT increased by 34.41%/17.72% and 19.96%/10.05% during Cu, Zn stress, respectively, whereas it decreased by 18.45%/34.42% when it was subjected to Cu+Zn stress. Compared with the WT, all of the OTs presented higher AsA level under the three stress conditions, and the greatest increase AsA content was detected in the Cu+Zn treatment, reached 21.78% ( Figure 5 ). As represented in Figure 5 , GSH and AsA contents had similar results, decreasing during Cu+Zn exposure, and increasing with Cu or Zn individual treatment. As shown in Figure 5 , the Dehydroascorbate (DHA), Glutathione disulfide (GSSG) contents of the OT and WT seedlings increased under Cu, Zn and Cu+Zn stressful conditions, and the greatest increase was found under the Cu+Zn treatment. However, compared with the WT seedlings, the OT seedlings presented significantly lower DHA and GSSG contents under all three stress conditions. The changes in the AsA, GSH, DHA and GSSG contents of the OT and WT seedlings were conductive to reducing the AsA/DHA and GSH/GSSG ratios ( Figure 5 ) after exposure to Cu, Zn single or combined stresses, except for the increased ratio of AsA/DHA in the OT seedlings under Zn stress compared with those under normal conditions.

The activities of antioxidative enzymes such as superoxide dismutase (SOD, EC1.15.1.1), peroxidase (POD, EC1.11.1.7), catalase (CAT, EC1.11.1.6), glutathioneperoxidase (GPX), ascorbate peroxidase (APX), glutathione reductase (GR), monodehydroascorbate reductase (MDHAR) and dehydroascorbate reductase (DHAR) were measured in the OTs and WTs exposed to Cu, Zn and Cu+Zn stress conditions ( Figure 6A ). In both the WT and OT rice seedlings, the activities of SOD and POD in response to the Cu, Zn and Cu+Zn treatments improved compared with those in the control, and significantly greater activities were detected in the OT than in the WT ( Figure 6A ) among the three stress treatments. Notably, there was a clear reduction in the CAT activity of the OT and WT plants under all three stress conditions ( Figure 6A ). This might be because the removal of H₂O₂ is carried out by APX and GPX, so an increase in CAT activity is not required. Compared with those of the control, the same increasing tendencies were observed for GPX, APX, GR, MDHAR and DHAR activities in both the OT and WT seedlings exposed to Cu, Zn stress, whereas a decreasing tendency was detected under Cu+Zn stress condition ( Figure 6A ). In conclusion, all of the results revealed greater activities of antioxidant enzymes in the OT lines than in the WT. These findings suggested that OT strengthened the ability of the plants to scavenge ROS and maintain ROS homeostasis by increasing the activities of antioxidant enzymes.

To evaluate the regulatory role of VvOPR1 in gene expression, through which it could improve tolerance to Cu, Zn stress, several antioxidative enzyme-encoding genes were selected for qRT-PCR analysis ( Figure 6B ). The expression of all of the genes showed no significant difference between the WT and OT seedlings under controlled conditions. Conversely, the expression levels of all antioxidative enzyme-encoding genes were significantly greater in the OT and WT seedlings than in the control seedlings when they were exposed to Cu, Zn, Cu+Zn stress. The expression levels of OsSOD, OsPOD, OsCAT, OsAPX4, OsGR3 and OsGPX1 in the OTs were significantly higher than those in the WT seedlings under either Cu, Zn or Cu+Zn stress conditions ( Figure 6B ). OsSOD and OsPOD showed the highest expression levels under Cu+Zn stress condition ( Figure 6B ), while the highest expression level of OsCAT was observed in the Cu-treated seedlings ( Figure 6B ). The expression pattern of antioxidative enzyme-encoding genes involved in the AsA-GSH cycle (OsAPX4, OsMDHAR2, OsDHAR1, OsGR3 and OsGPX1) was similar to that of OsPOD when exposed to Cu, Zn, Cu+Zn stresses ( Figure 6B ), and the highest expression of these genes was also induced by Cu+Zn stress. Compared with those in the control seedlings, the expression levels of OsMDHAR2 and OsGR3 were higher in the OT and WT seedlings under Cu, Zn stress conditions, but no significant differences were detected between the OT and WT seedlings ( Figure 6B ).

To detect whether the overexpression of VvOPR1 would induce ABA or JA in OT under Cu, Zn stress conditions, the endogenous ABA and JA levels in OT and WT seedlings were determined. Under control conditions, no significant difference in ABA or JA content was detected between the OT and WT seedlings, whereas the levels of ABA and JA were elevated in both plant types after exposure to Cu, Zn, Cu+Zn stress, and the highest elevation was detected in the Cu+Zn treatment ( Figure 7A ). Specifically, significant differences in ABA levels were detected between the OT and WT under all stressed conditions, whereas no significant differences in JA levels were detected under the same stress conditions.

To further elucidate the molecular mechanism underlying the resistance of OT to Cu, Zn stress, the expression levels of ABA biosynthetic genes (OsNCED1, OsZEP1 and OsAAO3), catabolic genes (OsABA8ox1), signaling genes (OsPP2C68 and OSRK1) and responsive genes (RAB21, LEA3, RAB16C and RAB16D) and JA-related (OsDAD1, OsPLA1, OsLOX2, OsAOS1, OsAOC, OsOPR7, OsJAR1 and OsCOI1b) genes were assayed. The expression levels of ABA-related genes (except for the expression of OsABA8ox1 under Zn stress conditions) were significantly higher in the OT lines than in the WT seedlings under Cu, Zn stress conditions ( Figure 7B ). As shown in Figure 7C , the expression level of OsCOI1b in the OT and WT elevated after they were exposed to Cu, Zn stress, but reduced when they were subjected to Cu+Zn stress. The OsPLA1 and OsJAR1 genes were downregulated under Cu, Cu+Zn stress conditions but upregulated with Zn treatment. Other JA-related genes, such as OsDAD1, OsLOX2, OsAOS1, OsAOC and OsOPR7 exhibited similar expression patterns: Cu, Zn and Cu+Zn stress all elevated the expression of these genes, with the smallest increase occurring in Zn-treated seedlings, then the expression of these genes increased in the Cu-treated seedlings and peaked under Cu+Zn stress conditions.

To further detect the role of ABA in VvOPR1-enhanced Cu, Zn stress, VvOPR1 was transformed into the Arabidopsis NCED1 mutant, which is defective in ABA synthesis to construct VvOPR1/NCED1 lines ( Figure 7D ). Similar results were detected in the WT, VvOPR1/NCED1 and NCED1 lines under nonstressed conditions. The survival rate of the NCED1 mutant was seriously restricted by Cu, Zn stress, and the survival rate was lower by more than 60% compared with that of WT seedlings subjected to Cu stress ( Figure 7D ). VvOPR1 overexpression significantly increased the survival rate of seedlings subjected to Cu, Zn stress, and the survival rate of VvOPR1/NCED1 line was 12.96% to 3.7%, 60.19% to 39.81% to those of NCED1 line under Cu, Zn stress, respectively ( Figure 7D ). However, no significant difference was detected between the NCED1 mutant and VvOPR1/NCED1 line under H₂O₂ stress.

After exposure to 0.75 mM Cu and 7.5 mM Zn for 5 days, the contents of heavy metals drastically increased, and the concentrations of Cu²⁺/Zn²⁺ were significantly lower in the OT than in the WT ( Tables 7 , 8 ). Compared with those in the control seedlings, the Cu²⁺ content in the 0.75 mM Cu-treated WT and OT seedlings was nearly 2444- and 2364-fold greater, respectively, than that in the control; whereas the Cu²⁺ content was 2190- and 206-fold greater, respectively under Cu+Zn stress condition. A great increase in Zn²⁺ content was also detected in the Zn-treated seedlings, compared with the control, 849- and 788-fold increases in Zn²⁺ content were detected in the WT and OT seedlings under 7.5 mM Zn stress condition; 724- and 679-fold increases in Zn²⁺ content were detected in the WT and OT seedlings under Cu+Zn stress condition.

The distribution of Cu²⁺/Zn²⁺ taken up was not homogeneous, and a much higher proportion remained in the roots than was transported into the aerial parts. The proportion of Cu²⁺ in the roots versus the aerial parts was 79.85% versus 20.15%, 91.28% versus 8.72% and 92% versus 8% in WT under the control, Cu and Cu+Zn stress conditions, respectively; 79.22% versus 20.78%, 92.28% versus 7.72% and 92.51% versus 7.49% in the OT under the same conditions. In contrast to the Cu²⁺ distribution, a higher Zn²⁺ proportion was detected in the aerial parts, ranging from 18.7% (the proportion of Zn²⁺ in the OT treated with Cu+Zn stress) to 35.6% (the proportion of Zn²⁺ in the WT under control conditions).

In terms of the translocation factor (TF), under Cu, Zn stress conditions, both the OT and WT lines presented the reduced TFs compared with those under the control conditions. As a consequence of VvOPR1 overexpression, the distribution of Cu²⁺, Zn²⁺ in rice seedlings was altered, reducing the proportion of Cu²⁺, Zn²⁺ accumulation in the aerial parts of the OT and WT seedlings under Cu, Zn stress. TF values of Cu²⁺, Zn²⁺ in the OT were lower than those in the WT, but the TF value of Cu²⁺ was far below the TF value of Zn²⁺ ( Tables 7 , 8 ). The tolerance index (TI) of OT, which is based on root length, was higher than that of WT under all three treatment conditions, indicating that the sensitivity of WT to Cu, Zn stress was greater than that of OT. When exposed to 0.75 mM Cu and 0.75 mM Cu+7.5 mM Zn, the accumulation rates (ARs) of Cu²⁺ in the OT lines were 1.7- and 1.87-fold higher than those in the WT, respectively. While AR in OT was 0.98 and 1.85 fold for Zn²⁺ compared to WT when these seedlings were treated with 7.5 mM Zn or 0.75 mM Cu+7.5 mM Zn, respectively.

In this study, nine VvOPRs were identified in grapevine. The rooted maximum-likelihood phylogenetic tree revealed two OPR subfamilies in grapevine following the classification system (Montanini et al., 2007), and all of the VvOPRs except VvOPR3 were clustered with sub.I ( Figure 1B ). VvOPR genes usually contain 4~6 exons and 3~5 introns according to the VvOPR exon/intron structure analyses ( Figure 1C ), and these results were similar to the gene structures of ZmOPR and TaOPR (Zhang et al., 2005; Mou et al., 2019). A combination of the phylogenetic ( Figure 1B ) and genetic structure analysis ( Figure 1C ) revealed that most of the VvOPR genes within a subfamily showed a similar exon/intron structures. VvOPR4 in sub.I contained the most introns ( Figure 1C ), indicating that the intron loss events occurred during the structural evolution of the OPR gene family (Li et al., 2009). As shown by the diversity of VvOPR protein-conserved motifs, the number and the position of VvOPR motifs in each subfamily are clearly conserved ( Figure 1C ), with VvOPR1 in sub.I lacking motif 9 and 10; analogously, the motif loss events were also detected in sub.II, and VvOPR3 lacks motif 7, and the absence of this motif is crucial for the secondary structure of proteins (Xin et al., 2017). In general, paralogous genes may have new biological functions relative to their ancestor genes (Guo et al., 2008), so we concluded that VvOPR1 might function in conjunction with the associated homologous genes AtOPR1,2.

The number of OPR members in the VvOPR family differed from that in other representative plants ( Table 1 ), possibly because grapevine has undergone whole-genome duplications during its evolutionary history. Gene duplication is one of the primary driving forces in the evolution of genomes (Moore and Purugganan, 2004; Cannon et al., 2004), hence, it is highly possible that the gene duplication led to VvOPR gene family expansion, and that new genes contributed to the new structures and new biological functions. Therefore, we deduce that VvOPR1 might develop some new functions that are different from AtOPR1,2.

The expression of OPRs in dicots and monocots was found to be tissue specific (Liu et al., 2020; Stintzi and Browse, 2000; Li et al., 2011; Biesgen and Weiler, 1999; Mou et al., 2019; Xin et al., 2017; Creelman and Mullet, 1995). Here, VvOPR1 showed the highest expression level in roots, followed by stems and leaves, and the expression profiles of VvOPR1 under abiotic stress treatments (Cu, Zn, H₂O₂ and ABA stresses) ( Figure 5 ) were much different from those of GhOPR (Liu et al., 2020), ZmOPR (Zhang et al., 2005) and TaOPR (Wang et al., 2016a; Dong et al., 2013). Therefore, based on the specific stress inductive patterns of VvOPR1 suggest that VvOPR1 is involved in the response to Cu, Zn stress.

Research on OPRs participating in development and response to various abiotic and biotic stresses in some plants has been conducted thoroughly (Bosch et al., 2014; Dave et al., 2011). In this study, the growth retardation was observed in the OT and WT seedlings under Cu, Zn stress, however, the OT seedlings showed obvious phenotypic changes compared with the WT seedlings, manifested as higher generation rates, longer shoot and root lengths and higher fresh weights ( Figure 4 ). The growth reduction raised with increasing levels of Cu, Zn, which induced toxicity at elevated concentrations, and the impact on root elongation was greater than that on shoot growth ( Figure 4 ); similar results have been reported for durum wheat (Michaud et al., 2007; Ghavri and Singh, 2012; Bibi et al., 2006; Shu et al., 2002). In Phaseolus vulgaris, Cu toxicity disrupted the capacity of cells to remove oxidatively damaged proteins by inhibiting the ubiquitin proteasome pathway in embryonic stages thus inhibiting seed germination (Karmous et al., 2014). Seed germination which is very sensitive to the external medium is regulated by changes in the cellular redox status (Nanda and Agrawal, 2016), and the addition of metal further aggravates the microenvironment, causing damage to proteins and leading to a reduction in germination (El-Maarouf-Bouteau and Bailly, 2008). In this study, the relatively high germination rate of OT under Cu, Zn stress conditions might have been due to decreased accumulation of oxidatively damaged proteins. The Cu, Zn-induced growth reduction has also been observed in Spirodela polyrhiza (Upadhyaya and Panda, 2010), Withania somnifera (Khatun et al., 2008) and Sorghum bicolor (Soudek et al., 2014).

Recently, JA has been shown to be effective in improving plant tolerance to heavy metal stress (Yan et al., 2015; Piotrowska et al., 2009; Ahmad et al., 2017). In this study, the increase in JA levels in the leaves of the OT and WT seedlings in response to the Cu, Zn treatments was consistent with the increase in JA levels after Cu and Cd treatment (Maksymiec et al., 2005). The increase in JA caused by Cu, Zn stress could involve up-regulation of some JA biosynthesis genes, such as OsDAD1, OsPLA1, OsLOX2, OsAOS1, OsAOC1 and OsOPR7 ( Figure 7C ). These results suggested that JA is involved in Cu, Zn stress, however, no significant difference was defected in JA content or the expression level of JA synthesis genes was detected between OT and WT seedlings, so we concluded that neither the expression of JA synthesis genes nor the endogenous JA level was dependent on VvOPR1 overexpression ( Figure 7C ). In addition, similar results were detected for the expression levels of the JA signaling related genes OsJAR1 and OsCoI1b ( Figure 7C ). These findings indicate that VvOPR1 dose not regulate JA synthesis or signaling pathways.

In combination, we like to conclude that VvOPR1 may be an effective gene for improving Cu, Zn stress tolerance. Here, we first synthesized and characterized the characterization the OPRI gene VvOPR1, which regulates the expression of ABA biosynthesis and catabolism genes thus leading to increased endogenous ABA accumulation under Cu, Zn stress conditions, subsequently promoting the ABA signaling pathway, and reducing Cu, Zn accumulation and translocation. Moreover, VvOPR1 overexpression reduced the production of ROS (such as reduced H₂O₂ and MDA) and increased the ROS-scavenging system to confer tolerance to Cu, Zn stress. Therefore, we concluded that VvOPR1 may be an effective gene for improving Cu, Zn stress tolerance in plants. In addition, the function of VvOPR1 may lie in the metabolism of trans-(+)-OPDA, with consequent effects on the activities of the (ABA-dependent responsive and/or ROS) signaling pathway which this molecule mediates. Moreover, the presence of VvOPR1 did not induce any changing in JA synthesis or signaling pathways.