Cucumber (Cucumis sativus L. cv. Jinyou No. 4) was used as experimental material. The seedlings were cultivated in half-strength Hoagland nutrient solution (pH 6.3 ± 0.2, EC=2.2 ± 0.2 mS·cm^-1) and cultivated in the artificial climate chamber. The growth temperature was 25°C/16°C (day/night). The photoperiod was 14 h/10 h (day/night) with light intensity at 600 μmol·m^-2·s^-1, and relative humidity was stabilized at 60-75%. To analyze the effects of hormones on cucumber seedlings, 100 μM abscisic acid (ABA), 100 μM salicylic acid (SA), 100 μM ethylene (ETH), and 100 μM methyl jasmonate (MeJA) were sprayed on the leaves when the seedlings had two true leaves. To analyze the effects of abiotic stress on cucumber seedlings, 75 mM NaCl and 20% (w/v) polyethylene glycol (PEG) 6000 were added to the Hoagland nutrient solution of cultivated seedlings to simulate salt stress and drought stress, respectively. The seedlings were placed in a 4°C light incubator to simulate cold stress. Leaves and roots were collected at 0, 3, 6, 9, 12, 24, 48, and 72 h after treatment and stored at -80°C for use. To analyze the tissue-specific expression pattern, different tissues of the cucumber reproductive growth period were stored at -80°C.

Wild-type tobacco (Nicotiana tabacum L.) and transgenic tobacco (CsSAMDC3-overexpressing) were cultured in an artificial climate chamber. When the tobacco grew to four ture leaves, irrigated with 200 mM NaCl. After 5 days of treatment, the leaves and roots were stored at -80°C.

The amino acid sequence of SAMDC was retrieved from the reported Cucurbit Genomics Database (http://cucurbitgenomics.org/) and verified in NCBI Database. The isoelectric point (pI) and molecular weight (MW) of the candidate SAMDC were calculated using ExPASy (https://www.expasy.org/). To study the evolutionary relationship of SAMDC between cucumber and other plant species, the SAMDC gene sequence and SAMDC protein sequence of 12 plants such as pumpkin, tobacco, and soybean were obtained by searching NCBI Genome Database. The SAMDC protein sequences of different species were compared by Clustal W, and the phylogenetic tree was constructed by NJ (neighbor-joining) method, Poisson correction, and 1000 bootstrap replicates in MEGA 11 software. The conserved motifs of different species and the conserved domains of SAMDC protein sequences were searched by the online MEME tool (https://meme-suite.org/meme/) and NCBI CDD (https://www.ncbi.nlm.nih.gov/cdd/), and visualized by TB Tools (Chen et al., 2020).

To construct CsSAMDC3-overexpressing tobacco, the primer pair PAC019-CsSAMDC3-F/R ( Supplemental Table 1 ) was designed. CsSAMDC3 was cloned from cucumber cDNA with PrimeSTAR^® Max DNA Polymerase (Takara, China) following the instructions. The cucumber SAMDC3 gene fragment was ligated to the digested PAC019 vector using ClonExpress^® II One Step Cloning Kit (Vazyme, China). The normal sequencing plasmid was transferred into Agrobacterium EHA105 by heat shock method. CsSAMDC3 was transformed into Nicotiana tabacum by Agrobacterium-mediated according to the previous research method (Wang et al., 2017). Homozygous transgenic plants were screened by kanamycin and identified by PCR analysis. Quantitative real-time PCR (qRT-PCR) analysis was also performed to validate the transformation of CsSAMDC3 further. T3 homozygous transgenic lines for further study.

RNA simple Total RNA Kit (Tiangen, China) was used to extract total RNA from samples. HiScript^® III Q RT Super-Mix for qPCR Kit (Vazyme, China) reverse transcribed 1μg total RNA into cDNA for gene cloning and qPCR. Based on the selected gene sequence, the primer pair ( Supplemental Table 1 ) was designed using Beacon Designer™ 8.10 (Premier Biosoft International, USA). qRT-PCR was conducted on Quant-Studio™ 5 Real-Time PCR System (Applied Biosystems) with ChamQ SYBR qPCR Master Mix (Vazyme, China), which included 10 μL of ChamQ SYBR qPCR Master Mix (2 ×), 0.4 μL of sense or anti-sense primer, 0.4 μL of ROX Reference dye (50 ×), 1 μL of cDNA and 7.8 μL of ddH₂O in a total volume of 20 μL. The PCR program was as follows: 95°C for 30 s; 40 cycles of 95°C for 10 s, 58°C for 10 s; and finally, 72°C for 30 s.

The relative expression levels of the selected genes were calculated by 2^-ΔΔCT method using cucumber or tobacco actin genes as internal controls (Livak and Schmittgen, 2001).

The electrolyte leakage rate was measured according to the description of He et al. (2019). 0.5 g fresh sample was immersed in a test tube containing 20 mL deionized water. After shaken at room temperature for 4~5 h, the initial conductivity (EC₁) was measured. Then, the sample was boiled at 95 °C for 20 min and cooled to room temperature, and the final conductivity (EC₂) was measured in the bath. We also measured the conductivity of deionized water (EC₀). EL (%) was calculated as [(EC₁-EC₀)/(EC₂-EC₀)] × 100.

MDA content was determined according to the description of Dhindsa et al. (1981). The leaves were ground on ice with 5% TCA and centrifuged at 4000 g for 10 min. Then 2 mL of the supernatant was mixed with an equal amount of 0.67% thiobarbituric acid (TBA). After being heated in a boiling water bath for 30 min, the mixture was centrifuged at 3000 g for 15 min. The absorbance of the supernatant was read at 450, 532, and 600 nm and calculated. The MDA content is represented as nmol·g^-1 FW.

The content of H₂O₂ was determined according to the description of Zhu et al. (2021). The 0.2 g leaf sample was ground on ice using 1.6 mL of pre-cooled 0.1% trichloroacetic acid (TCA) to a slurry and then centrifuged at 12000 g for 20 min. The 0.2 mL supernatant was mixed with 1 mL 1 M KI solution and 0.25 mL 0.1 M potassium phosphate buffer (pH 7.8) and placed in the dark for 1 h. The absorbance at 390 nm was read with 0.1% TCA as blank. The content of H₂O₂ in the sample was calculated according to the standard curve of known H₂O₂ concentration. The H₂O₂ content is expressed as μmol·g^-1 FW.

The leaves were ground into a slurry in 50 mM precooled phosphate buffer (pH 7.8), transferred into a 2 mL centrifuge tube, and centrifuged at 4°C, 12000 g for 20 min. The resulting supernatant was the crude enzyme solution for determining antioxidant enzyme activity.

Superoxide dismutase (SOD) activity was determined with reference to the description of Giannopolitis and Ries (1977) with minor modifications. The NBT photoreduction method was used. The reaction mixture contained 50 mM PBS (pH 7.8), 30 μM EDTA-Na₂, 14.5 mM methionine, 60 μM riboflavin and 2.25 mM nitroblue tetrazolium chloride (NBT). 3 mL reaction mixture was mixed with 40 μL crude enzyme solution, and the tubes were illuminated for 20 min. The absorbance was read at 560 nm. One unit of SOD activity was defined as the amount of enzyme required to cause 50% inhibition of NBT photoreduction.

Peroxidase (POD) and catalase (CAT) were determined according to the description of Wu et al. (2022). The enzyme extract was mixed with the reaction solution. The reaction solution to determine POD was 20 mM phosphate buffer (pH 6.0), 3.5 M guaiacol, and 30% H₂O₂. The reaction solution to determine CAT was 50 mM phosphate buffer (pH 7.0) and 30% H₂O₂. The absorbance changes at 470 or 240 nm within 40 s were measured. A unit of POD or CAT activity was expressed as a change of OD₄₇₀ or OD₂₄₀ value of 0.01 min^-1.

The net photosynthetic rate (Pn) of tobacco leaves at the same leaf position was measured by portable photosynthesis system (Li-6400 XT, Li-COR, Lincoln, NE, USA). The light intensity, leaf temperature, relative humidity, and ambient CO₂ concentration maintained at 800 μmol·m^-2·s^-1, 25°C, 70%, and 400 ± 10 μmol·mol^-1, respectively. Chlorophyll fluorescence was measured by fluorescence imaging system (IMAGING-PAM, Heinz Walz, Effeltrich, Germany). The measured data and collected fluorescence images were analyzed by ImagingWin software (Heinz Walz, Effeltrich, Germany). The chlorophyll fluorescence parameters were measured according to the method of Shu et al. (2016), and the maximal quantum yield of PSII photochemistry (Fv/Fm) was calculated.

PAs were measured according to the description of Shu et al. (2012). The sample was ground into a slurry in 1.6 mL 5% precooled perchloric acid (PCA) and then centrifuged at 12000 g at 4°C for 20 min. The supernatant was collected to measure free and bound PAs, while the particles were used to measure bound PAs.

For free PAs, 0.7 mL of supernatant, 1.4 mL of 2 M NaOH, and 15 μL of benzoyl chloride were mixed, then vortexed for 20 s and kept at 30 °C for 30 min. Then, 2 mL saturated NaCl solution, and cold colddiehtyl ether were added to the mixed solution to extract benzoyl polyamines. After centrifugation at 4 °C for 12000 g for 5 min, 1 mL of ether phase was evaporated to dryness and redissolved in 1 mL of 64% (v/v) methanol.

For conjugated PAs, 0.7 mL of supernatant was mixed with 5 mL of 6 M HCl and sealed in an ampoule at 110 °C for 18 h to convert conjugated PAs into free PAs. The hydrolysate was evaporated at 70°C, and the resulting residue was re-suspended in 1.6 mL of 5% PCA. The following steps are the same as free PAs extraction.

For bound PAs, the particles were washed four times with 5% PCA and centrifuged at 3000 g for 5 min. The resulting particles were suspended in 5 mL 6 M HCl and then subjected to the same steps as conjugated PAs extraction.

Samples redissolved in methanol are stored at -20°C and filtered with a membrane (0.45 μm) before testing. UPLC system (Thermo, UltiMate 3000), including ACQUITY UPLC HSS T₃ column, acetonitrile, and water (volume ratio of 44: 56) as the solvent, the flow rate of 0.45 mL min^-1 for the detection of PAs content. The sum of three forms of PAs is the total amount of endogenous PAs.

Subcellular localization prediction of CsSAMDC3 using the CELLO tool (http://cello.life.nctu.edu.tw/).

The agrobacterium strain EHA105 transformed with PAC019-CsSAMDC3 plasmid was transiently transformed into Nicotiana tabacum by injection infection method. A confocal laser scanning microscope (LSM 780, Zeiss, Germany) was used for imaging after incubation for 36-48 h under the light.

All data were analyzed by single factor analysis of variance (ANOVA) using IBM SPSS 26.0 software (SPSS Inc., Chicago, IL, USA). Duncan’s multiple comparison method was used to analyze the difference between different treatments at a P < 0.05 level of significance.

Four CsSAMDC gene sequences were identified by PCR amplification and sequencing, which were consistent with the search results of the database. As shown in Table 1 , CsSAMDC1 and CsSAMDC3 are located on chromosome 3, while CsSAMDC2 and CsSAMDC4 are located on chromosome 6 and 2, respectively. The length of the CsSAMDC1-4 gene sequence is relatively close, all of which are about 1100 bp, the encoding amino acid (AA) number is 340-389, the encoding protein isoelectric point (pI) is 4.56-6.88, molecular weight is about 40 KDa, hydrophilic average coefficient (GRAVY) is close to 0, and it is inferred to be amphoteric protein.

Phylogenetic analysis of the full-length SAMDC protein sequences of 13 species, including cucumber, showed that these SAMDC proteins could be divided into 3 groups, of which CsSAMDC2 and CsSAMDC3 were in group 1, while CsSAMDC4 and CsSAMDC1 were in group 2 and group 3, respectively ( Figure 1A ). Conserved motif analysis of SAMDC protein sequence by the online MEME tool identified 16 conserved motifs. The distribution was consistent with the phylogenetic tree ( Figure 1B ). The conserved domain of the SAMDC protein sequence was analyzed by NCBI CDD. A typical SAMDC protease domain was found ( Figure 1C ). By searching NCBI Genome Database, the entire sequence of the SAMDC genes was obtained for visual analysis. The results showed that the SAMDC genes of group 1 basically did not contain introns, while the introns of group 2 and group 3 had different lengths ( Figure 1D ). In addition, the CDS regions of most SAMDC genes are continuous ( Figure 1D ).

The response of cucumber SAMDC family genes to salt stress was analyzed by qRT-PCR. As shown in Figure 2 , there were differences in the expression patterns of SAMDC family genes in cucumber roots and leaves. The expression trends of CsSAMDC2 and CsSAMDC3 are similar, both of which firstly increased and then decreased. The expression levels of CsSAMDC2 and CsSAMDC3 in roots peaked after 9 h of salt treatment, which were 6 times and 8 times higher than that of the control, respectively. The expression of CsSAMDC2 in leaves also peaked at 9 h, which was 15 times higher than that of the control, while the expression of CsSAMDC3 peaked earlier at 6 h, which was 7 times higher than that of the control. It suggests that CsSAMDC3 is the most sensitive to salt stress and the first to respond to salt stress. In addition, the response of CsSAMDC1 to salt stress was relatively stable in leaves, while its expression level in roots was close to that of the control within 24 h of salt stress and then increased. The response of CsSAMDC4 to salt stress was more prominent in leaves, and its expression peaked at 48 h, which was 12 times higher than that of the control, while its expression in roots showed a trend of slight decrease and then increase. After that, we selected CsSAMDC3, which is sensitive to salt stress, as a typical case for further study.

CsSAMDC3 was expressed in different tissues of cucumber. The expression level of CsSAMDC3 in fruit and flower was 5-8 times higher than that in leaves. In comparison, the expression level in the root, stem, and leaf was lower ( Figure 3A ), indicating that CsSAMDC3 was mainly related to flower organ development.

Subcellular localization prediction of CsSAMDC3 showed that the protein was likely to be localized in the cell membrane and nucleus ( Supplemental Figure 1 ). To verify the results of CELLO subcellular localization prediction, the green fluorescence signal of GFP-CsSAMDC3 fusion protein was detected by ultra-high resolution laser confocal microscopy. Figure 3B showed that the green fluorescence signal of the fusion protein was on the cell membrane and nucleus, indicating that CsSAMDC3 was located on the cell membrane and nucleus. The results were consistent with CELLO prediction.

As shown in Figure 4 , under ABA treatment, the expression of CsSAMDC3 in cucumber leaves and roots increased and reached the peak value at 9 h, then decreased to the average level at 48 h. Under SA treatment, the expression of CsSAMDC3 in cucumber leaves fluctuated less compared with ABA treatment, and the peak value also appeared at 9 h, while the expression of CsSAMDC3 in cucumber roots reached the highest at 24 h, and the response was more intense. Under MeJA treatment, the expression of CsSAMDC3 in leaves increased at 3 h, decreased slightly, increased to the maximum at 12 h, and then decreased to the standard value. In roots, the response of CsSAMDC3 to MeJA was not strong, but the overall trend was upward. Under ETH treatment, the expression of CsSAMDC3 in leaves and roots was very prominent, reaching the peak at 24 h and 12 h, respectively, and the response of CsSAMDC3 to ETH was the strongest compared with other treatment groups. For 4 °C stress and PEG simulated drought stress, the overall response of CsSAMDC3 fluctuated less, especially in the leaves under cold stress and the roots under drought stress; The expression level has been maintained at a low level. Under salt stress, the expression levels of CsSAMDC3 in leaves and roots reached the peak at 6 h and 9 h, respectively, and the response in roots was more robust. Compared with other abiotic stress treatments, the expression of CsSAMDC3 reached the peak earlier under NaCl stress, indicating that CsSAMDC3 was more sensitive to salt stress.

To further study the function of CsSAMDC3 under salt stress, the genetic transformation of tobacco was carried out by the Agrobacterium-mediated leaf disc method. Positive plants were screened by kanamycin, and DNA was extracted and verified by PCR with PAC019-F/PAC019-CsSAMDC3-R ( Supplemental Table 1 ) (Figure 5A). The expression of CsSAMDC3 gene in two lines with good seed quality was detected by qRT-PCR. It can be seen from Figure 5B that the expression level of CsSAMDC3 in the transgenic line OE-1# was 19 times higher than that in the transgenic line OE-2#.

As shown in Figures 6A, B , after 5 days of 200 mM NaCl treatment, the growth of wild-type and overexpressed CsSAMDC3 tobacco was inhibited, showed different degrees of salt damage: compared with overexpressed plants, wild-type plants were shorter and smaller, and root growth was more significantly inhibited. Among them, the fresh and dry weight of the overexpression lines OE-1# and OE-2# decreased by 32.73%, 39.83% and 29.78%, 31.00%, respectively, compared with the control group. In contrast, the wild type decreased more significantly, 53.71% and 57.74%, respectively ( Figures 6D, E ). In addition, the Fv/Fm and Pn values of wild-type tobacco after salt treatment were significantly lower than those of CsSAMDC3-overexpressing tobacco, which were 36.17% and 13.00% lower, respectively ( Figures 6C, F, G ). The above results showed that the growth and ФPSII of wild-type tobacco were more severely inhibited under salt stress, and overexpression of CsSAMDC3 alleviated the growth inhibition induced by salt stress.

To verify the role of CsSAMDC3 in salt stress, we further analyzed the oxidative indexes of wild-type and CsSAMDC3-overexpressing tobacco under salt stress. As shown in Figure 7 , the electrical conductivity, MDA content, and H₂O₂ content of overexpressing plants after salt treatment were significantly lower than those of wild-type plants in both leaves and roots. Among them, the electrical conductivity of OE-1# and OE-2# decreased more in leaves, which were 19.75% and 21.99%, respectively ( Figure 7A ). MDA content decreased more in roots, 39.24% and 45.16%, respectively ( Figure 7D ). H₂O₂ content in leaves decreased by 27.64% and 11.91%, respectively ( Figure 7E ). In addition, we determined the activities of antioxidant enzymes in leaves and roots of wild-type and CsSAMDC3-overexpressing tobacco under salt stress. The results showed that for SOD, the activity of CsSAMDC3-overexpressing tobacco in roots was substantially higher than that of wild-type tobacco, whereas the activity in leaves was comparable to that of wild-type tobacco ( Figures 8A, B ). For POD, the activity of CsSAMDC3-overexpressing tobacco in roots and leaves was significantly higher than that of wild-type tobacco, and the activity of OE-1# in leaves and roots was 81.79% and 223.00% higher than that of WT, respectively ( Figures 8C, D ); For CAT, the activities of OE-1# and OE-2# in leaves were 29.44% and 22.38% higher than WT, respectively ( Figure 8E ). The activities of OE-1# and OE-2# in roots were 20.53% and 9.47% higher than WT, respectively ( Figure 8F ). Overall, overexpression of CsSAMDC3 in tobacco can reduce MDA and H₂O₂ content by increasing antioxidant enzyme activity, thereby enhancing the tolerance of tobacco to salt stress.

The expression levels of antioxidase-related coding genes (NtSOD, NtPOD, NtCAT) were detected. As shown in Figure 9 , compared with WT, the expression levels of NtSOD, NtPOD, and NtCAT in OE-1# and OE-2# were significantly decreased under salt stress. Among them, the expression of NtSOD was the most obvious difference, and OE-1# and OE-2# were 66.92% and 68.26% lower than WT, respectively ( Figure 9A ). Under the control conditions, except NtPOD, the expression difference of other antioxidant enzyme related coding genes was not significant between transgenic tobacco and wild type.

As shown in Figure 10 , transgenic tobacco plants had lower Put content and higher Spd and Spm content compared to wild type under the control conditions. Compared with WT, the Spd and Spm content in OE-1# increased by 12.78% and 42.21% respectively, while the Put content decreased by 29.78%; the Spd and Spm content in OE-2# increased by 4.40% and 33.98% respectively, while the Put content decreased by 18.88%. However, under salt stress, the content of Put, Spd, and Spm in OE-1# decreased by 16.71%, increased by 28.81%, and increased by 45.15% compared with WT, respectively. The content of Put, Spd, and Spm in OE-2# decreased by 2.87%, increased by 21.25% and 51.68%, respectively, compared with WT. The difference between Spd and Spm content of CsSAMDC3-overexpressing tobacco and wild type under salt stress was further widened ( Figures 10B, C ). The values of (Spd + Spm)/Put in OE-1# and OE-2# were increased under salt stress compared with WT ( Figure 10D ). This indicates that overexpression of CsSAMDC3 under salt stress can promote the synthesis of more Spd and Spm in transgenic tobacco.

As shown in Figure 11 , under control conditions, the expression levels of NtSAMS and NtSPMS in OE-1# were significantly higher than those in WT and OE-2#, and the expression levels of NtSPDS and NtPAO in OE-2# were significantly higher than those in WT and OE-1#. However, under salt stress, compared with WT, the expression levels of PAs metabolism-related coding genes (NtSAMS, NtSPDS, NtSPMS, NtPAO) in OE-1# and OE-2# were decreased.