Centipedegrass plants and a common bermudagrass (Cynodon dactylon) that was used in our previous study (Lu et al., 2008) were planted in 15-cm diameter plastic pots containing a mixture of peat and perlite (3:1, v/v) in a greenhouse for 2 months, with temperature ranging from 25 to 30°C, irrigating daily and fertilizing once a week with 15N–6.6P–12.5K fertilizer. Bermudagrass plants were placed in a growth chamber with a 12-h photoperiod under light of 200 μmol m^-2 s^-1 at 6°C for 4 days for cold treatment for analysis of CdSAMDC1 expression, while centipedegrass plants were used for physiological and molecular measurements. For treatment with chemicals, leaf fragments of transgenic plants and WT were placed in deionized water for 1 h to eliminate the potential wound stress, and then placed in beakers containing 1 mM methylglyoxal bis (guanylhydrazone) or MGBG, 1 mM ascorbic acid (AsA), 1 mM NaN₂, 100 mM tungstate, or 200 mM 2-phenyl-4,4,5,5-tetramethylimidazoline-1-oxyl3-oxide (PTIO) under light of 80 mmol photons m^-2 s^-1 for 12 h, while those treated with deionized water for 12 h under the same condition served as a control. After treatments the leaf fragments were sampled and immediately frozen under liquid N₂ for further analysis.

Total RNA was isolated from bermudagrass leaves by using TRIzol reagent (Life Technologies, Carlsbad, CA, United States) according to the manufacturer’s protocol. First-strand cDNA was synthesized from 1 μg of total RNA in a 20-μl reaction mixture, using M-MLV reverse transcriptase and Oligo (dT)₁₈ primer. For amplification of CdSAMDC1, polymerase chain reaction (PCR) was conducted in a reaction mixture containing the first-strand cDNA as the template, primers SAMDC_F (CCTGCTCCAATGGCTGTTCT) and SAMDC_R (CCCGTCTTACTCATCAAGCACTC), and KOD-Plus DNA polymerase (TOYOBO, Osaka, Japan).

Embryogenic calli were induced from sterilized mature seeds of centipedegrass and cultivated as previously described (Liu et al., 2008). The embryogenic callus was transformed using Agrobacterium tumefaciens strain EHA105 harboring pCAMBIA-35S-CdSAMDC1 construct as described previously (Liu et al., 2012). the calli were placed on callus induction medium without selection pressure for 1-week after co-cultured for 3 days, followed by placing on selection medium containing hygromycin B (50 mg l^-1) for 8-week. The hygromycin-resistant calli were subjected for regeneration on regeneration medium containing hygromycin (50 mg l^-1), illuminated with a 16 h photoperiod (80 μmol m^-2 s^-1). The regenerated shoots were transferred to half strength of MS medium containing sucrose for rooting. The plantlets were transferred to soil in 15-cm plastic pots growing in a greenhouse at temperatures of 30/25°C (day/night) under natural light.

One gram of leaves was used for extract genomic DNA using the hexadecyltrimethylammonium bromide (CTAB) method. DNA samples (20 μg) were digested overnight with HindIII, separated by electrophoresis on 0.8% agarose gel, and transferred to Hybond XL nylon membrane (Amersham, GE Healthcare Limited, Buckinghamshire, United Kingdom), sequentially. The coding sequence of hpt was labeled as DNA probes for hybridization using a PCR digoxigenin (DIG) probe synthesis kit (Roche Diagnostics, Basel, Switzerland). The DNA filter was washed sequentially with 2× SSC, 0.1% SDS; 1× SSC, 0.1% SDS for 10 min at room temperature; and 0.5× SSC, 0.1% SDS for 15 min at 65°C after prehybridization and hybridization. Hybridization signals were detected using a Lumivision PRO (TAITEC, Saitama, Japan).

Total RNA was extracted as described above. One μg of total RNA was used for synthesis of first-strand cDNA, using the PrimeScript RT reagent Kit With gDNA Eraser (Takara Bio, Inc., Otsu, Shiga, Japan). After diluted for 50-fold, the cDNA was used as template for real-time quantitative RT-PCR (qRT-PCR) analysis in a total of 10 μl PCR reaction, containing 15 ng of cDNA, 200 nM each of forward and reverse primers, and 5 μl SYBR Premix Ex Taq (Takara Bio, Inc., Otsu, Shiga, Japan), with three technical replicates and two biological replicates. A parallel reaction to amplify actin1 was used to normalize the amount of template. The PCR primers include: CdSAMDC1 forward primer ZG5547 (5′-CGCCATCGAAGCAATAGAAAAC-3′) and reverse primer ZG5548 (5′-CCGCGGCCAAGCAGGAA-3′);actin1 forward primer ZG1551 (5′-TCTTGCTGGTCGTGACTTGACGG-3′) and reverse primer ZG1552 (5′-ACCTGGCCATCAGGCAGCTCAT-3′). The primers were designed using the software tool Beacon Designer (Premier Biosoft International, Palo Alto, CA, United States), and the primer specificity was validated by melting profiles and showed a single product specific melting temperature. All PCR efficiency was above 95%.

The temperature that resulted in 50% lethal (LT₅₀) was calculated using a fitted model plot for evaluation of freezing tolerance (Pennycooke et al., 2008; Guo et al., 2014). Freezing tolerance was also evaluated by survival rate. Stolons were cut into segments with one node, and placed in a beaker incubating in a Programmable Freezer (model: Polystat cc1 & k6, Huber Unit, Offenburg, Germany), following freezing treatment by decreasing temperature from 25 to 0°C linearly within 6 h, and maintained at 0, -2, and -3°C for 1 h respectively. After thawing overnight at 0°C, the stolons were planted in soil for vegetative propagation until new shoots were regenerated. The regenerated plants were counted and survival rate were calculated. The experiments were performed three times using 40 stolons segments each line per replicate. For assessment of chilling tolerance, centipedegrass plants were moved into a growth chamber at 6°C with a 12-h photoperiod under light of 200 μmol m^-2 s^-1 for 30 days as described previously (Chen et al., 2013). Net photosynthetic rate (A) was measured by using a LI-6400P Portable Photosynthesis System (LI-COR, Inc., Lincoln, NE, United States) (Chen et al., 2013).

Leaves were stained in 1 mg ml^-1 of 3,3-diaminobenzidine (DAB) solution for 1 h, followed by decoloring in boiling ethanol (95%) for 20 min before photography (Orozco-Cárdenas and Ryan, 1999). In another case, H₂O₂ was determined spectrophotometrically as previously described (Zhou et al., 2006).

Free polyamines were extracted and measured as described previously (Chen et al., 2013). Leaves (0.5 g) were extracted in 5 ml of 5% (v/v) cold perchloric acid (PCA) and incubated on ice for 1 h. The homogenate was centrifuged at 20,000 ×g for 30 min. Aliquots (0.5 ml) of supernatant were mixed with 1 ml of 2 M NaOH and 7 μl of benzoyl chloride and incubated at 37°C for 20 min in dark for benzoylation. The benzoylated polyamines were extracted to diethyl ether, resuspended in 1 ml of mobile phase solution (64% methanol in an isocratic elution), and filtered (4.5 μm filter) before HPLC analysis. Twenty μl of sample was injected into a Waters chromatographic system (Waters, Mildford, MA, United States), supplied with a C18 column (4.6 mm × 250 mm), and detected at 254 nm using a 2487 dual UV detector (Waters, Milford, MA, United States). Polyamine levels were calculated based on standard curves of commercial standards in combination with a recovery of the extraction procedure.

Polyamine oxidase was extracted in 0.1 M phosphate buffer (pH 7.0), and the activity was measured as described previously (Chen et al., 2014). The reaction was initiated and incubated at 30°C for 30 min after addition of 20 μl of 20 mM Spd or Spd into reaction mixture (3 ml) that was consisted of 2.50 ml of 0.1 M phosphate buffer (pH 7.0), 0.1 ml of horseradish peroxidase (250 units), 0.2 ml of coloring solution (25 μl N, N-dimethylaniline and 10 mg 4-aminoantipyrine were dissolved in 100 ml of 0.1 M phosphate buffer, pH 7.0) and 0.2 ml enzyme extract or inactivated enzyme (by heating the enzyme for 20 min in a boiling water bath) as control. Absorbance at 550 nm was recorded. One unit of PAO activity was defined as the amount of enzyme required for catalyzing 1 μmol of Spd or Spm oxidation within 1 min.

Leaves were ground in 5 ml of 50 mM phosphate buffer (pH 7.8) and supernatants were recovered for determinations of SOD, CAT, and NR after centrifugation at 13,000 ×g for 15 min as previously described previously (Guo et al., 2006; Lu et al., 2014). SOD reaction solution (3 ml), which was consisted of 13 μM methionine, 1.3 μM riboflavin, 63 μM ρ-nitro blue tetrazolium chloride, and enzyme extract in 50 mM phosphate buffer (pH 7.8), was incubated for 10 min at room temperature under fluorescent light with 80 μmol m^-2 s^-1. Absorbance at 560 nm was determined with a spectrophotometer. One unit of SOD activity was defined as the amount of enzyme required for inhibition of photochemical reduction of ρ-nitro blue tetrazolium chloride (NBT) by 50%. CAT reaction that contained 15 mM H₂O₂ in 50 mM phosphate buffer (pH 7.0) was initiated by adding 50 μl of enzyme extract. The decreased absorbance at 240 nm within 1 min was recorded. One unit of CAT was defined as the amount of enzyme required for catalyzing the conversion of 1 μmol H₂O₂ within 1 min. NR reaction that contained 60 mM KNO₃ and 0.25 mM NADH in 1.6 ml of 50 mM phosphate buffer (pH 7.5) was started by adding 0.4 ml of enzyme extract. After incubation at 25°C for 30 min, 1 ml of 1% sulfanilamide in 1.5 M HCl and 1 ml of 0.01% N-(1-naphthyl)-ethylenediammonium dichloride were immediately added into the reaction solution. Absorbance at 540 nm was recorded after centrifugation, and nitrite production was calculated by comparison with a standard curve of NaNO₂. One unit of NR was defined as the amount of enzyme required for catalyzing the conversion of 1 mmol NO₂ within 1 h.

Nitric oxide content was determined using Griess reagent as described previously (Zhou et al., 2005). Leaves (0.6 g) were ground in a mortar with pestle in 3 ml of 50 mM cool acetic acid buffer containing 4% zinc diacetate (pH 3.6). The homogenates were centrifuged at 10,000 × g for 15 min at 4°C. The supernatant was added by 0.1 g of activated charcoal. The filtrate was collected after vortex and filtration. One ml of filtrate was mixed with 1 ml of the Greiss reagent, followed by incubation for 30 min at room temperature. Absorbance at 540 nm was determined. NO content was calculated based on a standard curve of NaNO₂.

Protein concentration in enzyme extracts was determined using Coomassie Brilliant Blue G-250 solution using albumin from bovine serum (BSA) as a standard (Bradford, 1976).

The experiments were arranged in a completely randomized design with three pots of plants as replicates. For measurements of A, five leaves in each pot were randomly chosen and used for assay independently. For all the biochemical and physiological measurements, a pooled material from several different plants in each pot was randomly collected and used for assay. Significance of differences in the various parameters was assessed by one-way ANOVA (P < 0.05) using an SPSS program (SPSS, Inc., Chicago, IL, United States).

A cDNA sequence of CdSAMDC1 (GenBank accession number JX878505) with 1212-bp length was cloned from bermudagrass leaves. It encodes a peptide of 403 amino acids with 43.8 kDa and isoelectric point (pI) of 4.72. Sequence blast showed that CdSAMDC1 was most homologous (94%) to a CsSMADC2 (ADC45378) in Cleistogenes songorica in amino acid sequence, and 84 and 83% identical to ZmSAMDC4 in maize (CAA69075) and OsSAMDC1 in rice (LOC_Os04g42095), respectively. A phylogenetic analysis based on the full-length amino acid sequences showed that CdSMADC1 has a high identity with ZmSAMDC4, 1, and 2 (Figure 1). There is no insert in DNA sequence of CdSAMDC1. CdSAMDC1 transcript was significantly induced after 6 h of cold treatment and reached to a maximum level at 48 h. Higher levels were maintained until to 96 h after cold treatment (Figure 2A).

Transgenic centipedegrass plants were molecularly detected. DNA hybridization signals by using hpt fragment as probe were shown in transgenic lines 23 and 24 with a unique integration pattern, while no signal was shown in the WT plants (Figure 2B). Compared to WT plants, CdSAMDC1 transcript was detected in transgenic plants (Figure 2C). The results indicated that the transgene was integrated into centipedegrass genomes with one copy and CdDAMDC1 was expressed in transgenic centipedegrass plants. Free polyamines were measured in transgenic centipedegrass in comparison with WT. Put level showed no difference between WT and transgenic plants. Compared to WT, Spd level was 2.3 to 2.9-fold higher and Spd level and 1.6 to 1.8-fold higher in transgenic lines (Figure 2D).

Freezing tolerance was evaluated using LT₅₀ and survival rate. Compared to a -3.2°C of LT₅₀ in WT plants, transgenic lines had lower level of LT₅₀ (-5.2°C) under non-acclimated condition. After 7 days of cold acclimation, LT₅₀ decreased to -5.8°C in WT and -6.5°C in transgenic lines, respectively (Figure 3A). In consistence, WT had a 40% of survival rate, while transgenic lines had 59–61% of survival rate under non-acclimated condition. After cold acclimation, WT had a 64% of survival rate, while transgenic lines had a 79% of survival rate (Figure 3B). Chilling tolerance was also evaluated after 30 days of chilling treatment. WT and transgenic plants had similar A under non-stressed condition. Chilling treatment resulted in decrease in A, while significantly higher levels of A were observed in transgenic plants than in WT (Figure 3C). More brown or dead leaves were observed in WT than in transgenic plants after chilling treatment (Figure 3D). Color and H₂O₂ of the second leaf from the top were compared. Chilling treatment resulted in leaf browning in WT, but the leaf was maintained green in transgenic plants (Figure 3E). DAB staining showed that more H₂O₂ were accumulated in the second leaf in WT than in transgenic plants after chilling treatment (Figure 3F).

Antioxidant enzymes protect plants against low temperature induced oxidative damage on photosynthetic apparatus and membrane system (Chen et al., 2013; Lu et al., 2013). Antioxidant enzyme activity was measured. Compared to WT, 37 to 39% and 2.1- to 2.2-fold higher activities of SOD and CAT were observed in transgenic plants under control condition. SOD and CAT activities were increased after 7 days of cold acclimation, and higher activities were still maintained in transgenic plant than in WT (Figures 4A,B).

Spermindine and Spm are catabolized to produce H₂O₂ in plants catalyzed by PAO, while H₂O₂ is signaling in induction of cold responsive genes including those encoding antioxidant enzymes (Wan et al., 2009; Guo et al., 2014). Higher activity of PAO was observed in transgenic plants as compared with WT when either Spd or Spm was used as substrate (Figures 5A,B), and H₂O₂ level was higher in transgenic plants than in WT (Figure 5C). Feeding with MGBG, inhibitor of SAMDC, decreased PAO activity and H₂O₂ level in transgenic plants (Figure 5), indicating that the increased PAO activity and H₂O₂ level in transgenic plants were resulted from expression of CdSAMDC1.

Hydrogen peroxide has been shown to induce expression of NIR1 and lead to increased NR activity and accumulated NO in plants, while NO induces expression of antioxidant enzyme genes (Wan et al., 2009; Zhang et al., 2009; Lu et al., 2014). Higher levels of NR activities and NO were observed in transgenic plants compared with WT, which were blocked by pretreatment with AsA, an antioxidant to scavenge H₂O₂, or MGBG (Figures 6A,B), while pretreatment with NaN₂ and tungstate, inhibitors of NR, inhibited the elevated NO in transgenic plants. The results indicated that the elevated NR and NO were associated with accumulation of H₂O₂ as a result of expression of CdSAMDC1, while the elevated NO was dependent upon NR.

To assess whether the elevated PAO activity was affected by NR dependent NO, PAO activity was measured after plants were treated by NR inhibitors and NO scavenger. The results showed that PAO activity was not altered by pretreatments with NaN₂, tungstate, and PTIO, a scavenger of NO, when either Spd or Spm was used as substrate (Figure 7).

Activities of SOD and CAT were measured after pretreatment with MGBG, AsA, NaN₂, tungstate, and PTIO to understand whether H₂O₂ and NO signaling were involved in the enhanced antioxidant enzyme activity in transgenic plants (Figure 4). Pretreatments with MGBG and AsA blocked the elevated activities of SOD and CAT in transgenic plants. Similarly, the elevated activities were also blocked by pretreatment with PTIO, NaN₂ and tungstate (Figures 8A,B). The results indicated that the elevated SOD and CAT activities were associated with accumulated NO as a result of H₂O₂-dependent induction of NR.