Comparative Physiological Analysis Reveals the Role of NR-Derived Nitric Oxide in the Cold Tolerance of Forage Legumes

The role of nitric oxide (NO) signaling in the cold acclimation of forage legumes was investigated in this study. Medicago sativa subsp. falcata (L.) Arcang. (hereafter M. falcata) is a forage legume with a higher cold tolerance than Medicago truncatula, a model legume. Cold acclimation treatment resulted in increased cold tolerance in both M. falcata and M. truncatula, which was suppressed by pretreatment with tungstate, an inhibitor of nitrate reductase (NR), and 2-phenyl-4,4,5,5-tetramethylimidazoline-1-oxyl 3-oxide (PTIO), a scavenger of NO. Likely, NITRATE REDUCTASE 1 (NIA1), but not NIA2 transcript, NR activity, and NO production were increased after cold treatment. Treatments with exogenous NO donors resulted in increased cold tolerance in both species. Superoxide dismutase (SOD), catalase (CAT), and ascorbate-peroxidase (APX) activities and Cu,Zn-SOD2, Cu,Zn-SOD3, cytosolic APX1 (cAPX1), cAPX3 and chloroplastic APX1 (cpAPX1) transcript levels were induced in both species after cold treatment, which was suppressed by tungstate and 2-phenyl-4,4,5,5-tetramethylimidazoline-1-oxyl 3-oxide (PTIO). Treatment with exogenous NO resulted in enhanced activities of SOD, CAT, and APX. Moreover, higher levels of NIA1 transcript, NR activity, NO production, and antioxidant enzyme activities and transcripts were observed in M. falcata as compared with M. truncatula after cold treatment. The results suggest that NR-derived NO production and upregulated antioxidant defense are involved in cold acclimation in both species, while the higher levels of NO production and its derived antioxidant enzymes are associated with the higher cold tolerance in M. falcata as compared with M. truncatula.


Introduction
Low temperature is one of the major abiotic stresses limiting plant growth and development. Temperate plants have evolved a mechanism known as cold acclimation, by which they respond to low, but non-freezing temperatures to increase their freezing tolerance [1,2]. Cold acclimation involves altered expression of thousands of genes that leads to metabolism rearrangement and physiological adaptation. A few hundred of the approximately 4000 cold-regulated (COR) genes are regulated by the CRT binding factors (CBFs) pathway, while the others are CBF-independent in Arabidopsis [2]. Some of the COR genes encode key enzymes for osmolyte biosynthesis and the antioxidant defense system that lead to an accumulation of cryoprotective proteins and soluble sugars for the stabilization of the cellular osmotic potential under low temperature [3,4] and activation of antioxidant defense for

Differential Cold Tolerance in M. falcata and M. truncatula
A temperature that results in 50% electrolyte leakage (TEL 50 ) was determined for the evaluation of cold tolerance [23]. Lower TEL 50 was observed in non-acclimated plants of M. falcata than in M. truncatula. TEL 50 was decreased in both M. falcata and M. truncatula during cold treatment at 5 • C, except for that in M. truncatula at 21 d, and lower TEL 50 was observed in M. falcata than in M. truncatula throughout the cold treatment (Figure 1), indicating that M. falcata had higher cold tolerance than M. truncatula. The increased TEL 50 at 21 d in M. truncatula indicates a decrease in the cold tolerance. To understand whether this case was associated with chilling injury as a result of the long time exposure to low temperatures, the ion leakage and F v /F m were measured. Compared to control plants, ion leakage was increased and F v /F m was decreased in M. truncatula after 21 d of cold treatment, while they were not altered in M. falcata (Figure 2A,B). The results indicated that M. truncatula plants were damaged by long time exposure to low temperatures as compared with M. falcata., which led to a decreased freezing tolerance in M. truncatula.

Differential Cold Tolerance in M. falcata and M. truncatula
A temperature that results in 50% electrolyte leakage (TEL50) was determined for the evaluation of cold tolerance [23]. Lower TEL50 was observed in non-acclimated plants of M. falcata than in M. truncatula. TEL50 was decreased in both M. falcata and M. truncatula during cold treatment at 5 °C, except for that in M. truncatula at 21 d, and lower TEL50 was observed in M. falcata than in M. truncatula throughout the cold treatment (Figure 1), indicating that M. falcata had higher cold tolerance than M. truncatula. The increased TEL50 at 21 d in M. truncatula indicates a decrease in the cold tolerance. To understand whether this case was associated with chilling injury as a result of the long time exposure to low temperatures, the ion leakage and Fv/Fm were measured. Compared to control plants, ion leakage was increased and Fv/Fm was decreased in M. truncatula after 21 d of cold treatment, while they were not altered in M. falcata (Figure 2A, B). The results indicated that M. truncatula plants were damaged by long time exposure to low temperatures as compared with M. falcata., which led to a decreased freezing tolerance in M. truncatula. Ten-week-old M. falcata and M. truncatula cv. A17 plants were irrigated with 15 mL of 1 mM tungstate or 100 μM PTIO solution or H2O as a control, followed by exposure to 5 C in a growth chamber. Ion leakage of leaves was measured to calculate the temperature that resulted in 50% lethality (TEL50). Means of three replicates and standard errors are presented; the same letter below the column indicates no significant difference at p < 0.05. Ten-week-old M. falcata and M. truncatula cv. A17 plants were irrigated with 15 mL of 1 mM tungstate or 100 µM PTIO solution or H 2 O as a control, followed by exposure to 5 • C in a growth chamber. Ion leakage of leaves was measured to calculate the temperature that resulted in 50% lethality (TEL 50 ). Means of three replicates and standard errors are presented; the same letter below the column indicates no significant difference at p < 0.05.

NO is Involved in Cold Acclimation of M. falcata and M. truncatula
Pretreatment with PTIO or tungstate suppressed the decrease of TEL 50 in both M. falcata and M. truncatula during cold treatment (Figure 1), implying that NR-derived NO is associated with cold acclimation in both species. The data also showed that TEL 50 was continuously reduced in PTIO-or tungstate-treated plants during cold treatment, indicating that PTIO or tungstate could not fully block cold acclimation. To confirm the role of NO, the effect of the exogenous generator of NO on cold tolerance was examined. Leaflets were incubated with SNP or diethylammonium (Z)-1-(N,N-diethylamino) diazen-1-ium-1,2-diolate (DEA), donors of NO production, followed by measurement of TEL 50 . The result showed that TEL 50 was significantly decreased in both species by 12 or 24 h of treatment with SNP or DEA ( Figure 3A,B), indicating that the NO level is associated with cold tolerance. The above results suggested that NR-derived NO was involved in the cold acclimation of M. falcata and M. truncatula, but other mechanisms except for NO are also essential for cold acclimation. Ten-week-old M. falcata and M. truncatula cv. A17 plants were irrigated with 15 mL of 1 mM tungstate or 100 μM 2-phenyl-4,4,5,5,-tetramethylimidazoline-1-oxyl 3-oxide (PTIO) solution or H2O as a control, followed by exposure to 5 C in a growth chamber for 21 d. Means of three replicates and standard errors are presented; the same letter above the column indicates no significant difference within each day at p < 0.05.

NO is Involved in Cold Acclimation of M. falcata and M. truncatula
Pretreatment with PTIO or tungstate suppressed the decrease of TEL50 in both M. falcata and M. truncatula during cold treatment (Figure 1), implying that NR-derived NO is associated with cold acclimation in both species. The data also showed that TEL50 was continuously reduced in PTIO-or tungstate-treated plants during cold treatment, indicating that PTIO or tungstate could not fully block cold acclimation. To confirm the role of NO, the effect of the exogenous generator of NO on cold tolerance was examined. Leaflets were incubated with SNP or diethylammonium (Z)-1-(N,Ndiethylamino) diazen-1-ium-1,2-diolate (DEA), donors of NO production, followed by measurement of TEL50. The result showed that TEL50 was significantly decreased in both species by 12 or 24 h of treatment with SNP or DEA ( Figure 3A, B), indicating that the NO level is associated with cold tolerance. The above results suggested that NR-derived NO was involved in the cold acclimation of M. falcata and M. truncatula, but other mechanisms except for NO are also essential for cold acclimation.   B), or H2O as a control. Ion leakage of leaves was measured to calculate the temperature that resulted in 50% lethality (TEL50). Means of three replicates and standard errors are presented; the same letter above the column indicates no significant difference within each day at p < 0.05.

NO Production was Induced during Cold Treatment
A weak DAF fluorescence was observed in the leaves of both species before cold treatment, with higher fluorescence in M. falcata than in M. truncatula. The DAF fluorescence was enhanced after 24 h of cold treatment in both species, which was suppressed by pretreatment with tungstate and PTIO ( Figure 4A,B), indicating that the cold-induced NO production was associated with NR. In addition,  , or H 2 O as a control. Ion leakage of leaves was measured to calculate the temperature that resulted in 50% lethality (TEL 50 ). Means of three replicates and standard errors are presented; the same letter above the column indicates no significant difference within each day at p < 0.05.

NO Production was Induced during Cold Treatment
A weak DAF fluorescence was observed in the leaves of both species before cold treatment, with higher fluorescence in M. falcata than in M. truncatula. The DAF fluorescence was enhanced after 24 h of cold treatment in both species, which was suppressed by pretreatment with tungstate and PTIO ( Figure 4A,B), indicating that the cold-induced NO production was associated with NR. In addition, higher fluorescence was observed in M. falcata than in M. truncatula ( Figure 4A,B).  Ten-week-old M. falcata and M. truncatula cv. A17 plants were irrigated with 15 mL of 1 mM tungstate or 100 μM 2-phenyl-4, 4, 5, 5-tetramethylimidazoline-1-oxyl 3-oxide (PTIO) solution or H2O as a control, followed by exposure to 5 C in a growth chamber. After 24 h of cold treatment, NO production was detected using the NO specific fluorescent probe, DAF-FM-DA (A), and NIA1 and NIA2 transcripts were detected using quantitative RT-PCR. Actin was used as a reference gene to calculate the relative expression (C, D). Scan bar is 500 μm. NR activity was detected as indicated in the figure (B). Means of three replicates and standard errors are presented; the same letter above the column indicates no significant difference within each day at p < 0.05.

Antioxidant Enzyme Activities were Induced by Cold and NO
Nitric oxide has been previously documented to induce antioxidant enzyme activity and gene transcript in tobacco [31], while the antioxidant defense system plays an important role in abiotic stress tolerance. To understand whether NR-derived NO regulates antioxidant defense in M. falcata as compared to in M. truncatula, antioxidant enzyme activities were determined after 14 d of cold treatment when both species showed increased cold tolerance in response to cold treatment. Higher activities of SOD and CAT were observed in M. falcata than in M. truncatula under the control condition ( Figure 5A,B). SOD, CAT, and APX activities were increased in both species after 14 d of cold treatment, and higher levels were observed in M. falcata than in M. truncatula, while pretreatment with tungstate or PTIO suppressed the increase in enzyme activities ( Figure 5A-C). For further confirmation of the involvement of NO in the induction of the enzyme activities, the effects of exogenous NO on antioxidant enzymes were examined. SOD, CAT, and APX activities were induced after 12 h of treatment with DEA in both species, with higher activities in M. falcata than in M. truncatula ( Figure 5D-F). The above results indicated that the induced antioxidant enzyme activities during cold treatment were associated with NR-derived NO production. H2O2 was measured after Ten-week-old M. falcata and M. truncatula cv. A17 plants were irrigated with 15 mL of 1 mM tungstate or 100 µM 2-phenyl-4, 4, 5, 5-tetramethylimidazoline-1-oxyl 3-oxide (PTIO) solution or H 2 O as a control, followed by exposure to 5 • C in a growth chamber. After 24 h of cold treatment, NO production was detected using the NO specific fluorescent probe, DAF-FM-DA (A), and NIA1 and NIA2 transcripts were detected using quantitative RT-PCR. Actin was used as a reference gene to calculate the relative expression (C,D). Scan bar is 500 µm. NR activity was detected as indicated in the figure (B). Means of three replicates and standard errors are presented; the same letter above the column indicates no significant difference within each day at p < 0.05.
Nitrate reductase activity was increased in both species during cold treatment, which was inhibited by pretreatment with tungstate. In addition, higher NR activity was maintained in M. falcata than in M. truncatula ( Figure 4C), which was consistent with the NO data ( Figure 4B). NIA1 transcript levels were induced by 99% and 44%, respectively, in M. falcata and M. truncatula after 24 h of cold treatment, which were not altered by pretreatment with tungstate or PTIO ( Figure 4D). The NIA2 transcript level was not responsive to cold in both species ( Figure 4E). The results indicated that the increased NR activity in M. falcata during cold treatment might be associated with the induced expression of NIA1.

Antioxidant Enzyme Activities were Induced by Cold and NO
Nitric oxide has been previously documented to induce antioxidant enzyme activity and gene transcript in tobacco [31], while the antioxidant defense system plays an important role in abiotic stress tolerance. To understand whether NR-derived NO regulates antioxidant defense in M. falcata as compared to in M. truncatula, antioxidant enzyme activities were determined after 14 d of cold treatment when both species showed increased cold tolerance in response to cold treatment. Higher activities of SOD and CAT were observed in M. falcata than in M. truncatula under the control condition ( Figure 5A,B). SOD, CAT, and APX activities were increased in both species after 14 d of cold treatment, and higher levels were observed in M. falcata than in M. truncatula, while pretreatment with tungstate or PTIO suppressed the increase in enzyme activities ( Figure 5A-C). For further confirmation of the involvement of NO in the induction of the enzyme activities, the effects of exogenous NO on antioxidant enzymes were examined. SOD, CAT, and APX activities were induced after 12 h of treatment with DEA in both species, with higher activities in M. falcata than in M. truncatula ( Figure 5D-F). The above results indicated that the induced antioxidant enzyme activities during cold treatment were associated with NR-derived NO production. The transcript levels of the genes encoding SOD and APX were further examined. Cu,Zn-SOD1 and cAPX2 transcripts were not induced by cold in both species (Figure 6A,E). Transcript levels of Cu,Zn-SOD2, Cu,Zn-SOD3, cAPX3, and cpAPX1 were induced after 24 h of cold treatment, which was inhibited by pretreatment with tungstate or PTIO in both species ( Figure 6B,C,F,G). Although cAPX1 transcript was induced in both species by cold treatment, but the induction was not altered by The transcript levels of the genes encoding SOD and APX were further examined. Cu,Zn-SOD1 and cAPX2 transcripts were not induced by cold in both species (Figure 6A,E). Transcript levels of Cu,Zn-SOD2, Cu,Zn-SOD3, cAPX3, and cpAPX1 were induced after 24 h of cold treatment, which was inhibited by pretreatment with tungstate or PTIO in both species ( Figure 6B,C,F,G). Although cAPX1 transcript was induced in both species by cold treatment, but the induction was not altered by pretreatment with tungstate and PTIO ( Figure 6D). The results suggest that the cold induced antioxidant enzyme activities were associated with the expression of Cu,Zn-SOD2, Cu,Zn-SOD3, cAPX3, and cpAPX1. and cpAPX1 (chloroplast APX, G) transcripts were detected using qRT-PCR, and actin was amplified as a reference gene to calculate the relative expression. Means of three replicates and standard errors are presented; the same letter above the column indicates no significant difference at p < 0.05.

Discussion
M. falcata is important because of its great cold tolerance and is used for crossing with alfalfa in alfalfa breeding [22], while M. truncatula is an annual legume without a cold acclimation mechanism [23]. TEL50 is commonly used to evaluate the cold tolerance of alfalfa [23]. Lower TEL50 was observed in both non-acclimated and acclimated plants of M. falcata than in M. truncatula. Compared to the continuous decrease of TEL50 in M. falcata within 21 d of cold treatment, M. truncatula showed a chilling injury, with increased ion leakage and decreased Fv/Fm at 21 d of cold treatment, which resulted in an increase in TEL50. The results indicated that M. falcata had higher cold tolerance than M. truncatula, which is consistent with a previous report [23]. The differences in the CAS gene copy numbers and CRT/DRE copy numbers in the CAS gene upstream regions are proposed to determine the differential cold tolerance between M. falcata and M. truncatula [23]. Given that NO is involved in cold-induced expression of CBF1/3 and CBF regulons, such as COR15a, LTI30, and LTI78, in Arabidopsis [8] and NO production is increased in diverse plant species when plants or plant organs are exposed to low temperature [7][8][9][10]12], the importance of NO in cold tolerance is documented in the present study.
Nitrate reductase is the major enzyme catalyzing production of NO from nitrite reduction [14]. The cold-induced NO production is impaired in NR-deficient mutants or by treatments with NR inhibitors in Arabidopsis and Brassica napus [8,12]. Similar results were observed in this study. NO production, NR activity, and NIA1 transcript were increased after low temperature treatment in both M. falcata and M. truncatula, while the increased NO production was impaired by pretreatment with an inhibitor of NR in both species. Cold acclimation led to increased cold tolerance in M. falcata and M. truncatula, which was blocked by pretreatment with an inhibitor of NR or scavenger of NO, while (chloroplast APX, G) transcripts were detected using qRT-PCR, and actin was amplified as a reference gene to calculate the relative expression. Means of three replicates and standard errors are presented; the same letter above the column indicates no significant difference at p < 0.05.

Discussion
M. falcata is important because of its great cold tolerance and is used for crossing with alfalfa in alfalfa breeding [22], while M. truncatula is an annual legume without a cold acclimation mechanism [23]. TEL 50 is commonly used to evaluate the cold tolerance of alfalfa [23]. Lower TEL 50 was observed in both non-acclimated and acclimated plants of M. falcata than in M. truncatula. Compared to the continuous decrease of TEL 50 in M. falcata within 21 d of cold treatment, M. truncatula showed a chilling injury, with increased ion leakage and decreased F v /F m at 21 d of cold treatment, which resulted in an increase in TEL 50 . The results indicated that M. falcata had higher cold tolerance than M. truncatula, which is consistent with a previous report [23]. The differences in the CAS gene copy numbers and CRT/DRE copy numbers in the CAS gene upstream regions are proposed to determine the differential cold tolerance between M. falcata and M. truncatula [23]. Given that NO is involved in cold-induced expression of CBF1/3 and CBF regulons, such as COR15a, LTI30, and LTI78, in Arabidopsis [8] and NO production is increased in diverse plant species when plants or plant organs are exposed to low temperature [7][8][9][10]12], the importance of NO in cold tolerance is documented in the present study.
Nitrate reductase is the major enzyme catalyzing production of NO from nitrite reduction [14]. The cold-induced NO production is impaired in NR-deficient mutants or by treatments with NR inhibitors in Arabidopsis and Brassica napus [8,12]. Similar results were observed in this study. NO production, NR activity, and NIA1 transcript were increased after low temperature treatment in both M. falcata and M. truncatula, while the increased NO production was impaired by pretreatment with an inhibitor of NR in both species. Cold acclimation led to increased cold tolerance in M. falcata and M. truncatula, which was blocked by pretreatment with an inhibitor of NR or scavenger of NO, while exogenous NO treatments increased the cold tolerance in both M. falcata and M. truncatula. The results suggest that NR derived NO is involved in the cold acclimation of M. falcata and M. truncatula. In addition, higher levels of both NR activity and NO production were observed in M. falcata than in M. truncatula under low temperature conditions. It is suggested that the difference in NR activity and NO production during cold acclimation is associated with the differential cold tolerance between M. falcata and M. truncatula. An analysis of the promoter region of NIA1 in M. falcata is worthwhile to reveal the regulation of cold on NO production in the future. In addition, our results showed that cold acclimation increased cold tolerance was not completely inhibited by pretreatment with an NR inhibitor or NO scavenger, indicating that NO is not the exclusive factor for cold acclimation. NO is signaling in gene expression. The expression of a total of 1023 cDNA fragments and 1932 genes are altered in response to NO in M. truncatula and upland cotton, respectively [32,33], compared to the altered expression of 4000 genes during cold acclimation [2]. Thus, mechanisms other than NO signaling are also associated with the differences between M. falcata and M. truncatula.
It is unavoidable for plants to accumulate ROS under low temperature conditions, as a result of an imbalance between the production and utilization of a photo-generated reductant that leads to an increased photosynthetic electron flux to O 2 for the production of ROS. Higher antioxidant enzyme activities, including SOD, CAT, and APX, are maintained in chilling tolerant mutants than in the wild type of centipedegrass and stylo under low temperature conditions to avoid oxidative damages [34,35]. SOD, CAT, and APX activities are induced in M. falcata during cold treatment, and an induced expression of the genes encoding antioxidant enzymes is associated with the enhanced cold tolerance in transgenic tobacco plants overexpressing MfSAMS1 [29]. SOD, CAT, and APX activities and transcripts of Cu, Zn-SOD2, Cu, Zn-SOD3, cAPX3, and cpAPX1 were induced in M. falcata and M. truncatula after cold treatment, which were impaired by treatments with an NR inhibitor and NO scavenger. Exogenous application of NO induced SOD, CAT, and APX activities, which was consistent with our previous observation in stylo and bermudagrass [31,36]. The results suggest that NR-derived NO is involved in cold induced antioxidant enzyme activities in M. falcata and M. truncatula. Moreover, M. falcata had higher activities of SOD, CAT, and APX and a lower accumulation of H 2 O 2 than M. truncatula during cold treatment, suggesting that the higher activities are associated with the higher cold tolerance in M. falcata compared with M. truncatula. NO induces antioxidant enzyme expression by activating mitogen-activated protein kinase (MAPK) cascades, which is downstream of the ABA and H 2 O 2 signaling pathway, in maize in response to water stress [37] and in bermudagrass, stylo, and tobacco plants [17,31,35,36,38]. Except for an involvement in the induction of antioxidant expression, NO also mediates cold induced expression of other genes, such as MfMIPS1, MfHyPRP, and MfSAMS1, and those genes confer cold tolerance in M. falcata [25,26,29]. NO is an important signal regulating cold acclimation in forage legumes by mediating the expression of multiple cold responsive genes. Our results did not exclude the possible role of NO in cold tolerance through S-nitrosylation. S-Nitrosoglutathione is an important in vivo S-nitrosylating agent that is formed by the reaction between NO and GSH. S-nitrosylation can activate or inhibit protein activity and affect protein translocation and function, for example, NO-mediated S-nitrosylation of iron-containing SOD is associated with chilling tolerance in B. juncea [11]. Cold-induced modifications of S-nitrosylation proteins have been identified in various plant species [7,11]. It remains to be determined whether S-nitrosylation is involved in the cold tolerance in M. falcata.
In conclusion, NO signaling plays an important role in the cold acclimation of forage legumes, such as M. falcata and M. truncatula. NR-derived NO production is involved in the cold acclimation of M. falcata and M. truncatula, by up-regulating antioxidant enzymes. Moreover, the higher levels of NR activity and NO production and its derived antioxidant enzyme activities are associated with the higher cold tolerance in Medicago falcata as compared with M. truncatula.

Plant Growth and Treatments
Plants of Medicago falcata cv. Hulunbeir and Medicago truncatula cv. A17 were grown in a mixture of peat and perlite (3:1, v/v) in plastic pots (15 cm diameter and 15 cm depth) under natural light in a greenhouse for 8-10 weeks as described previously [25,27]. Ten-week-old M. falcata and M. truncatula cv. A17 plants were divided into three groups and respectively irrigated with 15 mL of 1 mM tungstate, NR inhibitor [18], 100 µM 2-phenyl-4,4,5,5-tetramethylimidazoline-1-oxyl 3-oxide (PTIO) solution, NO scavenger [39], or H 2 O as a control, and then transferred to a growth chamber at 5 • C under light of 200 µmol photos m −2 s −1 for 1 d, 14 d, or 21 d for cold treatment before sampling for specific measurements. Plants were irrigated with the above solutions of 15 mL of the reagents once every three days. For treatment with exogenous NO generators, detached leaves of M. falcata and A17 were placed in a petri dish containing 100 µM diethylammonium (Z)-1-(N,N-diethylamino) diazen-1-ium-1,2-diolate (DEA), 200 µM sodium nitroprusside (SNP) [37], or H 2 O as a control, respectively, for 12 h under light of 200 µmol m −2 s −1 . Each pot contained five plants. Each leaf sample was harvested from one pot, and three samples were used for measurements as replicates.

Isolation of RNA and Real-Time Quantitative PCR (qPCR) Analysis
Total RNA was isolated from leaves using a HiPure Plant RNA Mini Kit (Magen, Guangzhou, China). 1 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). qPCR was conducted for the detection of gene transcripts in a Mini Option Real-Time PCR system (Bio-Rad, Hercules, CA). PCR solution (10 µL) contained 15 ng of diluted cDNA template, 200 nM each for forward and reverse primers, and 5 µL SYBR Premix Ex Taq (Takara Bio Inc.). The primers and sequences are listed in Table 1. A negative control without a cDNA template was always included. Parallel reactions to amplify actin1 were used to normalize the amount of template, as actin1 is reliable as z reference gene in M. falcata and M. truncatula [24]. Melting profiles were detected for validation of the primer specificity, showing a single product specific melting temperature. All PCR efficiencies were above 95%. Relative expression was calculated by 2 −∆∆Ct , which was done automatically by the instrument.

Evaluation of Cold Tolerance
Cold tolerance was estimated by the temperature that resulted in 50% ion leakage (TEL 50 ) [23]. After plants were exposed to low temperature at 5 • C for 0, 7, 14, and 21 d (Figure 1), or 0, 6, 12, and 24 h, leaflets were detached and placed in a tube, which was then placed on ice for 1 h of equilibrium, followed by the addition of ice chips to the leaflets in each tube. The tubes were then placed in an ethanol bath in a programmable freezer (model: Polystat cc1 and k6, Huber Unit, Offenburg, Germany), followed by equilibration for 1 h at 0 • C and the temperature was decreased at a rate of −2 • C h −1 to various freezing temperatures (0, −2, −4, −6, −8, −10, −12, and −14 • C) with 1 h of holding. Three tubes were used as a replicate at each temperature. After thawing overnight at 0 • C in a freezer, 6 mL of deionized water was added to each tube. Ion leakage was measured at room temperature and calculated as (C 1 /C 2 ) × 100, where C 1 and C 2 indicate the conductivity before and after heating as previously described [25]. The freezing temperatures and corresponding ion leakages were used for calculation of TEL 50 using a logistic sigmoid fitted model plot by software Origin 9.0 (OriginLab, Hampton, 01036). For the evaluation of the chilling tolerance, the ion leakage and the maximum photochemical efficiency of the photosystem II (F v /F m ) was measured after plants were exposed to a low temperature of 5 • C for 21 d. For measurement of the ion leakage, leaflets were placed in a tube, followed by addition of 6 mL of deionized water. Ion leakage was measured at room temperature and calculated as (C 1 /C 2 ) × 100, where C 1 and C 2 indicate the conductivity before and after heating as previously described [25]. F v /F m was measured as described previously using a pulse-modulated fluorometer (Model FMS-2, Hansatech Instruments) according to the manufacturer's instructions [35].

Detection of NO in Leaves
NO-specific fluorescent dye 4-amino-5-methylamino-2,7-difluorofluorescein diacetate (DAF-FM DA, Sigma) was used for the determination of NO production as described previously [12] with modifications. Leaflets were incubated in 10 mM Tris-HCl buffer (pH 7.0) containing 10 µM DAF-FM for 1 h in the dark at room temperature, followed by washing with Tris-HCl (pH 7.0) buffer. A confocal laser scanning microscope (LSM 510, Carl Zeiss, Jena, Germany) was used for imaging the leaflets, with excitation at 488 nm and emission at 515 nm. The images were processed and analyzed using Image J software (Wayne Rasband, NH, USA). Data are presented as the relative fluorescence intensity [40].

Determination of NR, SOD, CAT, and APX Activities
Nitrate reductase, SOD, and CAT were extracted from leaves (0.3 g) in 3 mL of 50 mM phosphate buffer (pH 7.8) containing 2% (w/v) PVP and 2 mM EDTA, while APX was extracted in 3 mL of 50 mM phosphate buffer (pH 7.0) containing 2% (w/v) PVP, 1 mM AsA, and 1 mM EDTA. After centrifugation at 15,000 × g for 15 min at 4 • C, the supernatants were recovered for determinations of NR, SOD, CAT, and APX activity as previously described [17]. One unit of NR was defined as the amount of enzyme required to catalyze the conversion of 1 µmol NO 3 − within 1 h, while one unit of CAT or APX was defined as the amount of enzyme required to catalyze the conversion of one µmol H 2 O 2 (extinction coefficient 0.00394 mM −1 cm −1 ) or ascorbic acid (extinction coefficient 2.8 mM −1 cm −1 ) within 1 min. One unit of SOD activity was defined as the amount of enzyme required for inhibition of the photochemical reduction of ρ-nitro blue tetrazolium chloride by 50%.

Statistical Analysis
The experimental data were subjected to an analysis of variances using an SPSS program (SPSS Inc., Chicago, IL). Duncan's t-test was used to evaluate the differences among the means of treatments and plant lines at the 0.05 probability level.