H2O2 and Ca2+-based signaling and associated ion accumulation, antioxidant systems and secondary metabolism orchestrate the response to NaCl stress in perennial ryegrass

Little is known about the interplay between Ca2+ and H2O2 signaling in stressed cool-season turfgrass. To understand better how Ca2+ and H2O2 signals are integrated to enhance grass acclimation to stress conditions, we analyzed the rearrangements of endogenous ion accumulation, antioxidant systems and secondary metabolism in roots, stems and leaves of perennial ryegrass (Lolium perenne L.) treated with exogenous Ca2+ and H2O2 under salinity. Ca2+ signaling remarkably enhanced the physiological response to salt conditions. Ca2+ signaling could maintain ROS homeostasis in stressed grass by increasing the responses of antioxidant genes, proteins and enzymes. H2O2 signaling could activate ROS homeostasis by inducing antioxidant genes but weakened Ca2+ signaling in leaves. Furthermore, the metabolic profiles revealed that sugars and sugar alcohol accounted for 49.5–88.2% of all metabolites accumulation in all treated leaves and roots. However, the accumulation of these sugars and sugar alcohols displayed opposing trends between Ca2+ and H2O2 application in salt-stressed plants, which suggests that these metabolites are the common regulatory factor for Ca2+ and H2O2 signals. These findings assist in understanding better the integrated network in Ca2+ and H2O2 of cool-season turfgrass’ response to salinity.

The impact of exogenous Ca 2+ on endogenous ion accumulation in salt-stressed perennial ryegrass. Treatment with Ca 2+ under salinity enhanced higher Ca 2+ content in the leaves than that treated with NaCl alone (Fig. 2a). In the roots, NaCl treatment decreased the Ca 2+ content compared with the controls, but exogenous application of Ca 2+ relieved the decrease (Fig. 2c). In the leaves, external application of Ca 2+ declined Mg 2+ content at 4, 8 and 12 DAT, but treatment with NaCl alone stimulated the content decline at 12 DAT (Fig. 2a). In the stem, Mg 2+ content increased under both salt treatment and salt + Ca 2+ treatment (Fig. 2b). The increased Mg 2+ content in root was just found in treatment line with NaCl alone (Fig. 2c). For K + , NaCl treatment decreased the content in leaf, stem and root, but exogenous application of Ca 2+ relieved the decrease in leaf. In the NaCl treated leaf, stem and root of perennial ryegrass, endogenous Na + content showed remarkable increase but the increase was inhibited under the exogenous application of Ca 2+ . In addition, the NaCl treatment enhanced Ca 2+ /Mg 2+ ratio in leaf but reduced it in stem and root. Exogenous application of Ca 2+ under salinity caused a higher Ca 2+ /Mg 2+ ratio in leaf and root compared with NaCl treatment alone. The K + /Na + ratio decreased in leaf, stem and root in response to NaCl treatment. Exogenous application of Ca 2+ relieved the decrease in leaf at 8 DAT and in NaCl-stressed root.

The impact of exogenous Ca 2+ on the antioxidant systems of salt-stressed perennial ryegrass.
To investigate the salinity response induced by Ca 2+ signal, the SOD, POD, CAT and APX activities, as well as oxidative stress-responsive isoenzymes and genes were analyzed (Fig. 3). During the 12 days treatment, salinity stress induced greater SOD activity in leaves compared with controls, however, Ca 2+ addition to NaCl-treated plants induced more pronounced increase at 8 and 12 DAT (Fig. 3a). An increase in POD and CAT activities was Scientific RepoRts | 6:36396 | DOI: 10.1038/srep36396 observed at 4 DAT, while the activities were depressed at 8 and 12 DAT. Application of Ca 2+ upon salinity imposition enhanced POD activity at 4 and 12 DAT, and CAT activity at 4, 8 and 12 DAT respectively in comparison with NaCl only treatment (Fig. 3b). Salinity increased leaf APX activity at 4 DAT but decreased the activity at 12 DAT. Exogenous application of Ca 2+ in salt-treated leaves induced higher APX activity compared with NaCl treatment alone at 4, 8 and 12 DAT. The changes of antioxidant isoenzymes for SOD, POD, CAT and APX confirmed these discovery on enzyme activity (Supplementary Figure S2).

The impact of H 2 O 2 and Ca 2+ signaling on the oxidative status in salt-stressed perennial ryegrass.
To elucidate further the integrated connection between exogenous H 2 O 2 /Ca 2+ and oxidative protection against salinity stress, endogenous H 2 O 2 and Ca 2+ steady state levels were monitored in perennial ryegrass roots at 8 DAT (Fig. 4). Treatment with NaCl alone decreased Ca 2+  − accumulation and lower Ca 2+ production compared with NaCl treatment alone (Fig. 4).
Meanwhile, to further assess the role of H 2 O 2 and Ca 2+ signaling in the ROS metabolism, expression profiles of calmodulin binding protein genes (CaM1 and CaM2) and ROS metabolism genes (ChlCu/ZnSOD, CytCu/ZnSOD, MnSOD, CAT, APX, GPX and GR) were analyzed in perennial ryegrass leaves and roots. When exposed to all treatments, the transcripts levels of APX remained down-regulated in leaves (Fig. 5). Under NaCl treatment, CytCu/ZnSOD, MnSOD, CAT, GPX and GR increased transcript abundant but ChlCu/ZnSOD was down-regulated in the leaves. All ROS genes were up-regulated in the salt-stressed roots Upon Ca 2+ addition, ChlCu/ZnSOD, CAT, APX and GPX were down-regulated compared with control. However, application of H 2 O 2 or depriving Ca 2+ from NaCl treatment induced all ROS genes expression in leaves, except for CytCu/ZnSOD and APX. In roots, most of ROS genes also showed increased transcripts levels compared with control. Upon Ca 2+ addition, CaM1 was induced but CaM2 inhibited in both leaves and roots. However, CaM2 showed higher transcripts than CaM1 in roots after application of H 2 O 2 . For other treatment lines, CaM1 and CaM2 were induced in both leaves and roots.
The impact of H 2 O 2 and Ca 2+ signaling on the metabolite profiles in salt-stressed perennial ryegrass. To identify candidate metabolites targets in salt-stressed plants, the H 2 O 2 and Ca 2+ treated samples were extracted and derivatized, and then analyzed by GC-MS. A total of 160 total peaks with fairly consistent retention times (RT) and excellent resolution were resolved from each polar extract. Based on the internal consistency of RT and retention indices (RI), 41 metabolites could be identified across all samples in this study (Supplementary Table S2). The identified metabolites included 15 organic acids, 12 sugars, 8 amino acids, 4 fatty acids and 2 polyols ( Fig. 6; Supplementary Table S3). Each metabolite in each treatment was given in Table S3. Under NaCl treatment, the total content of 41 metabolites increased in perennial ryegrass leaves. Upon Ca 2+ addition, more metabolites accumulated at 8 DAT (Fig. 6d). However, H 2 O 2 stress decreased the 41 metabolites total content after 2 days treatment (Fig. 6c,d). Both NaCl and H 2 O 2 chemical treatments stimulated greater metabolites accumulation compared with NaCl or H 2 O 2 application alone (Fig. 6c,d). Removing Ca 2+ from NaCl treatment induced lower accumulation at 2 DAT ( Fig. 6c) but higher metabolites content at 4 and 8 DAT (Fig. 6b-d) compared with NaCl treatment in leaves. However, treatments without Ca 2+ in NaCl solution decreased the metabolites content in roots (Fig. 6e). In addition, treatments with NaCl or H 2 O 2 increased the content of metobolites, but upon Ca 2+ addition, metabolite accumulation declined in roots (Fig. 6e).
The impact of H 2 O 2 and Ca 2+ signaling on sugars and sugar alcohol content in salt-stressed perennial ryegrass. The increase of fatty acid, sugars and organic acids content was the major contributor of total metabolites content increase in leaves (Fig. 6). Eleven sugars and one sugar alcohol accounted for more than 66.0% of total metabolites content except H 2 O 2 chemical treatments only 49.5% at 8 DAT. By further evaluating the effects of H 2 O 2 and Ca 2+ signaling on sugars and sugar alcohol content in detail, we found that all the twelve metabolites were highly accumulated after Ca 2+ application compared with control and NaCl treatment  lines (Fig. 7). With NaCl treatment, tagatose, idose, sucrose, talose, allose (5TMS) BP and psicose showed higher content compared with controls. However, NaCl treatment induced lower glucose and inositol content compared with controls. Tagatose, idose, glucose, talose and psicose content had higher content by depriving Ca 2+ from NaCl treatment than by NaCl treatment alone in leaves (Fig. 7). H 2 O 2 treatment induced lower accumulation of tagatose, idose, glucose, sucrose, cellobiose, psicose and allose (5TMS) BP, but greater content of talose, glucoheptose and allose (5TMS) MP in comparison with controls. In contrast, both NaCl and H 2 O 2 chemical treatments stimulated higher abundance of tagatose, idose, sucrose, psicose and allose (5TMS) BP, but lower glucoheptose abundance compared with controls (Fig. 7).    Vertical bars indicate standard error of each mean for 11 sugars and one sugar alcohol (Inositol) at a given day of treatment. Experiments were repeated at least three times, each one with three replicates, and P-value was calculated by LSD' t-test.

Discussion
In this study, we initially characterized the impact of external Ca 2+ or H 2 O 2 on the physiological status of salt-treated cool-season turfgrass. Exogenous administration with Ca 2+ alleviated the physiological damage induced by salt tress, as shown by the higher turf quality and lower EL, MDA and H 2 O 2 content. This confirmed previous observations that Ca 2+ treatment significantly improved the physiological response of stressed plants 5 . Similar results in the Ca 2+ induced physiological behavior in stressed plants were previously found in tall fescue 23 , which suggest that Ca 2+ acts on key signaling in the stress response pathways of cool-season turfgrass 29 . Furthermore, we found that the application of external H 2 O 2 could trigger Ca 2+ signals positively involved in the rearrangements of ROS accumulation and metabolite profiles.
To elucidate the mechanism of salt tolerance induced by external Ca 2+ application and transport, we monitored internal levels of Ca 2+ , Mg 2+ , K + and Na + in NaCl stressed root, stem and leaf. We found that Ca 2+ pool in leaves and roots is modulated in response to NaCl stress. NaCl treatment decreased Ca 2+ accumulation in root, but exogenous application of Ca 2+ relieved the decline. In the leaves, no significant difference was obtained between controls and NaCl treated line, however, a higher Ca 2+ content was found after treatment with Ca 2+ under salinity. These results suggest that Ca 2+ metabolism is regulated by salinity, which was similar with heat 23,30 and cold stresses 5 . In addition, the Ca 2+ pool influx into leaves under salt stress may be a crucial salinity acclimation mechanism for perennial ryegrass. Our previous study had confirmed that the changes of macronutrient cations content, such as Ca 2+ , Mg 2+ , Na + , and K + , are necessary for salinity adaptation of perennial ryegrass 12 . In this study, NaCl treatment decreased K + but increased Na + concentration in leaves, stems and roots, which is consistent with our previous studies 12, 31 . In roots, external Ca 2+ application significantly increased K + concentration but decreased Na + concentration. Even though Ca 2+ treatment did not increase K + concentration in leaves, Na + concentration was still lower compared with NaCl treatment alone. Moreover, external Ca 2+ application significantly induced higher K + /Na + ratio in roots at 8 days treatment leaves. Evidence has shown that regulation of K + uptake and prevention of Na + influx to maintain desirable K + /Na + ratios in the cytosol are vital strategies of improving plant salt tolerance 32 . These results suggest that increasing the K + /Na + ratio could be one regulation mechanism induced by Ca 2+ signaling in salt stressed perennial ryegrass. Meanwhile, Mg 2+ concentration decreased at 12 DAT in leaves, but increased in stems and roots. Interestingly, feeding salt-treated leaves with Ca 2+ induced lower accumulation of Mg 2+ and Na + compared with the NaCl treatment alone. A higher Ca 2+ / Mg 2+ ratio were observed in leaf and root than NaCl-stressed tissues. Low Ca 2+ /Mg 2+ quotient in culture media is often thought as a restriction factor to restrict species from growing in these culture conditions 33 . We could deduce that the increase of Ca 2+ /Mg 2+ ratio could be one adaption mechanism induced by Ca 2+ signaling in salt stressed perennial ryegrass. Some protective antioxidant enzymes/genes and metabolites are the backbone of the oxidative status in plants 34,35 . Although salinity conditions produced excessive H 2 O 2 and O 2 − in plants, the higher antioxidant enzymes activity including SOD, POD, CAT and APX inhibited the excessive ROS accumulation to maintain ROS homeostasis 12,36 , and our enzymes activity data support this. Ca 2+ signals perform functions upstream or downstream in the ROS signal transduction network in plants 10 . Notably, Ca 2+ application under salinity inhibited the decrease of enzyme activity, and greater SOD, POD, CAT and APX activity could be obtained when exogenous application of Ca 2+ in NaCl solution. These data are in agreement with that Ca 2+ signaling could improve the stress adaptation by stimulating higher antioxidant enzymes activity in plants 23 , which were further confirmed by greater isoenzymes intensity of SOD (3)(4)(5), POD (2-5), CAT (1-2), APX (1-5) (Supplementary Figure S2). The increased antioxidant enzymes activity might be attributed to up/down regulated expression of the candidate genes 17,34 . Consistently, this study also observed that the expression of ChlCu/ZnSOD, CytCu/ZnSOD, MnSOD, CAT, APX, GPX and GR changed under salt stress. These were involved in the regulation of antioxidant enzyme activities. NaCl direct treatment increased transcript abundance of CytCu/ZnSOD, MnSOD, CAT, GPX and GR in leaves. In contrast, Ca 2+ addition treatment induced the up-regulation of only three genes CytCu/ZnSOD, MnSOD and GR in the leaves. However, in roots, ChlCu/ZnSOD, CytCu/ZnSOD, MnSOD, CAT, APX and GR were up-regulated under exogenous Ca 2+ application in NaCl treatment, indicating that the positive role in the Ca 2+ dependent regulation of antioxidant enzyme gene was greater in root than in leaf under salt stress. These results were confirmed by the opposite changes detected in roots and leaves after removing Ca 2+ from NaCl treatment.
Both Na + and Ca 2+ were added into half-strength Hoagland's solution and first affected the normal functioning of the root in our experiment. H 2 O 2 forms the basis of the ROS signals mediated priming phenomenon in stressed plants 37 41 . Our previous study showed that NaCl stress increased the transcript abundance of CAT, GPX, APX or GR 12 . Achieving ROS homeostasis is necessary self-defense mechanism for plant in response to stress conditions 42 . Here, qPCR analysis showed that CAT and APX had higher transcript abundance than SOD genes such as ChlCu/ZnSOD, CytCu/ ZnSOD and MnSOD in NaCl-stressed root. The same result was obtained under H 2 O 2 treatment and H 2 O 2 + NaCl treatment, suggesting that self-regulation for ROS homeostasis could be activated due to the excessive accumulation of H 2 O 2 induced by NaCl stress. Together with exogenous Ca 2+ signal could increase SOD, POD, CAT and APX activity, as well as may function as a downstream signal activated by H 2 O 2 , we speculated that Ca 2+ and H 2 O 2 signals may exist one common regulation orientation that maintains ROS homeostasis in salt-stressed cool-season turfgrass.
Since the accumulation of various metabolites changed when plants were exposed to salinity conditions 43 , metabolite profile analysis was performed in leaf and root samples to characterize how Ca 2+ and H 2 O 2 signals affect metabolite patterns in salt-stressed cool-season turfgrass. Here, 41 metabolites exhibited different response patterns in plants exposed to NaCl, Ca 2+ or H 2 O 2 . Consistent with the previous research in barley (Hordeum vulgare L.) 43 and Arabidopsis thaliana 44 , the accumulation of amino acids, organic acids and sugars changed in salt-stressed perennial ryegrass. Metabolites may therefore play vital roles in salinity response pathway. In addition, the present study also characterized a number of modified metabolites that were not observed in previous salinity metabolic response studies, such as polyol and fatty acids. Compared with NaCl treatment alone, the total content of 41 metabolites showed upward trend upon Ca 2+ addition during the first 4 days exposure. In contrast, removing Ca 2+ or adding H 2 O 2 accelerated the upward trend. At 8 DAT, upon Ca 2+ addition treatment obtained the highest accumulation compared with other treatment, but H 2 O 2 alone treatment had the lowest metabolic accumulation during the 8 days treatment. The present data combined with previous findings involved in metabolites response under salt stress, may help to understand better the molecular basis of Ca 2+ and ROS signaling in salt-treated plants.
Sugars not only directly provide energy and solutes for osmotic adjustment 45,46 , but also act as sugar-sensing signal to regulate key candidate gene expression 47 . Our metabolomic data demonstrated that the sugar and sugar alcohol contents accounted for 49.5-88.2% of all 41 metabolites accumulation in treated leaves, which lead us to hypothesize that sugars profiles is involved in sensing Ca 2+ and ROS signaling in salt-treated leaves. First, NaCl treatment induced the production of tagatose, idose, sucrose, talose, allose (5TMS) BP and psicose but inhibited glucose and inositol content in leaves. However, we found that all eleven sugars and one sugar alcohol showed higher accumulation after Ca 2+ application compared with control or NaCl treatment lines in leaves. On the contrary, NaCl treatment induced higher sugar than that in Ca 2+ application lines in root. These results suggest that up-regulated sugars induced by Ca 2+ signaling in leaf could contribute to the stronger response to salt stress. Secondly, the lowest contribution rate (49.5%) for sugars in total metabolites content was just obtained at H 2 O 2 treatment alone at 8 DAT. Correspondingly, in leaves, we found that H 2 O 2 treatment alone induced lower accumulation of tagatose, idose, glucose, sucrose, cellobiose, psicose and allose (5TMS) BP compared with controls or NaCl treatment lines, and just the talose, glucoheptose and allose (5TMS) MP content increased, which indicated that the positive role induced by Ca 2+ signaling in leaf may be inhibited by H 2 O 2 signaling. However, converse results were observed in stressed roots, which may be in relation to the different carbohydrate allocation in source and sink tissue under salt stress 46 .
In conclusion, the present study reported here reinforces the understanding of salt tolerance involved in H 2 O 2 and Ca 2+ -based signaling response through coupling ion accumulation, antioxidant systems and secondary metabolism analysis in cool-season turfgrass. Increased Ca 2+ /Mg 2+ ratio but decreased Na + /K + ratio were observed following external Ca 2+ application in salt-stressed grass. Exogenous Ca 2+ signaling could induce higher antioxidant enzymes activity and the expression level of antioxidant gene and protein in plants to adapt salinity environment. However, H 2 O 2 application decreased Ca 2+ signaling but increased the transcript abundance of CAT and APX. In order to maintain ROS homeostasis, Ca 2+ and H 2 O 2 signals had one common regulation pattern in salt-stressed cool-season turfgrass. In addition, H 2 O 2 /Ca 2+ -mediated metabolites detected in this study could provide a dataset of common regulatory factors for signaling transduction and salinity acclimation in perennial ryegrass. These findings involved in overlapping roles of H 2 O 2 /Ca 2+ signaling in salinity response mechanisms supply one novel strategy for a cool-season turfgrass's adaptation under salt stress.

Methods
Plant materials and growth conditions. Perennial ryegrass 'Quick start II' seeds were planted into plastic pots (10 cm diameter, 15 cm deep) with sand as substrate. After seedlings had grown to 8 cm tall, they were mowed to a height of 5 cm and were irrigated twice weekly with half-strength Hoagland's solution 48 . Forty-dayold seedlings were rinsed thoroughly using distilled water, and transferred into 300 mL Erlenmeyer flasks filled with approximately 290 mL half-strength Hoagland's solution. The bottlenecks were closed with appropriate amount of absorbent paper twined using plastic food wrap. To prevent potential algal growth, the flasks were wrapped with aluminum foil. The nutrient solutions were completely replaced each week. All plants were grown under greenhouse conditions (daily temperature of 2471/20711C (day/night), photosynthetic active radiation (PAR) at 300 μ mol m −2 s −1 , and a 14 h photoperiod) throughout the experiment. Treatments Experiment 1. Four weeks after canopy and root system establishment, plants were separated into three groups (Group I, II, III) according to the similar transpiration rate calculated based on the weight difference at three-day intervals before salt treatment initiation described by Hu et al. 12 . Group I plants were supplied with half-strength Hoagland's solution throughout the whole experimental procedure as control line. Group II plants, as salt treatment line, were treated with 300 mM NaCl for 12 d. Group III plants, as Ca 2+ treatment line, were treated with 300 mM NaCl + 7 mM Ca(NO 3 ) 2 ·4H 2 O for 12 d. All treatment solutions were prepared in halfstrength Hoagland's solution. Shoots samples for physiological, gene expression and metabolic analysis were harvested at 0, 4 d, 8 d and 12 d after treatment (DAT), respectively. Roots samples were harvested at the end of experiment. Treatments were arranged as a completely randomized design and each treatment having four plantpot systems as four replicates.

Measurements
Cell membrane stability and turf quality. Cell membrane stability was determined by the electrolyte leakage (EL) level measured based on the method described by Hu et al. 17 . Fresh 0.1 g of uniform leaf segments were washed with deionized water, incubated in 15 mL deionized water, and then shaken for 24 h at room temperature. The initial conductance (C i ) was measured using a conductance meter (JENCO-3173, Jenco Instruments, Inc., San Diego, CA, USA). The samples were then autoclaved at 120 °C for 20 min to completely disrupt the tissues and to release all electrolytes. After cooling to room temperature, the conductance of the incubation solution with killed tissues (C max ) was determined. Relative EL was calculated using the formula: EL (%) = (C i /C max ) × 100.
Turf quality was rated visually based on turfgrass color (percentage green leaves), plant density and degree of leaf wilting on a scale of 0 to 9, where 0 score indicated withered, yellow, thin or dead grass, while 6 indicated minimum acceptable level. A 9 score indicated green, dense and uniform grass 49 17 . The SOD activity was measured according to the method of Jiang and Huang 23 . A 3 mL reaction mixture, composed of 50 mM phosphate buffer solution (PBS) (pH 7.8), 60 mM Riboflavin, 195 mM Met, 3mmM EDTA, 1.125 mM NBT and 0.1 mL enzyme extract, was placed under light at 3000 lux for 10 min, and was recorded at 560 nm absorbance by spectrophotometer (UV-2600, UNICO Instruments Co., Ltd., Shanghai, China). One unit of SOD activity was defined as the amount of enzyme that inhibited 50% photochemical reduction of NBT. POD activity was measured as an increase in absorbance at 470 nm for 1 min following the oxidation of guaiacol 50  Cell lipid peroxidation was determined by the MDA content measured as follows. One milliliter of enzyme extract was mixed with 2 ml of reaction solution containing 20% (v/v) trichloroacetic acid and 0.5% (v/v) TBA. The mixture was heated in a water bath at 95 °C for 30 min, then cooled quickly in ice-water bath to room temperature, and centrifuged at 14,000 rpm for 20 min. Absorbance of the supernatant was measured at 532 and 600 nm. MDA content was calculated based on subtracting the absorption at 600 nm from the absorption at 532 nm and calibrated with the extinction coefficient of 155 mM −1 cm −1 .
H 2 O 2 level was determined based on the method described by Lin and Kao 53 . One milliliter of supernatant was mixed thoroughly with 1 ml of 0.1% titanium sulphate in 20% H 2 SO 4 (v/v), and then centrifuged at 6000 × g for 15 min at room temperature. The supernatant was measured at 410 nm absorbance. The H 2 O 2 level was calculated according the standard curve generated with known concentrations of H 2 O 2 and calibrated by extinction coefficient of 0.28 μ mol −1 cm −1 .
RNA isolation, cDNA synthesis and real-time PCR analysis. Total RNA extraction was performed from the leaves and roots using Trizol reagent (Invitrogen, Paisley, UK) according to the manufacturer's instructions and then was incubated with RNase-free DNase (RQ1; Promega, Madison, WI, USA). RNA integrity was examined at 260 and 280 nm by spectrophotometer (UV-2600; UNICO Instruments, Shanghai, China) and checked on a gel electrophoresis in 1.5% agarose gels with 1 μ L RNA (= 0.5 μ g μ L −1 ). Subsequently, cDNA synthesis was performed from 2 μ g purified RNA using cDNA synthesis kit according to the manufacturer's protocol (Fermentas, Burlington, ON, Canada). The resultant cDNA was diluted six-fold and kept at − 20 °C for RT-PCR analysis.
The transcript levels of the target genes were analyzed using ABI StepOne Plus Real-Time PCR system (Applied Biosystems, Foster City, CA) and SYBR Green Real-Time PCR Master Mix (Toyobo, Japan) in 20 mL reactions. Each reaction mix contained 2 ng of total RNA, 0.5 μ L of each primer (Supplementary Table S1) and 10 μ L master mix. Reactions were carried out as follows: initial denaturation at 95 °C for 3 min, 38 cycles of 10 s at 94 °C, 20 s at 50-55 °C, and 20 s at 72 °C, followed by 5 min at 72 °C. For specific product verification, a melting curve was performed from 82 °C at 0.2 °C increments with a 10 s hold between observations. YT521-B gene was used as a standard control in the RT-PCR reactions. The relative expression of specific genes was quantitated with comparative Ct method as described earlier 54  , perennial ryegrass roots were stained with 3,3′ -diaminobenzidine (DAB) and nitroblue tetrazolium (NBT), respectively, based on the method described by Zhang et al. 56  Metabolite extraction and derivatization. Metabolites from controls, Ca 2+ , H 2 O 2 , NaCl treated leaves and roots were extracted based on the procedure reported by Roessner et al. 57 with some modifications. Frozen leaves and roots were ground to a fine powder in liquid nitrogen, transferred into 2 mL Eppendorf tubes, and then extracted in 1.4 mL of 80% (v/v) aqueous methanol under intensive oscillation (200 rmp) at 25 °C for 2 h. Then, a 50 μ L methyl nonadecanoate (2 mg mL −1 in chloroform) with 20 μ L internal standard (2 mg mL −1 ribitol in water) was added. Extraction was carried out at 70 °C in a metal bath for 15 min (each 5 min vortex 5 s). The tube was centrifuged for 5 min at 12 000 g. The supernatant was decanted to new 10-mL tubes, and 1.5 mL of water and 0.75 mL of chloroform were added. The mixture was vortexed thoroughly and subsequently centrifuged for 15 min at 8000 g. 300 μ L polar phase (methanol/water) was decanted into 1.5 mL HPLC vials and dried in a benchtop centrifugal concentrator (Labconco Corporation, Kansas City, MI) overnight. The dried residue was redissolved and derivatized with 80 μ L of 20 mg mL −1 methoxyamine hydrochloride in pyridine for 2 h under Scientific RepoRts | 6:36396 | DOI: 10.1038/srep36396 intensive oscillation (200 rmp) at 37 °C, and followed by a 2 h treatment at 37 °C with 50 μ L N-methyl-N-trimethylsilyltrifluoroacetamide (MSTFA).

Gas chromatography mass spectrometry (GC-MS) analysis.
For GC-MS analysis, a 1 μ L of the derivatization solutions was determined with a GC-MS (DSQII, Agilent 7890A/5975C, Hemel Hempstead, USA) system based on the method described by Hancock et al. 58 . Samples were injected into the DB5-MSTM column (15 m × 0.25 mm × 0.25 μ m; J&W, Folsom, CA, USA) with a split ratio of 1:25. Injection temperature was set to 280 °C, the interface temperature was set at 290 °C, and the ion source temperature was adjusted to 200 °C. The column temperature was 5 min isothermal at 70 °C, then the GC oven temperature was raised to 260 °C with 5 °C min −1 in 2 min after injection, and finally held at 260 °C for 10 min. Helium gas was used as carrier set with a constant flow set at 1 mL min −1 . The MS measurement were set at electron impact (EI) source, electron energy 70 eV, solvent delay 4 min and the scan range 30-650 m/z at 0.6 scan s −1 . Retention time (RT), retention indices (RI, http://gmd.mpimp-golm.mpg.de/search.aspx) and NIST Mass Spectral Database (version 11) were implemented to identify the target metabolites. Only metabolite detected at least three in five samples was considered true. Each metabolite was finally identified based on the internal consistency of RT and RI. For each sample in controls and Ca 2+ , H 2 O 2 , NaCl treated leaves and roots, we performed four biological replicate and two technical replicates per biological experiment. Statistical analysis. Statistical significance was performed by ANOVA using the Statistics Analysis System (SAS) (version 9.0 for Windows; SAS Institute, Cary, NC). Means for physiological measurements, gene expression and metabolites between treatments were separated using the Fisher's least significant difference test (LSD) at P < 0.05 level.