Enhanced accumulation of gibberellins rendered rice seedlings sensitive to ammonium toxicity.

Ammonium (NH4+) phytotoxicity is a worldwide phenomenon, but the primary toxic mechanisms are still controversial. In the present study, we investigated the physiological function of gibberellins (GAs) in the response of rice plants to NH4+ toxicity and polyamine accumulation using GA biosynthesis-related rice mutants. Exposure to NH4+ significantly decreased GA4 production in shoots of wild-type (WT) plants. Both exogenous GA application to the WT and increases in endogenous GA levels in eui1 mutants rendered them more sensitive to NH4+ toxicity. In contrast, growth of sd1 GA-deficient mutants was more tolerant to NH4+ toxicity than that of their WT counterparts. The role of polyamines in GA-mediated NH4+ toxicity was evaluated using WT rice plants and their GA-related mutants. The eui1 mutants with GA overproduction displayed a higher endogenous putrescine (Put) accumulation than WT plants, leading to an enhanced Put/[spermidine (Spd)+spermine (Spm)] ratio in their shoots. In contrast, mutation of the SD1 gene encoding a defective enzyme in GA biosynthesis resulted in a significant increase in Spd and Spm production, and reduction in the Put/(Spd+Spm) ratio when exposed to a high NH4+ medium. Exogenous application of Put exacerbated symptoms associated with NH4+ toxicity in rice shoots, while the symptoms were alleviated by an inhibitor of Put biosynthesis. These findings highlight the involvement of GAs in NH4+ toxicity, and that GA-induced Put accumulation is responsible for the increased sensitivity to NH4+ toxicity in rice plants.

Introduction

Ammonium (NH4+) and nitrate (NO3−) are two important inorganic nitrogen (N) sources for plant growth (Ford, 2014; Liu ). Since metabolism of NH4+ consumes less energy than that of NO3−, NH4+ can be a better choice for energy conservation. However, upon exposure to solution containing NH4+, most plant species exhibit severe toxic symptoms, including suppression of plant growth, changes in root architecture, decreases in the root to shoot ratio, and leaf chlorosis (Britto and Kronzucker, 2002). Several mechanisms have been proposed to explain NH4+ phytoxicity. These include increased proton efflux, futile transmembrane NH4+ cycling, depletion of carbon supply, deficiency in mineral cations, damaged chloroplast ultrastructure, and oxidative stress (Kronzucker ; Britto and Kronzucker, 2002; Bittsánszky ; Esteban ).

Polyamines (PAs) have been reported to be involved in the regulation of NH4+ toxicity (Houdusse , 2008). PAs are low molecular weight, aliphatic polycations, and are ubiquitously distributed in plants (Hussain ). Putrescine (Put), spermidine (Spd), and spermine (Spm) are the three major PAs in plants (Bouchereau ; Minocha ; Liu ). An increase in endogenous PAs in response to abiotic stress has been observed in plants (Murty ). However, the function of the elevated PAs in response to abiotic stress is under debate because of their dual effects. Abiotic stress-induced increases in PAs can either protect plants against stress or cause damage to plants (Kusano ; Liu ; Bachrach, 2010; Alet ; Hussain ; Minocha ; Shi and Chan, 2014). There are reports demonstrating that the synthesis of Spm and Spd from Put, but not Put accumulation, could be a key protective factor for stressed plants (Bouchereau ; Capell ). It has been well demonstrated that NH4+ nutrition induced an accumulation of PAs in plants, particularly Put, leading to an increased Put/(Spd+Spm) ratio (Houdusse , 2008). Accumulation of Put can have negative effects on plants, such as tissue necrosis, protein loss, potassium leakage, and depolarization of membranes (Houdusse ). Put content has been shown to be closely correlated with NH4+ toxicity in wheat and pepper plants, while NO3− addition alleviated NH4+ toxicity due to decreases in Put content (Houdusse ). These results indicate that Put is a negative regulator for plants in response to NH4+ toxicity, but little information is available about the regulatory mechanisms of Put production under NH4+ toxicity conditions.

Although much progress has been made, the primary mechanisms underlying NH4+ toxicity in higher plants still remain unclear (Britto ; Esteban ). Several phytohormones have been implicated in the regulation of response to NH4+ toxicity, including auxin and ethylene (Peret ; Li ; Li ). For example, NH4+ inhibited primary and lateral root growth, and the total root length of plants was positively correlated with the auxin content (Peret ). Ethylene production was increased linearly with tissue NH4+ concentrations, and NH4+-induced shoot ethylene production was shown to be involved in inhibition of lateral root formation in Arabidopsis (Li ). These studies have mainly focused on the roles of auxin and ethylene in the regulation of root architecture under NH4+ toxicity conditions, while little information is available about the effects of phytohormones on shoot growth and N metabolism.

In addition to auxin and ethylene, gibberellin (GA) is another important plant hormone with multiple functions in plant growth and development, particularly its role in the Green Revolution (Hedden, 2003; Pingali, 2012). The introduction of semi-dwarf rice led to large yield increases throughout Asia in the 1960s, and the major semi-dwarfing allele, sd-1, is still extensively used in breeding modern rice cultivars (Spielmeyer ; Hedden, 2003). For rice, the SD1 gene encodes a defective enzyme in the GA biosynthesis pathway, and the mutation of this gene markedly decreased the shoot length in rice (Sasaki ). In addition to the introduction of high-yielding semi-dwarfing crop varieties, the application of large amounts of fertilizers accompanied the Green Revolution (Hedden, 2003). There are reports demonstrating that GA is involved in the regulation of phosphorus (P) and iron (Fe) acquisition in plants (Jiang ; Wild ; Wang ). In a recent study, Li reported that N use efficiency of Green Revolution varieties and grain yield of rice were increased by modulating the abundance of DELLA protein, which is a master regulator in the negative regulation of GA signaling. Specifically, they found that growth and N uptake were suppressed by GA deficiency-induced DELLA accumulation (Li ). These results highlight the roles of GA in N use efficiency. Rice is often planted under anaerobic conditions in the irrigated rice paddy field, and displays a greater tolerance to NH4+ than other cereals (Sasakawa and Yamamoto, 1978). Although rice is regarded as an NH4+-tolerant species, it also suffers from NH4+ toxicity, and excessive use of N fertilizer has increased the NH4+ concentration in many paddy soils, thus exposing rice plants to toxic levels of NH4+ (Balkos ; Chen ). In the present study, we used the wild-type (WT), the GA overproduction mutant eui1, and the GA-deficient mutant sd1 to evaluate the physiological role of GA in the response of rice plants to NH4+ toxicity.

Materials and methods

Plant materials and growth condition

Rice plants of ZS97, Taichung65, eui1, and sd1 were used in this study. eui1 is a GA overproduction mutant and sd1 is a GA-deficient mutant (Sasaki ; Zhu ). ZS97 and Taichung65 are the corresponding WTs of the eui1 and sd1 mutants, respectively. EUI1 encodes a cytochrome P450 monooxygenase that epoxidizes GA in a novel deactivation reaction in rice, and mutation of EUI confers a higher active GA concentration in eui1 mutant plants (Luo ; Zhu ). The rice SD1 gene encodes a defective enzyme in the GA biosynthetic pathway, and the sd1 mutant is deficient in bioactive GA content (Sasaki ). Seeds were surface-sterilized by incubation for 3 min in 75% ethanol, followed by 10 min in 0.1% HgCl2, and then washed thoroughly with sterile water. The sterilized seeds were soaked in water for 24 h in the dark and then transferred to a nylon net floating on water for 1 week. The 7-day-old seedlings were then transferred to nutrient solution containing the macronutrients (mM): NH4NO3 (1.0), MgSO4·7H2O (1.64), K2SO4 (0.51), CaCl2·4H2O (1.0), and NaH2PO4·2H2O (0.32), and the micronutrients (µM): MnCl2·4H2O (1.0), H3BO3 (18.9), (NH4)6Mo7O24·4H2O (0.075), ZnSO4·7H2O (0.15), CuSO4·5H2O (0.15), and Fe (III)-EDTA (50) with a pH of 5.5. The pH of the nutrient solution was adjusted daily and the solution was renewed every 3 d. The hydroponic experiments were carried out in a growth room with a 16 h light (30 °C)/8 h dark (22 °C) photoperiod, and the relative humidity was controlled at ~70%.

Treatments with different N forms

In our preliminary experiments, we found that treatment with 2.5 mM NH4+ had no effect on ZS97 and the eui1 mutant. When NH4+ in the growth medium was increased to 10 mM, shoot growth of ZS97 was not affected, while eui1 mutants showed NH4+ toxicity symptoms. We thus chose two levels of NH4+ (2.5 mM and 10 mM) in the present study. Given that NO3− treatment is often used as a control in studies of NH4+ toxicity (Leleu and Vuylsteker, 2004; Houdusse , 2008; Garnica ), the corresponding NO3− concentrations were used as control in this study. After 2 weeks pre-culture in the above nutrient solution, NH4NO3 was replaced by NH4Cl or KNO3 at the final concentration of 2.5 mM and 10 mM. The pH of the nutrient solutions (adjusted with KOH) was kept constant by buffering at 5.5 with 2.0 mM MES. Potassium (K+) was supplemented to 10 mM with KCl across all the treatments. After treatments with NH4Cl or KNO3 for 2 weeks, the shoot length of ZS97, Taichung65, eui1, and sd1 plants was recorded using a ruler, seedlings were harvested and separated into shoots and roots, and they were dried for 2 d at 75 °C for determination of dry biomass.

Treatments with exogenous gibberellic acid (GA3) and paclobutrazol (PAC)

After a 2 week pre-culture, ZS97 and eui1 plants were transferred to nutrient solution containing 10 mM NH4Cl or 10 mM KNO3 with varying concentrations of GA3 (0, 1, 10, 100, and 1000 nM) for 2 weeks. Since 100 nM GA significantly increased the shoot length, 100 nM GA was used in the following exogenous GA application experiment. In addition to exogenous GA treatment, the GA biosynthesis inhibitor PACat 1 µM was used to treat WT (ZS97) and eui1 plants. GA or PAC was dissolved in a minimal volume of ethanol, and then made up to volume with nutrient solution. After treatment, the shoot length was recorded using a ruler, seedlings were harvested and separated into shoots and roots, and were dried for 2 d at 75 °C for determination of dry biomass.

Determination of endogenous bioactive GA

After a 2 week pre-culture, seedlings of the WT (ZS97) were transferred to solution supplemented with 2.5 mM or10 mM NH4Cl, or 2.5 mM or 10 mM KNO3 for 2 weeks. Endogenous concentrations of bioactive GA1 and GA4 in shoots of WT rice plants were determined by GC-MS following a modified protocols described by Wijayanti . Briefly, the collected shoots were instantly frozen in liquid nitrogen and ground with a pestle. To better preserve the GAs, the ground samples were dried under a vacuum (0.08 mbar) at low temperature (–50 °C). The dried shoot samples were weighed, recorded, extracted with MeOH:H2O=4:1 containing [2H2]GA1 and [2H2]GA4 as internal standards, and evaporated to dryness. The dried extract was re-dissolved in MeOH:H2O=4:1 and passed through a C18 Sep-pack column and HPLC column of ZORBAX SB-C18 (150 mm×4.6 mm, 5 μm). The fractions corresponding to GA1 and GA4 were collected and dried using a centrifugal concentrator. The GA fractions were trimethylsilylated with N-methyl-N-trimethylsilyltrifluoroacetamide (MSTFA) at 80 °C for 30 min. The samples were taken to dryness and dissolved in hexane prior to injecting into a fused silica glass capillary column DB-5 (0.25 mm×30 m, 0.25 μm film thickness, J&W). Injection and interface temperatures were 260 °C and 280 °C, respectively. The column temperature was maintained at 80 °C for 2 min, and then was increased to 180 °C by 10 °C min−1 and to 290 °C by 6 °C min−1. Electron energy was 70 eV. The identity of GAs was verified by monitoring diagnostic ions of both endogenous and deuterated GAs according to the reference of Fernández . The internal standards [2H2]GA1 (cat. no. 0322491) and [2H2]GA4 (cat. no. 0322531) were purchased from OlChemIm Company.

Determination of NH4+ and NO3− in plants

After a 2 week pre-culture, ZS97, Taichung65, eui1, and sd1 plants were exposed to solution supplemented with 10 mM NH4Cl or 10 mM KNO3 for 2 weeks. Seedlings were harvested, and samples of shoots and roots were ground to powder in liquid nitrogen. The samples were suspended in 5 ml of deionized water and incubated at 45 °C for 1 h. The suspension was filtered, and NO3− and NH4+ contents were determined by a continuous-flow auto-analyzer.

Quantification of polyamines

Methods used to quantify plant PAs were modified from previous methods (Naka ). In brief, 20 mg of lyophilized rice plant powder was extracted for determination of PAs with 1,7-diaminoheptane as internal standard, and then benzoylated. The benzoylated PAs were centrifuged for 10 min at 10 000 rpm, and the supernatant was collected for drying with a vaccum centrifuge concentrator (CV100-DNA, Aijimu, Beijing, China). The collected samples were stored at –20 °C for analysis the next day.

The identification and quantification of PAs were conducted according to PA standards as follows. In brief, the dried extracts were re-dissolved in MeOH immediately prior to the analyses, and examined using ultra-high performance liquid chromatography–tandem MS (UPLC-MS/MS; Waters, Milford, MA, USA) coupled to a triple-quadrupole mass spectrometer (XEVO®-TQ) with electrospray ionization (ESI). The separation was carried out with a ZORBAX Eclipse plus C18 (150 mm×3.0 mm) with particle size of 1.8 µm (Agilent Technologies, USA) at 40 °C. The solvents used were water containing 0.1% formic acid (solvent A) and acetonitrile (solvent B). The following gradient elution program at a flow rate of 0.4 ml min−1 was applied: 0–1 min (40% B), 1–6 min (40–100% B), 6–8 min (100% B), and 8–11 min (40% B). MS detection was performed with a positive ion ESI mode with q cone voltage of –30 V and a nebulizer gas flow rate of 800 l h−1. The mass spectrometer was operated in multiple reaction monitoring (MRM) mode. Product ion (MS2) spectra were generated using collision-induced dissociation (CID) in the collision cell with helium gas, and Masslynx NT version 4.1 (Waters) software was used to process the data. The most intense product ion signal was achieved by applying collision voltages of 25 eV for Put, Spd, Spm, and 1,7-diaminoheptane. The following transitions were selected for MRM: m/z 297→105 for Put, m/z 458→162 for Spd, m/z 619→497 for Spm, and m/z 339→114 for the internal standard.

Treatments with exogenous Put and dl-α-difluoromethylarginine (DFMA)

After a 2 week pre-culture, seedlings of ZS97 and eui1 were transferred to nutrient solution containing 10 mM NH4Cl supplemented with 2 mM Put or 1 mM DFMA, an inhibitor of Put biosynthesis (Kallio ; Yamamoto ). After treatments, shoot length was recorded using a ruler, seedlings were harvested and separated into shoots and roots, and then the tissues were dried for 2 d at 75 °C for determination of dry biomass.

Statistical analysis

One- or two-way ANOVA was conducted using the SAS statistical software. Significant differences among treatments were evaluated by LSD (least significant difference) multiple range tests (P≤0.05).

Results

Gibberellins are involved in responses to ammonium toxicity

NO3− has been reported to alleviate NH4+ toxicity, and NO3− treatment is often regarded as a control in studies of NH4+ toxicity (Leleu and Vuylsteker, 2004; Houdusse , 2008; Garnica ). We thus first compared growth responses of rice plants to GA application in the presence of 10 mM NH4+ and 10 mM NO3− in the growth medium. As shown in Fig. 1A, there was no difference in shoot length between NH4+ and NO3− treatments when there was no GA application. Exogenous application of GA at concentrations >100 nM significantly increased shoot length and shoot DW regardless of N forms (Supplementary Fig. S1 at JXB online). Shoot length and shoot DW with NH4+ treatment were lower than with NO3− treatment when exogenous GA was applied at concentrations >100 nM (Fig. 1A, B). In contrast to NO3−, the presence of NH4+ in the growth medium markedly inhibited root biomass production regardless of GA concentrations (Fig. 1C). Application of a high concentration of GA marginally enhanced root DW of plants grown in NH4+ medium, and it had no effect on the root DW of plants grown in NO3− medium (Supplementary Fig. S1). These results suggest that shoot growth of rice plants is sensitive to high NH4+ when exogenous GA is applied at concentrations >100 nM.

Fig. 1.

Growth responses of rice plants to exogenous GA application under different N forms. Two-week-old ZS97 rice seedlings were exposed to 10 mM NH4Cl or 10 mM KNO3 with varying concentrations of GA for 2 weeks. Shoot length and plant biomass were measured. Data are means ±SE (n=12 for shoot length, n=4 for plant biomass). Different lower case letters indicate a significant difference (P≤0.05) between different N treatments.

Since GA application changed the growth responses of rice plants to high NH4+ and NO3−, we further tested the effect of NH4+ and NO3− on endogenous bioactive GAs; that is, the concentrations of GA1 and GA4 in seedlings of rice plants. The endogenous GA1 concentration was not significantly different between seedlings treated with 2.5 mM and 10 mM NH4+ (Fig. 2A), while a significant increase in GA1 concentration was observed upon exposure to 10 mM NO3− (Fig. 2A). The endogenous GA4 concentration was markedly higher in the presence of 2.5 mM NH4+ in the medium than with 2.5 mM NO3−, and the GA4 concentration was reduced with increasing both NH4+ or NO3− levels in the medium, with the magnitude of reduction greater in higher NH4+ than in NO3− medium (Fig. 2B). The total bioactive GA (GA1+GA4) concentrations exhibited similar changes to GA4 in responses to increases in N in the growth medium (Fig. 2C). These results imply that GA4, but not GA1, may play an important role in NH4+ toxicity in rice plants.

Fig. 2.

Effect of different NH4+ and NO3− concentrations on GA1 and GA4 production. Two-week-old ZS97 rice plants were exposed to different concentrations of NH4Cl or KNO3 for 2 weeks. Plant bioactive GA1 and GA4 in shoots were measured. Data are means ±SE (n=4). Different lower case letters indicate a significant difference (P≤0.05) within the same N treatment.

To test whether GA is involved in NH4+-induced changes in physiological processes in rice plants, we studied the responses of plant growth to NH4+ toxicity using GA biosynthesis-related rice mutants (eui1 and sd1). There was no difference in shoot length, shoot DW, and root DW between treatments with 2.5 mM and 10 mM NH4+ in WT plants (ZS97) (Fig. 3B–D). Mutation of EUI1 in eui1 mutants with overproduction of GAs significantly increased plant growth and altered the response to NH4+ compared with their WT plants (Fig. 3B–D). There were no visible NH4+ toxicity symptoms in shoots of WT plants in the presence of 10 mM NH4+ in the growth medium (Fig. 3A). In contrast, exposure to 10 mM NH4+ led to chlorosis and necrosis of newly formed leaves in eui1 mutants (Fig. 3A), which is a typical NH4+ toxicity symptom. Shoot length and plant biomass of eui1 decreased significantly by 10 mM NH4+ when compared with those in the presence of 2.5 mM NH4+ (Fig. 3B–D). In addition, there were marked interactive effects on rice plant growth between genotypes (WT and eui1) and NH4+ treatments (Supplementary Table S1). These results indicate that endogenous GA overproduction caused by the mutation of EUI1 can enhance sensitivity of rice to NH4+ toxicity. In contrast to the WT, mutation of the SD1 gene caused a decrease in shoot length but not plant biomass (Fig. 3F–H), and conferred on the GA-deficient mutant sd1 greater tolerance to high NH4+ concentration in the medium. NH4+-induced leaf chlorosis and curling were observed in WT plants (Taichung65), but the symptoms were alleviated by mutation of the SD1 gene in the GA-deficient sd1 mutant plants (Fig. 3E). There was no difference in shoot length and shoot DW of the sd1 mutant between treatments with 2.5 mM and 10 mM NH4+ (Fig. 3F, G). Similar to eui1, exposure of seedlings to 10 mM NH4+ inhibited root growth of sd1 (Fig. 3H). In contrast to sd1 mutants, their WT Taichung65 plants were sensitive to NH4+ toxicity, such that exposure to 10 mM NH4+ led to significant decreases in shoot biomass (Fig. 3G). These results may suggest that GA is involved in NH4+ toxicity in rice plants.

Fig. 3.

The effect of an increase or decrease in endogenous GA on rice plant growth under different NH4+ concentrations. Two-week-old WT (ZS97 or Taichung65), eui1, and sd1 rice plants were exposed to 2.5 mM or 10 mM NH4Cl for 2 weeks. Shoot length and plant biomass were measured. Data are means ±SE (n=16 for shoot length, n=4 for plant biomass). Different upper case letters indicate a significant difference (P≤0.05) between the WT and mutants. Different lower case letters indicate a significant difference (P≤0.05) within the same rice genotype.

We further investigated the growth responses of eui1 and sd1 to NO3− supply in the growth medium. Mutation of EUI1 in eui1 mutants with overproduction of GAs significantly increased plant growth in NO3− treatment (Fig. 4A–C), and mutation of the SD1 gene only inhibited shoot length but not plant biomass (Fig. 4D–F). In contrast to NH4+, shoot length and plant biomass of the WT plants and mutants were relatively independent of NO3− concentrations in the medium (Fig. 4A–F), and there were no interactive effects on rice plant growth between genotypes and NO3− treatments (Supplementary Table S2), indicating that GAs play different roles in the regulation of NH4+ and NO3− nutrition.

Fig. 4.

Effect of an increase or decrease of endogenous GA on rice plant growth under different NO3− concentrations. Two-week-old WT (ZS9 or Taichung65), eui1, and sd1 rice plants were exposed to 2.5 mM or 10 mM KNO3 treatment for 2 weeks. Shoot length and plant biomass were measured. Data are means ±SE (n=16 for shoot length, n=4 for plant biomass). Different upper case letters indicate a significant difference (P≤0.05) between the WT and mutants. Different lower case letters indicate a significant difference (P≤0.05) within the same rice genotype.

To confirm the involvement of GA in NH4+ toxicity, we further investigated the response of the WT and eui1 mutant to exogenous GA and the GA biosynthesis inhibitor PAC in the presence of 10 mM NH4+ or 10 mM NO3− in the growth medium. Exogenous application of GA and PAC significantly increased and decreased the shoot length of the WT and eui1 (Fig. 5A, B), respectively, but the magnitude of the GA-induced increase of shoot length was less in NH4+ than in NO3− (Fig. 5A, B), and GA exacerbated NH4+-induced leaf chlorosis of the WT and eui1 (Supplementary Figs S2, S3). Application of GA and PAC significantly increased and decreased shoot DW of the WT, respectively, regardless of the N form (Fig. 5C). Both GA and PAC had a slight effect on shoot DW of eui1 under 10 mM NH4+ treatment (Fig. 5E). In contrast to NH4+, shoot DW of eui1 was significantly increased by GA and decreased by 40% by PAC in the presence of NO3− in the growth medium (Fig. 5E). Treatment with PAC did not affect root DW regardless of the N form (Fig. 5D, F). These results indicate that the eui1 mutant is more sensitive to NH4+ than WT plants, and that there is an additive effect between exogenous GA application and endogenous GA on NH4+ toxicity.

Fig. 5.

Effect of exogenous GA or the GA biosynthesis inhibitor PAC on rice plant growth of the WT and eui1 under different N forms. Two-week-old WT and eui11 rice plants were exposed to 10 mM NH4Cl or 10 mM KNO3 with 100 nM GA or 1 μM PAC for 2 weeks. Plant biomass was measured. Data are means ±SE (n=4). Different lower case letters indicate a significant difference (P≤0.05) within the same N treatment.

GA-induced Put accumulation enhanced ammonium toxicity

We further evaluated the effects of GA on NH4+ concentrations in roots and shoots of rice plants in the presence of NH4+. NH4+ concentrations in shoots were relatively insensitive to endogenous GA concentrations (Fig. 6A, B). However, NH4+ concentrations in roots were significantly increased in eui1 mutant plants, while a decrease in NH4+ concentration was observed in roots of sd1 mutant plants (Fig. 6A, B).

Fig. 6.

Effect of exogenous GA on NH4+ and NO3− concentration in shoots. Two-week-old rice plants were exposed to 10 mM NH4Cl for 2 weeks. NH4+ concentration in shoots and roots was measured. Data are means ±SE (n=4). Different lower case letters indicate a significant difference (P≤0.05) within the same organ.

Biosynthesis of PAs is closely associated with NH4+ metabolism, and Put accumulation and an increased Put/(Spd+Spm) ratio are positively correlated with NH4+ toxicity (Houdusse , 2008; Garnica ). To validate the involvement of PAs in GA-induced sensitivity to NH4+ toxicity, we quantified the concentrations of PAs in the WT, and eui1 and sd1 mutants in response to NH4+ supply. The Put concentration was similar in shoots of the WT and eui1 when grown in 10 mM NO3− medium (Fig. 7A). In contrast, Put concentration was markedly increased in WT and eui1 upon exposure to medium supplemented with 10 mM NH4+, and the magnitude of increase was greater in eui1 mutants than in their counterpart WT plants (Fig. 7A). In contrast, Spd and Spm concentrations were relatively constant in response to N forms and endogenous GA levels (Fig. 7B, C). There was no difference between the WT and eui1 in the Put/(Spd+Spm) ratio in the presence of 10 mM NO3− in the medium (Fig. 7D). In contrast to 10 mM NO3−, exposure to 10 mM NH4+ led to marked increases in the ratio in the WT and eui1 mutants (Fig. 7D). In contrast to the WT, EUI1 mutation enhanced the Put/(Spd+Spm) ratio by 46% under higher NH4+ conditions (Fig. 7D). These results suggest that the Put concentration and Put/(Spd+Spm) ratio are positively correlated with the GA concentration under high NH4+ growth conditions.

Fig. 7.

Effect of N forms and GA overaccumulation on PA production. Two-week-old WT (ZS97) and eui1 rice plants were exposed to 10 mM NH4Cl or 10 mM KNO3 for 2 weeks. Put, Spd, and Spm in shoots were measured. Data are means ±SE (n=4). Different lower case letters indicate a significant difference (P≤0.05) within the same N treatment.

There was no difference in Put concentration between sd1 and its corresponding WT in growth medium supplemented with high NH4+ (Fig. 8A). However, mutation of SD1 in the sd1 mutant led to significant increases in Spd and Spm concentrations, thus resulting in a decrease in the Put/(Spd+Spm) ratio in sd1 mutants (Fig. 8A, B).

Fig. 8.

Effect of GA deficiency on PA production. Two-week-old WT (Taichung65) and sd1 rice plants were exposed to 10 mM NH4Cl for 2 weeks. Put, Spd, and Spm in shoots were measured. Data are means ±SE (n=4). Different lower case letters indicate a significant difference (P≤0.05).

The above results suggest that Put is a key factor in NH4+ toxicity. To test this hypothesis, we investigated the effects of Put on growth of WT plants by treatment with exogenous Put application and inhibition of Put biosynthesis under 10 mM NH4+ conditions. No visible symptoms of NH4+ toxicity appeared in leaves of WT plants upon exposure to medium supplemented with 10 mM NH4+ (Fig. 9A). However, exogenous application of Put induced obvious symptoms of NH4+ toxicity, including suppression of shoot growth, leaf necrosis, and chlorosis (Fig. 9A). Further, Put application significantly decreased shoot length and shoot DW, but not root DW (Fig. 9B–D). In contrast, inhibition of Put production by its biosynthesis inhibitor DFMA improved shoot and root growth, as evidenced by increases in shoot and root DW (Fig. 9C, D). Similar to WT plants, treatments with Put and DFMA significantly inhibited and enhanced growth of eui1 mutants, respectively (Supplementary Fig. S4). These results indicate that Put is directly involved in the regulation of NH4+ toxicity.

Fig. 9.

Effect of Put on ZS97 plant growth under 10 mM NH4+. Two-week-old WT (ZS97) rice plants were exposed to 10 mM NH4Cl with 2 mM Put or 1 mM DFMA for 2 weeks. Shoot length and plant biomass were measured. Data are means ±SE (n=12 for shoot length, n=4 for plant biomass). Different lower case letters indicate a significant difference (P≤0.05).

Discussion

In the present study, we explored the physiological roles of GA in tolerance of rice plants to NH4+ toxicity using GA biosynthesis-related rice mutants. There has been increasing evidence to elucidate the mechanisms underlying NH4+ toxicity in higher plants (Britto ; Esteban ). Since NH4+ toxicity has multiple negative impacts on plant growth and development, the mechanisms are complicated and the primary mechanisms remain largely unclear. GA was the main candidate in the Green Revolution in the 1960s, and development of semi-dwarfism Green Revolution varieties of cereals has greatly increased crop yields (Hedden, 2003; Pingali, 2012). However, Green Revolution varieties often exhibit a decrease in N use efficiency (Hedden, 2003). Recently, Li found that GA deficiency reduced N responses and inhibited N uptake and metabolism in rice plants by regulation of DELLA protein. Although the Green revolution solved the low yield problem, it led to environmental damage due to a large amount of N fertilizer input. NH4+ is a preferred N source under anaerobic conditions in the irrigated rice paddy field. The excessive use of N fertilizers has increased NH4+ concentrations in many paddy soils, exposing rice plants to a toxic level of NH4+ in soils (Balkos ; Chen ). However, there is limited information on the interaction between GA and NH4+ toxicity in rice plants. Our results revealed that the symptoms associated with NH4+ toxicity were positively correlated with the shoot GA concentration in rice plants. We further demonstrated that Put overproduction was involved in the GA-induced sensitivity to NH4+ toxicity in rice plants.

Rice is often regarded as an NH4+-tolerant species in the literature (Sasakawa and Yamamoto, 1978). In the present study, we showed no effect of 10 mM NH4+ in the growth medium on shoot biomass of the WT (ZS97) (Fig. 3C), indicating that this genotype is tolerant to NH4+ toxicity. However, another WT (Taichuang65) was more sensitive to 10 mM NH4+ compared with ZS97 (Fig. 3G). It is common that substantial differences exist among plant genotypes in response to abiotic stresses. For example, indica and japonica rice differed in their responses to K deficiency and cadmium (Cd) toxicity (Uraguchi ; Zhou ; Chen ). In the present study, ZS97 and Taichuang65 belong to indica and japonica subspecies, respectively. It seems that variation in tolerance to toxic NH4+ may exist between rice subspecies. The difference in NH4+ toxicity between the two rice subspecies warrants further investigation. GA is a well-known phytohormone that promotes plant growth (Yamaguchi, 2008). We showed that both exogenous and endogenous GA enhancement promoted shoot length and biomass regardless of N form (cf. Figs 1, 3, 4). GA-deficient sd1 mutants only had decreased shoot length, but not shoot biomass (Figs 3, 4). This is possibly due to the SD mutation inducing thick and strong stalks to prevent lodging (Hedden, 2003). In the present study, we found that exogenous GA application led to less enhancement in shoot length and biomass of ZS97 in the presence of 10 mM NH4+ compared with 10 mM NO3− in the growth medium. NO3− is often used as a control in studies of NH4+ toxicity, and suppression of shoot growth is an important characteristic of NH4+ toxicity (Leleu and Vuylsteker, 2004; Houdusse , 2008; Garnica ). Our results suggest that there may be a close interaction between GA and NH4+ toxicity, and that GA is possibly involved in NH4+ toxicity in rice plants. Additionally, data obtained from the GA overproduction mutant eui1 exposed to varying levels of NH4+ support this proposition. For instance, eui1 mutants had a higher endogenous GA concentration than WT plants, and their shoot length and DW were greater than those of the WT in medium containing 2.5 mM NH4+, while their shoot DW was decreased to a similar level to WT plants when exposed to medium supplemented with 10 mM NH4+ (Fig. 3C). In addition to suppression of shoot growth, high levels of NH4+ in the growth medium also inhibited root growth (Britto and Kronzucker, 2002; Leleu and Vuylsteker, 2004). In this study, we found that root growth was inhibited by NH4+ relative to the same level of NO3− in the growth medium (Fig. 1C). We further found that rice root growth showed variable responses to GA under NH4+ toxicity conditions. Exogenous GA application had little effect on root growth of ZS97 in the presence of 10 mM NH4+ (Fig. 1C). In contrast to the WT, EUI1 mutation significantly suppressed root growth in the presence of 10 mM NH4+ (Fig. 3D). Root growth of the GA-deficient mutant sd1 showed a similar pattern to the WT when exposed to medium containing 10 mM NH4+ (Fig. 3H). The differences in root growth in response to a high level of NH4+ may result from a difference in GA concentration and sensitivity to NH4+ of these rice genotypes. Taken together, these findings indicate that a higher endogenous GA level in rice plants may make shoot growth more sensitive to high NH4+.

In contrast to the WT, exposure to high NH4+ inhibited growth of eui1, and induced leaf chlorosis and necrosis (Fig. 3A–D), which are typical symptoms of NH4+ toxicity (Britto and Kronzucker, 2002). Both exogenous GA application and overproduction led to more severe symptoms pf NH4+ toxicity in the WT and eui1 mutant, suggesting a negative regulation of NH4+ toxicity by GA. This conclusion is consolidated by the following lines of evidence. First, in contrast to eui1, GA-deficient sd1 mutants were more tolerant to NH4+ than their WT plants (Fig. 3E–G). Secondly, growth of eui1 and sd1 was not inhibited when grown in the same high concentrations of NO3− (Fig. 4), indicating that the NH4+ itself is toxic to the plants. Thirdly, WT plants showed an attenuated increase in plant biomass when exogenous GA application was >100 nM under 10 mM NH4+ conditions, but shoot biomass was positively correlated with exogenous GA doses under 10 mM NO3− (Fig. 1). Fourthly, in contrast to the WT and NO3−, there was an additive effect between exogenous GA application and endogenous GA on NH4+ toxicity in eui1 mutants under 10 mM NH4+, and depression of GA production by PAC alleviated NH4+ toxicity symptoms and offset PAC’s negative function on plant growth in eui1 (Fig. 5). Finally, plants can actively down-regulate GA production when exposed to stress environments (Jiang ; Wild ; Wang ). For instance, Fe deficiency inhibited active GA production in Arabidopsis and rice (Wild ; Wang ). Similarly, NH4+ toxicity inhibited bioactive GA production in WT plants (Fig. 2B, C), suggesting that down-regulation of endogenous GA production may be an effective pathway to enhance tolerance to NH4+ toxicity in rice plants. These results strengthen the conclusion that GA is a negative regulator of NH4+ toxicity, and that NH4+ toxicity is positively correlated with GA concentration in rice shoots.

GA has been reported to be involved in N uptake and metabolism in rice plants, such that a GA-deficient rice mutant has a lower N uptake rate (Li ). We found that GA overproduction and deficiency significantly decreased and increased NH4+ concentration in roots, respectively (Fig. 6). This result is in agreement with the results of Li . It has been reported that expression of genes and activities of enzymes related to NH4+ metabolism were suppressed by GA deficiency in rice (Li ). PAs are downstream products in the NH4+ metabolism pathway, and involved in the regulation of many abiotic stresses (Kusano ; Liu ; Bachrach, 2010; Alet ; Hussain ; Minocha ; Shi and Chan, 2014). The synthesis of Spm and Spd from Put, but not Put accumulation, has been suggested to be a key factor to protect stressed cells (Bouchereau ; Capell ). It has been well documented that NH4+ toxicity is often accompanied by Put accumulation in plants, leading to an increase in the Put/(Spd+Spm) ratio (Houdusse , 2008). Put accumulation induces leaf necrosis, and application of NO3− can alleviate NH4+ toxicity by inhibiting Put production (Houdusse , 2008). In our study, we found that NH4+ toxicity induced leaf necrosis in the GA overproduction mutant eui1 (Fig. 3). Similar to results in the literature, treatment with NH4+ induced Put accumulation and increased the Put/(Spm+Spd) ratio (Fig. 7), suggesting that GA-induced Put accumulation is a key factor in NH4+ toxicity. Although mutation of SD1 did not affect Put accumulation, it increased Spd and Spm concentration, leading to a decrease in the Put/(Spm+Spd) ratio (Fig. 8). The functions of Put accumulation in NH4+ toxicity have been well documented in the literatures (Houdusse , 2008; Garnica ). To further confirm that Put accumulation is a key mechanism responsible for NH4+ toxicity, we investigated the growth responses of rice plants to exogenous Put application and inhibition of Put biosynthesis. There were no visible NH4+ toxicity symptoms in the shoots of WT (ZS97) plants under 10 mM NH4+ (Fig. 3). However, exogenous Put application to ZS97 inhibited plant growth and induced obvious symptoms of NH4+ toxicity (Fig. 9). Furthermore, application of DFMA, an inhibitor of Put biosynthesis, enhanced rice plant growth under 10 mM NH4+ (Fig. 9; Supplementary Fig. S4). Despite enhancement of NH4+ toxicity symptoms in shoots by exogenous Put application, exogenous application of Put had no impact on root growth (Fig. 9). This may be accounted for by the different sensitivity to Put between roots and shoots. All the results suggest that Put may be directly involved in the regulation of NH4+ toxicity. Similar to our results, aluminum-inhibited root growth in rice seedlings has been reported to be mediated by Put accumulation (Wang and Kao, 2006). Therefore, our results highlight that the accumulated Put is likely to be an important regulator of NH4+ toxicity, and that excessive Put accumulation is responsible for GA-enhanced sensitivity to NH4+ toxicity in rice plants.

In addition to the regulation of NH4+ toxicity responses, we also found that GA possibly participated in the regulation of NO3− utilization in rice plants. However, the regulatory patterns of GA in NH4+ utilization differed from those of NO3−. NH4+ toxicity was negatively related to endogenous GA levels, while rice plant growth seemed to be positively related to endogenous GA levels under high NO3− conditions (Figs 1B, 4E, 5E). The results suggest that higher GA levels can enhance rice plant growth under higher NO3− conditions, and that the underlying mechanisms remain to be investigated in future studies.

In summary, a putative working model is shown in Fig. 10 to illustrate the GA-dependent physiological mechanisms that regulate NH4+ toxicity in rice plants according to our results and those reported in the literature. Variation of endogenous GA in rice plants modifies N uptake and metabolism of PAs, leading to different responses of rice plants to toxic NH4+. The molecular mechanisms underlying GA signaling cascades affecting NO3− and NH4+ uptake and metabolism in rice plants need further investigation.

Fig. 10.

A putative working model to illustrate the GA-induced physiological mechanisms regulating ammonium toxicity in rice plants according to this study and the literature (red arrow, increase; green arrow, decrease). Variation of GA in rice plants modifies N uptake and PA metabolism, leading to different sensitivity of rice plants to ammonium toxicity.

Supplementary data

Supplementary data are available at JXB online.

Table S1. Results of two-way ANOVA of the effect of genotypes and ammonium treatments on relative shoot length, shoot DW, and root DW.

Table S2. Results of two-way ANOVA of the effect of genotypes and nitrate treatments on relative shoot length, shoot DW, and root DW.

Fig. S1. Growth responses of rice plants to exogenous GA application under different forms of N.

Fig. S2. Effect of exogenous GA or the GA biosynthesis inhibitor PAC on WT rice plant growth under different forms of N.

Fig. S3. Effect of exogenous GA or the GA biosynthesis inhibitor PAC on eui1 rice plant growth under different N forms.

Fig. S4. Effect of exogenous Put on growth of eui1 plants under 10 mM NH4+ conditions.

Click here for additional data file.

Click here for additional data file.