Nitrate nutrition influences multiple factors in order to increase energy efficiency under hypoxia in Arabidopsis

Abstract

Background and Aims

Nitrogen (N) levels vary between ecosystems, while the form of available N has a substantial impact on growth, development and perception of stress. Plants have the capacity to assimilate N in the form of either nitrate (NO₃^–) or ammonium (NH₄⁺). Recent studies revealed that NO₃^– nutrition increases nitric oxide (NO) levels under hypoxia. When oxygen availability changes, plants need to generate energy to protect themselves against hypoxia-induced damage. As the effects of NO₃^– or NH₄⁺ nutrition on energy production remain unresolved, this study was conducted to investigate the role of N source on group VII transcription factors, fermentative genes, energy metabolism and respiration under normoxic and hypoxic conditions.

Methods

We used Arabidopsis plants grown on Hoagland medium with either NO₃^– or NH₄⁺ as a source of N and exposed to 0.8 % oxygen environment. In both roots and seedlings, we investigated the phytoglobin–nitric oxide cycle and the pathways of fermentation and respiration; furthermore, NO levels were tested using a combination of techniques including diaminofluorescein fluorescence, the gas phase Griess reagent assay, respiration by using an oxygen sensor and gene expression analysis by real-time quantitative reverse transcription–PCR methods.

Key Results

Under NO₃^– nutrition, hypoxic stress leads to increases in nitrate reductase activity, NO production, class 1 phytoglobin transcript abundance and metphytoglobin reductase activity. In contrast, none of these processes responded to hypoxia under NH₄⁺ nutrition. Under NO₃^– nutrition, a decreased total respiratory rate and increased alternative oxidase capacity and expression were observed during hypoxia. Data correlated with decreased reactive oxygen species and lipid peroxidation levels. Moreover, increased fermentation and NAD⁺ recycling as well as increased ATP production concomitant with the increased expression of transcription factor genes HRE1, HRE2, RAP2.2 and RAP2.12 were observed during hypoxia under NO₃^– nutrition.

Conclusions

The results of this study collectively indicate that nitrate nutrition influences multiple factors in order to increase energy efficiency under hypoxia.

INTRODUCTION

Plants are obligate aerobes. They need oxygen for the operation of the mitochondrial respiratory chain involved in the production of energy that is a prerequisite for their survival and development. A gradual or sudden decline in oxygen occurs during flooding and waterlogging. Moreover, plants also experience hypoxia during their development, since cells within the central regions of bulky tissues are highly hypoxic (Geigenberger et al., 2000; Rolletschek et al., 2002, 2004). During the course of evolution, plants developed several strategies to combat hypoxia-induced stress. These processes include a range of physiological, biochemical and anatomical changes (Bailey-Serres et al., 2012). Anatomical changes include internode elongation in plants such as deep-water rice (Hattori et al., 2009), petiole elongation and the hyponasty response in Arabidopsis (Polko et al., 2011) and Rumex palustris plants (Colmer and Voesenek, 2009), and adventitious root formation (Mergemann and Sauter, 2000) and lysigenous aerenchyma formation in wheat (Wany et al., 2017) and in rice (Abiko et al., 2012). Not only anatomical adaptations but also metabolic changes are needed for coping with and recovering from flooding (Geigenberger, 2003). For instance, hypoxia modulates regulation of expression of metabolism-associated genes, energy consumption, cellular metabolism and growth (Geigenberger, 2003; Huang et al., 2008; Igamberdiev and Hill, 2009; Igamberdiev and Kleczkowski, 2011). Many hypoxia-tolerant plants have evolved several adaptive responses to compensate for the energy loss caused by oxygen deprivation (Gupta et al., 2009). Examples of such metabolic adaptations to hypoxia include the inhibition of storage metabolism and shifts towards more energy-efficient metabolic pathways such as the switch from the invertase to sucrose synthase pathways of sucrose catabolism (Sousa and Sodek, 2002; Bologa et al., 2003). Fermentation in combination with an altered modal operation of the tricarboxylic acid (TCA) cycle can also aid in coping with hypoxic stress (Banti et al., 2013). Plants can additionally directly decrease their rate of oxygen consumption under hypoxia via fine-tuning of the cytochrome and alternative pathways of mitochondrial respiration (Gupta et al., 2009; Zabalza et al., 2009). A further example of metabolic reprogramming is the fact that under hypoxia, citrate-mediated induction of alternative oxidase (AOX) induction occurs via inhibition of aconitase (Gupta et al., 2012) and this aids in the reduction of reactive oxygen species (ROS) (Royo et al., 2015). In addition, nitric oxide (NO) is also known to increase internal oxygen levels and control ROS (Gupta et al., 2014). NO has, furthermore, been proposed to control its own generation and scavenging by modulating nitrate assimilation and S-nitrosoglutathione reductase (GSNOR1) activity (Frungillo et al., 2014).

Nitrogen (N) has a great impact on all of the processes mentioned above. Nitrogen is essential for the biosynthesis of amino acids and many secondary metabolites; its soil level varies greatly, while the form of available N has a substantial impact on growth, development and biotic stress (Stitt, 1999; Gupta et al., 2013). Plants have the capability to assimilate N in the form of either NO₃^– or NH₄⁺. NO₃^– is readily transported to leaves for reduction and assimilation, whereas NH₄⁺ is mostly directly assimilated by roots (Crawford and Forde, 2002). Previously, it was shown that NO₃^– nutrition leads to higher levels of proteins and amino acids, and faster growth (Crawford, 1995; Forde and Clarkson, 1999). Moreover, primary metabolism is affected by NO₃^–, and organic acid levels increase following the enhanced NO₃^– nutrition. It is, however, important to note that NO₃^– has a higher energetic requirement since NO₃^– must be reduced to nitrite in the cytosol (utilizing one molecule of NADH in the process) and subsequently to NH₄⁺ in the plastid (utilizing three NADPH equivalents). NH₄⁺, formed either from NO₃^– or directly taken up from the soil, is eventually incorporated into the amino acid glutamate by the glutamine synthetase/glutamine:2-oxoglutarate aminotransferase (GS/GOGAT) system (Crawford and Forde, 2002).

There is a direct link between NO₃^– nutrition and NO production. Recent evidence suggests that plants produce significant levels of NO under hypoxia. The cytosolic nitrate reductase (cNR), by taking nitrite as an alternative substrate, plays a role in the production of NO from nitrite under hypoxia (Planchet et al., 2005). Previously, it was shown that acidification caused by hypoxia leads to an increase in NR activity (Botrel and Kaiser, 1997; Stoimenova et al., 2003). Dephosphorylation of NR also leads to an increase in NR activity (Botrel and Kaiser, 1997; Shi et al., 2008). NO₃^– levels influence NO production via production of nitrite, which can be transported to the mitochondria where it can be reduced to NO at the sites of complex III and IV (Gupta et al., 2005). Hypoxically generated NO can trigger the operation of the non-symbiotic haemoglobin1 (phytoglobin)–nitric oxide (Pgb–NO) cycle, within which its active scavenging by oxyhaemoglobin leads to regeneration of NAD⁺ and apparently also to ATP production (Stoimenova et al., 2007).

Despite the considerable advances in our knowledge described above, it remains unclear as to how NO₃^– or NH₄⁺ can impact the operation of the Pgb–NO cycle and what is their role in NAD⁺ recycling. Recent elegant studies demonstrated the importance of N-end rule-mediated protein degradation in hypoxia survival (Gibbs et al., 2011, 2015; Licausi et al., 2011; Vicente et al., 2017). Furthermore, activation of the group VII transcription factors has been clearly demonstrated to induce fermentative genes (Gibbs et al., 2011). However, the impact of either NO₃^– or NH₄⁺ nutrition on activation of these transcription factors is currently unknown.

Here, we studied the impact of NO₃^– and NH₄⁺ nutrition on fermentation, the TCA cycle, mitochondrial respiration and N metabolism of Arabidopsis. We provide novel insights into the role of NO₃^– nutrition in diverse processes, including the induction of group VII ERF transcription factors and fermentative genes, the increase in TCA cycle activity and amino acid metabolism, and the upregulation of the AOX pathway. Additionally, our results suggest that NO₃^– nutrition may be able to suppress ROS formation.

MATERIALS AND METHODS

Plant growth

Wild-type (Col-0) Arabidopsis seeds were thoroughly rinsed in sterile distilled water 2–3 times. Sterilization was performed by incubating seeds in 1.5 % NaOCl solution for 5 min followed by rinsing 7–8 times in sterile distilled water. The seeds were kept in a cold room for 24 h in order to break dormancy and subsequently transferred, under sterile conditions, to vertical agar plates containing 3 mm NH₄Cl or 3 mm KNO₃, 1 mm CaCl₂, 1 mm MgSO₄, 1 mm KH₂PO₄,1 mm K₂HPO₄, 15 μm H₃BO₃, 3 μm MnCl₂·4H₂O, 0.25 μm ZnSO₄·7H₂O, 0.1 μm CuSO₄·5H₂O, 0.04 μm Na₂MoO₄, 25 μm NaFe-EDTA, 0.8 % (w/v) sucrose and 0.8 % (w/v) agar. The pH was adjusted to 6.3. Plants were grown for 15 d in a 16 h:8 h day/night regime at 23 °C using Phillips TLD 36 W 1/830 light bulbs [200 μmol m^–2 s^–1 photosyntheticaly active radiation (PAR) at the height of the growing seedlings]. Plants were treated with normal air containing 21 % O₂ (normoxia) or a gas mixture that contains 0.8 % oxygen and 99.2 % nitrogen gas (hypoxia) for 24 h. They were subsequently removed from plates, then seedlings and root tissues were separated with an autoclaved scalpel and either immediately used for assays or frozen in liquid nitrogen and stored at –80 °C.

Determination of DAF fluorescence

For diaminofluorescein (DAF) fluorescence, roots were excised from plates immediately after hypoxia treatment (while plates were still in the hypoxia chamber) and incubated in 10 μm DAF-FM DA (Molecular Probes, Life Technologies, USA) for 15 min in darkness. The NO scavenger cPTIO [2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (Molecular Probes, Life Technologies and Sigma Aldrich, USA)] at 200 μm was used as a control. The formation of DAF-2T following NO reaction with DAF-FM DA was visualized using a fluorescence microscope (Nikon80i, Japan) upon excitation at 488 nm and emission at 515 nm. Images were analysed using the ImageJ software, and fluorescence intensity was plotted in arbitrary units (AU).

Gas phase Griess reagent assay

The NO emitted over a period of 6 h by the seedlings grown on 3 mm NO₃^– or 3 mm NH₄⁺ was detected by placing them in 20 mm HEPES buffer containing 0.5 mm nitrite, pH 7.2. Seedlings were treated in a stream of air or 0.8 % oxygen mixed with 99.2 % nitrogen gas into a solution containing 1 % (w/v) sulphanilamide and 0.02 % (w/v) NED [N-(1-naphthyl) ethylenediamine dihydrochloride] solution. The NO was first oxidized to NO₂ in the air stream and then converted into nitrite in the solution where the nitrite produced was calculated by measuring absorbance at 546 nm.

Measurement of nitrite levels and NR activity

The pre-frozen seedlings (200 mg) were finely ground in liquid nitrogen and transferred to 2 mL of extraction buffer [100 mM HEPES pH 7.6, 3.5 mm mercaptoethanol, 15 mm MgCl₂, 0.5% polyvinyl pyrrolidone (PVP), 0.5 % bovine serum albumin (BSA) and 0.3 % Triton X-100). The suspension was centrifuged at 14 000 rpm for 30 min at 4 °C and clear supernatant was used. For determining NR activity, 100 μL of supernatant was added to 800 μL of reaction mixture (50 mM HEPES-KOH pH 7.6, 5 mm KNO₃ and 0.2 mm NADH). After 10 min of incubation at 25 °C, the reaction was stopped by adding 125 μL of zinc acetate (0.5 m) and Griess reagent (0.5 mL of sulphalinamide acid and 0.5 mL of NED solution). Purified NR (Sigma-Aldrich, USA) was used in the reaction to calculate the rate of nitrate reduction. The nitrite content was determined colorimetrically at 546 nm. The control reaction was carried out using 0.1 μm potassium nitrite solution in order to check colour development with Griess reagent (Planchet et al., 2005).

Metphytoglobin reductase assay

Metphytoglobin (methaemoglobin) reductase was measured at 550 nm with cytochrome c as a substrate as described by Igamberdiev et al. (2004). The same extract was used for NR measurements was added to 950 μL of reaction mixture containing 50 mm sodium phosphate buffer (pH 7.4), 50 μm cytochrome c, 0.2 mm NADH and 10 μm methylene blue. The spectra between 516 and 600 nm were recorded on a spectrophotometer and the enzyme activity was calculated from the obtained values using a standard NADH curve.

Measurement of respiration

Respiration was measured using a TBR-1025 single channel oxygen sensor (World Precision Instruments, Sarasota, FL, USA) calibrated by a simple two-point calibration (100 and 0 % air saturation) using nitrogen gas. Seedlings weighing 15 mg were placed in a 1.5 mL amber glass vial and continuously stirred with a magnetic stirrer. To estimate electron transport partitioning between cytochrome c oxidase (COX) and AOX, 10 μm myxothiazol and 2.5 mm salicylhydroxamic acid (SHAM) were added independently in order to inhibit the respective pathways.

Measurement of enzyme activities

Pyruvate decarboxylase (PDC) (EC 4.1.1.1) activity was determined by the continuous rate determination method described by Gounaris et al. (1971) with slight modifications. Approximately 200 mg of fresh tissues were completely homogenized in 200 mm citrate buffer (pH 7.0). Then, the homogenate was centrifuged at 14 000 rpm for 5 min at room temperature. The supernatant was transferred into new 1.5 mL tubes. The reaction mixture (300 μL) consisted of 10 μL of sample extract in 150 mm citrate buffer (pH 6.0), 50 mm sodium pyruvate, 100 μm NADH and 10 U mL^–1 alcohol dehydrogenase (ADH) enzyme solution. The decrease in absorbance at 340 nm was read for 5 min in a 96-well microplate (96-well U-Bottom, clear PS microplate, Greiner-Bio-One, Sigma-Aldrich, India) set in the kinetic mode at a discrete wavelength of 340 nm in the microplate reader (Polar Star Omega, BMG LabTech, Germany). The PDC activity was calculated by the change in absorbance at 340 nm, and the molar extinction coefficient of NADH (6.22 mmol^–1 cm^–1 at 340 nm) using Beer–Lambert’s law. PDC activity is represented as μmol NADH formed during conversion of pyruvate to acetaldehyde per minute per μg of protein at room temperature. The protein concentration was calculated by Bradford’s assay using a BSA standard curve (Bradford, 1971).

The same extract (as used for PDC activity) was used for measuring ADH activity. The 200 μL reaction mixture consisted of 100 μg of total protein in 50 mm sodium phosphate buffer (pH 8.0), 150 mm ethanol and 300 μm NAD⁺. The increase in absorbance at 340 nm was read at 340 nm set in a kinetic mode in the microplate reader. The ADH activity was calculated by the change in absorbance at 340 nm, and the molar extinction coefficient of NADH (6.22 mmol^–1 cm^–1 at 340 nm) using Beer–Lambert’s law. ADH activity is represented as μmol NADH formed during conversion of ethanol to acetaldehyde per minute per μg of protein at room temperature

Measurement of the ATP/ADP ratio was performed using the luciferin/luciferase assay kit (Sigma Aldrich, USA). The NAD⁺/NADH ratio was determined using a quantification kit (Life Science SOURCE, Bio Vision, USA). Approximately 100 mg of fresh samples were weighed and washed with NADH/NAD extraction buffer (provided in the kit) and then completely homogenized in 400 μL of the same extraction buffer. The homogenate was centrifuged at 14 000 rpm for 5 min at room temperature. The NADH/NAD supernatant was then transferred into new 1.5 mL tubes. From this, an aliquot (50 μL) was transferred into a labelled 96-well microplate. This was used for the measurement of total NAD (NADt; NADH and NAD). For the measurement of NADH, a 200 μL aliquot was transferred to new 1.5 mL tubes. NAD was decomposed from the extract by heating the extract in a thermoblock set at 60 °C for 30 min. The samples were cooled on ice and then 50 μL of the decomposed NAD samples were transferred to the labelled 96-well microplate. For the NAD/NADH assay, the reaction mixture was prepared and measured by taking the absorbance at 450 nm. The amount of NADt and NADH was calculated by applying the unknown sample readings to the NADH standard curve (described in the kit). The NAD/NADH ratio was calculated by the formula described in the kit.

Measurement of ethanol and lactate

Seedlings grown with NH₄⁺ or NO₃^– nutrition were treated in air or 0.8 % O₂ (hypoxia) for 24 h, and ethanol production was analysed enzymatically. Ethanol and l-lactate were determined using the quantification kit by Megazyme (Ireland) with a microplate reader by following the manufacturer’s instructions. The same sample extract (PDC activity) was used for the quantification. The amount of ethanol and l-lactate produced was calculated by the formula described in the kit and is presented in ng per gram fresh weight.

Lipid peroxidation assay

The level of lipid peroxidation was monitored by spectrophotometric determination of malondialdehyde (MDA), using thiobarbituric acid (TBA) according to Jambunathan (2010) with some modifications. The control and treated root tissues (200 mg) of Arabidopsis were homogenized in 2 mL of 0.1 % trichloroacetic acid. The crude homogenates were centrifuged at 10 000 g for 15 min at 4 °C. The supernatant was collected, 500 μL was mixed with 1 mL of 20 % tricarboxylic acid and 1 mL of 0.5 % TBA, and incubated for 30 min at 95 °C in a fume hood. The reaction was stopped by keeping the tubes in an ice bucket for a period of 5 min. The absorbance values of non-specific light dissipation were measured at 600 nm and subtracted from that measured at 532 nm; these final extracted values represented the volume of MDA-reactive TBA. The concentration of MDA-reactive TBA was calculated from the molar extinction coefficient of 155 mM^–1 cm^–1 using Beer– Lambert’s equation.

Measurement of ROS

The superoxide (O₂·^–) levels were measured by in vivo staining using nitroblue tetrazolium chloride (NBT; Sigma-Aldrich) as a substrate (Jambunathan, 2010). Control and treated roots were excised from the shoot using a sharp razor blade. These were incubated in 2 mL of 0.1 % (w/v) NBT dissolved in 50 mm potassium phosphate (pH 6.4) for 8–12 h under dark conditions at room temperature. Following these treatments, the excess of NBT stain was removed by rinsing the roots three times with 95 % ethanol. Samples were stored in 95 % ethanol prior to being visualized and photographed under a bright field microscope (Nikon 80i, Japan) for the presence of dark blue-violet spots (formazan complex).

Intracellular ROS levels were measured by monitoring the fluorescence using dichlorofluorescein (DCF), the oxidation product of H₂-DCF. H₂-DCF is converted to the fluorescent form when oxidized by hydrogen peroxide, superoxide and other reactive free radical products. Control and treated root segments were cut with a sharp razor blade into fine segments which were incubated in 10 μm H₂-DCF-DA for 5 min at room temperature under dark conditions. Excess H₂-DCF-DA was removed by washing 3–4 times with 10 mm Tris–HCl (pH 7.2). The samples were subsequently immersed in 1 mL of the same buffer and visualized using the Nikon 80i fluorescence microscope (488 nm excitation and 525 nm emission).

Total RNA isolation, cDNA preparation and real-time quantitative reverse transcription–PCR of carbon, nitrogen and antioxidant pathway genes

Total RNA was isolated from 14-day-old seedlings using Tri reagent (MRC, USA). A 2 μg aliquot of total RNA of each sample was used to synthesize first-strand cDNA by oligo(dT)₁₈ primer using a Revert Aid™ H-Minus first-strand cDNA synthesis kit (Fermentas, Thermofisher Scientific, USA). Real-time qPCR amplification was carried out using the ABI Prism 7900 sequence detection system (Applied Biosystems), and power SYBR-green master mix (Applied Biosystems) with the relative quantification of the expression of each individual gene being performed using the comparative threshold cycle method, as described in the ABI PRISM 7900 Sequence Detection System user bulletin number 2 (Applied Biosystems). Actin, tubulin and 18S rRNA were used as internal reference genes for data normalization. The amplification program used for qRT–PCR was: 95 °C for 2 min, 40 cycles at 95 °C for 30 s, 60 °C for 30 s, 72 °C for 30 s; 60 °C for 15 s and 95 °C for 15 s. All samples were amplified in triplicate, and the mean Ct values were considered. Primer sequences are given in Supplementary Data Table S1.

RESULTS

Nitrate and ammonium nutrition causes change in NO levels, NR activity, phytoglobin expression and metphytoglobin reductase activity

We first studied the effect of 3 mm NO₃^– and 3 mm NH₄⁺ nutrition on NO, NR activity and phytoglobin1 (Pgb1) gene expression under normoxia and hypoxia (0.8 %). For this purpose, we grew Arabidopsis plants on vertical plates containing either 3 mm NO₃^– or NH₄⁺ in the medium. After 2 weeks, plants were treated with 21 % oxygen (normoxia) and 0.8 % oxygen (hypoxia) for a period of 24 h, and NO levels were monitored using the cell-permeable fluorophore DAF-FM DA. The main advantage of DAF-FM DA is that it is able to diffuse close to NO-producing sites. DAF fluorescence was considerably higher in NO₃^–-grown roots in comparison with NH₄⁺-grown roots (Fig. 1A, B). Fluorescence was increased massively following hypoxic stress treatments in NO₃^–-grown but only slightly increased in NH₄⁺-grown roots. In order to confirm that the observed fluorescence is related to NO, roots were incubated in a solution with an NO scavenger (cPTIO) plus DAF-FM DA. cPTIO binds to NO and forms NO₂ and cPTI (Goldstein et al., 2003). DAF fluorescence from NO₃^–-grown roots was effectively complemented by the scavenging effect of cPTIO, indicating that fluorescence observed from roots was reflecting the NO levels (Fig. 1A, B).

Fig. 1.

Levels of nitric oxide, nitrite, nitrate reductase activity, phytoglobin expression and metphytoglobin reductase activity in the roots of Arabidopsis plants grown on NO₃^– and NH₄⁺ after normoxic and hypoxic treatments. (A) Visualization of DAF fluorescence from NO₃^–- and NH₄⁺-grown roots treated with either normoxia or hypoxia. For cPTIO treatment, 200 μm cPTIO + DAF-FM DA was added to the roots. The image is a representative of three biological replicates. (B) Quantified DAF fluorescence from NO₃^–- and NH₄⁺-grown roots treated with either normoxia or hypoxia. Fluorescence from images was quantified using Image J and plotted as arbitrary units (AU). Values are presented as average means ± s.e. (n = 4). ** indicates P < 0.01, a significant difference between NO₃^–- and NH₄⁺-grown plants under normoxia and hypoxia. (C) Gas phase Griess reagent assay for NO measurement from NO₃^–- and NH₄⁺-grown seedlings under normoxia and hypoxia. The NO emitted over a period of 6 h by the normoxic or hypoxia-treated seedlings (0.5 g FW in HEPES buffer containing 0.5 mm nitrite, pH 7.2) of NO₃^–- and NH₄⁺-grown plants were kept in a stream of air in a solution containing 1 % (w/v) sulphanilamide and 0.02 % (w/v) NED in the air stream, and then converted to nitrite in the solution, where the nitrite formed is detected by its absorbance at 540 nm. The data are presented as the mean ± s.e. (n = 4). (D) Nitrite levels in NO₃^–- and NH₄⁺-grown plants under normoxia and hypoxia. Values are presented as the means ± s.e (n = 4). ** indicates P < 0.01, a significant difference between NO₃^–- and NH₄⁺-grown plants under normoxia and hypoxia. (E) Nitrate reductase activity in NO₃^–- and NH₄⁺-grown plants under normoxia and hypoxia. Values are presented as the means ± s.e (n = 4). ** indicates P < 0.01, a significant difference between NO₃^–- and NH₄⁺-grown plants under normoxia and hypoxia. (F) Relative expression levels of class 1 phytoglobin (PGB1) transcript levels ± s.e (n = 4). Error bars indicate the s.e Asterisks indicate significant differences: *P < 0.1 **P < 0.05; ***P < 0.005. (G) Metphytoglobin reductase activity in NO₃^–- and NH₄⁺-grown plants under normoxia and hypoxia. Values are presented as the means ± s.e (n = 4). ** indicates P < 0.01, a significant difference between NO₃^–- and NH₄⁺-grown plants under normoxia and hypoxia.

To complement the DAF fluorescence method with a second method to measure NO, we used the gas phase Griess reagent assay. For this purpose, roots were kept in a small head space cuvette and then treated with normoxia or hypoxia, with the derived NO gas being trapped in a mixture of sulphanilamide and NED under acidic (orthophosphoric acid) conditions. As shown in Fig. 1C, NO emission levels were higher in NO₃^–-grown plants as compared with NH₄⁺-grown plants, and responded to hypoxia only in the case of the NO₃^–-grown roots. These results were, therefore, in close accordance with those obtained from the DAF measurements.

NO₃^– is assimilated via NR and then nitrite reductase (NiR), and nitrite (NO₂^–) accumulation thus becomes a limiting factor for NR-dependent NO production (Planchet et al., 2005). We therefore next evaluated nitrite NO₂^– levels and NR activity which revealed that under normoxia both NO₃^–- and NH₄⁺-grown roots displayed only residual NO₂^– content and NR activity, but under hypoxia, NO₂^– levels and NR activity increased 10- and 6-fold, respectively, in NO₃^–-grown plants. However, NO₂^– accumulation was only 3-fold increased and NR activity remained unaltered in NH₄⁺-grown plants (Fig. 1D, E).

We next checked the expression of PGB1 which has previously been documented to play a role in NO scavenging and hypoxic survival. Under normoxia, NO₃^–-grown plants had significantly higher PGB1 transcript levels in comparison with NH₄⁺-grown plants. However, under hypoxia, whilst PGB1 expression levels increased in NO₃^–-grown plants, they remained unaltered in NH₄⁺-grown plants (Fig. 1F). NO is oxidized to nitrate by oxyphytoglobin [Pgb(Fe²⁺)O₂] which is itself reduced to form metPgb [Pgb(Fe³⁺)]. MetPgb is further reduced by metphytoglobin (MetPgb) reductase (metPgb-R) (Igamberdiev et al., 2005, 2006). In order to gain information concerning the operation of the Pgb–NO cycle, we next assayed the activity of NADH-dependent cytochrome c reductase activity which reflects the metPgb reductase activity (Igamberdiev et al., 2004, 2006). We observed that NO₃^–-grown plants exhibited a higher activity of this enzyme under normoxia than NH₄⁺-grown plants as well as a greater proportional increase in this activity following exposure to hypoxia (Fig. 1F).

Differential activation of AOX in response to NO₃^– and NH₄⁺ nutrition under normoxia and hypoxia

Given that NO has a great impact on respiratory metabolism, we checked total respiration and the operation of COX and AOX pathways in NO₃^–- and NH₄⁺-grown plants. Whilst the total respiratory rate of NO₃^–-grown plants was essentially the same as in NH₄⁺-grown plants (Fig. 2A), the respiration rate was significantly higher in NH₄⁺- than in NO₃^–-grown plants following hypoxia treatment, indicating that it was less suppressed (Fig. 2B). The contributions of the COX and AOX pathways were next investigated using specific inhibitors. Interestingly, under normoxia, NO₃^–-grown roots apparently exhibited significantly greater flux through the cytochrome than through the alternative pathway (Fig. 2C); however, the reverse was true following hypoxia (Fig. 2D). In contrast, NH₄⁺-grown roots apparently exhibited significantly greater flux through the alternative pathway than through the cytochrome pathway under normoxia (Fig. 2E), whilst the reverse was true following hypoxia (Fig. 2F).

Fig. 2.

Respiratory rates and capacities of the cytochrome (Cyt) and alternative oxidase (Alt) pathways in the roots of plants grown on NO₃^– and NH₄⁺ treated with normoxia and hypoxia. Total respiratory rates measured without inhibitors under normoxia (A) and under hypoxia (B). Respiratory rate in the presence of 2 mm salicylhydroxamic acid (SHAM), indicating the capacity of the cytochrome pathway and respiratory rate in the presence of 10 μm myxothiazol, indicating AOX capacity under normoxia in NO₃^–-grown plants (C), under hypoxia in NO₃^–-grown plants (D), under normoxia in NH₄⁺-grown plants (E) and under hypoxia in NH₄⁺-grown plants (F). Error bars indicate the s.e. Asterisks indicate significant differences between responses to oxygen conditions: *P < 0.1 **P < 0.05; ***P < 0.005.

NO₃^– nutrition influences ROS production under hypoxia

Results from the above experiments indicate that NH₄⁺ and NO₃^– nutrition have an influence on NO levels in roots. Increased NO levels could be responsible for the inhibition of respiration. The question therefore arose as to whether changes in NO levels also reflected changes in the levels of ROS. First, we measured superoxide levels by using NBT staining. Under normoxia, there was no significant change in superoxide levels but, under hypoxia, there was a significant increase in superoxide levels in both NH₄⁺- and NO₃^–-grown roots (Fig. 3A). In comparison with NH₄⁺ roots, the NO₃^– roots produced lower levels of superoxide. Measurement of ROS by DCF fluorescence revealed a similar trend, but a significant increase in ROS levels was observed in NH₄⁺-grown roots under hypoxia (Fig. 3B). Lipid peroxidation was increased under hypoxia, but no significant difference was observed between the growth treatments (Fig. 3C). AOX1A transcript levels remained largely the same; however, they increased under NO₃^– nutrition under hypoxia (Fig. 3D).

Fig. 3.

Effect of NO₃^– and NH₄⁺ nutrition on superoxide, total ROS, lipid peroxidation and AOX1A gene expression. (A) Superoxide levels in response to normoxia and hypoxia determined by NBT staining. Images are representatives of four independent replicates. (B) Total ROS indicated by DCF fluorescence in NO₃^– and NH₄⁺ nutrition under normoxia and hypoxia. Images are representatives of four independent replicates. (C) The extent of lipid peroxidation determined by MDA levels in NO₃^– and NH₄⁺ nutrition under normoxia and hypoxia. Values are means ± s.e. (n = 3). * Indicates P < 0.05, a significant difference between NO₃^–- and NH₄⁺-grown plants in normoxia and hypoxia. (D) AOX1A gene expression in NO₃^–- and NH₄⁺-grown plants under normoxia and hypoxia. Values are means ± s.e. (n = 3). A significant difference between all treatments was analysed by t-test at *P < 0.05, **P < 0.01, ***P < 0.001 with nitrate normoxia as control.

Fermentative metabolism is increased under hypoxia in NO₃^– in comparison with NH₄⁺

The Pgb–NO cycle consumes NAD(P)H (Igamberdiev and Hill, 2009) and, as such, its operation leads to regeneration of NAD(P)⁺ under hypoxia. Under NO₃^– nutrition, plants generated more NO and the expression of PGB1 was higher (Fig. 1A, F). We postulated that NAD(P)⁺ recycling will be likely to occur in this cycle under hypoxia which will in turn lead to an increased glycolytic activity. In order to test this hypothesis, we first measured the NAD⁺/NADH ratio of the total fraction of the cell. We found that the ratio was slightly higher under normoxia in NH₄⁺-grown roots in comparison with NO₃^–-grown roots. In contrast, whilst hypoxia resulted in a decrease of this ratio under both N nutrition regimes, the NO₃^–-grown roots had a higher NAD⁺/NADH ratio in comparison with NH₄⁺-grown roots (Fig. 4A).

Fig. 4.

Effect of normoxia and hypoxia on fermentative metabolism of plants grown on NO₃^– and NH₄⁺. (A) NAD⁺/NADH ratio. (B) Extractable alcohol dehydrogenase activity. (C) Extractable pyruvate decarboxylase activity. (D) Concentration of endogenous ethanol levels. (E) Concentration of endogenous lactate levels. (F) ATP/ADP ratios. Values are means ± s.e. (n = 3). A significant difference between all treatments was analysed by t-test at *P < 0.05, **P < 0.01, ***P < 0.001 with nitrate normoxia as control.

Fermentative metabolism was next evaluated by measuring the extractable enzyme activities of ADH and PDC. After 24 h of hypoxia, ADH and PDC activities increased irrespective of N nutrition; however, this increase was notably higher in NO₃^–-grown roots (Fig. 4B, C). Measuring ethanol and lactate levels revealed that these were low under normoxia, with lactate being slightly higher in NO₃^–- than in NH₄⁺-grown roots (Fig. 4D, E). Interestingly, although ethanol levels increased following hypoxia in both NO₃^–- and NH₄⁺-grown roots (albeit to a lesser extent in NH₄⁺-grown roots), the levels of lactate only increased in NO₃^–-grown roots (Fig. 4D, E).

Finally, we evaluated the ATP/ADP ratio of the total fraction. We found that under normoxia, NH₄⁺-grown plants display a slightly lower ATP/ADP ratio in comparison with NO₃^–-grown roots. However, under hypoxia, the ratio has decreased in both NH₄⁺- and NO₃^–-grown roots, but was much more significantly decreased in the case of NH₄⁺-grown roots (Fig. 4F).

Gene expression of glycolytic and TCA cycle enzymes in NO₃^–- and NH₄⁺-grown plants

Glycolysis and the TCA cycle play important roles in survival under hypoxia via various mechanisms including recycling of NAD⁺. Hence it is essential to know whether the source of N modulates these pathways via induction of genes involved in these pathways (Rocha et al., 2010). In order to evaluate this, we first checked expression of the pyruvate kinase gene (At2g36580) (Fig. 5A). This enzyme catalyses the terminal reaction of the glycolytic pathway by converting ADP and phosphoenolpyruvate (PEP) to ATP and pyruvate (Podesta and Plaxton, 1992). Our results show that the pyruvate kinase gene has significantly higher expression under normoxia in NO₃^–-grown plants in comparison with NH₄⁺-grown plants, whereas under hypoxia the expression of this gene was invariant between the NO₃^–- and NH₄⁺-grown plants.

Fig. 5.

Relative expression of genes involved in carbohydrate metabolism and fermentative pathways: (A) PK (pyruvate kinase; AT2G36580), (B) PDC1 (pyruvate decarboxylase-1; AT4G33070), (c) LDH (lactate dehydrogenase; AT4G17260) and (D) ADH1 (alcohol dehydrogenase-1; AT1G77120) in wild-type (WT) plants grown on nitrate (dark grey bars) and ammonium (light grey bars) nutrition under normoxia and hypoxia for 24 h (0.8 % O₂). Values are means ± s.e. (n = 3). A significant difference between all treatments was analysed by t-test at *P < 0.05, **P < 0.01, ***P < 0.001 with nitrate normoxia as a control.

Transcript levels of LDH (At4g17260) and ADH (At1g77120) were also considerably higher in NO₃^–-grown roots than in NH₄⁺-grown roots under both normoxia and hypoxia; however, the proportional increase on hypoxia was essentially the same regardless of the type of N nutrition (Fig. 5C, D).

We further checked expression of genes encoding enzymes of the TCA cycle (Fig. 6). The expression of the pyruvate dehydrogenase gene PDH-E1 (At1g01090), the product of which converts pyruvate to acetyl-CoA, displayed higher expression in NO₃^– under normoxia and was significantly reduced in NH₄⁺-grown plants, and absolutely no expression was found under hypoxia in both NO₃^– and NH₄⁺ nutrition (Fig. 6A). Expression of the citrate synthase gene CS4 (At2g44350) was 3-fold higher under normoxia in NH₄⁺- in comparison with NO₃^–-grown plants; however, under hypoxia, there was an increased expression of CS4 up to 6.5-fold under NO₃^– nutrition whereas no increase was observed under NH₄⁺ nutrition (Fig. 6B). Expression levels of the aconitase gene ACO2 (At4g26970) increased up to 2-fold in NO₃^–-grown plants as compared with NH₄⁺-grown plants, while no increase was observed under hypoxia (Fig. 6C). Transcript levels of the succinyl-CoA synthetase gene SCS (At2g20420) were significantly higher in NO₃^–- than in NH₄⁺-grown plants under normoxia but were unaltered following hypoxia treatment (Fig. 6D). The transcript levels of the fumarase gene FUM1 (At2g47510) were unchanged in normoxia under both N conditions, but significantly upregulated to 5.8-fold under nitrate nutrition following hypoxia in comparison with NH₄⁺-grown plants (Fig. 6E). A similar pattern was observed in expression levels of both the succinate dehydrogenase gene SDH2 (At2g18450) and the malate dehydrogenase gene MDH1 (At1g04410) under hypoxia (Fig. 6F, G).

Fig. 6.

Relative expression of genes involved in the TCA pathway: (A) PDH-E1 (pyruvate dehydrogenase-1; AT1G01090), (B) CS4 (citrate synthase-4; AT2G44350), (C) ACO2 (aconitase-2; AT4G26970), (D) SCS (succinyl-CoA synthetase, beta subunit; AT2G20420), (E) FUM1 (fumarase-1; AT2G47510), (F) SDH2 (succinate dehydrogenase1-2; AT2G18450), (G) MDH1 (malate dehydrogenase-1; AT1G04410) in wild-type (WT) plants grown on nitrate (dark grey bars) and ammonium (light grey bars) nutrition under normoxia and hypoxia for 24 h (0.8 % O₂). Values are means ± s.e. (n = 3). A significant difference between all treatments was analysed by t-test at *P < 0.05, **P < 0.01, ***P < 0.001 with nitrate normoxia as a control.

Regulation of genes involved in nitrogen metabolism

We next evaluated the expression of genes involved in N metabolism which we expected to be much more affected by NO₃^– than by NH₄⁺ nutrition due to increased NR activity and increased TCA cycle gene expression. Expression of NIA1 (At1g77760), which plays an important role in NO production, was indeed higher in NO₃^–-grown plants in comparison with NH₄⁺-grown plants under normoxia. It was further elevated following hypoxia in NO₃^– plants (5-fold) but no increase was observed in NH₄⁺-grown plants (Fig. 7A). Expression of the glutamine synthetase gene GLN1 (At1g48470) was 6.5-fold higher in NH₄⁺-grown plants under normoxia which was drastically reduced to 1.5-fold following hypoxia treatment. Under NO₃^– nutrition, GLN1 expression significantly increased by up to 4-fold under hypoxia treatment in comparison with the levels observed under normoxia (Fig. 7B). The expression of the glutamate synthase gene GLU2 (AT2G41220) was strongly suppressed in NH₄⁺-grown plants under normoxia, but was upregulated in both NO₃^–- and NH₄⁺-grown plants following hypoxia (Fig. 7C). Under normoxia, expression of the 2-oxoglutarate dehydrogenase gene OGDH-E1 (At3g55410) was similar in both NH₄⁺- and NO₃^–-grown plants, and it was strongly induced (5-fold) in NO₃^– plants but was downregulated in NH₄⁺-grown plants under hypoxia (Fig. 7D). In contrast, AlaAT1 (At1g17290) expression was slightly higher under normoxia in NO₃^–-grown plants than in NH₄⁺-grown plants, whereas under hypoxia its expression was upregulated in NO₃^–-grown plants but downregulated in NH₄⁺-grown plants (Fig. 7E).

Fig. 7.

Relative expression of genes involved in nitrogen metabolism: (A) NIA1 (nitrate reductase-1; AT1G77760), (B) GLN1 (glutamine synthetase-1; AT1G48470), (C) GLU2 (glutamate synthase-2; AT2G41220), (D) OGDH-E1 (2-oxoglutarate dehydrogenase, E1 component; AT3G55410), (E) ALAAT1 (alanine aminotransferase-1; AT1G17290) in wild-type (WT) plants grown on nitrate (dark grey bars) and ammonium (light grey bars) nutrition under normoxia and hypoxia for 24 h (0.8 % O₂). Values are means ± s.e. (n = 3). A significant difference between all treatments is analysed by t-test at *P < 0.05, **P < 0.01, ***P < 0.001 with nitrate normoxia as a control.

Expression of group VII transcription factors and genes involved in the N-end rule-mediated protein degradation under hypoxia

It has recently been shown that group VII ERF transcription factors play a role in hypoxia survival via their induction of the expression of fermentation-associated genes (Gibbs et al., 2011; Licausi et al., 2011). Given that we found the increased expression of various genes involved in fermentation and energy metabolism, and a corresponding increase of enzyme activities (Figs 5 and 6), we next checked whether the expression of these transcription factors correlated with that of the above-mentioned genes (Fig. 8). The expression of the Hypoxia Response ERF-1 transcription factor gene HRE1 (At1g72360.3) was highly similar in NO₃^–- and NH₄⁺-grown plants under normoxia, while under hypoxia this gene was highly upregulated in NO₃^– plants (8.5-fold) as compared with only 2-fold upregulated levels in NH₄⁺-grown plants (Fig. 8A). Similarly, HRE2 (At2g47520.1) expression was highly downregulated in NH₄⁺-grown plants as compared with NO₃^–-grown plants under normoxia, but it was significantly increased up to 5.5- and 3.5-fold in NO₃^–- and NH₄⁺-grown plants, respectively, under hypoxia (Fig. 8B). RAP2.2 (At3g14230.1) (Fig. 8C) and RAP2.12 (At1g53910.1) (Fig. 8D) expression also increased under hypoxia in both NO₃^– and NH₄⁺. However, NO₃^–-grown plants exhibited slightly higher expression than NH₄⁺-grown plants. Next, we checked expression of the arginine-tRNA protein transferase 1 (ATE1; At5g05700.1) and arginine-tRNA protein transferase 2 (ATE2; At3g11240.1) genes which previously have been demonstrated to be involved in oxygen sensing via in N-end rule pathway. Expression of ATE1 was very high in NO₃^–-grown plants under normoxia but extremely low under hypoxia in both NO₃^–- and NH₄⁺-grown plants (Fig. 8E). A similar trend was observed in ATE2 expression levels under hypoxia (Fig. 8F). We also evaluated the expression of the ethylene response transcription factor gene ETR1 (At1g66340.1), observing that there was no change in the expression of ETR1 under NH₄⁺ nutrition in either normoxia or hypoxia, but there was a 4.5-fold induction in NO₃^–-grown plants following hypoxia treatment (Fig. 8G).

Fig. 8.

Relative expression of genes involved in oxygen sensing: (A) HRE1 (hypoxia hesponse ERF-1; AT1G72360), (B) HRE2 (hypoxia response ERF-2; AT2G47520), (C) RAP2.2 (ERF/AP2 transcription factor family; AT3G14230), (D) RAP2.12 (ERF/AP2 transcription factor family; AT1G53910), (E) ATE1 (arginine-tRNA protein transferase 1; AT5G05700), (F) ATE2 (arginine-tRNA protein transferase 2; AT3G11240). (G) ETR1 (ethylene response-1; AT1G66340) in wild-type (WT) plants grown on nitrate (dark grey bars) and ammonium (light grey bars) nutrition under normoxia and hypoxia for 24 h (0.8% O₂). Values are means ± s.e. (n = 3). A significant difference between all treatments is analysed by t-test at *P < 0.05, **P < 0.01, ***P < 0.001 with nitrate normoxia as a control.

DISCUSSION

For efficient crop productivity, adequate nutrients are essential, with N being one of the most important essential macronutrients. In agricultural soils, N is supplied in the form of nitrate, ammonium or a combination of both (Gupta et al., 2013), although the activity of soil microbes also determines the form of available N (Schlesinger, 2009). It is very important to understand how bioenergetic pathways are modulated both under normal growth conditions and also under stress. Here, using Arabidopsis as a model system, we investigated the role of NO₃^– and NH₄⁺ nutrition on energy metabolic pathways under normoxia and hypoxia.

Arabidopsis roots display a high capacity to produce NO under 0.8 % oxygen. The rate of NO emission was considerably higher under NO₃^– than under NH₄⁺ nutrition as assessed by DAF fluorescence and the gas phase Griess reagent assay. Previously, it was found that NR can produce NO when it uses nitrite as an alternative substrate (Rockel et al., 2002). NR undergoes a regulatory switch during the transition from normoxia to hypoxia, whereby nitrite becomes a limiting factor for NO production (Planchet et al., 2005). For this reason, we checked nitrite levels and found a high increase in NO₂^– levels in NO₃^–-grown plants, which is probably indicative of the increase in NR activity observed under NO₃^– nutrition under hypoxia (Fig. 1E). NO₂^– produced from NO₃^– can act as a substrate for NO production in NR reaction and by the mitochondrial electron transport chain. In order to check the contribution of NR and mitochondrial electron transport chain for NO production, we added myxothiazol to the seedlings and found that 80 % of NO production was abolished and the remaining 20 % was inhibited by SHAM (unpubl. data). These results suggest that nitrite can act as a substrate for the electron transport chain, thereby leading to the majority of NO production by mitochondria.

Cytosolic Pgbs are capable of scavenging NO and oxidizing it to NO₃^– (Igamberdiev et al., 2005). Operation of this pathway may thus play a role in the exchange of NO between the mitochondria and cytosol. This Pgb–NO cycle contributes both to NADH and NADPH oxidation as well as ATP synthesis under hypoxia (Stoimenova et al., 2007). In our study, we found an increased expression of the PGB1 gene encoding the class 1 phytoglobin and elevated methaemoglobin reductase activity under NO₃^– nutrition under hypoxia (Fig. 1F, G). Given that Pgb and methaemoglobin reductase are involved in the detoxification of NO, this finding suggests that the Pgb–NO cycle is very active under hypoxia and NO₃^– nutrition. The increased NAD⁺/NADH ratio of the cell suggests that NAD⁺ recycling takes place during operation of this pathway (Fig. 4A), leading to an increased ATP/ADP ratio under hypoxia in NO₃^–-grown plants in comparison with NH₄⁺-grown plants (Fig. 4F). NR is a component of the Pgb–NO cycle and a potential source of NO from nitrite under hypoxia. The increased expression of the NR gene NIA1 in NO₃^–-grown plants (Fig. 7) suggests that its induction can contribute to NO synthesis and to increased turnover of the Pgb–NO cycle (Fig. 1). It was previously suggested that the class 1 Pgb plays a role in NO detoxification (Dordas et al., 2003). Here, we postulate that Pgb not only plays a role in detoxification but also aids in increasing energy efficiency under hypoxic stress, which corresponds to the data on the involvement of the Pgb–NO cycle in redox balance and ATP synthesis (Stoimenova et al., 2007). NO has been demonstrated to inhibit COX and thereby it is involved in the regulation of respiration (Millar and Day, 1996).

Under nitrate nutrition, plants exhibit slightly reduced respiratory rates. Increased levels of NO could be responsible for reduced respiratory rates under NO₃^– nutrition. Previously, it was shown that the capacity of AOX is higher under NH₄⁺ nutrition (Escobar et al., 2006); however, these experiments were conducted solely under normoxia. Here, we found a similar trend in NH₄⁺-grown plants under normoxia whereas the hypoxia-treated roots displayed an increased AOX capacity. Higher levels of NO could be responsible for the increased AOX capacity via two different mechanisms, namely reversible inhibition of COX by NO in competition with oxygen (Millar and Day, 1996) and inhibition of aconitase by NO leading to an increase in levels of citrate which subsequently induces AOX capacity (Gupta et al., 2012). Under hypoxia, NH₄⁺-grown plants displayed a reduced AOX capacity, suggesting that high levels of NO are required for AOX induction under hypoxia. Lower levels of AOX could be responsible for the increased levels of superoxide and total ROS under hypoxia observed in NH₄⁺-grown plants (Fig. 3A, B). However, we found no change in lipid peroxidation between NO₃^–- and NH₄⁺-grown plants (Fig. 3C) despite increased ROS in NO₃^– nutrition. Another important beneficial role of the regulation of respiration by NO is in increasing levels of internal oxygen. Previously, it was shown that NO inhibits respiration effectively, leading to an increase in internal cellular oxygen levels which in turn reduces internal ROS levels (Gupta et al., 2014).

One of the primary functions of the fermentative pathway is that the recycling of NAD⁺ allows continuous operation of glycolysis for energy production. Here, we found increased levels of NAD⁺/NADH under hypoxia in NO₃^– nutrition (Fig. 4A), supporting our hypothesis that the intensity of fermentation and NAD⁺ recycling is increased under these conditions. Catabolism of pyruvate occurs in order to increase the carbon flux from pyruvate into nitrite-supported respiration via PDC (Zabalza et al., 2009). PDC mediates the conversion of pyruvate to acetaldehyde, and expression of the gene was highly induced under NO₃^– nutrition (Fig. 5) but poorly induced under NH₄⁺ nutrition. Acetaldehyde is further converted to ethanol by ADH. Here, we found a high upregulation of ADH1 under hypoxia in NO₃^– nutrition, suggesting that it somehow aids in the recycling of NAD⁺ (Fig. 4A). Furthermore, analysis of ethanol levels (Fig. 4D) revealed that this correlated with the induction of ADH1 and increased activity of the enzyme (Fig. 4B). The gene encoding lactate dehydrogenase (LDH) was also slightly induced under hypoxia in NO₃^– nutrition. However, it has previously been demonstrated that LDH is less important for regeneration of NAD⁺, given its considerably slower turnover rate (Smith and ap Rees, 1979; Zabalza et al., 2005). Consistent with this finding, the increase in lactate levels in NO₃^– nutrition under hypoxia is considerably smaller than the changes in ethanol production observed in these conditions. More important is alanine formation from pyruvate which follows from the upregulation of AlaAT during NO₃^– nutrition. This indicates that alanine formation becomes an important pathway linking fermentation to amino acid metabolism, thus substantially contributing to bioenergetics under hypoxia (Rocha et al., 2010).

Although, the demonstration of a correlated increased expression of ADH1, LDH and PDC in NO₃^–-grown plants under hypoxia is important in its own right, the fact that the expression of these genes can be mediated by the induction of group VII ERF transcription factors (Hinz et al., 2010; Licausi et al., 2011) provides further mechanistic details concerning their physiological regulation. RAP2.12 is also known to bind to ADH1 (Papdi et al., 2008); therefore, it was interesting to know the impact of NO₃^– and NH₄⁺ on induction of these genes. In our study, we found that HRE1 was marginally and HRE2 highly induced under NO₃^– nutrition, suggesting that their induction could underlie the induction of the fermentative enzymes (Figs 4 and 5). RAP2.2 and RAP2.12 were also induced in both NO₃^–- and NH₄⁺-grown plants but their induction was higher under NO₃^– nutrition. Consistent with this observation are the previous data demonstrating that plants lacking these group VII transcription factors display reduced anaerobic responses (Licausi et al., 2010). Our data also revealed that the ATE2 gene was downregulated under hypoxia; however, further experiments will be needed in order to define the exact mechanisms by which these factors interact.

Observation of the changes in TCA cycle-associated gene expression revealed some further insights into the role of NO₃^– (Fig. 6). Expression of the citrate synthase gene (CS4) was elevated under hypoxia in NO₃^–-grown plants and this increase could be responsible, via a citrate-mediated effect (Gupta et al., 2012), for the observed increased AOX1A gene expression under hypoxia (Fig. 3D). In contrast, the expression of the aconitase gene (ACO2) was decreased under hypoxia in both NO₃^–- and NH₄⁺-grown plants, but a higher decrease was seen in NO₃^–-grown plants. It was previously shown that the increased NO and ROS can inhibit aconitase activity and its gene expression (Morgan et al., 2008; Gupta et al., 2012). It was hypothesized that this process had a protective effect against the additional oxidative stress by acting as a reversible ‘circuit breaker’ (Gardner et al., 1997). Our results indicate that NO and/or ROS can additionally suppress aconitase at the gene expression level. This provides further evidence that AOX is induced here due to inhibition of aconitase and also via inhibition of the COX pathway. Expression of SCS and FUM1 also increased under hypoxia in NO₃^–-grown plants, suggesting that NO₃^– or NO could contribute to the increased TCA cycle flux via their up-regulation. Similarly, MDH1, whose product plays a role in regeneration of NAD⁺ given its requirement for NADH during the conversion of malate to oxaloacetate, was upregulated under hypoxia in NO₃^–-grown plants, whereas it was only marginally increased in NH₄⁺ plants. This observation suggests that NO₃^–-grown plants have higher capacity for NAD⁺ recycling. Upregulation of MDH1 could alternatively/additionally play roles in regulation of pH during hypoxia (Roberts et al., 1992). Furthermore, the reduction of NO₃^– to NO₂^– contributes to reduction of the cytosolic acidosis under hypoxia (Libourel et al., 2006). The operation of the TCA cycle supplying NADH for nitrate and nitrite reduction in the Pgb–NO cycle can be considered as an important pH stat. PDC was also strongly upregulated in NO₃^–-grown plants under hypoxia. When taken together, these co-ordinated changes collectively allow us to propose a model of increased TCA cycle flux under hypoxia in NO₃^–-grown plants.

In summary, in the work presented here, we demonstrate that NO₃^– nutrition improves energy efficiency under hypoxia via the induction of genes involved in oxygen sensing combined with an increased efficiency of the Pgb–NO cycle and fermentative pathways (Fig. 9).

Fig. 9.

Nitrate nutrition induces multiple regulatory pathways under hypoxia to improve energy efficiency. Under hypoxia, nitrate (NO₃^–) nutrition induces nitrate reductase (NR) activity leading to the accumulation of nitrite (NO₂^–). NO₂^– can act as a substrate for the mitochondrial electron transport chain to increase nitric oxide (NO) production. The produced NO is oxidized by class 1 non-symbiotic haemoglobin (nHb1,) leading to recycling of NAD⁺, which is evidenced by increased phytoglobin expression and increased met-phytoglobin reductase activity and increased energy efficiency. NO₃^– nutrition modulates alternative respiratory pathways evidenced by accumulation of AOX1A transcripts. Nitrate nutrition can reduce reactive oxygen species (ROS). Nitrate nutrition also increase activities of ADH to increase energy efficiency.