Coordinated nitrogen and carbon remobilization for nitrate assimilation in leaf, sheath and root and associated cytokinin signals during early regrowth of Lolium perenne

Abstract

Background and Aims The efficiency of N assimilation in response to defoliation is a critical component of plant regrowth and forage production. The aim of this research was to test the effect of the internal C/N balance on ${\text{NO}}_3^{–}$ assimilation and to estimate the associated cytokinin signals following defoliation of perennial ryegrass (Lolium perenne L. ‘Grasslands Nui’) plants.

Methods Plants, manipulated to have contrasting internal N content and contrasting availability of water soluble carbohydrates (WSCs), were obtained by exposure to either continuous light or short days (8:16 h light–dark), and watered with modified N-free Hoagland medium containing either high (5 mm) or low (50 μm) ${\text{NO}}_3^{–}$ as sole N source. Half of the plants were defoliated and the root, sheath and leaf tissue were harvested at 8, 24 and 168 h after cutting. The spatiotemporal changes in WSCs, synthesis of amino acids and associated cytokinin content were recorded after cutting.

Key Results Leaf regrowth following defoliation involved changes in the low- and high-molecular weight WSCs. The extent of the changes and the partitioning of the WSC following defoliation were dependant on the initial WSC levels and the C and N availability. Cytokinin levels varied in the sheath and root as early as 8 h following defoliation and preceded an overall increase in amino acids at 24 h. Subsequently, negative feedback brought the amino acid response back towards pre-defoliation levels within 168 h after cutting, a response that was under control of the C/N ratio.

Conclusions WSC remobilization in the leaf is coordinated with N availability to the root, potentially via a systemic cytokinin signal, leading to efficient N assimilation in the leaf and the sheath tissues and to early leaf regrowth following defoliation.

INTRODUCTION

Soil nitrogen (N) availability is a major limiting factor in perennial ryegrass (Lolium perenne L.) pasture production (Dinnes et al., 2002). In modern crop management systems, N fertilizer supplied as urea, nitrate (${\text{NO}}_3^{–}$) or ammonium (${\text{NH}}_4^{+}$) are used to sustain and increase plant growth and forage production (Robertson and Vitousek, 2009). Excessive application of N fertilizer may result in short- and long-term effects such as leaching of the highly mobile ${\text{NO}}_3^{–}$ anion from agricultural land into waterways and subsequent decrease in aquatic system biodiversity (Camargo and Alonso, 2006) and decrease in drinking water quality (Carpenter et al., 1998). In addition, denitrification of ${\text{NO}}_3^{–}$ and nitrification of ${\text{NH}}_4^{+}$ in the soil releases nitrous oxide, a greenhouse gas, and other ozone-depleting gaseous emissions to the stratospheric ozone layer (Bremner and Blackmer, 1978; Mosier et al., 1991; Ravishankara et al., 2009). Alterations in the soil microbial community structure and soil acidification are also associated with N fertilizer pollution (Marschner et al., 2003; Guo et al., 2010).

Perennial ryegrass is a temperate forage grass used globally to provide feed of high digestibility for grazing stock (Wilkins and Humphreys, 2003). This pasture system usually receives repeated N fertilizer application after bovine grazing (Miller et al., 2001). The efficiency with which the plant takes up applied N, referred to as nitrogen use efficiency (NUE), can be defined as the plant biomass produced per unit of exogenous N input (Xu et al., 2012; Norton et al., 2015). NUE corresponds to the sum of the efficiency of N uptake by plant roots and of N utilization efficiency, which converts N taken up into above-ground biomass production (Good et al., 2004; Hirel et al., 2007; Masclaux-Daubresse et al., 2010). In perennial ryegrass plants, most ${\text{NO}}_3^{–}$ is transported out of the root for assimilation rather than directly assimilated in the root (Roche et al., 2016).

The efficiency of ${\text{NO}}_3^{–}$ assimilation depends on the regulation of the different enzymes involved in amino acid synthesis and their ultimate conversion into more complex N-containing molecules supporting growth. Amino acid biosynthesis occurs as a result of the assimilation of ${\text{NO}}_3^{–}$ with C-accepting molecules (Delhon et al., 1995). In Arabidopsis thaliana, ${\text{NO}}_3^{–}$ regulates the expression of numerous genes including those involved in N reduction (NR, NiR, genes coding for enzymes involved in the GS-GOGAT pathway), as well as those involved in glycolysis, sugar metabolism and photosynthesis (Scheible et al., 2004; Sakakibara, 2006). Relative to plants supplied with 1 mm of ${\text{NO}}_3^{–}$, roots of perennial ryegrass plants deprived of N for 10 d had a lower amino acid concentration, especially for the N-rich amino acids glutamine and asparagine, but showed an accumulation of carbohydrate storage forms (Louahlia et al., 2008).

Carbon-accepting molecules can be obtained directly from photosynthetic CO₂ assimilation and/or from the breakdown of carbohydrate reserves, which releases reducing equivalents and ATP for N assimilation (Krapp and Traong, 2006). Over the diurnal cycle, the photosynthetic period is characterized by a decrease in ${\text{NO}}_3^{–}$ content, gradual accumulation of simple carbohydrates and increase in N assimilation products, especially glutamine, glycine and serine, whereas long-term carbohydrate stores are remobilized during the night period (Stitt and Krapp, 1999; Cairns, 2003). During the first part of the light period, the rate of ${\text{NO}}_3^{–}$ assimilation has been found to initially exceed the rate of ${\text{NO}}_3^{–}$ uptake in A. thaliana plants due to competition between root N uptake, root to shoot N translocation and shoot assimilation of photorespiration products in the GS-GOGAT pathway (Matt et al., 2001). Therefore, the availability of the internal carbohydrate reserve and the dynamics of C remobilization are key components of the plant response to changes in supply and demand of C and N (Nunes-Nesi et al., 2010).

The non-structural carbohydrate reserves [sucrose and water-soluble carbohydrates (WSCs)] in perennial ryegrass plants are mainly localized in the vacuoles of cells in the tiller bases and roots (Pollock and Cairns, 1991; Vijn and Smeekens, 1999; Cairns and Gallagher, 2004). The WSCs are linear or branched polymers of fructan of variable degrees of polymerization (DP). Fructans consist of a precursor of sucrose or other hexoses attached to fructose units (Danckwerts and Gordon, 1987). The dynamics of WSC accumulation and remobilization are critical components for N assimilation and involve fructosyltransferase isoenzymes for fructan biosynthesis and fructan exohydrolase for C-reserve breakdown (Hisano et al., 2004; Chalmers et al., 2005; Lothier et al., 2007). Previously, we have found that treatment with 5 mm${\text{NO}}_3^{–}$ to N-deprived perennial ryegrass plants was associated with an initial remobilization of the low-molecular-weight (LMW) WSCs within the first 24 h of treatment, followed in time by remobilization within the shoot of the high-molecular-weight (HMW) WSCs to sustain N assimilation over a week-long experimental period (Roche et al., 2016).

Improvement in the on-farm NUE of perennial ryegrass requires in-depth understanding of the mechanisms integrating C and N metabolism following fertilizer application during the grazing cycle. Defoliation causes an immediate decrease in root growth, photosynthesis, respiration and nutrient uptake (Clement et al., 1978; Parsons et al., 1983; Jarvis and Macduff, 1989). Louahlia et al. (2008) observed a decrease in the net uptake rate of ${\text{NO}}_3^{–}$ within 2 d of defoliation (Louahlia et al., 2008). Low N plants supplied for 10 d with an N-free nutrient solution presented a less severe decline in ${\text{NO}}_3^{–}$ uptake when compared with plants supplied with high N levels (1 mm KNO₃) following a single defoliation (Louahlia et al., 1999). This suggests that the N reserve status of the plants affects the dynamics of N uptake during recovery from defoliation (Louahlia et al., 1999). However, an inverse relationship between N uptake and shoot N assimilation has been noted during the first few days following defoliation (Louahlia et al., 2008). Louahlia et al. (2008) showed that the defoliation-induced reduction in N uptake could be transiently rescued over the initial 24 h of regrowth by supplying defoliated perennial ryegrass plants with 1 mm of glucose, fructose or sucrose, indicating a link between carbohydrate status and regulation of N uptake. In contrast, external application of amino acids may decrease ${\text{NO}}_3^{–}$ uptake (Louahlia et al., 2008). However, the mechanism by which the internal C/N balance regulates N assimilation is yet to be established.

The cytokinins are considered to act as local and long-distance signals able to control many facets of plant development (Takei et al., 2001; Forde, 2002; Sakakibara, 2006). They have been implicated in the regulation of the ${\text{NO}}_3^{–}$ transporter NRT gene and N assimilation genes (NR, AS and glutamine dehydrogenase), and C metabolism (Sakakibara et al., 2006). In addition, both ${\text{NO}}_3^{–}$ and cytokinin signals can influence photosynthetic capacity via regulation of genes coding for enzymes involved in trehalose metabolism (Brenner et al., 2005). In A. thaliana plants, ${\text{NO}}_3^{–}$ treatment up-regulated the isopentenyl transferase AtIPT3 gene coding for the rate-limiting step in cytokinin biosynthesis and resulted in the biosynthesis of the isopentenyladenine (iP)-type cytokinins in the root (Miyawaki et al., 2004; Takei et al., 2004). The iP-type cytokinins were subsequently converted to the trans-zeatin (tZ)-type cytokinins (Sakakibara, 2006). The tZ-type cytokinins are generally found in the xylem and could be a root to shoot signal of ${\text{NO}}_3^{–}$ availability, whereas the iP-type cytokinins are mainly detected in the phloem and might play a role as a shoot to root signal of whole plant N status (Hirose et al., 2008; Ruffel et al., 2011; Kiba et al., 2013). Consistently, the amount of cytokinin moving from root to shoot appears to be dependent upon the N status of the plant and this mechanism is conserved among higher plants, such as in Zea mays (Takei et al., 2001), the perennial herb Urtica dioica (Wagner and Beck, 1993) and Pisum sativum (Beveridge et al., 1997). Consequently, the cytokinins are considered to be strong signals involved in co-ordinating the C/N status of the plant with shoot and root growth (Roitsch and Ehneß, 2000; Wang and Ruan, 2016).

In this study, we investigated the effect of contrasting internal C/N balance on the efficiency of ${\text{NO}}_3^{–}$ assimilation in response to an increase in C demand, experimentally induced by a defoliation treatment of the perennial ryegrass plants. The interaction between C remobilization and ${\text{NO}}_3^{–}$ assimilation on biochemical and physiological traits was investigated by measurement of spatiotemporal changes in WSCs, amino acid profiling and growth responses over a period of 7 d following defoliation. To obtain plants with contrasting internal N content and WSC availability, plants were exposed to either continuous light or short days (8 : 16 h light/dark), and to either high (5 mm) or low (50 µm) ${\text{NO}}_3^{–}$ as a sole N source. Half of the plants were defoliated and the leaf tissue (leaf material above the sheath), sheath (pseudo-stem) and roots were harvested at 8, 24 and 168 h after cutting. Changes in cytokinin content were measured at 8 h following cutting to determine whether cytokinin can act as an early signal integrating C and N supply and demand, and to determine its involvement in the regulation of ${\text{NO}}_3^{–}$ assimilation after defoliation.

MATERIALS AND METHODS

Plant material

Perennial ryegrass (L. perenne L. ‘Grasslands Nui’) seeds were grown in individual pots at the University of Canterbury glasshouses (43°31′48″S, 172°37′13″E) for 12 weeks using unfertilized soil treated once with commercial NPK fertilizer to maintain plant growth until plant establishment. Seedlings were subsequently watered for 2 weeks with N-free Hoagland liquid medium (N-free Hoagland medium, BioWorld, Irving, TX, USA) to supply all nutrients in sufficient amounts with the exception of N. To obtain plants with different carbohydrate status and N content, perennial ryegrass plants were then transferred to controlled-environment rooms (University of Canterbury, New Zealand) under either continuous light (irradiance of 57–67 quanta µmol m^–2 s^–1), subsequently referred to as LD, or short days (8 : 16 h light/dark, irradiance of 70–98 quanta µmol m^–2 s^–1), subsequently referred to as SD, and watered with N-free Hoagland medium modified with either high (5 mm, HN) or low (50 µm, LN) ${\text{NO}}_3^{–}$ as sole N source for 2 weeks until the time of defoliation. Half of the plants were defoliated at 4 cm above the crown tissue at time 0 of the experiment and subsequently leaf, sheath and root tissues were harvested at 8, 24 and 168 h after cutting, flash frozen in liquid N and stored at –80 °C. The other half of the plants were not defoliated and were treated the same as the defoliated plants. Leaf tissue corresponded to the foliar/lamina material located above the sheath. Sheath tissues are here defined as the pseudo-stem tissue located between crown and ligule, including the leaf bases and mature leaf sheaths. Three biological samples were used for subsequent analysis. Each biological replicate comprised pooled tissue samples harvested from seven plants. Samples were ground into fine powder under liquid N for further analysis. Measurements of total N content and total C content were obtained by isotope ratio mass spectrometry at the Department of Soil and Physical Sciences, Lincoln University, Christchurch, New Zealand.

Fructan analysis

WSCs of low (LMW) and high molecular weight (HMW) were extracted and analysed using an Agilent 1290 Infinity LC System (Agilent Technologies, Waldbronn, Germany) coupled to an Agilent 6550 Accurate-Mass QTOF LC-MS system with a dual Agilent Jet Stream source operating in negative mode and a QTOF mass range selected to 70–1700 mass-to-charge ratio. Oligosaccharides of DP 2–20 were extracted from 25 or 10 mg of freeze-dried ground plant material mixed with 750 or 300 µL of boiling milliQ water, respectively. Samples were vortexed and placed on a heating block for 15 min at 90 °C. Samples were cooled to room temperature and the tubes centrifuged for 10 min at 20 000 g. The supernatant was transferred to a filter tube and centrifuged for 2 min at 20 000 g. Extracts were transferred to LC vials and aliquots of 2 µL were injected onto an Acquity UPLC HSS T3 C₁₈ column (2·1 × 50 mm, 1·8 µm) combined with a 2·1 × 5 mm, 1·8 µm VanGuard pre-column (Waters Corporation, Milford, MA, USA) held at 40 °C. The linear gradient elution consisted of 0·1–10 % of solvent B (0·1% formic acid, 75 : 25 acetonitrile/isopropanol) over 2 min, followed by an increase to 99 % solvent B for 5 min and then stabilized for a further 2 min. Subsequently, solvent B was decreased to 0·1% for 0·3 min and then the flow rate was increased to 0·8 mL min^–1 for 0·5 min and held for 0·5 min. Finally the flow rate was reduced to 0·5 mL min^–1 and held for 2 min before the next injection. Data were collected in centroid mode with an acquisition rate of 4 scans s^–1 and 1975 transients per spectrum. Data processing was performed on NetCDF files processed and analysed using in-house scripts in MATLAB 7.14.739 (R2012a) (Mathworks, Natick, MA, USA). WSCs of LMW were defined as oligosaccharides of DP from 3 to 9 and HMW WSCs corresponded to DP10 to DP20.

Amino acid determination

Free amino acids were derivatized with Waters AccQ-Tag, separated by HPLC, and quantified by fluorescence detection (excitation, 250 nm; emission, 395 nm) as previously described (Reverter et al., 1997). Samples were processed as follows: 300 µL of milliQ Ultrapure water was added to approximatively 25 mg d. wt of freeze-dried ground plant material. Samples were vortexed for 1 min and centrifuged for 10 min at 15000 g. The extraction was performed in a new tube by adding 250 μL of chloroform and 900 μL of methanol to 50 µL of supernatant. Samples were vortexed again for 30 s and centrifuged for 5 min at 13^000 g. The supernatants (50 µL) were transferred to a new tube, freeze-dried for 30 min, reconstituted in 20 µL of 20 mm HCl and derivatized with 20 µL of the δ-aminoquinolyl-N-hydroxysuccinimidyl carbamate (AQC) following the Waters AccQTaq Chemistry Package Instruction Manual (WAT05287 Rev. 1; Waters). Derivatized amino acids were separated by reversed-phase HPLC on an AccQ-Tag 60A column, 4 µm (Waters; 150 × 3·9 mm) and detected with a fluorescence detector (RF-10Ax, Shimadzu (Sydney, Australia); excitation: 250 nm, emission: 395 nm). Amino acids detected and quantified were: Ala, Asn, Asp, Cys, Gln, Glu, Gly, His, Ile, Leu, Lys, Met, Phe, Pro, Ser,Thr, Tyr and Val. Co-eluted peaks for His and Gln, and for Ser and Asn were subsequently referred to as His + Gln and Ser + Asn, respectively. Amino acids were identified and quantified by comparison with Waters amino acid hydrolysate standard. Glutamine and Asn standards were obtained separately from Sigma (St Louis, MO, USA).

Endogenous cytokinin quantification

Extraction was performed on 3–5·5 mg d. wt of prepared biological triplicates in 1 mL of modified Bieleski solution (60 % MeOH, 10 % HCOOH and 30 % H₂O). Purification was performed as described by Dobrev and Kamı´nek (2002) with minor modifications (Antoniadi et al., 2015). The samples were purified using a combination of C₁₈ (100 mg mL^–1) and MCX cartridges (30 mg mL^–1) and purification recovery was validated by using 18 stable isotope-labelled cytokinin internal standards (0·2 pmol of cytokinin bases, ribosides, N-glucosides, 0·5 pmol of O-glucosides and nucleotides). Sample analysis was performed using the LC-MS/MS system consisting of an ACQUITY UPLC System (Waters) and Xevo TQ-S (Waters) triple quadrupole mass spectrometer. A multiple reaction-monitoring (MRM) mode of selected precursor ions and the appropriate production were used for quantification (Svačinová et al., 2012). Full cytokinin profiling can be found in the Supplementary Data (Table S1).

Statistical analysis

One-way ANOVA with Tukey correction was used to test the significance of differences in dry weight of regrown leaves at 168 h after cutting between growth conditions. The following variables were subjected to a two-way ANOVA with Tukey correction to test the effect of the growth condition at time 0 including N concentration, C concentration, C : N ratio and total WSCs. Sidak correction was used for the individual amino acids to test relative differences in concentration between time 0 and 8 h after defoliation, and between uncut and cut plants at 24 h and at 168 h after defoliation. Cytokinin content was analysed by a T-test with two-tailed distribution and homoscedasticity to compare cut plants at 8 h after defoliation and control uncut plants.

RESULTS

Physiological responses to contrasting N supply and carbohydrate loads

To evaluate the effect of contrasting internal carbohydrate status and contrasting ${\text{NO}}_3^{–}$ supply on perennial ryegrass plants, physiological measurements, total internal C and N content, total WSC abundance and total amino acid concentration were determined at the start of the experiment (t = 0 h). Perennial ryegrass plants were initially grown under N-sufficient conditions, subsequently exposed to contrasting day length (continuous light, LD or 8 h light/16 h dark, SD) and contrasting ${\text{NO}}_3^{–}$ supply (5 mm, HN or 50 µm, LN) for 2 weeks, and then harvested over time following defoliation. Production of foliage, tillers and regrowth leaf material at 168 h after defoliation was more strongly related to C rather than to N supply (Fig. 1). Indeed, fewer green leaves (Fig. 1A), fewer tillers (Fig. 1B) and a lower dry weight of regrown leaf (Fig. 1C) were found under SD relative to LD, regardless of the N supply (i.e. SDHN or SDLN).

Fig. 1.

Variation over time in green leaf number (A) and tiller number (B) of the uncut perennial ryegrass plants grown under either continuous light (referred to as LD) or short days (8 : 16 h light/dark, SD), and watered with N-free Hoagland medium modified with either high (5 mm, HN) or low (50 µm, LN) ${\text{NO}}_3^{–}$ as sole N source. Dry weight (C) was measured 168 h after defoliation on regrown leaf material (foliar/lamina material located above the sheath). Data are means ± SE, n = 7–10. Different lower-case letters correspond to significant differences between growth conditions calculated by one-way ANOVA with Tukey correction (P ≤ 0·05).

The response of total N concentration across the four growth conditions correlated with the response of total amino acid concentration in the leaf and sheath tissues before defoliation (Fig. 2A, D). By contrast, total amino acid concentration in the root (Fig. 2D) did not vary across growth conditions. Total C concentration was not different across growth conditions in the leaf and in the sheath, whereas significant changes were observed in the root between LDHN and LDLN, and between LDLN and SDHN (Supplementary Data S1). Exposure to LD was associated with greater total WSC abundance (in C equivalents) than under SD (Fig. 2B). In addition, LDLN resulted in significantly greater WSC abundance in the leaves relative to LDHN (Fig. 2B). The C : N ratio was greatest under LDLN than it was under the three other growth conditions (Fig. 2C). Internal N concentration was relatively greater in the leaf, intermediate in the sheath and lowest in the root (Fig. 2A), whereas total WSC abundance and total amino acid concentration were greater in the sheath relative to the other tissues (Fig. 2B, D).

Fig. 2.

Total N concentration (A), relative abundance in C units of water soluble carbohydrate (WSC) content (B), C/N ratio (C) and total concentration in amino acids (D) in perennial ryegrass leaf, sheath and root at time 0 (before defoliation). Plants were exposed to continuous light (referred to as LD) or short days (8 : 16 h light/dark, SD), and watered with modified N-free Hoagland medium containing either high (5 mm, HN) or low (50 µm, LN) ${\text{NO}}_3^{–}$ as sole N source. Data are means ± SE, n = 3 pools of seven plants. Different lower-case letters correspond to significant differences between growth conditions for each tissue types and were calculated by two-way ANOVA with Tukey correction (P ≤ 0·05). NS, not statistically different.

Defoliation–induced remobilization of WSCs

To investigate the dynamics of remobilization of the internal carbohydrate availability in response to defoliation, WSCs were recorded over time at 8, 24 and 168 h after cutting under growth conditions of contrasting carbohydrate status and ${\text{NO}}_3^{–}$ supply. The disaccharides and the LMW WSCs forms of tri-, tetra- and penta-saccharides were the most abundant WSCs in terms of C unit equivalents (Fig. 3). Defoliation induced a general decrease in the concentration of WSCs in terms of C units, which was initiated in the sheath as early as 8 h after cutting (Fig. 3B). The remobilization of root WSCs was preceded by a slight increase in WSCs at 8 h (Fig. 3C). By 168 h after defoliation, the WSC profile differed across experimental conditions and across tissue types. In the leaf, most of the LMW WSCs, here defined as oligosaccharides of DP3 to DP9, had recovered to their pre-cut values 168 h after defoliation, regardless of growth conditions (Fig. 3A). At 168 h after cutting, the sheath LMW WSC levels had also recovered to pre-cut levels under LD. However, under the LDLN treatment, the sheath LMW WSCs had accumulated to greater levels than those recorded at 0 h (Fig. 3B). By contrast to LD, exposure to SD was associated with a progressive remobilization over time of the sheath WSCs, which resulted in the total depletion of most of the sheath HMW WSCs (DP10–20) under SD by 168 h (Fig. 3B). Similarly, most of the root WSC reserves were exhausted by 168 h after defoliation (Fig. 3C). The remobilization of the WSCs over time was relatively faster under conditions of imbalanced C/N ratio, such as LDLN and SDHN conditions: for example, for the LMW WSCs under LDLN and SDHN conditions, and for the HMW WSCs under SDHN (Fig. 3B).

Fig. 3.

Abundance in C units of WSCs measured over time after defoliation in the leaf (A), sheath (B) and root (C) of perennial ryegrass plants exposed to continuous light (LD) or short days (8 : 16 h light/dark, SD), and watered with modified N-free Hoagland medium containing either high (5 mm, HN) or low (50 µm, LN) ${\text{NO}}_3^{–}$ as sole N source. Oligosaccharides of degree of polymerization 2–20 were determined at 0 h (time of defoliation), and 8, 24 and 168 h after defoliation. Leaf material at 8 h after defoliation had not regrown yet and could not be harvested. Data are means ± SE, n = 3 pools of seven plants.

Amino acid profiles

Nitrate assimilation into organic compounds was estimated by measurement of the abundance of individual free amino acids (Fig. 4). Some of the amino acids varied significantly in response to defoliation and some also changed over time in the control uncut plants, although generally to a lesser extent than in response to defoliation (Supplementary Data S2). Overall, a significant increase was observed 24 h after defoliation and a subsequent decrease was noted at 168 h (relative to the same time point in control uncut plants). At 8 h after defoliation, under SDLN conditions the concentrations of Glu, Ile, Leu, Phe, Tyr and Val were significantly greater in the roots of cut plants than in uncut plants at time 0. At 24 and 168 h after defoliation, the defoliation-induced changes in amino acid concentrations became significant, regardless of the tissue type and growth conditions. In particular, at 24 h after cutting, Asp, Gly, His + Gln, Ile, Leu, Lys, Phe, Ser + Asn, Thr, Tyr and Val increased significantly, whereas Glu and Pro decreased compared to the uncut plants. The extent of the response to defoliation differed between growth conditions: it was greater under SDHN and SDLN conditions, and notably reduced under LDLN. By 168 h after cutting, most of the changes observed under LDHN conditions took place in the root, whereas a significant decrease in amino acids was recorded mainly in the sheath and leaf under SDHN and SDLN.

Fig. 4.

Heat map showing positive (red) or negative (blue) changes in concentration of amino acids relative to uncut plants in response to defoliation in the leaf, sheath and root of perennial ryegrass plants at 8, 24 and 168 h following defoliation. Plants were exposed to continuous light (LD) or short days (8 : 16 h light/dark, SD), and watered with modified N-free Hoagland medium containing either high (5 mm, HN) or low (50 µm, LN) ${\text{NO}}_3^{–}$ as sole N source. Data are means ± SE, n = 3 pools of seven plants. Asterisks at 8 h indicate a significant difference between plant materials harvested 8 h after cutting and amino acid concentrations measured at 0 h (pre-defoliation). Significance in individual amino acid concentration at 24 and 168 h in response to defoliation was relative to measurements at 24 and 168 h in control (uncut) plants. The colour gradient corresponds to the level of statistical significance calculated by two-way ANOVA with Sidak correction for the individual amino acids (*P ≤ 0·05, **P ≤ 0·01, ***P ≤ 0·001 and ****P ≤ 0·0001). NS, not statistically different.

Cytokinin content 8 h after cutting

Most of the significant changes in cytokinin concentration observed at 8 h after cutting took place in the sheath rather than in the root (Fig. 5). Leaf material at 8 h after defoliation had not regrown yet and could not be harvested. Defoliation was associated with an increase in the concentration of iP and iP nucleotide (iPRMP) in the sheath, by more than 60 and 40 %, respectively, regardless of growth conditions (Fig. 5A). In contrast, a significant decrease in tZ nucleotide (tZRMP) levels was recorded in sheath and root tissues under SDLN, and in the sheath of plants exposed to LDHN. More complex trends in the tZ riboside (tZR) profiling were recorded: a decrease in the sheath under SDHN, and an increase in root tZR under LDLN, LDHN and SDHN following defoliation. Most of the cis-cytokinin responses to defoliation were detected under SDHN. In particular, under SDHN, cZRMP levels accumulated in the sheath and the root, cZ levels decreased in the root and cZR O-glucoside (cZROG) accumulated in the root. Under SDLN a significant increase in cZR was observed in the sheath following defoliation.

Fig. 5.

Cytokinin concentration of perennial ryegrass sheath (A) and root (B) in control uncut plants (white bars) or in cut plants at 8 h after defoliation (grey bars). Plants were exposed to continuous light (LD) or short days (8 : 16 h light/dark, SD), and watered with modified N-free Hoagland medium containing either high (5 mm, HN) or low (50 µm, LN) ${\text{NO}}_3^{–}$ as sole N source. Data are means ± SE, n = 3 pools of seven plants. Lower-case letters indicate significant differences between cytokinin concentrations at 8 h after cutting and at 0 h (pre-defoliation) for each growth condition. Statistical significance was calculated using a T-test at P ≤ 0·05. Cytokinin concentrations below detection limit are referred to as <LOD. Yellow background highlights the active free-base cytokinins. The complete dataset can be found in Table S1.

DISCUSSION

Balance between carbon and nitrogen regulates nitrate assimilation in the leaf, sheath and root of uncut plants

Identifying the partitioning of C and N reserves in perennial ryegrass is a key factor towards understanding the regulation of N assimilation. Distinct physiological characteristics reflect functional differences between tissue types. Namely, the sheath is the main site of WSC accumulation (Prud’homme et al., 1992) and amino acid synthesis and/or accumulation (Bigot et al., 1991), whereas N accumulates to a greater overall level in the leaf relative to the other tissues (Moser et al., 1982). Leaf N accumulates from either: (1) N forms directly taken up by the root, transported via xylem and stored in the leaf vacuoles (Granstedt and Huffaker, 1982); or (2) complex N-containing molecules derived from the amino acids locally produced in the leaf or imported from the large pool of amino acids present in the sheath (Fig. 2A, D). The rate of ${\text{NO}}_3^{–}$ uptake by root cells is controlled by demand-driven regulatory mechanisms limiting discrepancies between N supply and growth rate (Imsande and Touraine, 1994). Growth under our experimental conditions resulted in contrasting internal N and C accumulation at time 0 (Figs 2 and 3). Previously, Morvan-Bertrand et al. (1999) showed that 4 d of exposure to contrasting photoperiodic conditions (continuous light or darkness, 16 : 8 h day/night cycle) as an experimental treatment resulted in differential accumulation of carbohydrate stores and rapidly affected NUE of perennial ryegrass plants. Our data indicate that WSC accumulation, and in particular DP3–6, was dependant on the extent of the photosynthesis but was not significantly affected by N supply in the sheath or root (Figs 2B and 3). By contrast, endogenous N concentration and amino acid concentration varied with the interaction between N supply and C demand within each tissue type, although the amino acid concentration in the roots was not responsive to growth conditions (Fig. 2A, D). This indicates a constitutive accumulation and/or assimilation of amino acids in the roots which occurred irrespective of the availability in N and C.

By comparing LDHN plants with LDLN plants, our data indicate that a greater N availability to the root correlated with a rapid consumption of photosynthetic C and with a lower WSC accumulation in the leaf, which resulted in a relatively greater amino acid synthesis in the leaf and in the sheath tissues (Figs 1A and 2B, D). This suggests a systemic integration of the C and N signals of supply and demand to support efficient amino acid synthesis. Consistently, plants grown under LDLN conditions were exposed to a greater imbalance between C and N availability relative to the other treatments, which resulted in greater WSC accumulation and reduced accumulation of amino acids relative to LDHN, SDHN and SDLN (Fig. 2B, C). Lattanzi et al. (2005) found that long-term WSC stores had little impact on leaf growth, which suggests that the pool size of the WSC recorded under our experimental conditions might result from N regulation of WSC accumulation. We agree that leaf growth relies mainly on the short-term supply of LMW WSCs and recent photosynthetic C in these uncut plants (Lattanzi et al., 2005). Therefore, the dynamics of WSC remobilization are coordinated with N partitioning to support N assimilation and subsequent growth response in perennial ryegrass.

Remobilization of LMW and HMW WSCs in the leaf and sheath support amino acid synthesis for rapid regrowth following defoliation

Under our experimental conditions, defoliation was associated with an overall decrease in WSC abundance in terms of C units (Fig. 3), suggesting that the WSC reserves were being remobilized. In addition, a decreased WSC abundance may be associated with a decrease in the synthesis of photosynthetically derived WSCs and with increased plant biomass during regrowth following defoliation. Our spatiotemporal profiling of WSCs and amino acids in defoliated plants indicates three distinct patterns at 8, 24 and 168 h following an increase in defoliation-induced C demand from stored pools. At 8 h after defoliation, roots in plants exposed to SDLN showed an accumulation of WSCs across all DPs relative to time 0 (Fig. 3C), whereas the LMW WSCs were consumed in the sheath (Fig. 3B). We suggest that, under low C and low N conditions, defoliation induced both a remobilization of the WSC reserves in the above-ground part of the plant and an export to the roots, where the C was rapidly assimilated into amino acids at the site of N uptake (Figs 3B, C and 4). It is possible that the significant response observed 8 h after defoliation under conditions of limited external N availability and high C demand corresponded to a critical survival response to herbivory/defoliation characterized by a greater rate of N assimilation and/or remobilization of initial N reserves relative to the other growth conditions (Figs 2A and 4).

Ourry et al. (1990) showed that N taken up after defoliation was not directly assimilated but stored mainly in the vacuole of the leaf, sheath and root tissues for up to 5 d after defoliation. This was followed in time by a period of N reduction throughout the plant between 5 and 12 d after defoliation (Ourry et al., 1990). This contrasts with our findings which indicate that by 24 h following defoliation a marked accumulation and storage of amino acids had occurred (Fig. 4). The discrepancy in the timing of the response to defoliation could be explained by greater initial C reserves under our experimental conditions stimulating an earlier assimilation into amino acids relative to the plants in the Ourry et al. (1990) study. In addition, our data indicate that the regrown leaf tissue and the sheath became the main sites for significant accumulation of amino acids in response to defoliation relative to uncut plants (Fig. 4). Notably, by 24 h the LMW WSCs and most of the HMW WSCs were reduced in the leaf and sheath to their lowest levels relative to the other time points (Fig. 3A, B). Together with the remobilized WSCs, C can be used directly from photosynthesis by newly regrown tissues, with a proportional increase in photosynthate use compared with C reserves over time following defoliation as more photosynthetic tissues develop (Morvan-Bertrand et al., 1999).

The amino acids Glu, Gln, Asp, Asn are direct products from the GS-GOGAT pathway and are the precursors of the other amino acids. The changes in Gln, Asp and Asn in response to defoliation were more significant in the leaf tissue than in the sheath and the root 24 h after defoliation (Fig. 4). This suggests that removal of leaf material might have shifted the main site of amino acid assimilation and/or accumulation from the sheath at time 0 to the growing leaf 24 h after defoliation (Figs 2D and 4). In addition, the significant response to defoliation in the leaf was extended to most of the individual forms of amino acids, whereas most of the changes in the sheath were observed for Ile, Leu, Lys, Phe and Tyr relative to uncut plants at 24 h (Fig. 4). This suggests that defoliation not only stimulated N assimilation in the leaf but also the subsequent conversion into complex N-containing molecules to support regrowth of photosynthetic leaf material in the 24-h period after defoliation.

Defoliation can result in the stimulation of transpiration rates, decrease in root hydraulic conductivity and increase in hydraulic conductance of the remaining green tissues by changes in aquaporin expression (Liu et al., 2014). Nitrate uptake and transport from root to leaves follow the transpiration stream (Tegeder and Rentsch, 2010). Therefore, changes in water transport induced by defoliation may have influenced the localization and rate of N assimilation recorded under our experimental conditions.

Rapid regrowth after defoliation relies on complementary fluxes of newly acquired and reserve-derived C and N and is associated with re-establishment of a balanced C/N ratio from complementary fluxes of reserve-derived and currently assimilated C and N (Richards and Caldwell, 1985; Ourry et al., 1989; Schnyder, 1999). Greater initial WSC reserves recorded under LDLN at time 0 (Fig. 2) were associated with relatively faster remobilization and recovery of LMW WSCs when compared with the other growth conditions in response to defoliation (Fig. 3) due to a limitation in N availability for C assimilation (Fig. 4).

Louahlia et al. (2008) suggested that the concentration of amino acids in defoliated perennial ryegrass increased for up to 24 h and then decreased. Indeed, at 168 h after defoliation, our data suggest that a negative feedback reduced the defoliation-induced accumulation of free amino acids observed 24 h after cutting (Fig. 3A). Nitrogen reduction during regrowth may originate from either root N uptake or from remobilization through proteolysis and/or amino acid hydrolysis in a species-dependent manner (Thornton et al., 1993). In addition, Prud’homme et al. (1992) identified two phases in WSC dynamics following defoliation: the WSCs were remobilized in all tissues for up to 6 d following defoliation to sustain foliage development, after which a second period was characterized by a recovery of the WSCs store levels for up to 29 d after defoliation.

In this study, two highly significant decreases in amino acids were recorded at 168 h after defoliation relative to uncut plants: under SDLN in the sheath and under LDHN in the root (Fig. 4). By 168 h after cutting, the root WSCs from DP2–3 had partially recovered and DP6–20 WSCs had reached their lowest levels relative to the other time points under LDHN (Fig. 3C), whereas the sheath WSCs from DP7–20 were depleted under SDLN (Fig. 3B). The response to defoliation in LDHN roots corresponds to an active remobilization of resources which could be associated with root growth. A limited availability of N and C under SDLN conditions resulted in a relative longer remobilization of the amino acids from the large N pool stored in the sheath at time 0 (Fig. 2D) when compared with LDHN, LDLN and SDHN at 168 h after defoliation. By contrast, the leaf and sheath of plants grown under LDHN presented stabilized levels of amino acids (Fig. 4) and partially recovered WSC levels (Fig. 3A, B) as well as a greater growth response after defoliation relative to the other experimental conditions (Fig. 1C). This indicates that by 168 h after cutting the leaf and sheath were almost fully recovered from the C perturbation created by the defoliation treatment. Our data agree with the observations of Schnyder (1999) that foliage production rate after defoliation is related to C rather than N supply; without C, N cannot be assimilated.

Cytokinins respond to defoliation

Maintenance of a C/N balance requires local and long-distance signals to coordinate C and N supply and use across tissue types. The iP-type nucleoside form of cytokinin is considered to be a shoot to root signal of the N status of the plant, whereas the tZRs could be a root to shoot signal of N availability (Hirose et al., 2008; Ruffel et al., 2011; Kiba et al., 2013). Our data indicate that defoliation resulted in significant accumulation of iP and iPRMP in the sheath across all growth conditions and tZR accumulation in the root under LDHN, LDLN and SDHN in response to defoliation in plants 8 h after cut (Fig. 5). This suggests long-distance sheath to root and/or root to sheath signalling mediated by cytokinins taking place 8 h after defoliation (Fig. 5A). Considering that the SDHN condition is characterized by limited C availability and high N supply, it is possible that cytokinins are induced by a decrease in the C/N ratio and act as early systemic signals for rapid regulation of WSC remobilization and amino acid synthesis under such growth conditions. Gajdošová et al. (2011) hypothesized that cZ-containing tRNA maintains minimal levels of cytokinins under growth-limiting conditions, and this could explain the significant changes in the concentration of cZ-type cytokinins measured under our experimental conditions at 8 h after defoliation (Fig. 5).

CONCLUSIONS

Our results highlight the integration of the N and C signals for amino acid assimilation and regrowth following defoliation. We suggest that cytokinins act as a mobile signal between the leaf sheath and the root system regulated by the C/N ratio 8 h after defoliation. Subsequently, an overall increase in amino acid synthesis took place, mainly in the leaf and sheath, to support rapid regrowth of photosynthetic material 24 h after cutting. Finally, a negative feedback brought the response back towards pre-defoliation levels within 168 h and this resulted in contrasting amino acid profiling depending upon the initial C and N reserves and availability during regrowth. Our study indicates that there are substantial differences between tissue types and this develops our understanding of the spatial and temporal dynamics in C and N assimilation taking place following the removal of above-ground tissue likely to be associated with grazing. Our data provide a better understanding of the underlying mechanisms controlling the efficiency of N uptake and assimilation, which is required if we are to limit N leaching in managed ryegrass pasture systems. We suggest that N-fertilizer application should be planned such that peak N availability does not occur immediately after grazing but is delayed until the C resources of the plant are reinstated.

SUPPLEMENTARY DATA

Supplementary data are available online at https://academic.oup.com/aob and consist of the following. Data S1: measurements of C concentration. Data S2: heat map of variation over the time course of the experiment for individual amino acids of uncut plants. Table S1: full cytokinin profiling. Data S3: list of abbreviations.