Planta (2006) 225:165–181 DOI 10.1007/s00425-006-0338-5
O RI G I NAL ART I C LE
Wheat nitrogen metabolism during grain Wlling: comparative role of glumes and the Xag leaf Marta S. Lopes · Nuria Cortadellas · Thomas Kichey · Frédéric Dubois · Dimah Z. Habash · José L. Araus
Received: 20 December 2005 / Accepted: 26 May 2006 / Published online: 28 June 2006 © Springer-Verlag 2006
Abstract The mobilization of nitrogen (N) compounds and the roles played by glumes and the Xag leaf during grain Wlling were studied in bread wheat (Triticum aestivum L. cv. Florida) grown under Weld conditions. Glumes lost twice as much of their total N content as that lost by the Xag leaf between the milk and early dough stages. In the Xag leaf, glumes and grains, Glu, Asp, Ser and Ala accounted for 85% of all the reductions in the free amino acid pool. Principal component analysis of free amino acid pools separated grains from the glumes and the Xag leaf, suggesting grain speciWc regulations in the use of free amino acids in protein synthesis. In all three organs, no decrease in Gln was detected, probably due to steady glutamine synthetase (GS; EC 6.3.1.2) activities per soluble protein in both the Xag leaf and glumes. Compared with the Xag leaf, glumes presented relatively smaller
M. S. Lopes · J. L. Araus (&) Unitat de Fisiologia Vegetal, Facultat de Biologia, Universitat de Barcelona, Diagonal 645, 08028 Barcelona, Spain e-mail:
[email protected] N. Cortadellas Unitat de Microscopia Electronica, Facultat de Medicina, Serveis Cientíticotècnics de la Universitat de Barcelona, Casanovas 143, 6ª planta, 08036 Barcelona, Spain D. Z. Habash Crop Performance and Improvement Division, Rothamsted Research, Harpenden, Hertfordshire, AL5 2JQ, UK T. Kichey · F. Dubois Unité de Biologie des Plantes et Contrôle des Insectes Ravageurs, Université de Picardie Jules Verne, 33 rue St Leu, 80039 Amiens, France
amounts of the chloroplast GS associated isoform. This we show is due to a lower relative number of mesophyll cells in glumes as supported by the diVerent anatomy and the cellular pattern of the GS immunolocalization. We argue that cellular distribution plays a key role in supporting metabolism to enable the various functions undertaken by glume tissue. Keywords Glumes · Grain Wlling · Glutamine synthetase · Nitrogen · Wheat Abbreviations C Carbon dw Dry weight Fd-GOGAT Ferredoxin dependent glutamate synthase fw Fresh weight GS1 and 2 Glutamine synthetase isoforms N Nitrogen PEPc Phosphoenolpyruvate carboxylase RLI Relative labelling index Rubisco Ribulose-1,5-bisphosphate carboxylase oxygenase
Introduction Substantial reductions in the nitrogen (N) content of wheat plant parts occur between Xowering and maturity. Wheat shoot proteins undergo hydrolysis and the resulting amino acids are exported to the grains (Feller and Fischer 1994). However, in cereals (Schjoerring 1991), including wheat (Maheswari et al. 1992), considerable N losses in the form of ammonia occur during this period. Namely, an entire leaf and ear can lose
123
166
around 30 and 77 nmol of ammonia per entire organ and per hour (ca. 0.03–0.04% of their N content), respectively (Maheswari et al. 1992). The design of new ideotypes that use N more eYciently, or which prevent N losses, is imperative for adaptation to sustainable crop practices. Thus, knowledge of N metabolism and assimilation is needed to establish the physiological traits that these ideotypes require. Of all the vegetative parts of wheat, the Xag leaves and glumes are metabolically active for the longest period of time and their relative contribution to the Wnal grain nitrogen yield is high (Simpson et al. 1983). In the period just prior to rapid grain Wlling, when there are no alternative sinks, the glumes act as a temporary sink for the N that will soon be remobilized to the grains (Waters et al. 1980). A number of studies have also examined the contribution of ear photosynthesis to Wnal grain yield (Simmons 1987; Araus et al. 1993a; Abbad et al. 2004). Reynolds et al. (2001) reported a ‘high spike photosynthesis’ as a relevant trait for drought tolerance in wheat (see also Tambussi et al. 2005). However, whereas the carbon (C) photosynthetic metabolism of the spike has been studied in some detail (Tambussi et al. 2005), the contribution of this organ to N metabolism during grain Wlling remains obscure. An assessment of the pool of free ammonium, total free amino acid content, and the activity of glutamine synthetase (GS; EC 6.3.1.2) and glutamate dehydrogenase (EC 1.4.1.2) in the Xoral parts reveals that glumes, awns, and grains are, as well as Xag leaves, active sites of ammonia turnover (Maheswari et al. 1992). More recent work in maize also showed that the cob performs signiWcant enzymatic interconversions among Gln, Ala, Asp, and Asn during early reproductive development (Seebauer et al. 2004). Ammonium is assimilated into glutamine and glutamate through the consecutive reaction of GS and glutamate synthase (GOGAT; EC 1.4.7.1). In plants, GS occurs in several isoforms located in both cytosol and plastids and is responsible for ammonia assimilation and re-assimilation (Hirel and Lea 2001; MiXin and Habash 2002). This key enzyme has been used as a marker to predict N status for selection of wheat genotypes with better N-use eYciency (Kichey et al. 2006). In wheat and tobacco leaves, cytosolic GS1 is simultaneously induced with chloroplastic protein degradation and it has been proposed that this GS form is involved in the re-assimilation of ammonium released during senescence (Peeters and van Laere 1994; Tercé-Laforgue et al. 2004). Although some interesting work has been published about GS immunolocalization and protein expression during senescence of the wheat Xag leaf and stem (Kichey et al. 2005), the role and expression of
123
Planta (2006) 225:165–181
diVerent GS isoforms and GOGAT in wheat glumes are unknown. In the present research, N depletion during the last part of grain Wlling was studied in bread wheat Xag leaves and glumes grown under Weld conditions. The Wnal stages of grain Wlling were chosen for the present study since ears and Xag leaves are the last active vegetative organs in the shoot. Further, the second half of grain Wlling corresponds to the stage when the most marked reductions in protein content from Xag leaves are observed (Waters et al. 1980). Thus, a better understanding of the contribution of both organs to grain N in these stages is needed. To this end, changes between the milk and early dough stages in total N, total soluble protein, ribulose-1,5-bisphosphate carboxylase oxygenase (Rubisco; EC 4.1.1.39), and the relative composition of free amino acid pools were assessed. Total N and free amino acid pools were also studied in grains to obtain a good understanding of the contribution made by the Xag leaf and glumes to Wnal grain N. Activity, protein amount and immunolocalization of GS, a key enzyme in the metabolism of amino acids, were recorded in both the Xag leaf and glumes in order to gain a better understanding of the importance of this enzyme in these organs. In this study, we also hypothesize that the amounts of GS2 and GS1 in the Xag leaf and glumes may be primarily correlated with the diVerent amounts of mesophyll present in both organs. Fd-GOGAT (protein level) was also studied to complete our understanding of the hitherto unknown GS/GOGAT cycle in glumes. Additionally, phosphoenolpyruvate carboxylase (PEPc; EC 4.1.1.31), together with Rubisco, were analysed at the protein level to determine the primary C metabolism status in both glumes and the Xag leaf. This research contributes to a better understanding of glumes, which have a major contribution in providing N to grains during the Wnal stages of grain Wlling.
Materials and methods Plant material and growth conditions Bread winter wheat plants (Triticum aestivum L.) from an awnless cultivar (cv. Florida) were used in the present study. The Weld trial was conducted during 2003– 2004 in Gimenells (41°38⬘N, 0°23⬘E, 200 m asl, Lleida Province, North-eastern Spain) in a CalcixerolicXerochrept soil (USDA). Agronomical practices were those usually recommended for irrigated wheat under conventional tillage conditions in this area. Planting took place in November with a seed density of approximately 500 seeds m¡2. The soil was fertilized before
Planta (2006) 225:165–181
sowing, except for N that was applied in the mid tillering and stem elongation phases, corresponding to stages 25 and 35 of Zadocks decimal code, respectively (Zadocks et al. 1974), with 75 kg NH4NO3 ha¡1 each. Irrigation was applied whenever needed. Flag leaves (blades) and ears were harvested in the milk and early dough phases, stages 75 and 87 of Zadocks decimal scale, respectively (Zadocks et al. 1974), around 3 and 4 weeks after anthesis. All organs were collected from the main shoot of nine plants in each of the four plots assayed and stored at ¡80°C until use.
167
Wcation of the diVerent amino acids was performed by measuring their retention time and comparing them with a standard (A¡2908 plus Asn, Gln, and Gaba; Sigma). The amount of each free amino acid was calculated in the whole organ at both developmental stages and the percentages of the diVerent amino acids within the total pool were calculated. The percentage changes in free amino acids between the milk and early dough stages in each organ were calculated. Protein extraction, GS synthetase assay and protein gel-blot analysis of GS, Rubisco, PEPc and GOGAT
Tissue dry matter and total N and C analysis Flag leaf blades, glumes (rachis not included), and grains were sampled and immediately frozen in liquid N and stored at ¡80°C until use. All organs were separated under liquid N and their fresh weight (fw) determined, keeping the material frozen all the time. These tissues were then lyophilized and ground to a Wne powder. The dry weight (dw) and speciWc dry weight (SDW, the ratio of dw to fw) of each organ were determined after lyophilization. SpeciWc N and C contents (% dw) were measured by elemental analysis (EA1108, Series 1, Carlo Erba Istrumentazione, Milan, Italy), using directly ca. 2 mg of ground and lyophilized sample. Measurements were taken at the facilities of the University of Barcelona. Further N and C contents per whole organ (Xag leaf blade, glumes, and grains) were calculated, and the changes in N and C per organ between the milk and early dough stages were assessed. Amino-acid analyses Around 20 mg of lyophilized powder (prepared as previously described) from Xag leaves, glumes, and grains were extracted in 1 ml of 80% (v/v) ethanol for 1 h at 4°C. During extraction, the samples were continuously agitated; then, they were spun for 5 min at 10,000g. The supernatant was recovered, and the pellet was subjected to further extractions in 1 ml 60% (v/v) ethanol and Wnally in 1 ml of water. All supernatants were combined to form the hydro-alcoholic extract with an internal standard (L-norleucine, Sigma). The ethanol–water fractions were stored at ¡80°C until use. Amino acids were separated at the facilities of the University of Barcelona using an auto analyser (Pharmacia LKB Biotechnology, Alpha Plus, Series 2) according to Spackman et al. (1958), with a polystyrene divinyl-benzene sulfonate ion exchange resin. DiVerent buVers of lithium citrate were used. Detection was performed with ninhydrin and detected at 570 and 440 nm. Identi-
Flag leaves and glumes previously stored at ¡80°C were separated and ground in liquid N. To the resulting powder, the GS synthetase assay extraction buVer was added (1 ml per 0.1 g of fresh tissue) containing 100 mM triethanolamine, 1 mM EDTA, 10 mM MgSO4, 5 mM glutamate, and 10% (v/v) ethylene glycol at pH 7.6. Then, 6 mM DTT and 10 M leupeptin were added and ground until homogeneous. The slurry was centrifuged at 21,000g for 10 min and GS was measured as described by Lea et al. (1990). The protein concentration of the tissue extracts was determined using the method outlined by Bradford (1976). Proteins were fractionated by SDS-PAGE [12% (w/v) polyacrilamide] according to Laemmli (1970). The denatured proteins were then electrophoretically transferred onto a nitrocellulose membrane (Schleicher and Schuell, Dassel, Germany) as described by Towbin et al. (1979). An antiserum raised in rabbit against GS1 puriWed from Phaseolus vulgaris L. root nodule tissue was used to detect GS1 and GS2 proteins (Rothamsted Research; Cullimore and MiXin 1984). Protein–antibody complexes were located using peroxidase-conjugated goat anti-rabbit IgG. The same procedure was applied for the large subunit (LSU) of Rubisco polyclonal antibody prepared from Spinacea olearacea L. (courtesy of Dr Guiamet), phosphoenolpyruvate carboxylase polyclonal antibody from Zea mays L. IgG fraction (PEPc, Polysciences; Uedan and Sugiyama 1976), and Fd-glutamate synthase (Fd-GOGAT, Rothamsted Research) from Hordeum vulgare L. All three antibodies were raised in rabbit. Detection was performed with 4chloro-1-naphtol as described elsewhere (Hong et al. 1986). Band signals were quantiWed by densitometric scanning using the Discovery Series Quantity One 1-D Analysis Software (Bio-Rad). Chlorophyll measurements For Xag leaves, 50 l, and for glumes, 100 l, of the complete crude extract used in the GS assay were
123
168
added to 80% (v/v) acetone and chlorophyll measured as described by Bruinsma (1961). Microscopy and GS immunolocalization For Wxation and embedding, Xag leaf and glumes were sampled during the milk stage. Three spikes were sampled and glumes were taken in the middle of the spike. Sections, 1–2-mm wide across the width of the leaf and glume, were cut from the centre of each organ, Wxed at reduced pressure in 2% (v/v) paraformaldehyde, 0.2% (v/v) glutaraldehyde in 0.1 M sodium phosphate buVer (pH 7.4) for 24 h at 4°C. The samples were subsequently rinsed in buVer, dehydrated through a graded ethanol series, and embedded in Lowikryl K4M at ¡20°C, as described by Carlemalm et al. (1982). For immunolocalization of GS by transmission electron microscopy, ultrathin sections (60–80 nm thick) were cut with an ultramicrotome Leika MFS (Vienna, Austria), mounted on 400-mesh nickel grids, and were allowed to dry at 37°C. Sections were Wrst incubated with 5% normal goat serum in T1 buVer (0.05 M Tris– HCl buVer; 2.5% NaCl, 0.1% BSA; and 0.05% Tween 20, pH 7.4) for 30 min at room temperature and then with anti-GS rabbit serum diluted 1:500 in T1 buVer for 2 h at room temperature. Sections were then washed Wve times with T1 buVer, twice with T2 buVer (0.02 M Tris–HCl buVer containing 2% NaCl, 0.1% BSA and 0.05% Tween 20, pH 8) and incubated with 15 nm colloidal gold goat anti-rabbit IgG (British Biocell Int., CardiV, UK) diluted 1:25 in T2 buVer for 1 h at room temperature (Brugière et al. 2000). Control sections were run by omitting the primary antibody. After several washes, the grids were treated with 2% uranyl acetate in water and observed with a transmission electron microscope JEOL 1010 (Japan). Images were digitalized with Bioscan 812 (Gatan UK, Abingdon, UK) at a magniWcation of 20,000. The relative labelling index (RLI) was determined according to Mayhew et al. (2002). For light microscopy, transverse-sections, 1 m thick, were cut with a microtome Leika MFS, stained with methylene blue and photographed with a light microscope (Olympus BHZ-UMA, Tokyo, Japan), coupled with a digital camera (JVC TK 1270). Tissue areas of mesophyll and phloem cells were determined in random Welds of three sections (excluding intercellular spaces) using the IMAT software developed at the experimental facilities of the University of Barcelona (Spain). Relative amounts of GS in chloroplasts in the whole organ were calculated by multiplying the percentage of mesophyll cells per organ (previously measured in
123
Planta (2006) 225:165–181
percentage of area occupied within the whole organ, as described above) and the GS RLI measured in chloroplasts. Relative amounts of GS in the cytosol of mesophyll cells were the product of the percentage of mesophyll cells and GS RLI associated with mesophyll cytosol. The same calculation was applied to phloem cells and the sum of both cytosol values (from mesophyll and phloem cells) was performed. In agreement with previous reports (Hirel and Lea 2001; Kichey et al. 2005), we found that GS is preferentially abundant in chloroplasts and cytosol of mesophyll cells and cytosol of phloem companion cells. Therefore, RLI measurements were only taken for mesophyll and phloem cells. Statistical analysis Analysis of variance was performed with the GLM procedure (SAS Institute Inc., Cary, NC, USA) in order to calculate organ and phenology (Wxed factors) eVects. Means were compared by applying the Duncan test and Least Square Means Test (P
