Introduction

Water is crucial for plant growth (Boyer 1982), and, as such, an increase in the prevalence of drought will have an important negative impact on the productivity of agriculture (Passioura 2007). In general, drought depresses nutrient uptake by the root and transport to the shoot due to a restricted transpiration rate affecting active transport and membrane permeability (Kramer and Boyer 1995). Drought-dependent nitrogen deficiency may contribute to growth limitation under water deprivation (Heckathorn and others 1997). Many researchers have shown a directly proportional relationship between nitrate (NO3 −) and yield, and also between yield and foliar nitrogen (N) content (Kim and others 2011; Li and Lascano 2011). For this reason, crops that maintain a high N content and productivity under water stress are indispensable.

Nitrate is the main nitrogen source in agricultural soils. However, the reduced nitrogen form available to plants for assimilation into amino acids and proteins is ammonium (NH4 +) (Miflin and Habash 2002). The reduction of NO3 − to NO2 − catalyzed by nitrate reductase (NR) is considered the limiting step in N assimilation (Fig. 1). Drought affects different steps of nitrogen metabolism, namely, ion uptake, nitrogen assimilation, and amino acid and protein synthesis. Water stress may also restrict the ability of plants to reduce and assimilate nitrogen through the inhibition of enzymes implicated in nitrogen metabolism, such as NR. Therefore, NR, the first enzyme in the pathway of nitrate assimilation, has proved to be one of the enzymes that exhibit declining activity in water-stressed leaves of several species (Fresneau and others 2007; Robredo and others 2011). As such, an increase in the NR activity leads to a corresponding increase in the potential for nitrate reduction and confers a greater capacity for general amino acid synthesis, protein synthesis, or total nitrogen (Singh and Usha 2003).

Fig. 1
figure 1

A schematic representation of nitrate reduction, the photorespiratory cycle, and ammonium assimilation. Photorespiration takes place in the chloroplasts, peroxisomes, and mitochondria, and is a consequence of oxygenation of ribulose-1,5-biphosphate (RuBP) catalyzed by RuBP carboxylase/oxygenase (Rubisco), which generates one molecule of glycerate-3-phosphate (3-PGA) and one of glycolate-2-phosphate (2-PG). This 2-PG is hydrolyzed by phosphoglycolate phosphatase to glycolate, which is transported to the peroxisome and oxidized to glyoxylate by glyoxylate oxidase (GO). Glyoxylate is transaminated to glycine by the reaction catalyzed by glutamate:glyoxylate aminotransferase (GGAT) and is transported to the mitochondria. Subsequently, glycine is transformed into serine by the action of the enzymes glycine decarboxylase and hydromethyltransferase. The serine formed in the mitochondria is transported to the peroxisome, where it is transformed by serine:glyoxylate aminotransferase (SGAT) to hydroxypyruvate, which is reduced to glycerate by hydroxypyruvate reductase (HR). Finally, glycerate moves to the chloroplast, where it is phosphorylated by glycerate kinase, giving rise to a molecule of 3-PGA, which enters the Calvin cycle. NH4 + assimilation occurred via glutamine synthetase (GS) and glutamate synthase (GOGAT). Nitrogen is incorporated into aspartate and other amides and amino acids by aspartate aminotransferase (AAT)

Full size image

Another known effect of water stress in plants is a closure of stomata to avoid further loss of water through transpiration (Lawlor 1995). This could cause photoinhibition by diminishing the use of electrons by photosynthesis (Roland and others 2006). As protection of the photosynthetic apparatus against such damage, this excess of photons can be used by photorespiration in C3 plants. Thus, this process has been viewed as a wasteful process, a vestige of the high CO2 atmosphere under which plants evolved (Wingler and others 2000). At best, according to current thought, photorespiration may mitigate photoinhibition under high light and drought stress (Wingler and others 2000), or it may generate amino acids such as glycine for other metabolic pathways (Noctor and others 1999).

Plants produce significant quantities of NH4 + through the reduction of NO3 − and photorespiration in the step from glycine to serine (Fig. 1). In fact, photorespiration can produce 20-fold more NH4 + than that generated by NO3 − reduction and is considered the largest source of this cation, especially in C3 plants (Hirel and Lea 2001). NH4 + is toxic to plants, causing proton extrusion associated with ammonium uptake, cytosolic pH disturbances, uncoupling of photophosphorylation, and so on (Kronzucker and others 2001). Therefore, it must be rapidly assimilated into nontoxic organic nitrogen compounds. This assimilation occurs via glutamine synthetase (GS) and glutamate synthase (GOGAT) (Fig. 1). Nitrogen is incorporated into aspartate and other amides and amino acids by aspartate aminotransferase (AAT). Alternatively, glutamate dehydrogenase (GDH) can also catalyze NH4 + incorporation into glutamate by reductive amination of 2-oxoglutarate (Cammaerts and Jacobs 1985). The function of the alternative GDH pathway remains unclear; it is proposed to play a complementary role under adverse environmental conditions (Lu and others 2005). Wang and others (2007) reported that the GS/GOGAT cycle does not play a major role in NH4 + assimilation under salinity stress in wheat plants. In cucumber plants under nitrate stress, the GS/GOGAT cycle decreased, possibly due to low water potential and NH3 toxicity (Yang and others 2010). Other authors have shown that resistance to water stress was increased to improve the activity of N metabolism in key enzymes (Xu and Zhou 2006; Sánchez-Rodríguez and others 2011a).

Grafting is a horticultural technique, practiced for many years and in many parts of the world, used to overcome many abiotic stresses (Estañ and others 2005; Venema and others 2008; Abdelmageed and Gruda 2009). Grafted plants usually show increased uptake of water and minerals compared with self-rooted plants as a consequence of the vigorous root system used as rootstock (Ruiz and others 2006). Ruiz and others (1997) concluded that N content was influenced more by the rootstock genotype than by the scion in melon plants. The utilization of certain rootstocks has been found to stimulate NR activity and nitrogen metabolism in roses, melon, and tobacco plants (Pulgar and others 2000; Ruiz and others 2006). The characteristics of the rootstocks could result in increased absorption, upward transport, and accumulation of NO3 − in the scion, thereby stimulating NR and NO3 − assimilation (Martínez-Ballesta and others 2010). However, little is known about the effect of grafting on the activity of enzymes involved in NH4 + assimilation.

The practical and horticultural aspects of grafting technology have been described in several reviews (Lee and Oda 2003; Martinez-Ballesta and others 2010), but less has been compiled about the physiological implications of rootstock–scion interactions as a barrier for the translocation of water and nutrients, or the effect of the rootstock–scion connection on N metabolism of the grafted plants. In a previous work, we selected the most drought tolerant (cv. Zarina) and sensitive (cv. Josefina) from among five commercial tomato cultivars (Lycopersicon esculentum) (Sánchez-Rodríguez and others 2010) and observed that cv. Zarina presented an improvement in N metabolism under water stress conditions (Sánchez-Rodríguez and others 2011a). Therefore, the aim of the present work was to examine the ways in which the grafting affects enzymes involved in N metabolism in response to moderate water stress associated with photorespiration as a mechanism to generate NH4 +, to determine the involvement of grafting in this process under stress conditions. We studied the response to moderate water stress with different combinations of grafted scion–rootstocks and self-grafted and ungrafted tomato plants using cv. Zarina and cv. Josefina to test the viability and efficiency of this grafting technique in terms of N metabolism.

Materials and Methods

Plant Material and Treatments

Two tomato (L. esculentum Mill) cultivars, Zarina and Josefina, were used as scion and rootstock (Fig. 2). The seeds of these cultivars were germinated and grown for 30 days in a tray with wells (each well was 3 cm × 3 cm × 10 cm) in the nursery Semillero Saliplant S.L. (Carchuna, Granada). Grafting was performed when seedlings had developed three to four true leaves. In the vermiculite trays used for germination, the seedlings were cut over the cotyledons, using the shoot as scion and the remaining plant part as rootstock. Grafts were made immediately after cutting the plants, and grafting clips were used to hold the graft union. Self-grafted plants were included as controls. After grafting, seedlings were covered with a transparent plastic lid to maintain a high humidity level and to facilitate graft formation and were left in the shade for 24 h. The lid was opened slightly every day to allow reduction in relative humidity; it was removed 6 days after grafting. Afterward, ungrafted and grafted plants were transferred to a cultivation chamber at the Plant Physiology Department of the University of Granada under controlled conditions, with relative humidity of 50 ± 10 %, 25 °C/15 °C (day/night), 16-h/8-h photoperiod, and a PPFD (photosynthetic photon flux density) of 350 μmol m−2 s−1 (measured with an SB quantum 190 sensor, LI-COR Inc., Lincoln, NE, USA). Under these conditions, the plants grew in individual 8-L pots (25-cm upper diameter, 17-cm lower diameter, and 25 cm in height) filled with a 1:1 perlite:vermiculite mixture. Throughout the experiment, the plants were grown in a complete nutrient solution (Sánchez-Rodríguez and others 2010). The water stress treatments began 45 days after germination and were maintained for 22 days. The control treatment received 100 % field capacity (FC) irrigation, whereas moderate water stress corresponded to 50 % field capacity. The experimental design was a randomized complete block with 12 treatments (Zarina ungrafted, Josefina ungrafted, Zarina self-grafted, Josefina self-grafted, JosxZar and ZarxJos well-watered 100 % FC and water stress 50 % FC) (Fig. 2) arranged in individual pots with six plants per treatment (one plant per pot) and three replications each.

Fig. 2
figure 2

Outline of the grafting design

Full size image

Plant Sampling

All plants were at the late vegetative stage when harvested. Leaves (excluding petioles) were sampled on day 67 after germination. The plant material was rinsed three times in distilled water after disinfection with 1 % nonionic detergent and then blotted on filter paper. A part of the plant material was used for the assay of fresh weight (FW), amino acids, and proteins, and of NR, nitrite reductase (NiR), glutamine synthase (GS), glutamate synthetase (GOGAT), aspartate aminotransferase (AAT), Rubisco, glyoxylate oxidase (GO), glutamate:glyoxylate aminotransferase (GGAT), hydroxypyruvate reductase (HR), and glutamate dehydrogenase (GDH) enzymatic activities. The rest of the plant material was lyophilized and used to determine NO3 −, NH4 +, and organic and total reduced N and total N.

Analysis of N Forms, Soluble Protein, and Free Amino Acid Concentration

NO3 − was analyzed from an aqueous extraction of 0.2 g of DW in 10 mL of Millipore-filtered water. A 100-μL aliquot was taken for NO3 − determination and added to 10 % (w/v) salicylic acid in sulfuric acid at 96 %, and the NO3 − concentration was measured by spectrophotometry as performed by Cataldo and others (1975). NH4 + was analyzed from an aqueous extraction and was determined by using the colorimetric method described by Krom (1980).

For total reduced-N determination, a sample of 0.1 g DW was digested with sulfuric acid and H2O2 (Wolf 1982). After dilution with deionized water, a 1-mL aliquot of the digest was added to the reaction medium containing buffer [5 % potassium sodium tartrate, 100 μM sodium phosphate, and 5.4 % (w/v) sodium hydroxide], 15 %/0.03 % (w/v) sodium silicate/sodium nitroprusside, and 5.35 % (v/v) sodium hypochlorite. Samples were incubated at 37 °C for 15 min, and total reduced N was measured by spectrophotometry according to the method of Baethgen and Alley (1989). Total N concentration was assumed to represent the sum of the total reduced N and NO3 −.

Amino acids and proteins were determined by homogenization of 0.5 g FW in 50 mM cold KH2PO4 buffer at pH 7 which was then centrifuged at 12,000g for 15 min. The resulting supernatant was used for the determination of total amino acids by the ninhydrin method (Yemm and Cocking 1955). Soluble proteins were measured with Bradford G-250 reagent (Bradford 1976).

Nucleotide Analysis

Pyridine nucleotides were extracted from liquid-N-frozen leaf material in 1 mL of 100 mM NaOH [for NAD(P)H] or 5 % TCA [for NAD(P)+]. The extracts were boiled for 6 min, cooled on ice, and centrifuged at 12,000g for 6 min. Samples were adjusted to pH 8.0 with HCl or NaOH and 100 mM bicine (pH 8.0). Nucleotides were quantified by the enzyme-cycling method (Matsumura and Miyachi 1980) with some modification (Gibon and Larher 1997).

Enzyme Extractions and Assays

Leaves were ground in a mortar at 0 °C in 50 mM KH2PO4 buffer (pH 7.5) containing 2 mM EDTA, 1.5 % (w/v) soluble casein, 2 mM dithiothreitol (DTT), and 1 % (w/v) insoluble polyvinylpolypyrrolidone. The homogenate was filtered and then centrifuged at 30,000 g for 20 min. The resulting extract (cytosol and organelle fractions) was used to measure the enzyme activity of NR, NiR, GOGAT, and GDH. The extraction medium was optimized for these enzyme activities so that they could be extracted together according to the same method (Singh and Srivastava 1986).

The NR assay followed the methodology of Kaiser and Lewis (1984). The NO2 − formed was colorimetrically determined at 540 nm after azocoupling with sulfanilamide and naphthylethylenediamine dihydrochloride according to the method of Hageman and Hucklesby (1971).

NiR activity was defined by the disappearance of NO2 − from the reaction medium (Lillo 1984). After incubation at 30 °C for 30 min, the NO2 − content was determined colorimetrically as above.

GOGAT activity was assayed spectrophotometrically at 30 °C by monitoring the oxidation of NADH at 340 nm, essentially as indicated by Singh and Srivastava (1986), always within 2 h of extraction. Two controls, without ketoglutarate and glutamine, respectively, were used to correct for endogenous NADH oxidation. The decrease in absorbance was recorded for 5 min.

GDH activity was assayed by monitoring the oxidation of NADH at 340 nm, essentially as indicated by Singh and Srivastava (1986). The reaction mixture consisted of 50 mM KH2PO4 buffer (pH 7.5) with 200 mM NH4 + sulfate, 0.15 mM NADH, 2.5 mM 2-oxoglutarate, and enzyme extract. Two controls, without ketoglutarate and NH4 + sulfate, respectively, were used to correct for endogenous NADH oxidation. The decrease in absorbance was recorded for 3 min.

GS was determined by an adaptation of the hydroxamate synthetase assay published by Kaiser and Lewis (1984). Leaves were ground in a mortar at 0 °C in 50 mL of maleic acid–KOH buffer (pH 6.8), containing 100 mM sucrose, 2 % (v/v) β-mercaptoethanol, and 20 % (v/v) ethylene glycol. The homogenate was centrifuged at 30,000g for 20 min. The resulting extract was used to measure the enzyme activity of GS. The reaction mixture used in the GS assay was composed of 100 mM KH2PO4 buffer (pH 7.5), with 4 mM EDTA, 100 mM l-sodium glutamate, 450 mM MgSO4·7H2O, 300 mM hydroxylamine, 100 mM ATP, and enzyme extract. Two controls were prepared, one without glutamine and the other without hydroxylamine. After incubation at 28 °C for 30 min, the formation of glutamylhydroxamate was colorimetrically determined at 540 nm after complexing with acidified ferric chloride.

AAT activity was assayed spectrophotometrically at 340 nm using the method published by Gonzalez and others (1995). AAT enzyme was extracted in conditions identical to those for GS. The reaction mixture consisted of 50 mM Tris–HCl buffer (pH 8), 4 mM MgCl2, 10 mM aspartic acid, and enzyme extract. The decrease in absorbance was recorded for 3 min.

Rubisco activity was measured spectrophotometrically by coupling 3-phosphoglyceric acid formation with NADH oxidation at 25 °C according to Nakano and others (2000). The total activity was assayed after the crude extract was activated in a 0.1-mL activation mixture containing 33 mM Tris–HCl (pH 7.5), 0.67 mM EDTA, 33 mM MgCl2, and 10 mM NaHCO3 for 15 min. Initial Rubisco activity measurements were carried out in a 0.1-mL reaction medium containing 5 mM Hepes–NaOH (pH 8.0), 1 mM NaHCO3, 2 mM MgCl2, 0.25 mM DTT, 0.1 mM EDTA, 1 U glyceraldehyde 3-phosphate dehydrogenase, 0.5 mM ATP, 0.015 mM NADH, 0.5 mM phosphocreatine, 0.06 mM RuBP, and 10 μL of extract. The change in absorbance at 340 nm was monitored.

For the GO determination, fresh leaf tissue (0.25 g) was ground in a chilled mortar with PVPP and 1 mL of 50 mM Tris–HCl buffer (pH 7.8) with 0.01 % Triton X-100 and 5 mmol 1,4-dithioerythritol (DTT). The homogenate was centrifuged at 30,000g for 20 min. The supernatant was decanted and immediately used for the enzyme assay. GO was assayed as described by Feierabend and Beevers (1972) with modifications. A volume of assay mixture containing 50 mM Tris–HCl buffer (pH 7.8), 0.009 % Triton X-100, 3.3 mM phenylhydrazine HCl (pH 6.8), 50 μL of plant extract, and 5 mM glycolic acid (neutralized to pH 7 with KOH) was used to start the reaction. GO activity was determined by following the formation of glyoxylate phenylhydrazone at 324 nm for 2 min after an initial lag phase of 1 min.

For determination of GGAT and HR, leaves were ground in a chilled mortar in 100 mM Tris–HCl buffer (pH 7.3) containing 0.1 % (v/v) Triton X-100 and 10 mM DTT. The homogenate was centrifuged at 20,000g for 10 min. The resulting extract was used to measure enzyme activity. The extraction medium was optimized for the enzyme activities such that they could be extracted together using the same method (Hoder and Rej 1983).

GGAT activity was measured by coupling the reduction of 2-oxoglutarate by NADH in a reaction catalyzed by GDH. The reaction was assayed in a mixture containing 100 mM Tris–HCl (pH 7.3), 20 mM glutamate, 1 mM glyoxylate, 0.18 mM NADH, 0.11 mM pyridoxal-5-phosphate, 83 mM NH4Cl, and 0.3 U GDH in a final volume of 0.6 mL (Igarashi and others 2006).

HR assay was performed with 100 mM Tris–HCl (pH 7.3), 5 mM hydroxypyruvate, and 0.18 mM NADH. Activity was assayed spectrophotometrically by monitoring NADH oxidation at 340 nm (Hoder and Rej 1983).

The protein concentration of the extracts was determined according to the method of Bradford (1976) using bovine-serum albumin as the standard.

Statistical Analysis

Data were subjected to a simple analysis of variance (ANOVA) at 95 % confidence using the Statgraphics 6.1 program (Statpoint Technologies, Warrenton, VA, USA). Means were compared using Fisher’s least-significant differences (LSD).

Results

NH4 + Production: NO3 − Reduction and Photorespiration

Nitrate levels showed a significant increase in ungrafted Zarina under water stress conditions; however, in cv. Josefina a decrease of 40 % was observed (Table 1). In self-grafting, ZarxZar and JosxJos, no significant differences were observed between well-watered and water stress. In reciprocal grafting, only ZarxJos showed a stronger increase in nitrate concentration under stress conditions, whereas in JosxZar no significant differences were observed (Table 1). The results of NR assays reflected significant differences in cv. Zarina and its self-graft, in which we observed an increase under water stress conditions (Table 1). Besides, Josefina ungrafted and JosxJos showed decreased NR activity over well-watered conditions (Table 1). In reciprocal grafting, NR activity increased only in ZarxJos. Moreover, for NiR activity, no significant differences were observed in any cultivars (Table 1). In the case of NH4 + concentration, ungrafted Zarina showed a decrease over control conditions (Table 1). In the reciprocal grafting, no significant differences in NH4 + content were observed under stress conditions (Table 1).

Table 1 Influence of moderate water stress on response of NO3 − reduction and NH4 + concentration in ungrafted, grafted, and self-grafted tomato plants
Full size table

With regard to the photorespiration process, only the initial activity and the total Rubisco showed significant differences in ungrafted Zarina and JosxZar, which presented an increase in Rubisco activity under water stress conditions (Fig. 3). The activity of enzymes that complete the cycle of photorespiration, that is, GO, GGAT, and HPR, showed a general increase under water stress conditions in ungrafted Zarina, ZarxZar, and JosxZar (Table 2). However, for cv. Josefina self-graft and ZarxJos, a general decrease was observed with respect to the well-watered condition (Table 2). With respect to the different forms of pyridine dinucleotides, our results showed a decrease in NADH in Josefina ungrafted, JosxJos, and ZarxJos under water stress (Table 3), whereas JosxZar presented an increase of 32 % in the NADH concentration under water stress with respect to well-watered plants. Zarina ungrafted and JosxZar showed an increase in NADP+ under water stress (Table 3). For the NADH/NAD ratio, only JosxZar showed a significant increase of 43 % under water stress (Table 3).

Fig. 3
figure 3

Response of Rubisco activity in ungrafted, grafted, and self-grafted tomato plants well-watered and subjected to moderate water stress. Columns are mean ± SE (n = 9), and differences between means were compared using Fisher’s least-significant difference test (LSD; p = 0.05). Asterisk (*) indicates significant difference with control groups

Full size image
Table 2 Influence of moderate water stress on some photorespiration enzymes in ungrafted, grafted, and self-grafted tomato plants
Full size table
Table 3 Influence of moderate water stress on pyridine dinucleotides concentration in ungrafted, grafted, and self-grafted tomato plants
Full size table

NH4 + Incorporation and Assimilation Products

The enzymes of the GS/GOGAT cycle increased under water stress conditions in Zarina ungrafted and in reciprocal grafting (JosxZar and ZarxJos) (Table 4). However, in cv. Josefina no significant differences were observed. Self-grafting showed a decrease in GS and GOGAT activities under water stress conditions (Table 4). With regard to AAT activity, cv. Zarina, JosxZar, and ZarxJos showed an increase of 47, 22, and 67 %, respectively. No significant differences were observed in other graft combinations (Table 4). Finally, by contrast, the GDH activity increased significantly only in cv. Josefina, JosxJos, and ZarxZar, whereas no significant differences were observed in the other cases (Fig. 4).

Table 4 Influence of moderate water stress on enzymes responsible for NH4 + assimilation in ungrafted, grafted, and self-grafted tomato plants
Full size table
Fig. 4
figure 4

Response of glutamate dehydrogenase (GDH) in ungrafted, grafted, and self-grafted tomato plants well-watered and subjected to moderate water stress. Columns are mean ± SE (n = 9), and differences between means were compared by Fisher’s least-significant difference test (LSD; p = 0.05). Asterisk (*) indicates significant difference with control groups

Full size image

Reduced N was increased in Zarina ungrafted and ZarxZar under water stress conditions (Fig. 5a). However, a significant decrease was observed in cv. Josefina and its self-graft. There was no significant difference with respect to well-watered conditions in the reciprocal grafts (Fig. 5a). With regard to total N, only cv. Zarina and ZarxJos showed an increase under stress conditions (Fig. 5b). For soluble amino acids, no significant differences were observed in different grafting combinations (Fig. 5c). Also, soluble proteins increased in the cv. Zarina and the reciprocal grafts, whereas in the rest of the cases the values were not affected or decreased after water stress (Fig. 5d).

Fig. 5
figure 5

Response in foliar concentration of organic (a), total N (b), total amino acids (c), and soluble proteins (d) in ungrafted, grafted, and self-grafted tomato plants well-watered and subjected to moderate water stress. Columns are mean ± SE (n = 9), and differences between means were compared using Fisher’s least-significant difference test (LSD; p = 0.05). Asterisk (*) indicates significant difference with control groups

Full size image

Discussion

NH4 + Production: NO3 − Reduction and Photorespiration

In general, drought can depress nutrient uptake by the root and transport to the shoot as a result of a restricted transpiration rate (Kramer and Boyer 1995). However, water and nutrient uptake could be increased in grafted plants as a result of the enhancement of vigor by the rootstock’s root system and its effects on plant yield (Ruiz and others 1997). Indeed, in our previous work we studied the effects of grafts in uptake fluxes and found that the use of cv. Zarina as rootstock (ZarxJos) improves the NO3 − uptake flux under stress conditions (Sánchez-Rodríguez and others 2011b). According to these data, our results showed an increase in NO3 − concentration and NR activity under water stress only in cv. Zarina ungrafted and in ZarxJos (Zarina used as rootstock) (Table 1). The characteristics of the rootstocks could result in increased absorption, upward transport, and accumulation of NO3 − in the scion, thereby stimulating NR and NO3 − assimilation. Similar results were obtained by Ruiz and Romero (1999) in melon plants; NR activity and NO3 − accumulation were conditioned significantly by the scion–rootstock interaction and by rootstock genotype, whereas the scion genotype did not show any such effect. Many researchers have shown a directly proportional relationship between NO3 – and yield (Kim and others 2011; Li and Lascano 2011). Also, Ruiz and others (1997, 2006) have shown the essential role of NO3 − assimilation in the yield increase. Our result showed that cv. Zarina, JosxZar, and ZarxJos had greater biomass and a relative growth rate (RGR) associated with high leaf relative water content (LRWC) under water deficit conditions, indicating that these cultivars are more tolerant to this growth situation (Sánchez-Rodríguez and others 2011c).

It has been estimated that the production of NH4 + by photorespiration is much greater than the primary assimilation of NH4 + resulting from nitrate reduction (Wingler and others 2000). Our results showed that only cv. Zarina and JosxZar (cv. Zarina-like scion) had an increase in initial and total Rubisco activity under water stress (Fig. 3) and in the activities of the enzymes GO, GGAT, and HPR (Table 3). Thus far, the effects of drought on Rubisco activity were inconsistent in studies in which different plant species and stress durations were used (Parry and others 2002). Rubisco activity varies with plant species and cultivars that differ in drought tolerance; drought-tolerant plants typically exhibit higher Rubisco activity (Galmés and others 2011; Carmo-Silva and others 2012). These results agree with those of Ferreira-Silva and others (2010), who observed that the higher stability shown by Rubisco in cashew BRS/BRS grafted plants could indicate that this combination can be more resistant under salinity. Moreover, it was found that the grafted bitter melon seedlings had a higher Rubisco activity than ungrafted seedlings under flooding stress (Liao and Lin 1996). On the other hand, photorespiration serves as an important redox mechanism that increases the cytosolic NADH/NAD ratio (Lee and Oda 2003). In JosxZar, which increases the photorespiration, we observed an increase in NADH concentration and the NADH/NAD ratio, whereas in Zarina ungrafted and ZarxZar, no significant difference was observed (Table 3). Because the first step of NO3 − assimilation occurs in the cytosol and uses NADH, this may explain why we observed NO3 − assimilation to be greater when photorespiration was highest in cv. Zarina ungrafted and its self-graft. However, in JosxZar, photorespiration is high although nitrate uptake is not enhanced by the root and NO3 − assimilation is therefore lower.

NH4 + Incorporation and Assimilation Products

The reassimilation of NH4 + produced by the photorespiratory nitrogen cycle is essential for maintaining nitrogen status (Wingler and others 2000). In higher plants, NH4 + is assimilated mainly through the concerted action of GS and GOGAT. Several authors have shown that the decline in GS activity is correlated with water stress (Robredo and others 2011), and in our study, drought stress provoked a marked decrease in GS activity in cv. Josefina and the self-graft (Table 4). The GS/GOGAT cycle increased in cv. Zarina ungrafted, JosxZar, and ZarxJos only under stress conditions (Table 4). In ZarxJos, increased ammonium assimilation could result from increased NO3 − assimilation, whereas in JosxZar, NH4 + in the water-stressed plant might result mainly from increased photorespiration. These results agree with those of Masclaux-Daubresse and others (2006) from their study of tobacco plants. They provide strong evidence that the GS/GOGAT cycle is the primary route of ammonium assimilation and that GDH plays a minor role. In fact, GDH activity showed an increase only in Josefina ungrafted, its self-graft, and ZarxZar under water stress (Fig. 4). In this sense, it has been demonstrated that NH4 + might be a signal responsible for the induction of GDH activity (Masclaux-Daubresse and others 2006). Thus, the increased NH4 + observed in Josefina ungrafted, its self-graft, and ZarxZar under water stress (Table 1) might be responsible for the higher GDH activity (Fig. 4), which has been previously shown in wheat seedlings exposed to salinity (Wang and others 2007). Moreover, NH4 + is toxic to plants which might result in decreased biomass in these cultivars (Sánchez-Rodríguez and others 2011c). Besides, it has been demonstrated that the combined action of GS and GOGAT is the principal pathway for assimilating ammonia, and the amination activity of GDH functions only when the GS/GOGAT cycle pathway is inhibited under stress conditions such as salinity or drought (Mena-Petite and others 2006). Finally, the glutamate and glutamine generated in the GS/GOGAT cycle are allocated to the synthesis of aspartate and asparagine, which are produced in the reactions catalyzed by AAT and asparagine synthetase (Hodges 2002). Our results showed an increase in AAT activity in cv. Zarina, JosxZar, and ZarxJos under water stress (Table 4). This could be related to the increase in the GS/GOGAT cycle in these cultivars and grafts (Table 4).

The result of the incorporation of NH4 + can be quantified by the analysis of reduced N, which is generally the product of N assimilation and is formed mainly by amino acids and proteins. The total N, the sum of the total reduced N and NO3 −, is considered a critical parameter in the determination of the nutritional state of plants (Ruiz and Romero 1999). The greater efficiency in NO3 − reduction and NH4 + reassimilation in Zarina ungrafted, JosxZar, and ZarxJos plants was confirmed by the results for protein and total N, which were higher under water stress (Fig. 5). Ruiz and others (2006) had similar results with grafted tobacco plants, observing higher amino acids, protein, and total N in all grafted plants with respect to nongrafted plants. These results confirm those previously found by other authors, who observed that an increase in the NR activity leads to a corresponding increase in the potential for NO3 − reduction and confers a greater capacity for general amino acid synthesis, protein synthesis, or total N assimilation (Singh and Usha 2003).

The increase in N metabolism displayed by the ZarxJos and JosxZar combinations, which was associated with other physiological factors such as maintenance of leaf relative water content (LRWC) in ZarxJos or higher photorespiration in JosxZar, strongly suggests that these combinations are the most capable of coping with moderate drought stress. In addition, the results demonstrate that interactions between the rootstock and scion may exert a strong effect on the N metabolism responses of tomato plants under water stress. In this study, when Zarina (tolerant cultivar) was used as a rootstock grafted with Josefina (ZarxJos), these plants showed an improved N uptake and NO3 − assimilation (Table 1). On the other hand, when Zarina was used as the scion (JosxZar), these grafted plants showed an increase in the photorespiration cycle (Table 2), which may generate amino acids and proteins (Fig. 5c, d).

Our study offers promising results that could improve the understanding of some physiological mechanisms involved in scion and rootstock interactions under water stress conditions. However, further studies are needed to better elucidate biochemical, molecular, and genetic traits that might exert control on N metabolism associated with drought resistance in grafted plants. Such traits could be utilized in plant breeding programs by means of the selection of improved genotypes of rootstocks and scions using molecular marker-assisted techniques. In conclusion, grafting improves NO3 − photoassimilation or photorespiration in water stress experiments on tomato. Consequently, complex interactions between photorespiratory metabolism and NO3 − assimilation may be more important than previously recognized in plant leaves.