Polyamine oxidase activity contributes to sustain maize leaf elongation under saline stress

The possible involvement of apoplastic reactive oxygen species produced by the oxidation of free polyamines in the leaf growth of salinized maize has been studied here. Salt treatment increased the apoplastic spermine and spermidine levels, mainly in the leaf blade elongation zone. The total activity of polyamine oxidase was up to 20-fold higher than that of the copper-containing amine oxidase. Measurements of H 2 O 2 , (cid:1) O 2 2 , and HO (cid:1) production in the presence or absence of the polyamine oxidase inhibitors 1,19- bis -(ethylamine)-5,10,15 triazanonadecane and 1,8-diamino-octane suggest that, in salinized plants, the oxidation of free apoplastic polyamines by polyamine oxidase by would be the main source of reactive oxygen species in the elongation zone of maize leaf blades. This effect is probably due to increased substrate availability. Incubation with 200 l M spermine doubled segment elongation, whereas the addition of 1,19- bis -(ethylamine)-5,10,15 triazanonadecane and 1,8-diamino-octane to 200 l M spermine attenuated and reversed the last effect, respectively. Similarly, the addition of MnCl 2 (an (cid:1) O 2 2 dismutating agent) or the HO (cid:1) scavenger sodium benzoate along with spermine, annulled the elongating effect of the polyamine on the salinized segments. As a whole, the results obtained here demonstrated that, under salinity, polyamine oxidase activity provides a signiﬁcant production of reactive oxygen species in the apoplast which contributes to 25–30% of the maize leaf blade elongation.


Introduction
Reactive oxygen species (ROS), namely the superoxide radical (ÁO À 2 ), the hydroxyl radical (HOÁ), and H 2 O 2 are the major apoplastic ROS (aROS) in plants . aROS are necessary in many plant developmental processes (Foreman et al., 2003;Demidchik and Maathuis, 2007), particularly in the elongation zone (EZ) of maize leaves during leaf extension (Rodríguez et al., 2002). In these plants, the salt-induced decrease of aROS contributes to the reduction of leaf elongation (Rodríguez et al., 2004). On the other hand, the diminution of the aforementioned aROS has been attributed to the inhibitory effect of NaCl on the NADPH oxidase (NOX) complex . It has been shown that non-enzymatic processes involving reactive oxygen species (ROS) cause wall polysaccharide scission in vitro (Miller, 1986;Fry, 1998;Schweikert et al., 2000;Fry et al., 2001) and in vivo (Schopfer, 2001). It has also been suggested that a delicate equilibrium between cleavage and cross-linking activities by ROS may take place in the apoplast (Cosgrove, 1999). Under optimal conditions, NOX is the main source of apoplastic ÁO À 2 , which dismutes to H 2 O 2 through superoxide dismutase (SOD) activity. Remarkably, despite the fact that NaCl inhibits NOX activity, plants continue producing aROS in low concentrations and growing at a reduced rate (Rodríguez et al., 2004). Up to now, the origin of those aROS is unknown, and whether such a low aROS amount may still contribute to plant growth under salt stress conditions is uncertain.
Polyamines (PA) are small organic polycations, naturally found in eukaryotic and prokaryotic cells, which have been associated with cell growth and development (Bais et al., 1999;Steiner et al., 2007). In plant cells, the most abundant PA are putrescine (Put), spermidine (Spd), and spermine (Spm, Kaur-Sawhney et al., 2003). Although PA are detected in both symplastic and apoplastic compartments (Torrigiani et al., 1986;Pistocchi et al., 1988;Slocum, 1991;Tiburcio et al., 1997), their biosynthesis takes place only in symplastic subcellular localizations (Slocum, 1991;Borrell et al., 1995;Tiburcio et al., 1997). In turn, PA cross the plant cell membrane towards the apoplast via a still unknown mechanism (Cona et al., 2006a), where they are catabolized by amine oxidases (AOs), enzymes associated with apoplastic compartments (Federico and Angelini, 1991;Angelini et al., 1995;Tavladoraki et al., 1998;Cona et al., 2006a). The copper-containing amine oxidase (CuAO) catabolizes the oxidation of lower PA, such as Put and cadaverine, on primary amino groups, whereas plant polyamine oxidase (PAO) oxidizes higher PA, Spd, and Spm on their secondary amino groups (Federico and Angelini, 1991). PA oxidation produces hydrogen peroxide (H 2 O 2 ) among other products (Lim et al., 2006). Interestingly, some reports have demonstrated that, unlike NOX, PAO activity is stimulated by NaCl in Brassica campestris (Das et al., 1995). Therefore, in the present work, the possibility is addressed that maize AOs maintain or even increase their activity under saline conditions, thus contributing to keep the basal ROS level needed to uphold leaf growth under saline stress. For this purpose, PA levels and AOs activities upon salinization were determined, as well as the effect of PA concentration on growth of the most actively elongating region of the salinized maize leaf.

Plant material
Maize seeds (Zea mays cv. Prozea 30, Produsem, Pergamino, Argentina) were sown on moist vermiculite contained in plastic net frames placed over 4.5 l black plastic trays with aerated water. Trays were kept at 25°C under a light panel of fluorescent and incandescent light bulbs providing 95 lmol photons m À2 s À1 illumination, with a 16 h photoperiod. When the second leaf emerged, 6 d after sowing, the water was changed to half-strength Hoaglands solution (Hoagland and Arnon, 1950), which included 25 mM NaCl in the saline treatment. This solution was changed daily, increasing the NaCl concentration from 25 mM to 50 mM, 100 mM, and finally, 150 mM NaCl. Solutions were thereafter refreshed every 2 d. At harvest, 14 d after seeding, the region spanning 10-20 mm from the ligule was sectioned. This segment was used throughout the experiments.

Segment elongation measurements
The elongation of the leaf blade segments was measured as previously described by Rodríguez et al. (2004). Segments were gently vacuum infiltrated for 1.5 min and incubated for 7 h in different solutions. Digital images of segments were obtained before and after the incubation period, using a scanner (HP PSC 1510, Hewlett Packard Company, Palo Alto, CA). Segment length was measured with image processing software (Optimas 6.1, Optimas Corporation, Bothell, WA) and segment growth was expressed as a percentage of length increase, with respect to the control in that period.
Extraction of free PA, 1,3-diaminopropane (Dap), and apoplastic Na + To extract free PA from the cell extracts (cPA), 30 leaf blade segments were frozen in liquid N 2 and homogenized. The homogenate (300 mg) was resuspended in 1 ml of PCA 5% (v/v), incubated in ice for 30 min and centrifuged at 15 000 g (15 min). The pellet was discarded and the supernatant was kept at -20°C (solution A). Maize apoplastic fluid extraction was performed according to Rodríguez et al. (2002) with modifications. Segment pools were introduced within a net bag, which was, in turn, placed inside a plastic tube. To extract free PA and Dap from apoplastic extract (aPA and aDap, respectively), tubes were centrifuged for 10 min at 2000 g and the fraction collected was lyophilized, resuspended in 200 ll perchloric acid (PCA) 5% (v/v) (solution B) and used for free PA extraction according to Marina et al. (2008). For apoplastic Na + extraction, plastic tubes were centrifuged for 1 min at 1000 g to discard the washing solution and centrifuged again for 10 min at 2000 g to collect the apoplastic fluid. All centrifugations were done at 4°C and the apoplastic fluids obtained were kept at -20°C. Thirty segments from 30 leaves were used for free aPA and aDap extractions, and 120 segments from 24 leaves for Na + extraction. In addition, 30 leaf blade segments from 10 unsalinized plants were pooled, washed, transferred to net bags, and gently vacuum infiltrated for 2 min with water or 100 mM NaCl. The resulting apoplastic fluid was used to determine the apoplastic peroxidase (POX) levels and to check for variations in free aPA contents derived from the presence of salt in the apoplastic environment. Glucose 6-phosphate dehydrogenase activity, a marker of cytosolic contamination, was determined in every apoplastic fluid fraction according to Rodríguez et al. (2002).

Determination of free PA and Dap
Maize free PA were determined according to Jiménez-Bremont et al. (2007). For dansylation, 200 ll of solution A or B (see above) were added to 10 ll of 0.1 mM heptanodiamine (internal standard, ICN) plus 200 ll saturated Na 2 CO 3 and 400 ll dansyl chloride-acetone 1% (w/v). After 16 h at 25°C in the dark, 100 ll of proline 100% (w/v) was added to stop the reaction. Dansyl-derived PA were extracted with 500 ll toluene. The organic phase (400 ll) was evaporated under vacuum and resuspended in 400 (cPA) or 200 (aPA and aDap) ll acetonitrile. Dansylderived PA were separated by HPLC (ISCO 2350, ISCO Inc, Lincoln, NE) with a reverse phase column Sephasil C18 (Amersham Pharmacia) and detected with a spectrofluorometer (Variant Fluorichrom). The solvent mix was obtained with a gradient programmer ISCO 2360, with a flow of 1.5 ml min À1 as follows: 0-4.5 min, acetonitrile:H 2 O 70:30 v/v; 4.5-9 min, acetonitrile 100; 9-15 min, acetonitrile:H 2 O 70:30 v/v). Peak areas were integrated, normalized to heptanodiamine and interpolated into a PA standards calibration curve.

POX enzyme level
The reaction mixture (1 ml) contained 15 ll of apoplastic fluid, 20 ll 0.02 M guaiacol, and 0.1 M potassium phosphate pH 6.4. The reaction was started by adding 35 ll 88 mM H 2 O 2 and activity was measured as an increase in A 560 after 30 s with a spectrophotometer (Beckman DU Series 600, Beckman Instruments, Fullerton, CA). The specific activity calculation was based on the protein content of each sample, determined according to Bradford (1976).

Determination of the apoplastic Na + content
Apoplastic Na + concentration was determined by atomic emission spectrophotometry analysis of the apoplastic fluid fraction, using a Perkin-Elmer AA 100 spectrophotometer in emission mode.

H 2 O 2 production by amine oxidase activity
The AO activity level was determined according to Cona et al. (2006b) with some modifications. Segments were washed in water (control) or 100 mM NaCl (salinized) for 6 min in order to remove symplastic contamination. For in vivo measurements, pools of five segments were introduced in 1 ml solutions containing 100 lM 4-aminoantipyrine (4-AAP), 1 mM 3,5-dichloro-2-hydroxybenzenesulphonic acid (DCHBS), 20 mM potassium phosphate pH 6.5, plus or minus 0.5 mM Spm or Put, and 100 mM NaCl for the saline treatment. Segments were subsequently infiltrated for 2 min and further incubated for 5 h at room temperature. Then 1 ml of the incubation medium was collected and the resultant pink adduct was measured at A 515 with a spectrophotometer (HITACHI U-2000, Hitachi, Tokyo, Japan) and transformed into an H 2 O 2 molar concentration with a molar extinction coefficient at 515 nm (2.6310 4 M À1 cm À1 ). PAO and CuAO activities were calculated as the difference in H 2 O 2 produced between treatments containing and lacking substrate.

Extraction of apoplastic PAO
Extraction of apoplastic proteins was performed as described by Li (1993) with slight modifications . Plant material (40 g) was cut in 2 mm pieces, washed in distilled water and vacuum-infiltrated with 100 ml 5 mM potassium phosphate pH 6.5 added with 200 mM NaCl. The vacuum was broken and re-established every 5 min, for three successive times. The apoplastic fluid was collected, cooled at 4°C and added with 1 vol. of pre-cooled (-20°C) Me 2 CO. The resulting solution was incubated at 4°C, for 30 min and centrifuged at 15 000 g for 15 min. The supernatant was discarded and the pellet resuspended in 20 mM bis-tris-propane buffer pH 6.5 and applied to a DEAE-Sephacell column (13 2 cm) equilibrated with the same solution. The eluted solution (Solution D) was kept at 4°C until used.

In vitro PAO activity assay
For in vitro PAO activity measurement, pools of 20 segments were frozen with liquid N 2 , homogenized in 1 ml of 0.1 M potassium phosphate pH 6.5 at 4°C, and centrifuged at 15 000 g for 15 min. The pellet was discarded and the homogenate was kept at 4°C (sSolution C). PAO activity was determined according to Cona et al. (2006b). Previous tests determined that the optimal pH for PAO activities was 6.5. The reaction mixture contained 1 ml, 50 ll solution C or 80 ll D (see above) plus 100 lM 4-AAP, 1 mM DCHBS, 0.06 mg ml À1 horseradish POX, and 100 mM potassium phosphate pH 6.5. The mixture was incubated at 30°C for 2 min. The reaction was started by adding 5 ll of 10 mM Spm and the activity was measured for 1 min at A 515 with a spectrophotometer and transformed into an H 2 O 2 molar concentration. PAO K i for SL-11061 was calculated (Lineweaver and Burk, 1934). The K m value obtained for this preparation was K m ¼17.7 lM.
Extraction and purification of plasma membrane for NOX activity determination Plasma membrane was prepared according to Larsson (1985) with some variations. Leaves (70 g) from 7-d-old plants were homogenized with an omnimixer by giving three 20 s pulses at full speed. The extraction solution (200 ml) contained 50 mM TRIS-HCl pH 7.5, 0.33 M sucrose, 1 mM EDTA, 0.1 mM MgCl 2 , 1 mM ascorbate, 1 mM DTT, 1 mM phenylmethylsulphonyl fluoride, and 0.6% (w/v) polyvinylpoly pyrrolidone. The homogenate was filtered through four layers of cheesecloth, and the filtrate centrifuged at 10 000 g for 10 min. Microsomes were pelleted from the supernatant by centrifugation at 140 000 g for 45 min and resuspended in 10 ml 5 mM potassium phosphate pH 7.8 containing 0.33 M sucrose and 3 mM KCl. The suspension was fractionated by the aqueous two-phase partitioning method (Larsson, 1985). Phase separations were carried out in a series of 10 g phase systems with a final composition of 6.2% (w/w) dextran T500 (Sigma), 6.2% (w/ w) polyethylene glycol 3350 (Sigma), 0.33 M sucrose, 5 mM potassium phosphate pH 7.8, and 3 mM KCl. Three successive partitioning rounds yielded an upper phase (U3) and a lower phase (L3). U3 was 3-fold diluted in 10 mM TRIS-HCl buffer (pH 7.4) containing 0.33 M sucrose. The solution was centrifuged at 140 000 g for 60 min and the resulting pellet resuspended in 2 ml 10 mM TRIS-HCl buffer pH 7.4 and 0.33 M sucrose. All procedures were carried out at 4°C. The enrichment in plasma membranes of the upper phase was monitored by the percentage of V-ATPase inhibition (Serrano, 1978). U3 was enriched in plasma membrane up to 90%. U3 was kept at 4°C and used for enzyme activity immediately.

H 2 O 2 production by NADH-dependent POX
Reactions were carried out in 0.1 M potassium phosphate pH 4.5 containing 3 lg ml À1 horseradish peroxidase and 0.2 mM NADH (Frahry and Schopfer, 1998) with some modifications. Reactions were initiated by adding NADH and, 5 min later, aliquots of 500 ll were removed from the reaction mixture. NADH was eliminated with 0.1 M HCl followed by 0.1 M NaOH. H 2 O 2 was measured by fluorescence of 55 lM homovanillic acid at 407 nm (EM) and 305 nm (EX), in the presence of 12 lg ml À1 horseradish peroxidase and 0.2 M potassium phosphate pH 4.5 in a final 1 ml volume. The calibration curve was linear in the range of 0.5-3 lM H 2 O 2 . The calibration curve was not affected by 100 lM or 200 lM SL-11061, 50 lM or 200 lM DPI, and 100 lM or 200 mM NaCl.

Detection of ÁO À
2 accumulation in the whole leaf ÁO À 2 accumulation was detected by blue formazan precipitation (Hernández et al., 2001). For this purpose, 0.01% (w/v) nitro-blue tetrazolium (NBT) was added with nutrient solution to control and salinized 13-d-old plants. One day later, plants were harvested and the third leaf was boiled in 80% (v/v) ethanol for 10 min. Leaves were mounted on a glass slide and scanned.
In vivo HOÁ production HOÁ release to the medium was determined by the hydroxylation of sodium benzoate (BZ) by HOÁ. Hydroxyl BZ was detected by spectrofluorometry according to Schopfer et al. (2001) with modifications. Pools of six salinized segments were gently infiltrated and incubated for 7 h in 1 ml of aqueous solutions containing 2.5 mM BZ and 100 mM NaCl in the presence or absence of 100 lM SL-11061 at 30°C in the dark. Fluorescence was determined at 407 nm emission after excitation at 305 nm in a spectrofluorometer (Bio-Tek Kontron SFM 25, Kontron Instruments, Zü rich, Switzerland).

Effect of NaCl on elongation and apoplastic Na + concentration and free PA levels in leaf segments
Previous results showed that elongation in unsalinized and salt-treated leaves is maximal at the second blade EZ segment, that is, the region spanning 10-20 mm from the ligule (Fig. 1A). Elongation of excised second blade segments from salt-treated plants incubated in 100 mM NaCl was 50% compared with unsalinized segments (Fig. 1B), confirming the previous results by Rodríguez et al. (2004). Atomic emission spectrophotometry (AES) analysis of the segment apoplastic fluids revealed a 76.462.4 and 1.460.5 mM Na + content in salinized and unsalinized leaf blades, respectively (no cytosolic contamination was detected in the apoplastic fluid; see Supplementary Table S1 at JXB online). Therefore, as the saline content of the incubation solution was comparable with that existing in the apoplast of salinized leaves, it was decided to add 100 mM NaCl to the incubation mixture in the next in vivo experiments, in the case of salinized plants, as a means to avoid changes in the osmotic potential of the apoplastic environment. Free PA levels were measured by HPLC. Salinity lowered Put and increased Spd and Spm level of the cell extracts ( Fig.  2A-C), whereas it greatly increased Spm and Spd and slightly increased Put in the apoplast (Fig. 2D-F), suggesting a role for PA accumulation in the elongation zone of the maize leaf blade under saline stress.

Evaluation of the mechanisms involved in aPA increment in salinized leaves
It has been hypothesized that after being synthesized in the cytoplasm, PA cross the plasma membrane towards the apoplast, where they are catabolized by AOs. Thus, the observed rise in free aPA in salinized plants is probably the result of: (i) promotion of PA passage towards the apoplast or (ii) a decrease in the amount of AOs enzymes (or AOs activities) leading to free aPA accumulation. Given that the mechanisms of PA passage to the apoplast are unknown, it was decided to assess the last possibility. For this, the effect of salt addition on maximal in vivo and in vitro AO activity was determined. In this approach, H 2 O 2 production by PA oxidation was estimated through an oxidative POXdependent reaction that produces a pink adduct measurable by spectrophotometry (Cona et al., 2006b). Blade segments from salinized and unsalinized plants were infiltrated and incubated in the reaction mix with the addition of 0.5 mM of exogenous substrates Put and Spm, for CuAO and PAO determination, respectively (POX addition to the reaction mixture was not necessary since no variation was observed in the apoplastic oxidative POX activity between treatments, see Supplementary Fig. S1 at JXB online). As result, it was observed that (i) there were no differences in the maximal CuAO and PAO activities (achieved under saturating substrate conditions) due to salt treatment (Fig. 3A), (ii) PAO activity levels were up to 20-fold higher than those of CuAO (therefore further studies will be performed only on PAO activity). A second in vitro analysis using cell-free extracts from segment homogenates (and saturating substrate conditions) confirmed former results on PAO activity levels (Fig. 3B). Taken together, these results led us to: (i) reject the possibility of a negative saline effect on the total activities of AOs enzymes and (ii) to assume that salinity somehow promoted PA passage from the symplastic compartment towards the apoplast. Having in mind that aPA may interact with cell wall components, a third possibility is that the presence of NaCl in the apoplast causes the dissociation of pre-existent aPA from the cell wall. Blade segments from unsalinized plants were then infiltrated either with water or with 100 mM NaCl, resulting in the absence of any effect of this salt on free aPA contents (see Supplementary Table S2 at JXB online). Thus, any dissociating action of NaCl on aPA putatively associated to cell wall components of the apoplast was ruled out.

NaCl increases inherent PAO activity
So far, it has been demonstrated that aPA levels increased as a result of plant salinization (Fig. 2). This result encouraged us to examine whether the inherent PAO activity, which depends on the concentration of its endogenous substrate (aPA), correlates with that result. Therefore, PAO activity was evaluated by measuring the Dap content using HPLC. Dap is a product of the Spd and Spm oxidation, formed in the same molar quantities as H 2 O 2 (Cona et al., 2006a). The measurement of aDap levels indicated that salinity led to increased PAO activity in the leaf blade region under study (Fig. 4A). Alternatively, PAO activity in vivo was determined as previously described (Fig. 3A), without the exogenous Spm supplement. For this purpose, H 2 O 2 levels were measured in the presence or absence of 1,19-bis-(ethylamine)-5,10,15 triazanonadecane (SL-11061), a tobacco PAO inhibitor , which has also been found to inhibit oat PAO in vivo . The results showed that SL-11061 had no effect on apoplastic H 2 O 2 content in the absence of NaCl (Fig. 4B), showing that the contribution of PAO activity to the total apoplastic H 2 O 2 level was negligible under control conditions. Conversely, a 50% lowered H 2 O 2 content was found in SL-11061-treated segments under saline conditions. As a whole, these results showed that, under salt stress conditions, the contribution of PAO to the observed apoplastic H 2 O 2 pool in the elongation zone of the maize leaf blade is relevant, whereas in the absence of salt treatment, the formation of the main apoplastic H 2 O 2 would rely on mechanisms different from aPA oxidation.
On the other hand, under in vitro conditions NaCl did not affect PAO activity at all (Table 1), reinforcing that this enzyme keeps its activity in salinity. In addition, a low K i Fig. 3. PAO and CuAO activities under substrate saturating conditions. (A) In vivo AOs activities were determined according to Cona et al. (2006b). Pools of five segments were introduced in 1 ml of solution containing 100 lM 4-AAP, 1 mM DCHBS, 20 mM potassium phosphate pH 6.5 with and without 0.5 mM Spm or Put, and 100 mM NaCl for salinized segments. Segments were subsequently infiltrated for 5 min, incubated for 5 h and AOs activities determined by pink adduct production at A 515 . Data were transformed into H 2 O 2 molar concentrations with a molar extinction coefficient at 515 nm (2.6310 4 M À1 cm À1 ). PAO and CuAO levels were calculated as the difference in H 2 O 2 amounts between treatments with and without substrate. (B) In vitro PAO measurement. Pools of 20 blade segments were homogenized in 1 ml of 0.1 mM potassium phosphate pH 6.5 at 4°C, and centrifuged at 15 000 g for 15 min. PAO activity was determined according to Cona et al. (2006b). The experiment was conducted twice, yielding similar results. Abbreviations: c, unsalinized; s, salt-treated. Results are means 6SE (n¼6).  value (8.7310 À7 M) was found for the polyamine analogue SL-11061, indicating its high efficiency as a maize PAO inhibitor. This K i is comparable to that found for oat PAO (K i ¼1.5310 À9 M; Maiale et al., 2008).
PAO activity provides ÁO À 2 and HOÁ radicals under saline stress When leaves of salt-treated plants were stained with NBT, they showed a strong decrease in precipitate intensity as compared with control leaves (Fig. 5), demonstrating a reduction of the ÁO À 2 level due to the saline treatment. However, a certain amount of ÁO À 2 was still observable in these conditions. The fact that NaCl substantially decreases NOX activity, the main source of apoplastic ÁO À 2 and H 2 O 2 in non-stressed maize plants Rodríguez et al., 2007) as well as H 2 O 2 production by NADHdependent POX (Table 1), suggests the occurrence of some salt-tolerant mechanism for aROS production, alternative to POX and NOX. Interestingly, it has been proposed that AO activity is involved in the production of extracellular ÁO À 2 and HOÁ radicals (Kawano et al., 2000a). In order to test whether this process takes place in vivo in the apoplast maize leaf under saline condition, ÁO À 2 formation was determined by incubating salt-treated blade segments in XTT solution, in the presence or absence of SL-11061 (Fig.  6A). Our results showed that the addition of the PAO inhibitor resulted in a highly diminished ÁO À 2 formation. These results (in addition to the fact that SL-11061 does not scavenge ÁO À 2 radicals produced by NOX; see Table 1) support the idea that PAO activity indirectly produces most of the ÁO À 2 in this zone under saline conditions. Similarly, when HOÁ was determined by the BZ method , PAO activity represented around 70% of the production of that free radical in the salt-treated segments (Fig. 6B). These results reinforced the hypothesis that, under saline conditions, PA oxidation by PAO would be the main source for aROS production in the elongation zone of maize leaf blades. Table 1. Effects of inhibitors on apoplastic ROS-producing enzymes PAO was extracted and purified from apoplastic fluids of leaf blade segments according to Li (1993) with some modifications. PAO activity was determined according to Cona et al. (2006b). NOX was extracted and purified according to Larsson (1985) with some variations. NOX activity was determined according to Sagi and Fluhr (2001). H 2 O 2 production by NADH-dependent POX was determined according Frahry and Schopfer (1998) 5. ÁO À 2 production in the whole leaf blade. ÁO À 2 was detected by formazan precipitation. Control and salt-treated plants were incubated in the presence of 0.01% (w/v) NBT for 24 h. Plants were harvested and the third leaf was boiled in 80% (v/v) ethanol, mounted on a glass slide, and scanned. Abbreviations: c, unsalinized; s, salt-treated. Fig. 6. In vivo PAO activity-derived ÁO À 2 and HOÁ production under salinity. (A) ÁO À 2 was detected by XTT according to Frahry and Schopfer (2001). Pools of eight salinized second blade segments were gently infiltrated and incubated for 7 h in the dark in 1 ml of aqueous solutions containing 0.5 mM XTT and 100 mM NaCl in the presence or absence of 100 lM SL-11061. The incubation medium (1 ml) was centrifuged at 10 000 g and the supernatant was subjected to measurement with a spectrophotometer at A 470 . Data were transformed into ÁO À 2 molar extinction coefficient at 470 nm (2.16310 4 M À1 cm À1 ). (B) HOÁ was detected by hydroxylation of BZ according to Schopfer et al. (2001). Pools of six salinized segments were gently infiltrated and incubated for 7 h in 1 ml of aqueous solutions containing 2.5 mM BZ and 100 mM NaCl plus or minus 100 lM SL-11061 at 30°C in the dark. Fluorescence was determined in a spectrofluorometer (EX: 305 nm EM: 407 nm). Experiment was conducted twice, yielding similar results. Results are means 6SE (n¼4). Asterisks indicate a difference from the control (P <0.05).
In order to gain further insight into the involvement of higher polyamines catabolism by PAO on aROS generation, the effect was examined of adding different PAO modulators to the incubation buffer, on ÁO À 2 production in the elongation zone of salt-treated plants, detected by XTT (Fig. 7). The addition of 100 or 200 lM Spm increased the in vivo ÁO À 2 production, whereas treatment with SL-11061 (without Spm) or 1,8-DO, a commercial competitive PAO inhibitor (K i ¼3310 À7 M; Cona et al., 2004), either separately or in combination with Spm showed the opposite effect (Fig. 7). As expected, Dap addition did not change the ÁO À 2 levels. Similarly, ÁO À 2 production in segments treated with diphenylene iodonium (DPI), reported as a NOX  and NADH-dependent POX inhibitor (Frahry and Schopfer, 1998) was similar to that of the control treatment, evidence that, under saline conditions, the enzymes mentioned were inhibited. Such in vivo inhibition was in line with that observed in the in vitro assay (Table 1) and in previous results concerning NOX . Based on these results, it is concluded that PAO is the main contributor to apoplastic ÁO À 2 production in salinized maize leaves.
Possible source for the observed ÁO À 2 and HOÁ H 2 O 2 may be consumed to generate ÁO À 2 and HOÁ through the Haber-Weiss reaction (Haber and Weiss, 1932) and POX activity in the Fenton-like reaction Liszkay et al., 2004;Carol and Dolan, 2006). Therefore, the possibility that, under saline conditions, these free radicals originate in a reaction from the H 2 O 2 produced by PAO (which remains active under salinity), through a chain reaction catalysed by Fe 2+ or Cu + (Fry, 1998;Kawano et al., 2000b) was tested. Salinized segments were treated with Spm plus the Fe 2+ -specific chelator FZ (Kosegarten et al., 1999) or the Cu + -specific chelator NC (Kunapuli and Vaidyanathan, 1983). Results demonstrated that chelators significantly decreased ÁO À 2 production, indicating a probable involvement of a Fenton-Haber-Weisslike reaction in this process.

Effect of PAO and ROS modulators on segment elongation of salinized plants
So far it has been shown that aPA oxidation is mostly responsible for the presence of certain aROS amounts in leaf blades of salinized maize plants. However, since the amount of aROS (H 2 O 2 and H 2 O 2 -derived ÁO À 2 and HOÁ) produced in leaf segments is much lower in the presence of salt, compared with that of the control without NaCl, the question remains as to whether the observed amounts of these aROS may still contribute to leaf elongation. To answer this question, the effect of diverse PAO and ROS modulators on segment elongation of salinized plants was tested (Fig. 8). Incubation with 200 lM Spm doubled segment elongation, whereas the addition of SL-11061 attenuated and 1,8-DO reversed the last effect. Interestingly, Fig. 7. In vivo effect of PAO activity modulators on ÁO À 2 production under salinity. Salinized second blade segments were used to detect ÁO À 2 by XTT. Pools of eight segments were gently infiltrated and incubated for 7 h in the dark in 1 ml of aqueous solutions containing 0.5 mM XTT, 100 mM NaCl, and modulators of PAO activity. When indicated, the following concentrations were used: 200 lM SL-11061, 200 lM 1,8-DO, 100-800 lM Spm, 200 lM DPI, 200 lM Dap, 1 mM FZ, or 1 mM NC. Incubation medium (1 ml) was centrifuged at 10 000 g and the supernatant subjected to measurement with a spectrophotometer at A 470 . Data were transformed into ÁO À 2 molar extinction coefficient 470 nm (2.16310 4 M À1 cm À1 ). The experiment was performed twice, yielding similar results. Results are mean 6SE (n¼8). Bars with the same letter are not significantly different (P <0.05). Fig. 8. Effects of PAO activity modulators, ROS scavengers, and a Ca 2+ chelator on segment elongation. Second blade segments from salinized plants were incubated for 7 h in the dark with 100 mM NaCl. When indicated, the following concentrations were used: 200 lM SL-11061, 200 lM 1,8-DO, 50-800 lM Spm, 10 mM MnCl 2 , 5 mM BZ, and 10 mM EGTA. Segments were scanned before and after the incubation period and their length measured with an image processing software. The results are the percentage of length increase, with respect to the control during a 7 h incubation period. Absolute growth rate for control salinized second blade segments was 0.22160.008 mm h À1 . The experiment was conducted twice, yielding similar results. Results are means 6SE (n¼20). Bars with the same letter are not significantly different (P <0.05). 800 lM Spm had no effect on segment elongation. In the absence of exogenous Spm addition, both PAO inhibitors diminished segment length, as compared with the control. Moreover, when 1,8-DO was added to the plant nutrient solution from the beginning of salt treatment, it produced reduced growth and Dap accumulation, as well as increased Spd and Spm contents of the entire maize leaf (Table 2), compared with plants not treated with the inhibitor. The addition of MnCl 2 , a ÁO À 2 dismutating agent (Hernández et al., 2001) or BZ along with Spm, nullified the elongating effect of the polyamine on salinized segments. Finally, the incorporation of the specific Ca 2+ chelating agent ethylene glycol bis (b-aminoethylether)-N, N, N#, N#-tetra-acetic acid (EGTA), reduced segments length, even in the presence of 200 lM Spm.

Discussion
Evidence has accumulated over recent decades demonstrating that polyamines play an important role in many plant developmental processes (Evans and Malmberg, 1989;Walden et al., 1997) and in plant responses to salinity and other abiotic stress conditions in diverse plant species (Krishnamurthy and Bhagwat, 1989;Galston and Sawhney, 1990;Aziz et al., 1998;Bouchereau et al., 1999;Simon-Sarkadi et al., 2002;Sanchez et al., 2005;Kusano et al., 2007). Several biotechnological approaches like overexpressing PA-synthesizing enzymes (Kumria and Rajam, 2002;Capell et al., 2004;Kasukabe et al., 2004;Wi et al., 2006) or antisense and mutant generation (Kasinathan and Wingler, 2004;Yamaguchi et al., 2006), allowed the generation of plants with increased and decreased stress tolerance, respectively. Recent studies using transgenic plants overexpressing or downregulating apoplastic polyamine oxidase, revealed the importance of the H 2 O 2 derived from aPA catabolism in the induction of either salinityinduced tolerance or programmed cell death in tobacco (Moschou et al., 2008). The present work was focused on the possible involvement of the ROS produced from aPA oxidation in leaf growth processes of maize plants, grown under saline conditions. The first results showed that salt treatment reduced elongation in the region spanning 10-20 mm from the leaf ligule and, in parallel, it provoked a remarkable increment of higher apoplastic polyamines concentration in that region. This observation is in agreement with recent results showing that Spd, which is synthesized in the cytoplasm, is secreted into the apoplast upon salt treatment in tobacco (Moschou et al., 2008). In the present work, any other possibility was ruled out in order to confirm that salinity stimulates the passage of these substrates from the cytoplasm to the apoplast in maize plants. Similarly, the salt-induced decrease observed in Put levels, concomitant with the increase of total Spm levels of cell-free extracts in the segments, are consistent with the Spm accumulation described by other authors in salinized rice (Maiale et al., 2004), several vegetables (Zapata et al., 2004), Lotus glaber (Sannazzaro et al., 2007), and maize (Jiménez-Bremont et al., 2007).
In vivo and in vitro measurements of H 2 O 2 levels in the presence of saturating substrate conditions revealed the maximum feasible AO activity and showed, on the one hand, that PAO was the main enzyme contributing to the total PA oxidation level in maize leaves. Consequently, further studies were performed only on PAO activity, leaving CuAO activity aside. Biochemical, histochemical, and immunocytochemical studies allowed the localization of PAO, showing that it is specially abundant in the primary and secondary cell walls of xylem parenchyma, the endodermis, and epidermis of maize seedlings (Cona et al., 2006a). On the other hand, it was shown that salt treatment does not affect maximal PAO activity, suggesting that the enzyme is tolerant to this stress. Furthermore, the results obtained without exogenous substrate (that is to say, based on the actual polyamine cell content in the tissue) via the detection of the PAO product, Dap (Fig.  4A), demonstrated that salinity enhanced PAO activity. The last result consistently reflected the high Spd and Spm levels in that region ( Fig. 2D-F) and suggested that the observed increase of inherent PAO activity under saline stress was a consequence of the rise in its substrate. These results are in line with in vitro results obtained by Smith (1977). Interestingly, the fact that the aDap amount was two and three orders higher than those of Spd and Spm, respectively, suggests that higher PA were actively oxidized to Dap and H 2 O 2 in the apoplast, once they crossed the plasma membrane.
Although out of the scope of the present work, the possibility that polyamine metabolism in the root (the first organ sensing salinity) behaves upon salt treatment in similar manner as the leaf blade is intriguing. As far as we know, the information regarding root PAO activity and salinity is limited to one report by Zhao et al. (2003), who reported that 0-200 mM NaCl increased Put, Spd, and PAO activity in the roots of barley seedlings. Unfortunately, this information was not discussed in terms of root growth or elongation. Evidence of reduced ÁO À 2 amounts in salt-treated leaves by NBT staining (Fig. 5), revealed the occurrence of some mechanism for its production, alternative to that of NOX and NADH-dependent POX activities, which (unlike PAO) was strongly inhibited by salinity. The remarkable increase in in vivo ÁO À 2 production by Spm addition, along with the substantially lowered in vivo ÁO À 2 and HOÁ generation in the salt-treated segments by both PAO inhibitors or the Fe 2+ and Cu + -specific chelators FZ and NC (Figs 6, 7), support the notion that ÁO À 2 and HOÁ generation could occur from H 2 O 2 production through PA oxidation and a further reaction catalysed by Fe 2+ or Cu + , such as a Fenton-Haber-Weiss chain reaction (Kawano et al., 2000a, b).
It is noteworthy that in vitro, cadaverine, putrescine, spermidine, and spermine do not scavenge superoxide radicals, but were found to be scavengers of hydroxyl radicals (Das and Misra, 2004) and unpublished results from our group have confirmed those results. However, such a ROS-scavenger effect was observed only when polyamines were used in concentrations of 0.5 mM or higher. As in the present work, polyamine concentration has been always much lower than that amount, we may discard any ROS-scavenging effect in our results.
AO activity has formerly been related either to cell elongation in roots and hypocotyls of soybean seedlings (Delis et al., 2006) or to cell wall maturation in tobacco (Paschalidis and Roubelakis-Angelakis, 2005;Cona et al., 2006a). The purpose of this work was to evaluate the possible involvement of ROS production by PA oxidation in the leaf growth of maize plants grown under saline stress conditions. Our results suggest that tetramine oxidation contributes 25-30% of segment elongation under salinity (Fig. 8). Furthermore, when applied systemically along with NaCl, the PAO inhibitor 1,8-DO caused a 25% reduction in the elongation of whole leaves, compared with the control treatment without the inhibitor. Bearing in mind that the blade region analysed has 90% of the leaf elongation, it is deduced that, under a salt-stress situation, the minor contribution of PAO activity could still mean a significant yield improvement from an agronomical viewpoint. As a whole, these facts generate the expectation that biotechnological approaches like overexpressing enzymes responsible for PA biosynthesis or catabolism may be used to overcome reductions in the productivity of maize plants caused by salinity. In turn, the elimination of Spmstimulated elongation by the specific Ca 2+ chelating agent EGTA (Fig. 8), suggests that such elongation could be mediated by the activation of non-selective cation channels (NSCCs), through the HOÁ produced by Spm oxidation (a possibility that should be addressed in future research). This proposal is supported by the bulk of the evidence that has appeared during the last decade, which showed transient increases in cytosolic Ca 2+ ([Ca 2+ ] cyt ) as a second messenger, suggesting that there are ROS/[Ca 2+ ] cyt signalling pathways in several developmental processes. For example, guard cells and stomatal closure has been reported in Commelina communis and A. thaliana (McAinsh et al., 1996;Pei et al., 2000), as well as (2)-catechin-induced ROS production followed by ROS-induced Ca 2+ increases in Centaurea diffusa and Arabidopsis thaliana roots (Bais et al., 2003) or the growth stimulation of A. thaliana roots (Foreman et al., 2003) and pollen tubes (Demidchik and Maathuis, 2007) by the aROS activation of Ca 2+ -permeable NSCCs that induce inward Ca 2+ currents. Recently, it was shown that the lack of Spm in the Arabidopsis acl5/spms mutants caused hypersensitivity to NaCl, possibly due to impaired Ca 2+ -homeostasis (Yamaguchi et al., 2006) and that H 2 O 2 generated by CuAO activates NSCCs in the abscisic acid-induced stomatal closure process in Vicia faba .
ROS could also act on growth through a promotion of cell wall polysaccharide cleavage in vivo (Schopfer, 2001), such as that shown to operate in vitro (Miller, 1986;Fry, 1998;Schweikert et al., 2000). However, the action of ROS in the apoplast should be viewed as a delicate balance between cleavage and cross-linking activities (Cosgrove, 1999). An increase in PAO immunolabelling was observed inside secretory cytoplasmic organelles, suggesting the need for the intraprotoplasmic production of H 2 O 2 for polymer cross-linking in the secretory pathway (Fry et al., 2000;Cona et al., 2003). Also, the balance between cleavage and cross-linking activities may be associated with a differential activity of cell wall peroxidases because different soluble peroxidase isozymes characterize the expanding and expanded regions in maize leaves (de Souza and MacAdam, 2001) and in Festuca arundinacea (MacAdam et al., 1992).
The effectiveness of SL-11061 as inhibitor towards PA oxidation was formerly demonstrated in vivo, in an experiment using leaf blade segments in the presence of Spd . In the present work, the inhibitory effect of DPI on NOX and NADH-dependent POX (Table  1) reported previously is also confirmed: 50 lM DPI inhibited NOX activity by 77%  and the H 2 O 2 -producing activity by NADH-dependent POX by 94% . In addition, strong NOX inhibition by DPI is in congruence with the K i ¼5.6310 À6 M reported by O'Donnell et al. (1993). Reductions in ÁO À 2 production and elongation of salinized segments treated with SL-11061 in vivo (Figs 7,8), in addition to the fact that this polyamine analogue is efficient as a PAO but not as a NOX or NADH-dependent POX in vitro inhibitor (Table  1), demonstrated that PAO is not repressed by salinity. These facts also supported the use of this inhibitor to distinguish the PAO contribution to aROS production from that of the other two enzymes, under saline conditions. On the other hand, DPI treatment did not diminish either ÁO À 2 production or elongation of salinized segments (Figs 7, 8), in agreement with in vitro observations (Table 1), showing that NOX and NADH-dependent POX activities are inhibited in vivo by salinity.
Apoplastic Na + concentration varies among and within plant species. Under salt treatment, apoplastic ion concentrations of 164 mM and 56 mM were reported in pea and spinach, respectively (Speer and Kaiser, 1991), whereas it approached 600 mM in salt-stressed rice plants (Flowers et al., 1991). Dissimilar apoplastic Na + contents have been reported in salt-stressed maize leaves: 4-5 mM (Lohaus et al., 2000), 25 mM (Neves-Piestun and Bernstein, 2001), and 76 mM (own results). Compared with earlier reports, the higher apoplastic Na + content registered in the present work may be ascribed to a more concentrated NaCl solution used for salinization (150 versus 100 or 80 mM) or to variations in other experimental conditions. Such a diversity of results on apoplastic Na + contents highlights the importance of having assessed the actual apoplastic ion content in the salinized plant material under study in order to set a realistic experimental condition.
Finally, variations in the effect of different Spm concentrations on ÁO À 2 production and segment elongation (Figs 7, 8) gave evidence of a concentration-dependent Spm effect on ROS production and segment elongation. Unfortunately, previous reports describing changes in apoplastic PA levels have not measured cell elongation (Yoda et al., 2003;Angelini et al., 2008;Marina et al, 2008;Moschou et al., 2008).
Taken together, our results demonstrated that, under saline stress, PAO might still provide the necessary H 2 O 2 to generate ÁO À 2 through an increased substrate availability and thus sustain leaf elongation. These results allowed us to propose the model depicted in Fig. 9. Thus, in the scenario where NOX is inhibited by non-lethal NaCl stress and the ROS produced by this enzyme is substantially reduced, oxidation by PAO of Spm and Spd accumulated in the apoplast of the EZ would result in an alternative source to generate ROS, partially counteracting the growth-inhibiting effect caused by salinity.

Supplementary data
Supplementary data are available at JXB online.
Supplementary Table S2. Effect of the NaCl presence in the infiltration solution on the level of extracted apoplastic PA.