• Hydrogen peroxide (H2O2), ethylene, 1-aminocyclopropane-1-carboxylic acid (ACC) and salicylic acid (SA) concentrations and ACC synthase (ACS) gene expression were measured to establish whether the high sensitivity of the Populus deltoides × maximowiczii clone Eridano to ozone (O3) exposure, compared with the O3-resistant Populus deltoides × euramericana clone I-214, is attributable to differences in the modulation of signal transduction pathways.
• In a time-course experiment, Populus deltoides (poplar) clones were exposed to acute fumigation with 150 nl l−1 O3 for 5 h.
• The two poplar clones showed differences in ethylene evolution, I-214 displaying earlier and less pronounced ethylene emission than Eridano. In both clones, ethylene evolution was accompanied by increased ACS transcript levels and enhanced emission of free ACC. I-214 exhibited a greater basal concentration of free SA and a lower concentration of the conjugated pool. However, a slight accumulation of free SA at the end of the 5-h exposure was found only in Eridano, together with an earlier minimal increase in the concentration of conjugated SA.
• The results show that both clones react to O3 by producing H2O2, ethylene and SA, but the difference in sensitivity to the pollutant is probably attributable to differences in the kinetics and magnitude of this response.
Tropospheric ozone (O3) is believed to be the most important phytotoxic air pollutant in industrialized countries and one of the major anthropogenic stresses contributing to forest decline, the dramatic phenomenon affecting forest areas in Europe and North America (Schmeiden & Wild, 1995). During the past 40 years, the concentration of O3 has increased 2- to 5-fold (Kley et al., 1999) and its ambient concentration is still rising, with an annual increment of 1.6% (Marenco et al., 1994; Inclán et al., 1999). At present, in many regions of Europe, the critical concentrations established for sensitive plants are greatly exceeded. Recent results from the Italian CONECOFOR programme have shown that AOT40 values (sum of the hourly mean concentrations above 40 ppb) are around 5–10 ppm h in the Alps, usually higher than the critical concentrations for forests across the Italian peninsula (10 ppm h), and very high in the south of Italy (26 ppm h) (Gerosa et al., 2003).
Although ROS, and hydrogen peroxide (H2O2) in particular, are acknowledged to be important signal molecules, their concentration needs to be carefully tuned to avoid uncontrolled oxidative damage. H2O2 itself is only slightly toxic, except at very high concentrations, but can be decomposed by transition metals to form extremely reactive hydroxyl radicals (in the Fenton reaction). Thus, because lower H2O2 concentrations than those required to trigger cell death seem to be sufficient to activate defensive genes (Levine et al., 1994), the ability to achieve optimal H2O2 concentrations is of major importance in determining the fate of the plant cell. Of the different H2O2 scavengers, ascorbate peroxidase (APX) is recognized as one of the most efficient, because of its high affinity for H2O2 and its presence in different subcellular compartments (Foyer et al., 1994). In addition, H2O2 may be reduced by phenol-dependent peroxidase (POD) which catalyses the polymerization of lignin precursors and the cross-linking between cell wall proteins and polysaccharides.
Although ROS are recognized as key factors in determining the fate of a cell, recent studies underline the importance of the complex interactions between ROS and several signal molecules such as ethylene, salicylic acid (SA) and jasmonic acid (JA) in modulating the plant response to the pollutant (Koch et al., 2000; Overmyer et al., 2000; Rao et al., 2000, 2002). The interplay amongst these signal molecules probably involves ROS as a primary messenger acting upstream of the other signal transducers, although different and even opposite causal and temporal relationships amongst the various signalling routes have been reported in different species in response to different kinds of stress (Chamnongpol et al., 1993; Rao et al., 2000, 2002; Moeder et al., 2002; Vahala et al., 2003a,b).
Although the physiological responses of trees to O3 have been well described and characterized, the mechanisms involved in stress tolerance have not yet been completely clarified. We previously demonstrated that the basis for the different tolerances to the pollutant shown by two hybrid poplar clones (O3-sensitive Populus deltoides × maximowiczii clone Eridano and O3-resistant Populus deltoides × euramericana clone I-214) was not related to different fluxes of O3 reaching the plasma membranes (Ranieri et al., 1999). In the present report, we investigated whether the high sensitivity of clone Eridano is attributable to differential modulation of the signal transduction pathways. To this end, H2O2, ethylene, ACC and SA concentrations and the expression of the ACC synthase gene were measured in a time-course experiment. Some enzymatic activities involved in H2O2 production and scavenging were also measured, with the aim of elucidating which mechanisms are responsible for the O3-induced oxidative burst. We show that the attenuated response of the tolerant clone I-214 is related to differences in the timing and magnitude of signalling molecule production, which probably results in optimal induction of cell protective mechanisms without inducing cell death.
Materials and Methods
Cuttings of two poplar clones (O3-sensitive Populus deltoides× maximowiczii clone Eridano and O3-resistant Populus deltoides × euramericana clone I-214, kindly provided by the Poplar Research Institute, Casale Monterrato, Alessandria, Italy) were placed in plastic pots filled with a soil and expanded clay mix [1 : 1, volume/volume (v/v)] and grown for 2 months in the open air. Uniform plants were selected when 10 leaves were fully expanded. All the analyses were carried out on the sixth leaf, collected from three individual plants (for assessment of cell viability, ethylene evolution and in situ localization of H2O2) or pooled from different plants (for measurement of enzyme activities, SA and ACC quantification and RNA isolation).
Ozone fumigation was performed in air-conditioned chambers (0.48 m3). The temperature was maintained at 20 ± 1°C and relative humidity at 85 ± 5%. A photon flux density at plant height of 530 mol photons m−2 s−1 (photosynthetic active radiation: 400–700 nm) was provided by incandescent lamps. O3 was generated by electric discharge, passing pure oxygen through a Fisher 500 air-cooled generator (Fisher Labor und Verfahrenstechnik, Meckenheim, Germany). The O3 concentration in the fumigation chambers was continuously monitored with a Monitor Laboratories Analyzer model 8810 (Monitor Laboratories, San Diego, CA, USA) operating on the principle of ultraviolet (UV) absorption and interfaced with a personal computer. Plants were pre-adapted to the chamber conditions for 48 h and then exposed to an acute fumigation with 150 nl l−1 O3 for 5 h (from 08:00 to 13:00 h). Leaves were collected before (0 h) and during (1, 2 and 5 h) exposure to the pollutant. To check for lesion development and loss of cell membrane integrity, some plants were left to recover in pollutant-free air for up to 24 h.
Times of measurement refer to hours after the onset of fumigation. Untreated plants were kept in charcoal-filtered air chambers, under the same conditions, and used as controls.
Cell viability measurement
Cell viability was assessed by Evan's Blue dye, which is usually excluded from cells with intact membranes. Leaf discs of known area (1.13 cm2) were vacuum-infiltrated for 1 h with 0.25% Evan's Blue solution with continuous shaking (Schraudner et al., 1998). After infiltration, leaf discs were rinsed with distilled water to remove any excess dye and examined with a Leica microscope (Leica Microsystems, Wetzlar, Germany).
In situ localization of H2O2 accumulation
H2O2 production was assessed cytochemically via determination of cerium perhydroxide formation after reaction of cerium chloride (CeCl3) with endogenous H2O2 (Bestwick et al., 1997). Freshly harvested leaves were cut into slices (1–2 mm2) which were incubated for 1 h in 5 mm CeCl3 in 50 mm 3-(N-morpholino) propane sulphonic acid (MOPS), pH 7.2, fixed in 1.25%[weight/volume (w/v)] glutaraldehyde and 1.25% (w/v) paraformaldehyde in 50 mm sodium cacodylate buffer (CAB), pH 7.2, for 1 h and washed twice in CAB for 10 min (Bestwick et al., 1997). After postfixation for 2 h in 1% (w/v) osmium tetroxide in 50 mm CAB, pH 7.2, samples were washed twice in the same buffer, dehydrated in a graded ethanol series (25, 50, 75, 90 and 100%, v/v), transferred into propylene oxide and gradually embedded in Epon-Araldite. Thin sections of embedded tissues were obtained using a Reichert-Ultracut microtome (Reichert Microscope Services, Depew, NY, USA), mounted on uncoated copper grids and observed using a transmission electron microscope (Hitachi 300; Hitachi, Tokyo, Japan) at 75 kV.
Enzyme extraction and activity
Frozen leaves were ground in liquid nitrogen with 10% (w/w) polyvinylpolypirrolydone (PVPP) and homogenized in 100 mm Tricine-KOH buffer (pH 8.0), 20 mm MgCl2, 50 mm KCl, 10 mm ethylenediaminetetraacetic acid (EDTA), 0.1% (v/v) Triton X-100, 1 mm dithiothreitol (DTT) and 0.50 mm phenylmethylsulfonyl fluoride (PMSF). After centrifugation at 12 000g for 30 min at 4°C, the supernatant was collected and dialysed overnight against diluted extraction buffer. For APX assay, 50 mm sodium ascorbate was added to the extraction and dialysis buffers to avoid APX inactivation during the extraction procedure.
POD activity was tested using two different reducing phenolic substrates, o-dianisidine and syryngaldazine. The reaction medium for dianisidine-POD contained 20 mm sodium acetate (pH 5.0), 3 mm H2O2, 0.06% (w/v) dianisidine and an appropriate amount of enzyme extract. The rate of dianisidine oxidation was recorded at 460 nm and the activity was calculated using the extinction coefficient of 11.3 mm−1 cm−1 for dianisidine. The activity of syringaldazine-POD was determined by measuring the increase in absorbance at 530 mm of the reaction mixture containing 100 mm Na-K phosphate buffer, pH 6.0, 2.5 mm H2O2, 2 mm syringaldazine and the protein extract (Pandolfini et al., 1992).
APX activity was determined following the decrease in absorbance at 290 mm resulting from the oxidation of ascorbic acid in the first 30 s from the start of the reaction, using the extinction coefficient of 2.8 mm−1 cm−1 for ascorbate. The reaction medium contained 0.5 mm sodium ascorbate, 0.1 mm H2O2, 1 mm EDTA and 0.1 m Hepes-KOH buffer (pH 7.8) (Ranieri et al., 1996). One enzymatic unit is equivalent to 1 mol of ascorbic acid oxidized min−1 cm−1. To discriminate between APX and POD activities, 50 mmp-chloromercuribenzoate (pCMB), which is known to inactivate APX, was added to the enzymatic reaction mixture (Miyake & Asada, 1992).
The rate of NAD(P)H oxidation was measured by following the decrease in A340 of a reaction medium containing 40 mm sodium acetate (pH 5.5), 250 mm sucrose, 1 mm MnCl2, 100 m salicylhydroxamic acid, 100 m NADH and the extract aliquot (Vianello et al., 1997). The activity was calculated using the extinction coefficient of 6.22 m−1 cm−1 for NADH.
The protein content of the extracts was measured spectrophotometrically at 595 nm according to Bradford (1976), using bovine serum albumin as standard.
Extraction and quantification of SA
Leaf frozen tissue (100 mg) was ground in 5 ml of 100% methanol and centrifuged at 29 000g for 10 min at 4°C. The pellet was re-extracted with 3 ml of 100% methanol following the same procedure. Supernatants from both extractions were combined and dried by rotary flash evaporation (RFE) at 4°C. The residue was resuspended in 2 ml of 20% (v/v) formic acid and 50 l of 32% (v/v) HCl. The sample was centrifuged at 3000g for 5 min at 4°C after the addition of 5 ml ethyl acetate : cycloexane (1 : 1, v/v). The aqueous phase was re-extracted with 3 ml ethyl acetate:cycloexane (1 : 1, v/v) and the two resulting top organic phases were collected and dried by RFE. The residue was then resuspended in 0.5 ml of 27 mm trihydrate sodium acetate buffer (pH 3.5), 30 mm citric acid and 20% (v/v) methanol and centrifuged at 20 000g for 10 min at 4°C. To determine the β-glucosyl-salicylic acid content, 1.3 ml of 32% (v/v) HCl was added to the aqueous phase and samples were hydrolysed for 1 h at 80°C. Salicylic acid was quantified by reverse-phase high-performance liquid chromatography (HPLC) (Thermo Electron Corporation, Rodano Milanese, Italy) on a C18 column (Phenomenex Prodigy LC-18 ODS; Phenomenex Ltd Deutschland, Aschaffenburg, Germany; 250 × 4.6 mm; 5 m) equilibrated with a mixture of 90% 20 mm sodium acetate buffer (pH 5) and 10% methanol. A gradient of methanol (10–100%) was applied over 20 min. The flow rate was 1 ml min−1. SA concentrations were calculated based on fluorescence (excitation wavelength 290 nm; emission wavelength 407 nm) compared with a standard curve using authentic SA.
Determination of ACC content
Leaf material was ground in liquid nitrogen and extracted according to Langebartels et al. (1991). Free ACC and total ACC released by acid hydrolysis (2 N HCl for 3 h at 120°C) were determined according to Lizada & Yang (1979) as described by Langebartels et al. (1991). The amount of conjugated ACC was calculated by subtracting the amount of ACC from that of total ACC.
Fifteen minutes after excision, leaves were incubated within sealed containers at room temperature, and, after 1 h, 2-ml samples were withdrawn with a hypodermic syringe. Ethylene evolution was measured by injecting samples into a gas chromatograph equipped with a dual flame ionization detector and a metal column (150 × 0.4 cm internal diameter) packed with alumina (70–230 mesh). Column and detector temperatures were 70°C and 350°C, respectively. Nitrogen gas (N2) was used as a carrier at a flow rate of 40 ml min−1 (Mensuali Sodi et al., 1992).
RNA isolation and reverse transcription of RNA
Total RNA was extracted from frozen, homogenized leaf tissue as described by Kiefer et al. (2000). The RNA yield and quality were determined by spectral photometry at 260 and 280 nm. The integrity of RNA was checked by electrophoresis on 1% agarose gel. For cDNA synthesis, 5 g of total RNA was reverse-transcribed using SUPERSCRIPT™ II according to the manufacturer's instructions (Life Technologies, Karlsruhe, Germany).
The cDNA was amplified by polymerase chain reaction (PCR) with forward primers ACC-F (CAA GTT GAA TAG CCA CTT G) and ACC-FM (CTG ATT TCA TTA GCG TAC GGA GG), and the reverse primer ACC-R (GAA TAC CTA TCC TGA GCT AGG). The primers were made on the basis of conserved amino acid sequences of Populus nigra ACCs. The PCR conditions were as follows: 94°C for 5 min; then 39 cycles of 95°C for 1 min, 52°C for 45 s and 72°C for 30 s. The amplified cDNAs were subcloned into a pGEM-T vector (Promega, Madison, WI, USA). The sequences of several clones were determined and compared with the known sequences of ACS.
Semiquantitative PCR was carried out in the presence of GeneAmp RNA109 (Applied Biosystems, Darmstadt, Germany). The cDNA amplification was performed with the primers ACC-FM and ACC-R as follows: 94°C for 1 min; 52°C for 30 s; 72°C for 2 min for 29 cycles. The artificial cDNA was amplified using the primer pair DM151/DM152 (Applied Biosystems). PCR products were separated by electrophoresis on 1% agarose gel. For the quantification of the different bands, multianalyst software (Biorad, München, Germany) was used. An induction factor was calculated from the ratio of the specific gene to an internal artificial standard (geneAmp RNA pAW 109; Applied Biosystems).
The fumigation experiment was performed in duplicate. A minimum of 12 plants per treatment were used in all the experiments. Values shown in the figures are the means of six determinations ± standard error (SE), except for ethylene (n = 3). Comparison between means was carried out using the t-test and the P = 0.05 level of error.
The O3-sensitive poplar clone Eridano developed foliar spot-like lesions at the end of the 5-h pulse treatment with 150 nl l−1. Injury spread rapidly, resulting in coalescing large necrotic areas within 24 h of the start of the 5-h O3 exposure (Fig. 1a). Microscopic examination after staining with the vital nonpermeable dye Evan's Blue revealed large clusters of blue-stained cells in leaf discs of clone Eridano at the end of the 5-h O3 exposure, indicating the loss of cell membrane integrity. The number of dead cells further increased during post-cultivation in pollutant-free air, leading to large blue-stained areas 24 h after the onset of fumigation (Fig. 1b).
No visible sign of O3 damage was observed on the leaf surface of the tolerant clone I-214, neither during the 5-h O3 exposure nor during the recovery period in filtered air (Fig. 1a). Similarly, cell permeability to the Evan's Blue dye was not affected by O3 at the end of the fumigation period, as indicated by the almost complete lack of blue-stained cells. However, some individual cells or small clusters of cells that had died following O3 exposure were present during the post-fumigation period (Fig. 1b).
In situ localization of H2O2 from electron microscope cytochemical analysis
The reaction of H2O2 with CeCl3 produced electron-dense insoluble precipitates of cerium perhydroxide at sites where H2O2 accumulated.
In control leaves of both clones, no CeCl3 precipitates were detectable in the cell walls of palisade parenchyma or spongy mesophyll cells (Fig. 2a). Electron-dense precipitates were evident only in the cell walls of vascular tissue undergoing lignification (Fig. 2b). Fig. 2b shows control cells of clone I-214 only, but similar results were obtained for clone Eridano.
In both clones, CeCl3 staining was observed as early as 1 h after the beginning of the fumigation, when electron-dense precipitates were clearly visible in the cell walls of both spongy mesophyll and palisade parenchyma cells of treated Eridano leaves (Fig. 2c and e, respectively), whereas the O3-tolerant clone I-214 showed faint CeCl3 staining only in the cell walls of the spongy mesophyll tissue near the epidermis (Fig. 2d; cf. Fig. 2f). H2O2 accumulation reached its maximum concentration at the end of the 5-h O3 exposure, although in clone I-214 the CeCl3 precipitates were less evident than in clone Eridano. Staining was particularly intense at the sites of connection between adjacent cell walls close to the intercellular spaces, mainly in the palisade parenchyma tissue (Fig. 2g,h). In the O3-sensitive clone Eridano, H2O2 accumulation was also visible in the cell walls of the spongy mesophyll, although to a lesser extent than in the palisade tissue (data not shown).
Free and conjugated SA concentrations
To determine whether the different O3 sensitivities shown by the two poplar clones were related to different abilities to accumulate SA, the concentrations of free and conjugated SA were measured in control and O3-exposed plants. Both the clones exhibited constitutive concentrations of free SA which were more elevated than those recorded in herbaceous species such as Arabidopsis and tobacco. This was particularly evident for the O3-tolerant clone I-214, which was found to have a higher endogenous free-SA concentration than the O3-sensitive clone Eridano (+56%). No significant change was observed in response to O3 fumigation in clone I-214, while in clone Eridano the time-course experiment revealed a slight, but significant, increase in free-SA concentration (+36%) at the end of the 5-h O3 exposure (Fig. 3a,b). In clone Eridano, the conjugated pool of SA was found to be significantly enhanced starting from the second hour of fumigation (+13% in comparison to the control) and the percentage of increase remained almost constant until the end of the treatment (+19% at 5 h; Fig. 3d). A significant increase in the concentration of conjugated SA was also observed in clone I-214, but only 5 h after the beginning of the O3 exposure (+29%, Fig. 3c).
Free and conjugated ACC (1-aminocyclopropane-1-carboxylic acid) concentrations
Both clones accumulated free ACC at the end of the 5-h O3 treatment. In particular, in O3-exposed plants, the concentrations of free ACC were about 56-fold and 7-fold higher than in the respective controls in clone I-214 and in clone Eridano, respectively (Fig. 4). A difference in behaviour between the two clones was observed regarding conjugated ACC. While the concentration of conjugated ACC did not change during the exposure in the tolerant clone I-214, it was found to significantly increase in clone Eridano within the first hour of fumigation (+56%; Fig. 4), reaching a maximum concentration 5 h after the onset of the O3 exposure (+127%, Fig. 4).
ACC synthase gene expression
Primers were constructed based on the database gene sequence for ACC synthase (ACS) of Populus nigra, and were used to generate PCR products, which were cloned and sequenced. The PCR products of the two poplar clones were 750 bp in size, and there was 100% identity between the products of the two clones and 98% homology with the gene sequence of P. nigra. In both clones, ACS transcripts showed a transient increase following O3 exposure, reaching a maximum 2 h after the onset of the fumigation and decreasing thereafter to the control concentrations (Fig. 5).
In both poplar clones, O3 exposure induced ethylene evolution, but the timing and the extent of the response were different. In I-214, ethylene emission was earlier than in Eridano, being significantly stimulated after 2 h of exposure and reaching a maximum value at the end of the 5-h O3 treatment (about 8.6-fold and 48-fold higher than in the control, respectively; Fig. 6a). Meanwhile, in Eridano, ethylene evolution was found to be enhanced only 5 h after the onset of the O3 exposure, although the extent of the increase (about 154-fold higher than the control) was much greater than in I-214 (Fig. 6b).
The two poplar clones showed similar behaviour following O3 exposure regarding dianisidine-POD activity, although the constitutive levels of enzyme activity were lower in Eridano than in I-214 leaves. In both clones, dianisidine-POD activity was found to be quickly stimulated by O3 treatment (+96% and +156% in I-214 and Eridano, respectively, after 1 h of fumigation), remained significantly higher than the respective control activities during the second hour of exposure (+104% and +87% in I-214 and Eridano, respectively) and returned to control values at the end of the 5-h O3 treatment (Fig. 7a,b).
Syringaldazine is a synthetic substrate analogue to the syringilic residue of lignin, commonly used to measure the activity of peroxidases involved in the lignification process. Syringaldazine-POD activity was found to differ from that of dianisidine-POD. In I-214, syringaldazine-POD activity was significantly decreased by O3 exposure up to the second hour of treatment (−42% and −22%, respectively, after 1 and 2 h of fumigation; Fig. 7c). This trend was reversed at the end of the exposure, when syringaldazine-POD activity significantly increased over control values (+25%). In contrast, no significant change in syringaldazine-POD activity was found in Eridano at any time point in the O3 fumigation period (Fig. 7d).
In contrast to results for lignifying PODs, in I-214 leaves O3 treatment induced a marginally significant stimulation of APX activity as early as 1 h after the start of fumigation (+11%; Fig. 7e). APX activity reached a maximum value 2 h after the onset of the exposure (+30%), while at the end of the fumigation no difference between treated and control samples was recorded (Fig. 7e). Conversely, in Eridano, APX activity was negatively affected by O3 up to the second hour of treatment (–28% and −25%, respectively, after 1 and 2 h of fumigation; Fig. 7f), while a significant stimulation of this activity was found at the end of the exposure (+19%; Fig. 7f).
The two poplar clones also behaved quite differently regarding NAD(P)H oxidation rate. In I-214 leaves, the activity was significantly stimulated by O3 after the first hour of fumigation (+84%), when it reached a maximum concentration, and thereafter it began to diminish. After 2 h it was still higher than the control value (+28%) but at the end of the 5-h exposure it had returned to the control values (Fig. 7g). In Eridano, the NAD(P)H oxidation rate remained unchanged until the second hour of O3 exposure, when the activity was found to increase by 56% relative to the control. A slight further increase was observed at the end of the treatment (+65% relative to the control) (Fig. 7h).
Ozone is known to induce a variety of stress responses in plants, the type and extent of which may reflect the different sensitivities to the pollutant shown by different species, clones and cultivars (Kangasjärvi et al., 1994; Rao et al., 2000). Recently, attempts have been made to correlate O3 sensitivity with the activation of multiple signalling pathways involving different messengers interacting with each other in a rather complex manner (Sandermann et al., 1998; Rao et al., 2000; Vahala et al., 2003a,b).
O3-derived ROS and the subsequent oxidative burst have been receiving increasing attention as early signals capable of eliciting the diverse effects that constitute the plant response to O3 (Pellinen et al., 1999, 2002; Langebartels et al., 2002; Pasqualini et al., 2002; Wohlgemuth et al., 2002; Ranieri et al., 2003), leading to the suggestion that O3-induced damage could be mechanistically similar to the pathogen-induced response. The cytochemical localization of H2O2 by CeCl3 staining revealed that in both clones extracellular H2O2 accumulation was one of the earliest detectable responses to O3, being already evident 1 h after the beginning of the treatment. A spatial relationship between the sites of H2O2 accumulation and lesion development has been reported in the O3-sensitive tobacco cultivar Bel W3 (Schraudner et al., 1998; Wohlgemuth et al., 2002), in some tomato cultivars (Wohlgemuth et al., 2002), in some Arabidopsis genotypes (Rao & Davis, 1999) and in a sensitive birch (Betula pendula) clone (Pellinen et al., 1999, 2002). This is consistent with the behaviour of the O3-sensitive clone Eridano, which in fact displayed visible signs of injury on the leaf surface. However, similar to the evidence reported for sunflower (Helianthus annuus) (Ranieri et al., 2003), H2O2 also accumulated in the O3-tolerant clone I-214, although the small size and number of precipitates in comparison to Eridano could explain the absence of leaf lesions. As reported by Levine et al. (1994), concentrations of H2O2 required to induce cell death are usually higher than those capable of inducing defence gene activation. Moreover, the two clones also differed in tissue location of CeCl3 staining: in Eridano CeCl3 precipitates were always evident in both the spongy mesophyll and the parenchyma palisade cells, while in I-214 H2O2 accumulated, at the beginning of the exposure, only in the spongy mesophyll and seemed to diffuse into the palisade parenchyma only in the later stages of fumigation.
The origin of the O3-induced oxidative burst is still a matter of debate. It is probable that more than one mechanism is operating not only in different species but even in the same plant. Spontaneous O3 degradation in the apoplast, and O3 reactions with cell wall compounds such as phenolics, olefins and amide proteins and with the unsaturated lipids present in the plasma membranes can generate ROS, although, as demonstrated by electron paramagnetic resonance studies, their half-life is very short (Runeckles & Vaartnou, 1997). In addition to direct apoplastic generation, H2O2 accumulation can originate from the activity of H2O2-producing enzymes, such as plasma membrane NAD(P)H oxidase complex, extracellular pH-dependent PODs, oxalate oxidase and diamine and polyamine oxidases, and/or from insufficient H2O2-scavenging capacity of antioxidant mechanisms (Bolwell & Wojtaszek, 1997; Sebela et al., 2001; Langebartels et al., 2002; Ranieri et al., 2003).
In both poplar clones, O3-induced H2O2 accumulation seems to be a complex process involving the time-dependent regulation of both H2O2-generating and H2O2-detoxifying enzymes, the combination of which can at least partially account for the differences in the extent of H2O2 accumulation between the two clones. Apart from dianisidine-POD activity, which exhibited a similar initial increase at 1 and 2 h followed by a return to the control values at 5 h in both Eridano and I-214 leaves, the other enzymes involved in H2O2 turnover behaved in almost opposite ways in the two clones. In I-214, at the beginning of the O3 exposure, enhanced H2O2 production by NAD(P)H oxidizing PODs and reduced H2O2 consumption by decreased syringaldazine-POD activity contributed to the observed H2O2 accumulation. Because PODs involved in the lignification process are known to require an acidic environment (Otter & Polle, 1997), the rapid transient decrease in syringaldazine-POD activity, also observed in O3-treated sunflower plants (Ranieri et al., 2003), could depend on the O3-induced shift of apoplastic pH towards basic values. At the same time, such an increase in pH will provide the optimal conditions for NAD(P)H oxidizing PODs, whose maximum activity is in fact reported to occur at neutral to basic pH, depending upon the isoform studied (Bolwell & Wojtaszek, 1997). Conversely, in the later stages of the O3 exposure, the activity of NAD(P)H oxidizing PODs declined to the control values, so that H2O2 accumulation was mainly a result of the lower activity of H2O2-scavenging enzymes (dianisidine-POD and APX), although the possible involvement of other H2O2-producing enzymes, such as diamine and polyamine oxidase, which were not measured in the present experiment, cannot be excluded. However, at the same time, a parallel increase in syringaldazine-POD activity occurred, thus contributing to limitation of H2O2 accumulation.
A quite different scenario was observed in Eridano, where NAD(P)H oxidizing PODs did not make any contribution to H2O2 production up to the second hour of exposure, when the H2O2 concentration started to increase, leading to the intense accumulation of CeCl3 precipitates detected at 5 h. At the beginning of the exposure, in contrast, H2O2 accumulation seemed rather to originate from the significant decrease in the activity of APX, a key enzyme in the process of H2O2 scavenging (Foyer et al., 1994). While in O3-treated I-214, consistent with results reported by many authors (Kubo et al., 1995; Ranieri et al., 1996, 2000, 2003; Rao & Davis, 1999; Pellinen et al., 2002), APX activity quickly increased, thus contributing to limitation of H2O2 accumulation despite increased production, Eridano seemed unable to stimulate such a defensive mechanism at any point in the fumigation period. It should also be noted that I-214 displayed higher constitutive activities of syringaldazine-POD and dianisidine-POD, which could protect cells from primary ROS attack in the first stages of O3 exposure and could make this clone more efficient in counteracting the build-up of toxic H2O2 concentrations.
Although H2O2 is recognized as a messenger molecule centrally involved in the signalling cascade, its usually rapid detoxification by antioxidant enzymes and metabolites makes it unfeasible that H2O2 alone could account for the complex response to O3 at the whole-plant level. Other molecules are believed to cooperate with H2O2, either synergistically or antagonistically, to trigger the plant defence responses (Rao et al., 2000). Among these signal molecules, ethylene is believed to play an active role in mediating O3-induced lesion formation (Kangasjärvi et al., 1994; Wellburn & Wellburn, 1996; Vahala et al., 2003b; Sinn et al., 2004). In accordance with this theory, the O3-sensitive clone Eridano, which developed severe leaf injuries, exhibited an intense ethylene evolution at the end of the 5-h O3 exposure. However, although the timing and the extent of ethylene emission differed between the two clones, the tolerant clone I-214, which did not show any visible symptoms on the leaf surface, was also found to significantly evolve ethylene. In O3-treated sunflower leaves, increased ethylene evolution was also not accompanied by either enhanced lipid peroxidation or leaf damage (Ranieri et al., 2003), suggesting the existence of a threshold below which ethylene does not trigger lesion development. Indeed, besides the well-known effect of ethylene in promoting cell death, depending on the temporal pattern of biosynthesis, a ‘pro-survival’ role has been proposed for this molecule (Mehlhorn, 1990; Vahala et al., 2003b). Mung bean (Vigna radiata) and pea (Pisum sativum) plants treated with ethylene before O3 exposure were found to be even more tolerant to the pollutant (Mehlhorn, 1990). This pro-survival role of ethylene may require an early transient increase of ethylene evolution sufficient to induce the protective response by the plant cell but below the threshold responsible for the spreading of cell death. While this threshold was crossed in the sensitive poplar clone Eridano, the earlier and less pronounced ethylene emission in I-214 could have ensured adequate signal transduction without inducing lesion formation. Thus, consistent with the findings of Vahala et al. (2003b), who observed different kinetics and degrees of ethylene emission in three differently O3-sensitive birch clones, different roles for ethylene could also be hypothesized in these two poplar clones.
Ethylene is synthesized from S-adenosyl-L-methionine via ACC (Kende, 1993). The conversion of S-adenosyl-L-methionine into ACC, catalysed by ACS, is generally considered the rate-limiting step in ethylene biosynthesis. ACC is then converted to ethylene by ACC oxidase. Increased ACS transcript concentrations, accompanied by enhanced ACC emission, have been reported in different O3-treated species (Langebartels et al., 1991; Tuomainen et al., 1997; Moeder et al., 2002; Nakajima et al., 2002), and in some Solanaceae species a sequential expression of ACS genes has been observed in response to both biotic and abiotic stresses (Schlagnhaufer et al., 1997; Nakajima et al., 2001; Moeder et al., 2002). In particular, in tomato, the transcripts of LE-ACS1A and LE-ACS6 increased as early as 1 h after the start of the O3 exposure and, after a further increase at 2 h, decreased rapidly. However, the transcripts of LE-ACS2 increased at 2 h and remained at high concentrations up to 6 h (Nakajima et al., 2001). Similar behaviour for LE-ACS2 was also observed by Tuomainen et al. (1997). In agreement with the findings reported by the above-mentioned authors, in the present study O3 exposure led to a significant increase of ACS in the two poplar clones Eridano and I-214. The observed transient increase at 2 h, followed by a decline in the transcript concentration at the end of the 5-h O3 exposure, seems to suggest that this ACS isoform could be a ‘fast responding’ isoform.
As a result of the increase in the transcript concentrations of ACS, a marked increase in the pool of free ACC was detected at the end of the O3 exposure in both poplar clones. Subsequent ACC oxidation by ACC oxidase might then lead to ethylene formation, as indeed occurred in both clones at 5 h. The earlier ethylene evolution observed in I-214 at 2 h could derive from the pool of conjugated forms of ACC, among which N-malonyl ACC (Kende, 1993) and g-L-glutamyl ACC (Martin et al., 1995) have been identified. In contrast to Eridano, in I-214 the conjugated forms did not accumulate following O3 treatment. However, it is also possible that other members of the ACS multigene family are activated earlier than the form identified in the present experiment, resulting in ethylene production from de novo synthesized ACC.
In addition to ethylene, SA has also been shown to be an important modulator of O3-induced responses (Rao & Davis, 1999; Rao et al., 2002; Vahala et al., 2003a). Studies with Arabidopsis mutants revealed a central role for SA in lesion initiation and progression in response to O3 (Rao et al., 2002). In particular, greater constitutive amounts of free and total SA were measured in the O3-sensitive mutants overproducing ethylene, suggesting that high SA basal concentrations could potentiate O3-induced plant responses, including cell death, and demonstrating the synergistic action of SA and ethylene in lesion development (Rao et al., 2002). In addition, because transgenic or mutant plants compromised in SA signalling failed to produce ethylene in response to O3, these authors suggested that SA is required for stress-induced ethylene production (Rao et al., 2002). Similarly, the constitutive concentrations of free and total SA were found to be higher in the O3-sensitive poplar clone NE-388 compared with NE-245 (Koch et al., 2000) although in both clones, in contrast to the findings reported for Arabidopsis (Sharma et al., 1996; Rao et al., 2002), no significant increase was found following O3 exposure.
In our case, the sensitive poplar clone Eridano showed an approximately 2-fold higher basal concentration of conjugated pools of SA than I-214, but the opposite was found for the free form. This latter result is consistent with the observation reported in aspen clones by Vahala et al. (2003a), who suggested that the greater SA concentration of the tolerant clone might have depressed ethylene production, thus preventing lesion formation. SA has indeed been shown to inhibit ethylene production in pear (Pyrus communis) cell suspension culture (Leslie & Romani, 1988), in tomato fruits (Li et al., 1992), in rice (Oryza sativa) leaves (Huang et al., 1993) and in mung bean hypocotyls (Lee et al., 1999). O3-induced ethylene evolution by the O3-tolerant clone I-214, which exhibited higher constitutive concentrations of free SA, was much lower than in the O3-sensitive clone Eridano.
Independently of basal concentrations, however, only the sensitive poplar clone Eridano displayed a slight but significant accumulation of free SA following O3 exposure and showed an earlier increase in the conjugated form compared with I-214. In addition, although the extent of the O3-induced increase in the pool of conjugated SA was similar in the two clones, the absolute concentration was markedly higher in the sensitive clone, consistent with findings reported by Rao et al. (2002) in Arabidopsis mutants. As already discussed for ethylene, the final response to O3 seems to depend on the timing and magnitude of SA production, as well as on cross-talking with other signalling molecules. Thus, optimal SA concentrations are required to fine-tune the plant response in order to achieve the maximum stimulation of defence responses with minimal induction of cell death. In fact, while plants defective in SA production or perception undergo weak induction of antioxidant defence responses, ultimately leading to ROS accumulation, elevated concentrations of SA are known to potentiate the feedback amplification loop responsible for ROS production and PCD induction (Rao & Davis, 1999). SA is known to induce H2O2 accumulation by inhibiting catalase activity through specific binding to the enzyme (Chen et al., 1993), or by inducing H2O2 formation by PODs (Kawano & Muto, 2000).
In conclusion, both poplar clones seemed to react to O3 exposure by producing signalling molecules such as H2O2, ethylene and SA, but differences in the kinetics and magnitude of their production were clearly evident between the two clones. The lower sensitivity to O3 of the tolerant clone I-214 probably derives from its ability to maintain the concentration of these messengers below a toxic threshold, thus avoiding the excessive amplification of the defence response, ultimately leading to cell death. However, the complexity of the mechanisms involved in the plant response to O3 requires further analyses to clarify the role played by the signalling molecules, and especially how cross-talking amongst these different signalling pathways could determine the fate of the cell.
We are indebted to Dr Lorenzo Vietto (Poplar Research Institute, Casale Monferrato, Alessandria, Italy) for providing the poplar clones. This research was supported by a grant from MIUR (National Project), Rome, Italy and by funds of the University of Pisa. C.D. was supported by a PhD grant of the Sant’Anna School of University Studies and Doctoral Research.