A role for oxalic acid generation in ozone-induced signallization in Arabidopis cells



    Corresponding author
    1. Université Paris Diderot, Sorbonne Paris Cité, Institut des Energies de Demain (IED), Paris, France
    2. Institut de Biologie des Plantes, Bât 630, 91405 Orsay, France
    Search for more papers by this author

    1. Université Paris Diderot, Sorbonne Paris Cité, Institut des Energies de Demain (IED), Paris, France
    2. Graduate School of Environmental Engineering, University of Kitakyushu 1-1, Hibikino, Wakamatsu-ku, Kitakyushu 808-0135, Japan
    Search for more papers by this author

    1. Facultad de Agronomía, Universidad Nacional de La Pampa. Ruta 35 km 334, CC 300, 6300 Santa Rosa, La Pampa, Argentina
    2. Université Paris Diderot, Sorbonne Paris Cité, Paris Interdisciplinary Energy Research Institute (PIERI), Paris, France
    Search for more papers by this author

    1. Université Paris Diderot, Sorbonne Paris Cité, Institut des Energies de Demain (IED), Paris, France
    Search for more papers by this author

    1. Université Paris Diderot, Sorbonne Paris Cité, Institut des Energies de Demain (IED), Paris, France
    2. Institut de Biologie des Plantes, Bât 630, 91405 Orsay, France
    Search for more papers by this author

    1. Université Paris Diderot, Sorbonne Paris Cité, Institut des Energies de Demain (IED), Paris, France
    2. Institut de Biologie des Plantes, Bât 630, 91405 Orsay, France
    Search for more papers by this author

    1. Graduate School of Environmental Engineering, University of Kitakyushu 1-1, Hibikino, Wakamatsu-ku, Kitakyushu 808-0135, Japan
    2. Université Paris Diderot, Sorbonne Paris Cité, Paris Interdisciplinary Energy Research Institute (PIERI), Paris, France
    Search for more papers by this author

    1. Université Paris Diderot, Sorbonne Paris Cité, Institut des Energies de Demain (IED), Paris, France
    2. Institut de Biologie des Plantes, Bât 630, 91405 Orsay, France
    Search for more papers by this author

D. Tran. Fax: +33 169153424; e-mail: daniel.tran@univ-paris-diderot.fr


Ozone (O3) is an air pollutant with an impact increasingly important in our industrialized world. It affects human health and productivity in various crops. We provide the evidences that treatment of Arabidopsis thaliana with O3 results in ascorbate-derived oxalic acid production. Using cultured cells of A. thaliana as a model, here we further showed that oxalic acid induces activation of anion channels that trigger depolarization of the cell, increase in cytosolic Ca2+ concentration, generation of reactive oxygen species and cell death. We confirmed that O3 reacts with ascorbate in the culture, thus resulting in production of oxalic acid and this could be part of the O3-induced signalling pathways that trigger programmed cell death.


cytosolic calcium


actinomycin D


acorbic acid






fresh weight




oxalic acid


programmed cell death


plasma membrane


reactive oxygen species


Ozone (O3) is a gaseous air pollutant that causes damages to animals and plants. In animals, exposure to O3 affects mainly respiratory systems, especially functions of lungs (Enami, Hoffmann & Colussi 2008). In plants, exposure to O3 often induces a disturbance in photosynthesis and metabolism, and therefore finally resulting in the loss of crop yields and forest declines (Pell, Schlagnhaufer & Arteca 1997). An acute exposure to O3 causes lesions formation on the leaves of sensitive plants due to localized progress in programmed cell death (PCD) (Kangasjärvi et al. 1994; Sandermann et al. 1998). Following the uptake of O3 from stomata, the first biochemical barriers against O3 within plant tissues would be the events in apoplast. As O3 readily breaks down into various reactive oxygen species (ROS) (Sandermann 2008), the pool of ascorbic acid (AsA) within the apoplast, as in fluid films covering the pulmonary epithelium, acts as a part of the initial line of defence against oxidative stresses (Enami et al. 2008; Foyer & Noctor 2011). Such plant-protective roles of AsA in relation to the resistance to O3 have been well documented (Polle, Wieser & Havranek 1995; Luwe 1996; Eller & Sparks 2006). When AsA in the apoplast reacts with oxidants, it is oxidized, yielding short-lived dehydroascorbate (DHA). For supporting the AsA-mediated protection against oxidative stress, AsA must be recycled from its oxidized forms (DHA) through reducing events inside the plant cells after transportion across the plasma membrane (PM). Thus, AsA can be continuously supplied to the apoplast to react with oxidants (Foyer & Noctor 2011). In accordance, plant cells also respond to oxidative stress by rapidly exporting AsA into the apoplast (Parsons & Fry 2010). AsA is highly abundant in plants, often attaining millimolar concentrations in most of cell compartments (Smirnoff & Wheeler 2000; Foyer & Noctor 2011). Accordingly, O3 sensitivity generally correlates with the whole AsA status (Sandermann 2008) and several Arabidopsis thaliana mutants with reduced AsA content (vtc mutants) were isolated and characterized as O3 sensitive mutants (Conklin et al. 2000).

In most plants, AsA degradation can occur via DHA, yielding oxalic acid (OA) (Yang & Loewus 1975; Green & Fry 2005). The reaction of O3 with AsA was thus suggested to lead to production of OA (Sandermann 2008) which is a potent inducer of PCD in plants (Errakhi et al. 2008; Kim, Min & Dickman 2008). We recently showed that O3 induces the activation of a PM anion channel which was an early prerequisite of O3-induced PCD in cultured cells of A. thaliana (Kadono et al. 2010), similarly to the observed effect of OA treatment leading to PCD (Errakhi et al. 2008). We therefore search for a putative role for OA generation during O3 exposure similar to the one encountered during episode of peak pollution in urban areas (Rao, Koch & Davis 2000; IPPC 2007) by using cultured cells of A. thaliana with a working hypothesis connecting the actions of O3, OA and signalling components.

We previously showed that the oxidative cell death observed in response to O3 was mediated by the interplay between anion channel activation, Ca2+ influx and ROS generation (Kadono et al. 2010). In the present study, we analysed the impact of OA on these different key factors of the signalling pathways leading to O3-induced PCD. OA is effectively produced in cultured cells in response to O3, and only in the AsA-rich seedlings. Furthermore, the level of OA correlates with activation of anion channel, increase in cytoplasmic Ca2+ concentration ([Ca2+]cyt), generation of ROS and cell death, highlighting a possible role for OA, as a secondary toxicant, in the O3-induced PCD.


Cell culture conditions

Suspension-cultured cells of A. thaliana (ecotype Col-0) were grown in Gamborg medium (pH 5.8) (Kadono et al. 2010). They were maintained at 22 ± 2 °C, under continuous white light (40 µE m−2 s−1) and continuous shaking at 120 r.p.m. All experiments were performed at 22 ± 2 °C using the cells in log-phase (4 d after sub-culturing).

Plant culture conditions

A. thaliana seeds (ecotype Columbia and vtc 1-1 mutant obtained from NASC mutant collections) were sown in vitro on solid Gamborg medium and grown in a growth chamber under a light cycle of 12 h light and 12 h dark with 40 µE m2 s−1 at 22 °C.

Ozone exposure

Ozone exposure of the cultured cells was performed with air ozonized passed on the surface of the cell suspensions (10 mL; packed cell volume, 75%). Cells were continuously exposed to 30 ppb min−1 (flow, 0.005 L min−1; 0.1 mg O3 h−1) in order to mimic the episodes of pollution encountered in urban areas (Rao et al. 2000; IPPC 2007). For assays using seedlings, 15 to 20-day-old seedlings were exposed to O3 in an enclosed hermetic chamber and were exposed to 300 ppb of O3 (flow, 0.005 L min−1; 0.1 mg O3 h−1 for 10 min). Seedlings were collected 16 h after O3 exposure for analysis.

Phosphotungstate reagent

To prepare periodically the reagent, 15 g of Na2WO4 (Sigma, St Louis, MO, USA) and 6 g of Na2HPO4 (Sigma) were mix with heating in 24 mL deionized water. Then 14.5 mL of H2SO4 (3.7 m) were added slowly. The solution was heated for 2 h with reflux condenser not allowing it to boil. After cooling the solution, the pH was adjusted to 1.0 with H2SO4 dropwise.

Oxalate assay

For oxalate determination, soluble OA was isolated from 500 mg fresh weight (FW) of tissues, and cells prepared from 15 to 20-day-old seedling tissues and 4-day-old cultured cells, respectively, were ground in a mortar and pestle in 2 mL of distilled water (Guo et al. 2005). The homogenates were boiled for 15 min, kept statically overnight and centrifuged at 15 000 g for 15 min. Phosphoric-tungstate reagent (500 µL) were added in 1 mL of supernatant, and kept statically for 5 h prior to centrifugation for 10 min at 15 000 g. Obtained solution was adjusted to pH 4.5 with ammonium hydroxide dropwise, supplemented with 500 µL of calcium chloride reagent (0.45 m CaCl2 in 1 m acetic acid, pH 4.5) and kept at 4 °C for over 16 h. Again, the mixture was centrifuged for 10 min as above, and the supernatant liquid was removed. The precipitate was dissolved in 500 µL of 10% sulfuric acid and incubated in water-bath at 100 °C for 2 min, and titration of OA with 2.5 mm potassium permanganate (KMnO4) was performed (Guo et al. 2005). Data are expressed as nmol g−1FW with means ± SE.

Ascorbic acid assays

Total AsA was determined according to the method previously described (Huang et al. 2005). Briefly, cells or seedlings were harvested and weighed, then ground with a mortar and pestle in 4 mL of 5% trichloroacetic acid at 4 °C. The homogenates were centrifuged at 2 °C for 15 min at 13 000 g, and the supernatant was neutralized with saturated Na2CO3. The reduced ascorbate was assayed spectophotometrically at 265 nm in 1 m NaH2PO4 buffer, pH 5.6, with 1 U of ascorbate oxidase (Sigma). The total ascorbate was assayed after incubation in the presence of 10 mm DTT (Sigma). DHA concentrations were determined by subtracting the reduced AsA to the total AsA. Data are expressed as nmol.g−1 FW with means ± SE.


Individual cells were impaled and voltage clamped in the culture medium using an Axoclamp 2B amplifier (Axon Instruments, Foster City, CA, USA) for discontinuous single electrode voltage clamp experiments as previously described (Reboutier et al. 2007; Errakhi et al. 2008; Meimoun et al. 2009). Voltage and current were digitized using a computer fitted with a Digidata 1320A acquisition board (Axon Instruments). The electrometer was driven by pClamp software (pCLAMP8, Axon Instruments). Experiments were conducted on 4-day-old cultures at 22 °C (main ions in the medium after 4 d of culture: 9 mm K+, 11 mm NO3-) (Reboutier et al. 2002).

Aequorin luminescence measurements

Variations in [Ca2+]cyt were recorded using freshly generated A. thaliana cell suspensions expressing apoaequorin (Kadono et al. 2010). For [Ca2+]cyt measurements, aequorin was reconstituted by 3 h incubation of the cell suspension in Gamborg medium containing 2.5 µm native coelenterazine. Cell culture aliquots (500 µL) were transferred carefully to a luminometer glass tube, and the luminescence counts were recorded continuously at 0.2 s intervals with a FB12-Berthold luminometer (Titertek Berthold, Pforzheim, Germany). At the end of each experiment, the residual aequorin was discharged by the addition of 500 µL of a 1 m CaCl2 solution dissolved in 100% methanol. The resulting luminescence was used to estimate the total amount of aequorin for each condition. Calibration of calcium levels was performed using the equation: pCa = 0.332588(−logk) + 5.5593, where k is a rate constant equal to luminescence counts per second divided by total remaining counts. Data are expressed as µm with means ± SE.

Monitoring of ROS production

The production of singlet oxygen (1O2) and superoxide (O2•−) was monitored by the chemiluminescence of the Cypridina luciferin analog (CLA) as previously described (Kadono et al. 2010). CLA-chemiluminescence mainly indicates the presence of O2•−, and of 1O2 to a lesser extent (Nakano et al. 1986). Chemiluminescence from CLA was monitored using a FB12-Berthold luminometer (with a signal integrating time of 0.2 s). For the statistical analysis of the data, the luminescence ratio (Linduced/Lbasal) was calculated by dividing the induced luminescence intensities of CLA-luminescence (Linduced) with the luminescence intensity before stimulation (Lbasal).

Cell viability assays

Cell viability was checked using fluorescein diacetate (FDA) as previously described (Reboutier et al. 2007). Briefly, after the appropriate treatment, 1 mL of cell suspension was gently stirred with a magnetic stirrer before FDA was added to a final concentration of 12 mm. The fluorescence increase was monitored over a 120 s period using a F-2000 spectrofluorimeter (Hitachi High-Technologies Corporation, Tokyo, Japan). Results are presented as the percentage of cell death = (slope of treated cells/slope of non-treated cells) • 100 ± SE. Cell viability was also checked using the vital dye, Evans blue (Kadono et al. 2010). Cells (50 µL) were incubated for 5 min in 1 mL phosphate buffer pH 7 supplemented with Evans blue to a final concentration of 0.005%. Cells that accumulated Evans blue were considered dead. At least 1000 cells were counted for each independent treatment. The experiment was repeated at least 4 times for each condition.

Statistical analysis

Data were analysed by analysis of variance (ANOVA), and the mean separation was achieved by Newman and Keuls multiple range test. All numeric differences in the data were considered significantly different at the probability level of P ≤ 0.05.


O3 exposure induce ROS generation and cell death in cell suspension

As O3 is supposed to break down in various ROS in apoplast (Sandermann 2008), we first checked if it occurs in our model system. Here, ROS generation was monitored by chemiluminescence of CLA. Despite the fact that O3 efficiency transfer in water phase is around 20 to 30% at ambient temperature, this exposure induced a large and sustained generation of ROS in the suspension culture (Fig. 1a). Such treatment induced a time-dependant increase in cell death, reaching a plateau of about 80% in 10 h (Fig. 1b). In order to check if another reactive molecule such as reactive nitrogen species (RNS) could modify cellular response, we have treated cell suspension with pure oxygen ozonized (flow, 0.001 L min−1; 0.1 mg O3 h−1) for 16 h. The rate of cell death was around 73 ± 5%, in the same range to the one treated with air ozonized (Fig. 1c). This indicates that putative RNS production from air through ozonizer was not involved in cell death extent as supported by the absence of effect of PTIO, as scavenger of NO, on ozonized air-induced cell death of A. thaliana cells we have previously reported (Kadono et al. 2010).

Figure 1.

O3 exposure induced ROS generation and cell death in Arabidopsis thaliana cells. (a) Production of ROS, monitored by the chemiluminescence of the Cypridina luciferin analog (Kadono et al. 2010), upon treatement of the cultured cells with ozonized air at 30 ppb min−1. (b) Time-dependent development of O3-induced cell death in A. thaliana cells continuously exposed to ozonized air estimated with Evans blue (30 ppb min−1). (c) Cell death extent in cultured cells exposed to ozonized air, or ozonized pure oxygen at 30 ppb min−1 during 16 h. Control corresponds to cells treated with air only. The data correspond to means of at least four independent replicates and error bars correspond to SE. *Significantly different from controls, P < 0.05. O3, ozone; ROS, reactive oxygen species.

OA is generated from AsA upon O3 treatment

While L-ascorbic acid could be converted to OA in plants and in suspension cells (Yang & Loewus 1975; Green & Fry 2005), to our knowledge, generation of OA in response to O3 has not been demonstrated to date. Moreover, O3 sensitivity correlates with the AsA status of the plants (Sandermann 2008). Based on this hypothesis, A. thaliana suspension cells were used for the measurements of both AsA and OA contents after being continuously exposed to O3 (30 ppb min−1). A significant decrease in total AsA content in the cells was observed after 60 min of O3 exposure (Fig. 2a). This decrease in AsA pool was characterized by a reduction of the oxidized DHA state, ca. around 63% of AsA molecules in the O3-treated cell culture compared with the air-treated cell (ca. around 80%) (Fig. 2a). We observed that an increase in OA content coincides with the decrease in pool of AsA during the O3 treatment (Fig. 2b). The OA content in the medium attained 236 ± 13 µm after O3 exposure. It is most likely that AsA, DHA and OA levels are all maintained at relatively low level as a consequence of their dilution from apoplast as the liquid phase of the cell wall is in almost free contact with the medium and thus the induced change accompanying the oxidative stress could be buffered.

Figure 2.

Oxalic acid is produced from AsA in response to O3 in Arabidopsis thaliana cells. Metabolic contents in ascorbate and DHA (a) and oxalic acid (b) in cells after exposure to air with or without O3 (30 ppb min−1) during 30 and 60 min. Data correspond to mean values ± SE of at least three independent experiments. *Significantly different from controls, P < 0.05. AsA, ascorbic acid; DHA, dehydroascorbate.

We further checked in planta if OA content could be increased upon O3 treatment or not. Using 15 to 20-day-old Arabidopsis seedlings, a large increase in OA level was observed, reaching 338 ± 18 nmol mg FW−1 after O3 exposure compared with the air-treated seedlings (46 ± 12 nmol mgFW−1) (Fig. 3a). This increases in OA content was accompanied by a diminution of the AsA pool (Fig. 3b). It is noteworthy that in the same manner, the oxidized DHA state was diminished after O3 treatment (ca. 50%) compared with the air-treated seedling (ca. 73%). To assess the putative role of AsA in OA generation, we used the A. thaliana vtc 1-1 mutant which has merely 10–30% of AsA content compared with wild-type plants (Conklin et al. 2000). Upon O3 treatment of vtc 1-1 seedlings, no significant increase in OA level was observed (Fig. 3b), but a significant increase in AsA pool was observed reaching 2.1-fold (Fig. 3b). Taken together, the said results clearly suggest that rapid generation of OA in response to O3 treatment is due to degradation of AsA and notably DHA in A. thaliana. We thus further checked for the impact of OA in the course of cellular responses to O3, especially the early events in the O3-responsive signal transduction model previously proposed by our group (Kadono et al. 2010).

Figure 3.

Oxalic acid is produced from AsA in response to O3 in Arabidopsis thaliana seedlings. Oxalic acid contents (a) and contents in ascorbate and DHA (b) in 15-day-old seedlings of A. thaliana (Col-0), wild-type and vtc 1-1 mutant, after exposure to air with or without O3 (30 ppb min−1). Data correspond to mean values ± SE of at least three independent experiments. *Significantly different from controls, P < 0.05. AsA, ascorbic acid; DHA, dehydroascorbate; WT, wild type.

OA induces activation of anion channels

In cultured cells of A. thaliana, we previously showed that millimolar concentrations of OA reportedly induces PCD by involving the activation of anion channel (Errakhi et al. 2008), as observed upon O3 treatment (Kadono et al. 2010). Here, we therefore investigated if micromolar concentrations of OA, in the range inducible in O3-treated cells and tissue as we observed previously, are capable of regulating cell polarization and ion currents.

We recorded a rapid depolarization of about +15 ± 6 mV (n = 16) in the cells treated with 100 µm OA (Fig. 4a), from a mean membrane potential of −44 ± 4 mV (n = 16) recorded prior to the treatment. The observed depolarization was correlated with an increase in ion current to −1.48 ± 0.2 nA (after 6 s for a pulse of −200 mV, n = 16), from a mean control value of −0.59 ± 0.3 nA (n = 16) (Fig. 4b). In these conditions, addition of 10 mm choline chloride induced a negative shift of the reversal potential of this current of −15 ± 3 mV (n = 5, not shown) as expected for an anion current, considering an increase of the main permeant anions (NO3-, Cl-) in the 4-day-old cell culture from 11 mm (Reboutier et al. 2002) to 21 mm, with an internal concentration estimated at 5 mm. In our model, this type of current was shown to be highly sensitive to a series of structurally unrelated anion channel inhibitors (Reboutier et al. 2002; Kadono et al. 2010). Accordingly, the increase in ion current after OA addition was effectively reversed after addition of 200 µm of glibenclamide, an anion channel blocker, confirming the anionic nature of these currents (Fig. 4b). Although this current presents the features of slow anion channels (Schroeder & Keller 1992), some extent of the instantaneous current could be carried out by fast activating anion channels as described for guard cells (Hedrich, Busch & Raschke 1990). However, these data suggest that OA, as a secondary toxicant, could participate in the early O3-responsive signalling events by increasing anion currents and subsequent cell depolarization.

Figure 4.

Oxalic acid-induced depolarization and anion current increase. (a) Typical depolarization of a cultured cell in response to OA (left). Mean values of PM potentials before and after a 100 µm OA treatment (right). (b) Anion currents measured before and after 100 µm OA treatment, and after subsequent glibenclamide (200 µm) addition. The protocol was as illustrated, holding potential (Vh) was Vm (left). Corresponding current–voltage relationships at 6 s (right up). Mean values of anion currents recorded at −200 mV and 6 s, after the 10 min pulse of 10 min of air with or without O3 and with 200 µm of glibenclamide (right bottom). Recordings were done in the culture medium, main ions after 4 d of culture: 9 mm K+, 11 mm NO3- (Reboutier et al. 2002). Data correspond to mean values ± SD of at least 16 independent experiments. *Significantly different from controls, P < 0.05 and **significantly different from the OA-treated cells alone, P < 0.05. Gli, glibenclamide; O3, ozone; OA, oxalic acid; PM, plasma membrane; inline image, 100 µm OA; ▵, Control; ○, 200 µm Gli.

OA triggered cytosolic calcium increase

The changes in [Ca2+]cyt in response to O3 have also been reported for the cells of A. thaliana (Kadono et al. 2010) and tobacco (Kadono et al. 2006). We thus monitored the [Ca2+]cyt variations in the transgenic cells of A. thaliana expressing apoaequorin exclusively addressed in the cytosol. A rapid and transient increase in [Ca2+]cyt was induced in response to OA in a dose-dependant manner (Fig. 5a,b). The [Ca2+]cyt variation was even inducible by micromolar concentrations of OA (50, 100 and 200 µm), and the maximal increase was attained within 3–4 min after addition of OA (Fig. 5a). After peaking of [Ca2+]cyt increase, the level of [Ca2+]cyt returned to the basal level after 20 min (not shown). Taken together, OA production via ozonolysis of AsA could be one of efficient path for stimulating the increase in [Ca2+]cyt.

Figure 5.

Oxalic acid (OA)-induced variations of cytosolic Ca2+ in Arabidopsis thaliana cells. (a) Mean kinetics of [Ca2+]cyt variation and (b) mean value of increase recorded at peak, of aequorin-expressing cells in response to OA at different concentrations. Data correspond to mean values of at least five independent experiments and error bars correspond to SE. *Significantly different from controls, P < 0.05.

Oxalate-induced ROS generation in A. thaliana cultured cells

The mechanism for generation of ROS, such as 1O2, O2•− and hydrogen peroxide (H2O2), mediated via O3 degradation in the apoplast, has been well documented and beleaved to be involved in the O3-induced damages to biological systems (Sandermann et al. 1998). The case of plant cell damages via identical model was also reported in our previous study (Kadono et al. 2010). Exposure of Arabidopis cells to the micromolar concentrations of OA resulted in a rapid- and dose-dependent enhancement of the yield of CLA chemiluminescence (Fig. 6a,b). Because CLA is responsive to both O2•− and 1O2, Tiron and DABCO, scavengers of O2•− and 1O2, respectively, were used to identify the member(s) of ROS involved (Kadono et al. 2010). Pretreament of the cells with DABCO (5 mm) allowed a significant decrease in CLA-chemiluminescence while Tiron (5 mm) failed to give any impact (Fig. 6c). Therefore, the rapid increase in CLA chemiluminescence induced in the presence of OA appeared to be reflecting the production of 1O2 rather than O2•− (Fig. 6b,c).

Figure 6.

Oxalic acid (OA)-induced variations of reactive oxygen species (ROS) generation in Arabidopsis thaliana cells. (a) Typical trace of ROS generation in cells after 200 µm OA treatment and (b) mean value of ROS generation after OA at 0, 50, 100, 200 and 500 µm. The production of ROS was monitored by the chemiluminescence of the Cypridina luciferin analog (Kadono et al. 2010). (c) Effect of DABCO and Tiron, scavengers of 1O2 and O2- respectively, on ROS induced by 200 µm oxalate. Data correspond to mean values of at least five independent experiments and error bars correspond to SE. *Significantly different from controls, P < 0.05 and **significantly different from the OA-treated cells alone, P < 0.05.

Oxalate-induced cell death in A. thaliana cultured cells

Millimolar concentrations of OA were shown to induce the PCD in A. thaliana cells (Errakhi et al. 2008). We checked if OA at micromolar concentrations was able to induce cell death. Indeed, a dose-dependent cell death reach about 20% of the cell population, with 500 µm OA after 16 h (Fig. 7a,b). In order to check whether this OA-induced cell death was due to an active mechanism requiring gene expression and active cellular metabolism as previously reported for higher concentrations (Errakhi et al. 2008), A. thaliana cells were pretreated with actinomycin D (AD), an inhibitor of RNA synthesis, or with cycloheximide (Chx), an inhibitor of protein synthesis. Although AD and Chx induced a slight cell death themselves, both these inhibitors reduced the OA-induced cell death (Fig. 7b). These data support the hypothesis that OA can contribute to active cell death increase in response to O3.

Figure 7.

Oxalic acid (OA)-induced cell death in Arabidopsis thaliana cells. Effect of increasing concentrations of OA on (a) FDA and (b) Evans blue estimated cell viability after 16 h of treatment with or without pretreatment with actinomycin D (AD, 20 µg mL−1) or cycloheximide (Chx, 20 µg mL−1). Data reflect the means and SE of at least four independent experiments. *Significantly different from controls, P < 0.05 and **significantly different from the OA-treated cells, P < 0.05. FDA, fluorescein diacetate.


Our working hypothesis was that OA could be produced upon AsA degradation during O3 stress (Sandermann 2008), and then participates to signalling pathways leading to O3-induced PCD. The AsA pool of the plant is currently considered to be an important part of the initial line of defence against oxidative stress (Foyer & Noctor 2009) like those generated by O3 (Sandermann 2008). According to our hypothesis, we observed after the O3 treatment a concomitant decrease of the total AsA pool and an increase of OA levels. It is interesting to note that the increase of OA level occurred at cellular level but was more significant in the extracellular medium of A. thaliana cultured cells (Supporting Information Fig. S1a). Moreover, determination of OA performed on seedlings has demonstrated the O3-dependent accumulation of OA in wild-type seedlings but not in vtc 1-1, a low-AsA maintaining mutant (Conklin et al. 2000), reinforcing the hypothesis on the role for AsA in the generation of OA upon O3 exposure (Fig. 3). On one hand, an increase in total AsA pool as well as DHA content was observed in vtc 1-1 mutant in response to O3. On the other hand, no significant modification in OA content was observed in this mutant, probably due to the threshold detection; however its production cannot be excluded and could serve as a signalling molecule in O3-induced PCD as vtc1-1 is extremely O3 sensitive. Nonetheless, Green & Fry (2005) showed that OA was generated from extracellular DHA degradation in cultured Rosa cells; it is reasonable to suppose that this reaction could occur in Arabidopsis cells as oxidized DHA states are reduced 60 min after O3 exposure. However, we failed to detect a decrease in DHA in the external medium, but an increase in the ratio oxidized/reduced state of AsA pool (Supporting Information Fig. S1b). Although the transport of DHA to cytoplasm for reduction to AsA and transport back could be regarded as a limiting step, it is noteworthy that in cultured cells the export of AsA from cytoplasm following H2O2 treatment could occur in a few minutes (Parsons & Fry 2010). Furthermore, in our cellular model, as expected from the role of AsA as antioxidant, we observed the effect of oxidative stress, that is, decrease in AsA pool and increase in OA content, after 60 min of a continuous exposure to O3, although it can be observed in the same trend at 30 min. This phenomenon could be explained by the recovery of AsA level, that is, a possible combination of DHA reduction and a new synthesis of AsA, in the loss of the O3-induced degradation content during the first 30 min of O3 challenge, suggesting that the AsA/DHA turnover was efficient and quite rapid. Thus, in our living cells continuously challenged by O3, the dynamic fluxes of AsA/DHA did not allow a quantitative estimation of the OA production from apoplastic DHA degradation. Because we used young seedlings and culture cells, it remains an open question as to whether O3 reacts directly with AsA in the complex environment of the apoplast of mature leaves, particularly in plant species where the extracellular ascorbate pool is primarily oxidized. Moreover, an efflux of OA from cytoplasm could also participate to increase in apoplastic OA because under abiotic stress condition, OA efflux was also reported (Zheng et al. 2005). Indeed, different OA biosynthetic pathways exist within plants with glycolate oxidase or lactate dehydrogenase for example (Libert & Franceschi 1987). However, taken together, these data support the view that the degradation of the pool of AsA by O3 results in micromolar concentrations of OA in A. thaliana cultured cells.

We thus analysed how this generation of OA could be involved in early cellular events induced by O3. We recently showed that an interplay between anion channel activation [Ca2+]cyt increase, and ROS generation was an early prerequisite to the morphological and biochemical events participating to O3-induced PCD in A. thaliana cells (Kadono et al. 2010). A rapid increase in anion current and a depolarization of the PM were recorded after treatment of cells with 100 µm OA, a concentration of the same range that the amount of OA released in our cultured cells upon O3 treatment. The anion current increase was lower than the one previously reported with 6 mm OA (Errakhi et al. 2008), suggesting a dose-dependent effect of OA on anion channel activity. This increase in anion currents was not an effect of pH shift due to OA, as the addition of OA adjusted to pH 5.8 (i.e. cell culture medium pH) induced the same increase in anion current (data not shown). The recorded anion currents shared characteristic kinetic features of S-type anion channels, as shown previously in response to O3 treatment (Kadono et al. 2010). The addition of glibenclamide, an anion channel inhibitor effective in A. thaliana suspension cells (Kadono et al. 2010), strongly reduced the OA-induced increase in anion current and PM depolarization as previously observed for O3 (Kadono et al. 2010), thus suggesting the participation of same type of anion channel in both OA-induced process and O3-induced process. The S-type anion channels were characterized and identified in guard cells (Schroeder & Keller 1992; Vahisalu et al. 2008) where they provide the mechanism responsible for guard cell shrinkage leading to stomatal closure in response to O3 (Vahisalu et al. 2008). As for guard cells, these channels may participate to O3-induced shrinkage of cultured cells (Kadono et al. 2010). Cell shrinkage is a hallmark of animal apoptosis and PCD response in plant, and the early regulation of channels allowing long-term efflux of anions seemed to be a conserved event in both plant (Errakhi et al. 2008; Kadono et al. 2010) and animal (Okada et al. 2006).

We further could record a [Ca2+]cyt increase in response to OA in micromolar range (Fig. 5). Thus, despite the high affinity of OA for Ca2+ (Cessna et al. 2000), micromolar concentrations of OA could participate in O3-induced increase in [Ca2+]cyt previously reported in various models (Kadono et al. 2006, 2010). We could also detect an increase in ROS generation upon OA treatment in the micromolar range (Fig. 6). The impact of OA on ROS generation is complex because OA could inhibit the production of H2O2 in tobacco and soybean cultured cells (Cessna et al. 2000) and increase ROS levels, this last event being correlated with PCD (Kim et al. 2008). The generation of ROS was also described as an early cellular event induced by O3 in our model (Kadono et al. 2010) and, it is noteworthy that ROS generation upon O3 challenge could occur through various chemical and biological processes (Sandermann 2008). According to the effect of DABCO, a scavenger of 1O2, the main ROS produced in response to 200 µm OA was 1O2. The generation of 1O2 in reponse to O3 pulse was suggested in Sedum album L. leaves (Kanofsky & Sima 1995) and observed in our model (Kadono et al. 2010). Finally, PCD being observed in response to O3 as in response to millimolar concentrations of OA (Errakhi et al. 2008; Kim et al. 2008), we checked the impact of micromolar concentration of OA on cell viability. Our data are reminiscent of the dose-dependent OA-induced cell death observed for millimolar concentrations in A. thaliana cells (Errakhi et al. 2008). We observed a 20% increase in dose-dependent cell death in response to micromolar concentrations of OA (Fig. 7). In this range of concentration, the pH shift due to OA addition did not seem to participate to extent of cell death, as addition of OA adjusted at the pH 5.8 induced equivalent cell death in A. thaliana cells (Errakhi et al. 2008).

Previously, we suggested some O3-induced signalling pathways leading to cell death (Kadono et al. 2010). Collectively, our present data strongly suggest that AsA reacting with O3 could yield OA and, as a secondary toxicant, could thus induce different cellular events already described in O3 responses. Therfore, by activating anion channel, producing ROS and favouring [Ca2+]cyt increase, OA could fuel a cellular network of events participating to O3-induced PCD.


T.K. was supported by a grant from the JSPS, and M.L.M. by a grant from the Ministero de Educacion de La Nacion Argentina.