Plant stress responses are a key factor in steering the development of cells, tissues, and organs. However, the stress-induced signal transduction cascades that control localized growth and cell size/differentiation are not well understood. It is reported here that oxidative stress, exerted by paraquat or alloxan, induced localized cell proliferation in intact seedlings, in isolated root segments, and at the single cell level. Analysis of the stress-induced mitotic activity revealed that oxidative stress enhances auxin-dependent growth cycle reactivation. Based on the similarities between responses at plant, tissue, or single cell level, it is hypothesized that a common mechanism of reactive oxygen species enhanced auxin-responsiveness underlies the stress-induced re-orientation of growth, and that stress-induced effects on the protoplast growth cycle are directly relevant in terms of understanding whole plant behaviour.
The production of reactive oxygen species (ROS) is normally carefully controlled by the plant. A dynamic equilibrium exists between the formation of ROS and the activity of the antioxidant scavenging systems (Hancock et al., 2001; Mittler, 2002). However, when plants are exposed to abiotic stress, high levels of ROS commonly accumulate (for review, see Vranova et al., 2002), causing oxidative damage at the cellular and molecular level (Mittler et al., 1999; Mittler, 2002), which, in turn, leads to a decrease in plant productivity. ROS can rapidly oxidize membrane lipids, proteins, and other cellular organelles, leading to their dysfunction. To protect themselves from ROS, plants possess a number of free-radical scavenging enzymes such as ascorbate peroxidase (APX), catalase, and superoxide dismutase, and low-molecular weight antioxidants, like ascorbate and tocopherols (Hancock et al., 2001; Vranova et al., 2002).
The ability of plants to counteract stress conditions depends on the efficiency and speed at which they recognize the stress, generate signal molecules, and activate stress-protective mechanisms. Following the recognition of a particular stress condition, downstream signalling events are activated and these, in turn, initiate activation of genes encoding ROS-scavenging enzymes (Hancock et al., 2001). Indeed, ROS play a vital role as messengers in plants and vertebrates (Hancock et al., 2001; Neill et al., 2002). In plants, various ROS, including H2O2 and nitric oxide, are involved in signalling pathways leading to alternations in ion fluxes, activation of protein kinases, and changes in gene expression (Hancock et al., 2001). ROS-induced kinases include the mitogen-activated protein kinases, which play an essential role in cell cycle progression and cell division (Hirt, 2000; Kovtun et al., 2000). However, ROS can also play a role in inducing apoptosis in plant cells (Dat et al., 2003), or impact on developmental processes like leaf elongation (Rodriguez et al., 2002; Schopfer et al., 2002) and embryogenesis. Oxidative stress can serve as a modulator of somatic embryogenesis in plants by inducing autonomous cell division (Cui et al., 1999; Luo et al., 2001; Pasternak et al., 2002). However, relatively little is known about how ROS regulate plant growth and development under stress conditions, and how they interact with other signal molecules, including phytohormones. Auxin and H2O2 have antagonistic effects on cell cycle progression and gene activation (Hirt, 2000; Kovtun et al., 2000). ROS may block the cell cycle at specific checkpoints (S- and M-phase transition) both in plant and in vertebrate cells (Paulovich et al., 1997; Reichheld et al., 1999). Moreover, the expression of several stress response genes and the G1–S transition are both controlled by the same transcriptional activators, the E2F gene products (De Veylder et al., 2003).
A possible and emerging candidate for an intermediate function in both stress and growth responses seems to be the phytohormone auxin. Mild oxidative stresses mimic auxin stimuli in somatic embryogenesis (Dudits et al., 1995; Pasternak et al., 2002). Moreover, ROS seem to be involved in a typical auxin-mediated phenomenon like root gravitropism (Joo et al., 2001). On a whole plant level, a mild stress generates phenotypical changes reminiscent of a 2,3,5-triiodobenzoic acid (TIBA)-like disturbance of auxin distribution (Pasternak et al., 2005).
Using Arabidopsis, the aim of this study was to analyse, at the physiological and molecular level, the interaction between oxidative stress-inducing agents and auxin with respect to growth reorientation. This study integrates responses at the level of the whole plant (germinated seedlings), the isolated organ (root segments), and the isolated single cell (leaf protoplasts). Oxidative stress is exerted by the well-characterized photosystem I electron acceptor paraquat, or by the H2O2-generating compound alloxan (Takasu et al., 1991; Washburn and Wells, 1997; Davis et al., 1998). As presented here, oxidative stress can induce a transition from lateral root expansion to the formation of compact globular structures. In conclusion, oxidative stress enhances auxin-driven cell division, cluster formation, and/or lateral root formation at the level of the whole plant, the isolated organ, or the isolated cells.
Materials and methods
Whole plant experiments
Seeds of Arabidopsis thaliana (L.) Heynh. var. Columbia were surface-sterilized for 10 min in 20% commercial bleach and rinsed three times in sterile distilled water. They were then placed on half-strength Murashige and Skoog medium (Murashige and Skoog, 1962) containing 1.5% sucrose and 0.8% agar, and kept under a light regime of 16 h light (40 W m−2) and 8 h dark at a temperature of 24 °C. For each experiment and treatment, 40 seeds were cultured in a plastic 100 mm Petri dish, which contained 20 ml of the medium. Where indicated, paraquat and alloxan (Sigma) were dissolved in water, filter sterilized, and added to the medium. Auxins [indole-3-acetic acid (IAA), naphthyl acetic acid (NAA), or 2,4-dichlorophenoxyacetic acid (2,4-D)] were dissolved in DMSO (stock at 10 mg ml−1) and added to the sterilized medium.
Plantlets (12–14-d-old) were decapitated just above the hypocotyl. The remaining root and hypocotyl sections were cultured on liquid half-strength Murashige and Skoog medium (Murashige and Skoog, 1962) containing 1.5% sucrose, and kept under a light regime of 16 h light (40 W m−2) and 8 h dark at a temperature of 24 °C. Where indicated, IAA, alloxan, or paraquat were added to the liquid medium.
Root segment experiments
Roots from 2–3-week-old plants were cut in 5–7-mm-long segments and placed in liquid T39 medium (Pasternak et al., 1999, 2000) containing 0.2 mg l−1 6-benzylaminopurine. Where indicated, IAA, alloxan, or paraquat were added to the liquid medium. After 72 h of cultivation, segments were washed and transferred to proliferation medium (T39 medium supplemented with 2 mg l−1 kinetin and 0.1 mg l−1 IAA).
Leaf protoplasts were isolated from 3-week-old Arabidopsis plantlets raised in a greenhouse. Leaves were sterilized for 1 min in 20% commercial bleach, rinsed four times using sterile distilled water, and finely cut and incubated in a solution containing 0.48 M sucrose, 0.05 M glycine, 2 mM MES (pH 5.4), 5 mM CaCl2, 0.4% Cellulase Onozuka R10, and 0.2% Driselase. Digestion at 24 °C lasted 4–5 h. Isolated protoplasts were washed and resuspended in DPTK-1 culture medium (Kao and Michayluk, 1975) with appropriate growth regulators and stress agents.
Determination of morphological parameters
Arabidopsis seedlings and root segments were viewed under a Zeiss stereomicroscope. Photographs were taken using an Olympus digital camera under appropriate magnification (×3 to ×20). Primary root, cotyledon, and leaf sizes were determined using SPOT advanced software following imaging and processing. Results are expressed as mean ±standard error (n=20 explants). Each measurement was performed in at least three biological replications (each comprising about 40 seedlings). Cell sizes in petioles and leaves were measured under a Zeiss Standard 25 microscope using an ocular-micrometer.
Determination of cell division frequency and viability
Cell viability was determined using Evan's blue, a non-permeating dye, that leaks through ruptured plasma membranes and stains the contents of dead cells (Baker and Mock, 1994). Cell size was determined under a light microscope equipped with an ocular micrometer using a ×40 objective. Cell size is expressed as the average of the length and width of the cells measured individually. Elongation represents the ratio of cell length to cell width. At least 30 randomly chosen cells were measured per experiment. The system was calibrated using an object micrometer. Cytokinesis frequency was determined microscopically. The frequency of dividing cells was determined by inspecting at least 500 cells. For nuclear staining, plants were fixed with 2% formaldehyde in PBS for 30 min, washed twice with PBS, and stained with 4,6-diamidino-2-phenylindole (DAPI; 0.1 mg l−1) for 10 min. DAPI staining was observed under a fluorescent Axiovert 135 M microscope (Zeiss) at an excitation wavelength of 365 nm.
Ascorbate and glutathione determination
Cells were counted, collected by centrifugation, resuspended in 3% meta-phosphoric acid, and snap frozen in liquid N2. Ascorbate (ASC) and glutathione (GSH) were extracted by subjecting the cells to three cycles of freezing and thawing. Samples were then centrifuged at 10 000 g for 30 min at 4 °C and subsequently stored at −20 °C until analysis. During extraction and analysis, samples were kept in the dark and at 4 °C to minimize oxidation.
ASC/GSH determination was carried out on reverse phase HPLC. Separation occurred over a reverse-phase type C-18 column (3 μm particle diameter, Polaris 3 C18 kolom, Chromsep SS 100×4.6 mm; Varian Europe), which was kept at a constant temperature of 40 °C in a column oven (CTO-10AVP, Shimadzu). The flow was driven by an isocratic pump (LC-10ADVP, Shimadzu). The mobile phase consisted of 2 mM KCl, set at pH 2.5 by drop-wise addition of concentrated o-phosphoric acid. The volume flow rate of this mobile phase was set at 0.8 ml min−1. Oxygen was removed from the mobile phase by passing the flow through a degasser (DGU-14A, Shimadzu). Injection was performed via an auto-sampling unit (SIL-10ADVP, Shimadzu). Detection occurred via diode array (SPD-M10AVP, Shimadzu), measuring between 180 and 250 nm, and set in tandem with a home-made amperometric detection system (glassy carbon working electrode, calomel reference electrode, reference potential 1000 mV). The latter was kept at a constant temperature within the column oven. The electrochemical detector was connected to a personal computer via an SS420 board (Shimadzu). Chromatogram analysis was performed with the ClassVP software package (Shimadzu HPLC Class VP 612 SP5; Shimadzu). The whole system was controlled via the personal computer, in conjunction with a separate system controlling unit (SCL-10AVP; Shimadzu). Concentrations of antioxidants were expressed in pmol 10−6 cells.
Polymerase chain reaction (PCR) analysis
Protoplasts were frozen in liquid nitrogen, homogenized in Eppendorff tubes using a micro-pestle. Total RNA was extracted using Concert RNA isolation reagent (Invitrogen). First-strand cDNA templates for PCR were prepared using M-MULV-reverse transcriptase (RT) (Fermentas) with oligo-dT 18 primers. The expression of several genes was studied using RT-PCR as detailed (Aloni et al., 2003). To study gene expression, the following primers were used:
Indole-3-glycerol-phosphate synthase (IGS), sense: 5′-CAGCGTTTTGACAGACCAGA-3′ and antisense: 5′-CCAACAAGCTCGATTCCTTC-3′, yielding a 300 bp amplificate
Nitrilase (NIT1, NIT2, and NIT3), sense: 5′-ATCCCCGTTTACGACACT-3′ and antisense: 5′-ACGAAACATCCACCTTC-3′, yielding a 185 bp amplificate
Cytosolic ascorbate peroxidase (cAPX), sense: 5′-CCAACCGTGAGCGAAGATT-3′ and antisense: 5′-TAAGCATCAGCAAACCCAAG-3′, yielding a 689 bp amplificate
Dehydroascorbate reductase (DHAR), sense: 5′-TGGCTCTAGATATCTGCGTGA-3′ and antisense: 5′-CATTCACCTTCGATTCCCAA-3′, yielding a 639 bp amplificate
Indole-amino acid hydrolase (ILL1 and ILL2), sense: 5′-TCACTGGGAAAGGAGGTCAT-3′ and antisense: 5′-CAAAAGCTTCTTGACCCAACA-3′, yielding a 402 bp amplificate
PIN1, sense: 5′-GGTGGTGGTCGGAACTCTAA-3′ and antisense: 5′-TCGTAGCGGTGGAGTAATCG-3′, yielding a 400 bp amplificate
PIN3, sense: 5′-ACGTTTTCGGCGGAGCACCTG-3′ and antisense: 5′-TGCCACTGAATTCCCACAAGC-3′, yielding a 552 bp amplificate
Cytosolic glyceraldehyde-3-phosphate dehydrogenase (GAPDH), sense: 5′-GAATCAACGGATTCGGAAGA-3′ and antisense: 5′-AACAACCTTCTTGGCACCAC-3′, yielding a 347 bp amplifcate
Gluthatione transferase (GST1), sense: 5′-AAGGGCGTTGCCTTCGAGACC-3′ and antisense: 5′-TGGGAATCCCATGACTGATGC-3′, yielding a 313 bp amplificate
Glutathione synthase 1 (GSHS1), sense: 5′-TCTGACAGCTGACTGGACTCC-3′ and antisense: 5′-TACAGCAGCTCTTCGAACACG-3′, yielding a 287 bp amplificate
Glutathione synthase 2 (GSHS2), sense: 5′-TGGCTAAACCAGGTGTTCTCG-3′ and (antisense: 5′-ACTCCAAAACCAGCTGCAACG-3′, yielding a 458 bp amplificate
All RT-PCR reactions were run for 34 cycles (Eppendorf MasterCycler personal); conditions were as follows: NIT (95 °C for 2 min); 94 °C for 1 min, 42 °C for 40 s, 72 °C for 40 s, (72 °C for 7 min); GAPDH, IGS, PIN1, and ILL (95 °C for 2 min), 94 °C for 1 min, 49 °C for 40 s, 72 °C for 40 s, (72 °C for 7 min); PIN3, GSHS1, GSHS2, and GST (95 °C for 2 min), 94 °C for 1 min, 56 °C for 40 s, 72 °C for 40 s, (72 °C for 7 min). RT-PCR fragments were separated on 2% agarose gels. The expression of the housekeeping gene, cytosolic GAPDH, was used as an internal control.
Growth and development of Arabidopsis seedlings exposed to oxidative stress
Arabidopsis thaliana seeds were germinated on solid medium supplemented with either paraquat (0.2 μM) or alloxan (0.75, 1, or 1.5 mM). Effects of oxidative stress were studied in terms of testa rupture and root and leaf development. Testa rupture was accelerated when seeds were exposed to either paraquat or alloxan. For example, 48 h after imbibition, 54% of the 0.2 μM paraquat-treated seeds, 70% of the 1 mM alloxan-treated seeds, and just 47% of the control population showed testa rupture (Table 1).
0.75 mM alloxan
0.2 μM paraquat
|% Testa rupture at t = 48 h||46.9±1.9||70.1±3.3**||54.0±2.4|
|No. of leaves after 12 d||4.25±0.15||ND||4.31±0.12|
|No. of leaves after 14 d||5.18±0.22||5.26±0.17||ND|
0.75 mM alloxan
0.2 μM paraquat
|% Testa rupture at t = 48 h||46.9±1.9||70.1±3.3**||54.0±2.4|
|No. of leaves after 12 d||4.25±0.15||ND||4.31±0.12|
|No. of leaves after 14 d||5.18±0.22||5.26±0.17||ND|
The percentage of seeds with visible testa ruptured 48 h after imbibition, as well as the number of leaves formed after 12 or 14 d of treatment with either 0.2 μM paraquat or 0.75 mM alloxan. Standard errors are indicated. **, P <0.01; ND, not determined.
Paraquat or alloxan did not retard significantly the dynamics of seedling cotyledon-opening and leaf-forming reactions in the shoot apex, i.e. neither compound affected the rate at which new leaves formed. The first true leaves emerged 4 d after germination irrespective of alloxan or paraquat exposure. Control and paraquat-treated plants had formed 4.25 and 4.31 leaves, respectively, 12 d after germination (Table 1). In another experiment, control and alloxan-treated plants had formed 5.18 and 5.26 leaves, respectively, 14 d after germination. However, cotyledons of plants treated with paraquat (0.2 μM) or alloxan (0.75 mM and 1.5 mM) are significantly smaller compared with control ones (Fig. 1A). As cotyledon cells do not demonstrate mitotic activity, it was concluded that the oxidative stress treatment inhibited cotyledon cell expansion/elongation. Exposure of Arabidopsis seedlings to auxin, IAA (0.5 mg l−1) also yielded a phenotype with small cotyledons (Fig. 1A).
The lack of cell expansion in alloxan- and paraquat-treated plants is also reflected in the delayed emergence of the inflorescence. Nearly all 1-month-old control Arabidopsis plants were flowering under the present growth conditions. However, only 25% of the plants exposed to 0.5 mM alloxan, just 56% of the plants exposed to 0.1 μM paraquat, and 29% of the plants exposed to 0.2 μM paraquat flowered at that time.
Both paraquat and alloxan inhibited significantly (P <0.005) elongation of the main Arabidopsis seedling root (Fig. 1A). In parallel, new lateral roots were formed and emerged in both paraquat- and alloxan-treated Arabidopsis seedlings (Fig. 1B–D). These newly formed, lateral root meristems could be detected microscopically after 4–5 d exposure to either paraquat or alloxan. In more extreme cases, exposure of seedlings to 1.5 mM alloxan resulted in a cessation of root tip growth, disappearance of the root tip meristematic zone, concentration of root hair growth in a zone near the tip, and the formation of multiple, newly formed, lateral roots in the same root zone (Fig. 1C, D). In another extreme case, 2 mM of alloxan was found to induce the formation of adventitious roots, i.e. roots formed directly at the hypocotyl end (Fig. 1D). These effects of paraquat and alloxan can be mimicked by germinating plants on medium containing IAA (0.5 mg l−1). The phytohormone inhibits root and cotyledon elongation (Fig. 1A), and is known to boost the formation of lateral root meristems.
The presence of a visible nucleolus is correlated with the transcriptional activity in a cell (Kononowicz and Janick, 1988; Warner, 1989; Medina et al., 2000). DAPI only stains DNA and provides an adequate way of visualizing these differences in transcriptional activity. Actively dividing cells in the apical root meristem, each showing a large nucleolus (Fig. 2A, B) also transcribe much more actively than root hair cells (Fig. 2C, D). These actively transcribing nuclei are also rather round instead of elongated (Fig. 2A, C). Upon leaving the division zone in the root tip and entering the elongation zone and, later, the differentiation zone, the cell's nucleolus will disappear (Fig. 2D). A comparison of control and alloxan-treated cells in the vascular cylinder and the pericycle shows that the nucleus and the nucleolus were differentially structured. After alloxan treatment, nucleoli became visible in the pericycle cells (Fig. 2E, F), whereas this did not occur in control cells (Fig. 2G, H). Moreover, the nucleus became round in the pericycle cells. These observations suggest that the cells in the pericycle of alloxan-treated plants are much more transcriptionally active than their control counterparts. The formation of lateral root meristems, induced by low concentrations of paraquat or alloxan, was initiated in the pericycle, but not in the cortex or epidermis of Arabidopsis seedlings. The observed chromatin decondensation may thus also be interpreted as a renewed start of growth and development.
Development of lateral root primordia in decapitated Arabidopsis seedlings exposed to oxidative stress
The stress-induced formation of lateral roots in Arabidopsis seedlings may be related to alterations in the auxin gradient (Pasternak et al., 2005). Auxin is a prerequisite for mitotic activity and meristem formation (Casimiro et al., 2001; Marchant et al., 2002). To test the link between auxin gradients and oxidative stress, lateral root formation was studied in decapitated 14-d-old Arabidopsis plants. These plantlets were decapitated just above the hypocotyl. The endogenous IAA present in the roots of Arabidopsis seedlings has its origin in the shoot (Bhalerao et al., 2002). Decapitated plants were either left untreated or, alternatively, exposed to 0.2 mg l−1 NAA or IAA and/or 0.75 mM alloxan. In these decapitation experiments similar results were obtained with NAA and IAA. However, somewhat more stable NAA yielded more pronounced results in long-term experiments. Only results obtained with NAA are presented in this paper.
Alloxan exposure (in the absence of exogenous auxin) did not induce lateral roots, not even after 3 weeks of culturing. Similarly, decapitated plants kept on medium without either alloxan or auxin, did not form any lateral roots.
Both control and alloxan-treated decapitated plants (still in the absence of any auxin) went through gradual degradation (as confirmed by DAPI staining). However, in the alloxan-treated seedlings, epidermal cells became very large and vacuolated in the zone close to the root tip (Fig. 3A, B), while pericycle cells in the central cylinder started to degenerate (loss of subcellular structure) (Fig. 3C). Moreover, alloxan exposure led to a decrease in the length of the root elongation zone (from 130 mm under control conditions without NAA, to 67.5 mm upon alloxan application; Table 2), similar to that observed in whole seedlings (Fig. 1A). These morphological changes correlated with a reactivation of the cell cycle in the pericycle (data not shown), as described for intact plants (see previous part).
No. of root primordia
Length of root elongation zone
|0.2 mg l−1||0.75 mM||7.9±0.5||ND|
|0.2 mg l−1||0 mM||3.6±0.4**||ND|
|0 mg l−1||0.75 mM||0||67.5±1.5|
|0 mg l−1||0 mM||0||130.0±5***|
No. of root primordia
Length of root elongation zone
|0.2 mg l−1||0.75 mM||7.9±0.5||ND|
|0.2 mg l−1||0 mM||3.6±0.4**||ND|
|0 mg l−1||0.75 mM||0||67.5±1.5|
|0 mg l−1||0 mM||0||130.0±5***|
Formation of root primordia and size of the root elongation zone in Arabidopsis plants after addition of 0.2 mg l−1 NAA and/or 0.75 mM alloxan. The root elongation zone in NAA-treated plants was impossible to define and its size was therefore not determined (ND). Standard errors are indicated. **, P <0.01; ***, P <0.005.
If decapitated seedlings were exposed to NAA (0.2 mg l−1), numerous lateral root primordia emerged during the first 2–3 d of culturing (Fig. 3D, E) and these grew out to form short lateral roots. The formation of lateral roots was strongly accelerated when decapitated seedlings were exposed to both NAA and alloxan (7.9 lateral primordia per centimetre upon addition of alloxan and NAA versus 3.6 primordia per centimetre when only NAA had been supplied) (Table 2). Under these conditions there is widespread mitotic activity throughout the pericycle. These observations have led to the belief that the oxidative stress exerted by alloxan stimulates auxin-dependent lateral root formation.
Stress changes cell-differentiation status in isolated Arabidopsis root segments
Significant changes in auxin metabolism occur during seed germination (Bhalerao et al., 2002). To separate stress-induced changes in hormonal metabolism from changes in auxin metabolism during plant development and outgrowth, lateral root formation was studied on isolated root segments (without a root meristem). Root segments do not synthesize IAA, and are dependent on exogenous auxin (Bhalerao et al., 2002). Arabidopsis root segments were cultured on medium supplemented with 2,4-D and/or alloxan, or without any addition. In the absence of supplemental 2,4-D, mitotic activity was not observed in the Arabidopsis segments, irrespective of the presence of either alloxan or paraquat. However, a low 2,4-D concentration (0.5 μM) induced some callus formation (Fig. 4A). Root segments kept on a medium containing a low 2,4-D concentration supplemented with 1 mM alloxan or 0.2 μM paraquat formed compact globular cell clusters (Fig. 4B, C), originating in the pericycle close to the central cylinder (Fig. 4C). The compact clusters are considered to be morphogenetic in nature for two reasons. First, the cell clusters show a clear internal organization pattern (Fig. 4B). Secondly, the clusters are able to develop subsequently into plants, when put on appropriate regeneration medium (Fig. 4D). Segments kept on medium containing a low concentration of 2,4-D, but neither alloxan nor paraquat, formed only a few globular structures, and those that were formed were late in time. A 10-fold increased 2,4-D concentration (5 μM) strongly stimulated the formation of morphogenic cell clusters, and these were morphologically similar to those formed under low 2,4-D supplemented with alloxan or paraquat (not shown). When exposed to a high 2,4-D concentration, clusters were formed even without stress supplementation. The frequency at which these dividing cell clusters were formed was rather similar in root segments exposed to high 2,4-D and segments exposed to a combination of low 2,4-D and oxidative stress treatments. In order to quantify the formation of morphogenic cell clusters, the isolated root segments were transferred to proliferation medium and the frequency of plantlet regeneration was quantified. Very few viable plantlets were regenerated from segments kept on 2,4-D only. By contrast, 30–45% of the segments exposed to both oxidative stress and a low 2,4-D concentration formed viable plantlets (Fig. 4D). Segments deprived of 2,4-D did not yield plantlets, but started to degrade.
Oxidative stress enhances growth cycle activity in isolated Arabidopsis leaf protoplasts
To investigate the molecular mechanism underlying the effects of oxidative stress on plant growth reorientation, the effects of alloxan on isolated leaf protoplasts were studied. Isolated Arabidopsis leaf protoplasts comprise a homogenous population of single cells, which are mainly in the G0 phase of the cell cycle (Zhao et al., 2001). Arabidopsis leaf protoplasts were kept on medium with or without exogenous NAA (1 mg l−1). NAA stimulated re-entry into the growth cycle, vis-à-vis a reactivation of the proliferative cell cycle mechanisms. By contrast, cell division was not observed in protoplasts maintained on medium without auxin, rather, the viability of these cells dropped steadily and most cells (80–90%) died within 4–5 d. Under these conditions (no auxin), the main alloxan effect is to accelerate the decrease in viability. However, in the presence of exogenous auxin, alloxan had a very different effect. A concentration of 0.5 mM alloxan strongly accelerated the growth cycle of isolated protoplasts, as indicated by the number of cells entering cytokinesis (Fig. 5A). Cytokinesis commonly results in small, cytoplasm-rich, viable cells (25–28 μm diameter for alfalfa protoplasts; see Pasternak et al., 2002). Consistently, alloxan/NAA treatment of Arabidopsis protoplasts results in a population of relatively small, non-elongated cells with a high viability (Fig. 5B–E), when compared with the auxin control (Fig. 5F).
Alloxan does not induce differences in anti-oxidative status
To evaluate whether alloxan induces changes in the stress defensive capacity of the protoplasts, ASC and GSH contents and redox state (i.e. the ratio between the amount of reduced and the total amount of reduced and oxidized ASC or GSH) were measured throughout the experiment. The cellular concentration of ASC was observed to fall dramatically within 24 h after protoplast isolation (Fig. 6A). This drop was followed by a very gradual recovery of the ASC levels. Overall, the ASC status of control and alloxan-treated cells followed a similar pattern, although, notably, alloxan-treated cells contained significantly less ASC than the control cells (P <0.05) at 72 h. The internal redox status of ASC also did not differ between both treatments.
GSH contents (Fig. 6B) rose steadily during the experiment under both conditions, but by 72 h, they were significantly higher in the alloxan-treated cells than in the control (P <0.005). However, the GSH redox status did not differ between treatments, and was constant throughout the experiment.
Auxin-related gene expression after alloxan treatment
Gene expression in freshly prepared protoplasts was compared with expression in protoplasts kept for 24 h on auxin (2.5 μM NAA) or auxin plus alloxan (0.8 mM) (Fig. 7). Several genes were expressed in 24-h-old protoplasts, and these include the housekeeping gene, GAPDH (Aloni et al., 2003), the IAA biosynthesis gene NIT (tryptophan-dependent pathway), the IAA biosynthesis gene IGS (TRP-independent pathway), the IAA conjugate hydrolase gene ILL, the ascorbate metabolism genes cyt-APX and DHAR, and the glutathione transferase gene GST1 (an auxin responsive gene). The expression of the polar auxin transport genes PIN3 and PIN1 (Galweiler et al., 1998) was decreased in the presence of alloxan.
Alloxan induces oxidative stress-linked morphogenic responses in plant cells
The data presented here demonstrate that abiotic stress-inducing agents (alloxan and paraquat) induce a reorientation of growth in seedlings. Treatment of Arabidopsis thaliana seedlings with two distinct, oxidative stress-inducing agents led to a range of physiological responses, including faster testa rupture, reduced root elongation, and reduced cotyledon and leaf expansion. These data extend previous work showing that the heavy metal Cu, the free-radical producer paraquat, salicylic acid, and a hydrogen peroxide analogue all induced the formation of lateral and axillary roots in A. thaliana seedlings. In addition, these treatments inhibit root and flower meristem formation and result in a decrease in cell elongation (Pasternak et al., 2005).
Alloxan is widely used in diabetes studies. It is believed to induce the formation of ROS (Takasu et al., 1991; Brömme et al., 2000). Supposedly, the compound is being reduced to dialuric acid by glutathione-dependent thioltransferases (as would be dehydroascorbate; Davis et al., 1998), which then interacts with oxygen to produce superoxide and hydrogen peroxide (Washburn and Wells, 1997). Given the presence of similar GSH-dependent DHARs in the plant cell (Potters et al., 2002), it is plausible that this mechanism may also take place in the plant cell. Consistent with a role as inducer of oxidative stress, there are similar effects of alloxan and paraquat on seedling development (Fig. 1) and root segment outgrowth (Fig. 3). As demonstrated before (Pasternak et al., 2005), plants demonstrate a common morphogenic response to a wide range of oxidative stressors. The morphological changes observed in alloxan-treated plants are reminiscent of the changes observed after addition of Cu, paraquat, salicylic acid, or tert-butyl hydrogen peroxide (Pasternak et al., 2005).
Accelerated germination in response to oxidative stress has been reported by a number of groups (Duval and NeSmith, 2000; Narimanov, 2000; Ogawa and Iwabuchi, 2001) and may be related to the oxidation of endogenous inhibitors of germination (Ogawa and Iwabuchi, 2001). In the data reported here, oxidative stress-inducing agents stimulated lateral root formation in seedlings, and enhanced several auxin-induced processes in decapitated seedlings, isolated protoplasts, and root segments. The similarities between growth activation at the seedling, root-segment, and protoplast level strongly suggest common regulatory mechanisms. By encompassing the responses at the plant, organ, and protoplast level, the data presented here create a link between the activation of mitosis/cell cycle activity and whole plant responses.
Synergy between auxin and oxidative stress
The phytohormone auxin plays a central role in the control of cell and plant growth. It can stimulate or inhibit cell expansion, stimulate cell division, promote differentiation of vascular tissues, inhibit shoot branching, and promote lateral root formation (Casimiro et al., 2001; Marchant et al., 2002; Aloni et al., 2003). The synthetic auxin analogue, 2,4-D, is especially active in inducing cell division, and indeed somatic embryogenesis. Other plant growth regulators, as well as environmental signals, also impinge on growth and development, and there is considerable crosstalk between the various signal transduction chains. Polar auxin transport is thought to play a key role in lateral root formation (Bhalerao et al., 2002), and interference with auxin gradients impacts on the formation of lateral roots (Casimiro et al., 2001; Marchant et al., 2002). The formation of lateral roots by seedlings comprises de novo formation of meristems following de-differentiation of pericycle cells (Fig. 1E, F). Those pericycle cells give rise to a lateral root meristem remaining in the G1/G2 phase, while other pericycle cells remain in the G0 phase (Beeckman et al., 2001). Exposure to an appropriate environmental signal will initiate cell division activity in the G2 cells. In the course of studies on the abiotic stress-induced reorientation of growth, only pericycle cells were observed to recover mitotic activity, while Arabidopsis cortex and epidermis cells appear to be irreversibly differentiated (Fig. 1E, F). Moreover, exposure to low doses of oxidative stress and to auxin is particularly effective in initiating cell division in the pericycle cells. In the absence of externally added auxin, neither alloxan nor paraquat have an effect on lateral root formation and/or cell division activity in decapitated plants, isolated root segments, or isolated protoplasts. Thus, the effect of the oxidative stress, exerted by either alloxan or paraquat, is to enhance an auxin-driven process, leading to cell division and the formation of morphogenic cell clusters (Fig. 3B, C). Note that Fe-stress boosts cell division rates of alfalfa cells on medium containing auxin (Pasternak et al., 2002). Similarly, salicylic acid and H2O2 promoted embryogenesis in callus cultures of Astragalus adsurgens (Luo et al., 2001). Inhibition of catalase activity, as well as direct additions of H2O2 promoted embryogenesis in Lycium barbarum callus cultures (Cui et al., 1999). The H2O2 scavenger dimethylthiourea also impedes embryogenesis in callus cultures of Astragalus adsurgens (Luo et al., 2001) and Medicago sativa (T Pasternak, unpublished results). A wide variety of abiotic stresses can induce formation of ROS. Consistently, the embryonic potential of tissue cultures is enhanced by various stress treatments, including osmotic, heavy metal, and dehydration stress (Ikeda-Iwai et al., 2003; Kumria et al., 2003). These observations lead to the hypothesis that the auxin and oxidative stress complement each other in common stress responses.
This study shows that oxidative stress-inducing agents have a very complex effect on mitotic activity. In seedlings, a reorientation of growth involved a decrease of cell division activity in the apical root meristem, but a stimulation of division in the pericycle. To understand the specificity of these morphogenic processes, local gradients in auxin and ROS levels need to be appreciated. Oxidative stress levels are a function of a dynamic balance between the oxidative stress-inducing agent (including uptake, activity, and stability), and ROS scavengers and antioxidants. The late responses of both ASC and GSH pools suggest that both compounds are governed by a developmental control, rather than involved in a direct antioxidative response.
Oxidative stress seems to play a specific role in regulating cell cycle progression (Paulovich et al., 1997; Reichheld et al., 1999), and they might in part be mediated through ascorbate (Potters et al., 2002). However, irrespective of the existence of any direct effect of oxidative stress on the cell cycle, warranting the use of the term ‘checkpoint’ (Reichheld et al., 1999), the data presented here indicate that oxidative stress may indirectly affect mitotic activation by enhancing auxin effects on the growth cycle. Moreover, the late responses of both ASC and GSH pools (Fig. 6) suggest that both compounds are more likely to be under developmental control, than involved in a direct anti-oxidative response, causing a developmental phenomenon.
An oxidative stressor, like alloxan or paraquat, also seems to influence physiological processes beyond the cellular level. The observation in Fig. 4B (where a stressor seems responsible for the induction of an embryo-like structure) points towards the existence of stress-related changes in plant morphology and differentiation status. This has been demonstrated before by Pasternak et al. (2002); application of an oxidative stress makes plant cells acquire a less differentiated status. Clearly this phenomenon needs further research, both on a physiological and on a molecular level.
Cultured plant cells produce substantial amounts of IAA, and high levels of this native auxin are associated with embryonic potential (Jimenez and Bangerth, 2001). Notably, the exposure of protoplasts to Fe stress induced significant increases in IAA levels in isolated alfalfa protoplasts (Pasternak et al., 2002), and these led to increased mitotic activity. Work on Arabidopsis seedlings exposed to abiotic stress indicated the importance of changes in auxin gradients for plant stress responses, including morphogenesis (Pasternak et al., 2005). In that study, many stress-induced developmental effects could be mimicked by exposing plantlets to the polar auxin transport inhibitor TIBA (Pasternak et al., 2005). In the present study, oxidative stress induced a broad spectrum of auxin-like effects in seedlings (including the inhibition of root elongation, and inhibition of cotyledon and leaf elongation), and these effects are consistent with alterations in auxin levels and/or distribution. To elaborate this point further, the expression of genes involved in IAA and stress metabolism was evaluated. Gene expression in freshly prepared protoplasts was compared with expression in protoplasts kept for 24 h on auxin (2.5 μM NAA) or auxin plus alloxan (0.8 mM) (Fig. 7). That the expression of the PIN1 and PIN3 genes is depressed in the presence of alloxan is consistent with a role for auxin gradients in stress responses. Further experimental work is necessary to ascertain the physiological mechanism responsible for changes in IAA levels, such as those observed in Fe-stressed protoplasts (Pasternak et al., 2002). The similar effects of a wide variety of stresses on growth (reorientation) (Pasternak et al., 2005) suggest a common mechanism for control of IAA levels in cells and/or tissues. Moreover, the similarity of stress responses at the plantlet, organ, or cellular level suggested that the molecular response taking place in isolated protoplasts is directly relevant for understanding whole plant behaviour. The combination of an inhibitory effect of oxidative stress on the cell cycle (Reichheld et al., 1999), and a stimulatory effect on auxin levels (Pasternak et al., 2002), potentially yields a highly responsive control system for the cell cycle, which might facilitate the differential reorientation of growth observed in stressed seedlings.
GP is a post-doctoral researcher at the FWO-Flanders. Mrs Inge Van Dyck is gratefully acknowledged for her help in preparing the figures.