Ethylene and reactive oxygen species are involved in root aerenchyma formation and adaptation of wheat seedlings to oxygen-deficient conditions

1The principal cause of damage to plants grown in waterlogged soil is inadequate supply of oxygen to the submerged tissues. Oxygen deficiency in roots occurs with poor drainage after rain or irrigation, causing depressed growth of dryland plant species. Under natural conditions, roots are exposed to a gradual transition from normoxia to hypoxia, providing an opportunity for acclimation before conditions become anoxic. Exposing plants to hypoxic conditions greatly improves anoxic stress tolerance. Exposure of roots to hypoxic conditions substantially increased the activities of two ethanol fermentation enzymes, alcohol dehydrogenase (ADH) and pyruvate decarboxylase (PDC), in wheat (Triticum aestivum L.; ), maize (Zea mays L.; ), and barley (Hordeum vulgare L.; ). Transient induction of ADH1 transcripts and only a small induction of ADH enzyme activity were observed in root tips of aerobically grown maize seedlings subjected to anoxic conditions, whereas acclimation of root tips of seedlings by hypoxic pre-treatment enhanced induction of both ADH1 mRNA and ADH enzyme activity under subsequent anoxic conditions. Hypoxia also increased the levels of PDC transcripts in root tips of maize seedlings.

2Internal transport of oxygen from shoots to roots is essential to survival and functioning of roots. To adapt to waterlogging in soil, some gramineous plants develop lysigenous aerenchyma, which is formed by the creation of gas spaces as a result of death and the subsequent lysis of some cells, in the root cortex. In some wetland plants, root lysigenous aerenchyma is constitutively formed under drained soil conditions, and its formation is enhanced upon soil waterlogging. On the other hand, in dryland plants (e.g. wheat), lysigenous aerenchyma does not generally form under well-drained soil conditions, but is induced by poor aeration.

3Ethylene is involved in inducible aerenchyma formation. The treatment of maize roots with inhibitors of ethylene action or ethylene biosynthesis effectively blocks aerenchyma formation under hypoxic conditions. Reactive oxygen species (ROS) are key factors that transduce signals stimulated by abiotic stresses in plants. Respiratory burst oxidase homologue (RBOH), a plant homologue of gp91phox in mammalian NADPH oxidase, has an important role in ROS-mediated signalling, such as the defence response, programmed cell death, and development in plants. In rice (Oryza sativa L.), ethylene-induced hydrogen peroxide (H2O2)-mediated epidermal cell death during the emergence of adventitious roots is regulated by RBOH. On the other hand, application of H2O2 promotes lysigenous aerenchyma formation in internodes of rice stems. In cortical cells of maize roots, where lysigenous aerenchyma is developed, a gene encoding RBOH is strongly up-regulated under waterlogged conditions, suggesting that the ROS generation mediated by RBOH at least partly contributes to the lysigenous aerenchyma formation in maize roots.

4In this study, the effects of an ethylene precursor and an NADPH oxidase inhibitor on the tolerance of wheat seedlings to stagnant deoxygenated conditions (which mimic oxygen-deficient conditions in waterlogged soils), on the formation of lysigenous aerenchyma, and on the expressions of genes encoding ethanol fermentation enzymes in roots were examined. The results show that ethylene and ROS have important roles in each of these processes.

5Spring wheat cv. Bobwhite line SH 98 26 was used for all experiments. Seeds were surface sterilized in 0.5% (v/v) sodium hypochlorite for 30min. These seeds were rinsed thoroughly with deionized water and subsequently germinated on moist filter paper in Petri dishes in a growth chamber at 23 °C under light conditions. After 1 d, germinated seeds were placed on a meshfloat with an aerated half-strength nutrient solution [23 °C, light conditions, photosynthetically active radiation (PAR), 200–250 μmol m–2 s–1] for 4 d. The composition of the nutrient solution was described by.

6To assess the effect of an ethylene precursor, 1-aminocyclopropanecarboxylic acid (ACC), on wheat adaptation to oxygen-deficient conditions, 5-day-old aerobically grown seedlings were transferred to 5 litre pots (9–12 plants per pot, 250mm height×120mm length×180mm width) containing an aerated full-strength nutrient solution with 20 μM ACC (Sigma-Aldrich, St. Louis, MO, USA). For a control, wheat seedlings were transferred to an aerated nutrient solution without ACC treatments. After 2 d, these seedlings were transferred to 5 litre pots containing an aerated full-strength nutrient solution (aerated conditions) or stagnant solution (stagnant conditions; Supplementary Fig. S1 available at JXB online). Stagnant solution contained 0.1% (w/v) dissolved agar and was deoxygenated (dissolved oxygen, <0.5mg l–1) prior to use by flushing with N2 gas. To assess the effect of an NADPH oxidase inhibitor, diphenyleneiodonium (DPI), on the ACC treatment-promoted tolerance to stagnant conditions, 5-day-old seedlings were transferred to aerated conditions with 0, 0.1, or 1 μM DPI (Sigma-Aldrich) together with 20 μM ACC (Supplementary Fig. S1).

7Plants (14 d old) were harvested at 7 d after transfer to aerated conditions or stagnant conditions. The length and numbers of shoots, seminal roots, and adventitious roots were measured. Chlorophyll content of leaves was measured using a Soil Plant Analysis Development (SPAD) meter (SPAD-502, Konica Minolta, Tokyo, Japan). Chlorophyll meter readings were taken at the middle part of the leaves. Plants were divided into shoots and roots and dried for 7 d at 50 °C, and then shoot and root dry weights were measured. For the analysis of lateral root numbers and lengths, plants were harvested at 0h and 72h after transfer to stagnant conditions. Lateral root numbers were counted under a microscope (SZX16, OLYMPUS, Tokyo, Japan), and the lengths of lateral roots were measured with a ruler. Adventitious root numbers were counted each day during ACC pre-treatments and during subsequent treatment with aerated or stagnant conditions.

8Root cross-sections were prepared from 4mm long root segments excised from seminal roots of wheat seedlings grown in either aerated or stagnant conditions with or without ACC and DPI pre-treatments. Root segments were prepared at 10, 30, 50, and 70mm from the root tips, and 10 and 30mm from the root–shoot junctions of the seminal roots. Cross-sections were prepared by hand-sectioning with a razor blade. Each section was photographed using an optical microscope (BX60, OLYMPUS) with a CCD camera (DP70, OLYMPUS). The percentage of each cross-section occupied by aerenchyma was determined using ImageJ software (Ver. 1.43u, US National Institutes of Health, Bethesda, MD, USA).

92,3,5-Triphenyltetrazolium chloride (TTC) is normally colourless, but turns red when reduced by dehydrogenases in living cells. TTC (Tokyo Chemical Industry Co., Ltd., Tokyo, Japan) was dissolved in 0.1M sodium phosphate buffer (pH 7.0) to a final concentration of 0.6% (w/v). Root segments prepared at distances from 0 to 30mm or from 30mm to 50mm from root tips of the first seminal roots were weighed (3.0–5.0mg) and transferred to 100 μl of TTC solution. After 30min incubation at 40 °C, TTC solution was removed and root segments were rinsed once with deionized water. For the extraction, the root segments were incubated overnight at room temperature in 150 μl of 95% (v/v) ethanol. The reduction of TTC was expressed as the absorbance of the extracted solutions at 520nm in a spectrophotometer (DU800, Beckman Coulter Inc., Brea, CA, USA). For TTC staining, root segments prepared at distances from 0 to 30mm from root tips of the first seminal roots were transferred to 500 μl of TTC solution and incubated at 40 °C for 30min. After staining, the root segments were photographed using an optical microscope (SZX16, OLYMPUS) with a CCD camera (DP71, OLYMPUS).

10Total RNA was extracted from frozen fixed tissues from two sequential regions of the first seminal roots (i.e. 0–30mm and 30–50mm from the root tips, respectively) using an RNeasy Plant Mini Kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions. For qRT-PCR, 2ng of total RNA extracted from the first seminal roots were used as a template. Transcript levels were measured using a StepOnePlus Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) and One Step SYBR PrimeScript RT-PCR Kit II (Takara Bio Inc., Otsu, Japan) in accordance with the manufacturers’ protocols. A PCR fragment of each gene was purified and quantified, and then was used to draw standard curves for absolute quantification. Subsequently, the quantified mRNA levels of each gene were normalized to mRNA levels of the 18S ribosomal RNA gene as a control. qRT–PCR was performed using total RNA from three biological replicates. Primer sequences used for the experiments are shown in Supplementary Table S1 at JXB online. For TaADH1 and TaADH2 genes, two paralogues of TaADH1 and three paralogues of TaADH2 were archived in the database, and thus primers were designed to be able to amplify all of each paralogue.

11Ethylene was measured by a modification of the method of. The aerial parts were excised from wheat seedlings, and the remaining underground parts were placed in a container with saturated NaCl solution. The gas in the container was deaerated with a vacuum pump, and the gas released from the wheat seedlings was collected in a test tube using a funnel. The collected gas was transferred to a gas chromatography vial, and the vial was fitted with a rubber stopper while held upside down in saturated NaCl solution. Then the vial was righted, an aliquot of the headspace gas in the vial was withdrawn with a syringe, and the ethylene content was measured by gas chromatography (GC 353, GL Sciences, Tokyo, Japan).

12Data are presented as the means ±SD. Statistical differences between means were calculated using two-sample t-test. For multiple comparisons, data were analysed by one-way analysis of variance (ANOVA) and post-hoc Tukey’s test using SPSS Statistics Version 19 (IBM Software, New York, NY, USA).

13To assess the effect of ethylene on adaptation to oxygen-deficient conditions, 5-day-old wheat seedlings were pre-treated with an ethylene precursor, ACC (20 μM), under aerated conditions for 2 d. After 2 d of the ACC pre-treatment, reduction of the growth of shoots and first seminal roots was observed, while the numbers of leaves and the seminal roots were not affected (Supplementary Table S2 at JXB online). Subsequently, 7-day-old wheat seedlings were transferred to aerated or stagnant conditions. After 7 d (14 d old), growth of seedlings treated under each condition was measured (Fig. 1A, B; Table 1). Stagnant conditions reduced the shoot length of seedlings without the ACC pre-treatment (the ‘–ACC’ seedlings) by 25%, whereas the reduction was only by 8% in seedlings with the ACC pre-treatment (the ‘+ACC’ seedlings), compared with seedlings grown under aerated conditions (Table 1). The number of leaves was comparable (four leaves) between with and without ACC pre-treatment after 7 d growth under aerated conditions. Although stagnant conditions reduced the number of leaves to three in the ‘–ACC’ seedlings, the number of leaves (four leaves) was retained in the ‘+ACC’ seedlings (Table 1). Stagnant conditions reduced shoot dry weight of the ‘–ACC’ seedlings by 25%, but did not affect the dry weight of the ‘+ACC’ seedlings when compared with seedlings grown under aerated conditions (Table 1). The chlorophyll content (SPAD values) of leaves was significantly lower in the ‘–ACC’ seedlings grown under stagnant conditions when compared with seedlings grown under aerated conditions, while chlorophyll content was unaffected in the ‘+ACC’ seedlings (Fig. 1C, D; Table 1).

14The lengths of the longest adventitious roots of the ‘–ACC’ seedlings and the ‘+ACC’ seedlings grown under stagnant conditions were reduced by 39% and 27%, respectively (Table 1). In contrast, the numbers of adventitious roots in the ‘–ACC’ seedlings and the ‘+ACC’ seedlings grown under stagnant conditions were increased by 91% and 123%, respectively (Table 1). Interestingly, in the first seminal roots of the ‘–ACC’ seedlings grown under stagnant conditions, the root tips died, whereas they did not die in the ‘+ACC’ seedlings. After 3 d, root tip cells were densely stained with TTC (indicating that the cells were alive) in the ‘+ACC’ seedlings grown under stagnant conditions, whereas little or no staining was observed at root tips of the ‘–ACC’ seedlings (Fig. 1E). Moreover, emergence and elongation of lateral roots were almost completely inhibited in the ‘–ACC’ seedlings grown under stagnant conditions (Supplementary Fig. S2 at JXB online). The root dry weights of the ‘–ACC’ and ‘+ACC’ seedlings under stagnant conditions were 58% less and 18% less, respectively, than the root dry weight of the seedlings grown under aerated conditions (Table 1).

15Aerenchyma formation in seminal roots of wheat seedlings grown under aerated or stagnant conditions for 7 d was measured. With ACC pre-treatment, aerenchyma formation was significantly increased in response to stagnant conditions, whereas aerenchyma formation hardly occurred under aerated conditions (Fig. 2). The percentage of the cross-section of the root occupied by aerenchyma in the ‘+ACC’ seedlings increased from 3.1% to 7.8% (Fig. 2B). The percentages of aerenchyma were comparable between the ‘–ACC’ seedlings grown under stagnant conditions and those grown under aerated conditions, except for the most basal position (Fig. 2B). Similar results were obtained in the second and third seminal roots (Fig. 2B).

16To understand further the relationship between aerenchyma formation and root tip death, the short-term time-courses of aerenchyma formation and TTC reduction were analysed in the first seminal roots of wheat seedlings treated with stagnant conditions with or without ACC pre-treatment. The lengths of the first seminal roots immediately after pre-treatments (0h) were 100–120mm without ACC pre-treatment and 80–100mm with ACC pre-treatment. As a result, aerenchyma formation in each position of the first seminal roots was enhanced during ACC pre-treatment (0h; Fig. 3A). The percentages of aerenchyma in the first seminal roots of the ‘+ACC’ seedlings gradually increased during growth under stagnant conditions and plateaued at 48h (Fig. 3A). The percentages of aerenchyma at 72h after growth under stagnant conditions reached 2.9, 3.9, and 3.7% at 10, 30, and 50mm from the root tips, respectively, and 3.8% at 10mm from the root–shoot junction (Fig. 3A). In contrast, aerenchyma formation hardly occurred at any of the positions of the first seminal roots of the ‘–ACC’ seedlings immediately after pre-treatment (0h; Fig. 3A). The percentages of aerenchyma increased to 1.0% and 1.1% at 30mm and 50mm from the root tips, respectively, and to 2.2% at 10mm from the root–shoot junction at 72h after growth under stagnant conditions (Fig. 3A). It should be noted that the cross-sections of the roots at 10mm from the root tips were hardly obtained because of the root tip death.

17The TTC reduction assay has been used to evaluate viability of plant tissues quantitatively by detecting absorbance of reduced TTC, a red formazan derivative. At 0–30mm from the root tips, the values of TTC reduction (OD520) in the ‘–ACC’ seedlings were gradually reduced under stagnant conditions, whereas the value of TTC reduction in the ‘+ACC’ seedlings was kept higher (Fig. 3B). At 30–50mm from the root tips, no significant changes in the values of TTC reduction were observed between the ‘–ACC’ seedlings and the ‘+ACC’ seedlings (Fig. 3B).

18The aerenchyma formation and TTC reduction analyses gave similar results when shorter first seminal roots (70–90mm) of 4-day-old wheat seedlings were used for ‘–ACC’ conditions (Supplementary Fig. S3 at JXB online).

19To investigate whether the ACC pre-treatment induces expression of genes encoding ethanol fermentation enzymes in the first seminal roots, expression of two PDC genes, TaPDC genes, (accession nos AK332508 and BT009420), and three ADH genes, TaADH1 (accession nos EF122847 and EF122848), TaADH2 (accession nos EF122843, EF122844, and EF122845), and TaADH3 (accession no. EF122842) (Supplementary Table S1 at JXB online), was analysed by qRT–PCR (Fig. 4A, B). The expression levels of the TaPDC genes and the TaADH genes at 0–30mm from the root tips in the first seminal roots immediately after ACC pre-treatment (0h) were significantly higher in the ‘+ACC’ seedlings than in the ‘–ACC’ seedlings, and thereafter their expression was further increased in the ‘+ACC’ seedlings (Fig. 4A). It should be noted that the lower expression levels of the genes at 0–30mm from the root tips in the first seminal roots of the ‘–ACC’ seedlings grown under stagnant conditions could be largely due to the lower cell viability (Fig. 3B). At 30–50mm from the root tips, the expression levels of the TaPDC (BT009420) gene, and the TaADH1, 2, and 3 genes immediately after pre-treatment (0h) were significantly higher in the ‘+ACC’ seedlings than in the ‘–ACC’ seedlings (Fig. 4B). Expression of all genes encoding ethanol fermentation enzymes in the ‘+ACC’ seedlings and the ‘–ACC’ seedlings were induced under stagnant conditions, but their expression levels were higher in the ‘+ACC’ seedlings than in the ‘–ACC’ seedlings (Fig. 4B). The expression level of the TaPDC (BT009420) gene was higher than that of the TaPDC (AK332508) gene (Fig. 4), and the expression level of the TaADH1 gene was highest among the three TaADH genes immediately after ACC pre-treatment (0h), whereas TaADH2 gene expression was highest at 72h after initiation of the treatments (Fig. 4).

20The mitochondrial pyruvate dehydrogenase (PDH) complex plays a pivotal role in the tricarboxylic acid (TCA) cycle and following oxidative phosphorylation by the conversion of pyruvate to acetyl-CoA in aerobic cells. The activity of the mitochondrial PDH complex depends on the expression of the gene encoding PDH E1α, which is one of two subunits of the mitochondrial PDH E1 component. To assess whether the induction of aerenchyma formation in the first seminal roots of wheat seedlings affects the expression level of the wheat PDH E1α (TaPDHE1α) gene, qRT–PCR analysis of the TaPDHE1α gene (accession no. GU563379) (Supplementary Table S1 at JXB online) was performed (Fig. 4C, D). Although the expression level of the TaPDHE1α gene was reduced under stagnant conditions, reduction of the expression in the first seminal roots was significantly lower in the ‘+ACC’ seedlings than in the ‘–ACC’ seedlings (Fig. 4C, D). Interestingly, the expression level of TaPDHE1α at 30–50mm from the root tips of the first seminal roots of the ‘+ACC’ seedlings started to increase at 48h after initiation of growth under stagnant conditions (Fig. 4D), implying that the internal transport of oxygen from shoot to roots by forming the aerenchyma leads to the recovery of TaPDHE1α expression.

21To assess whether ethylene accumulated in roots of wheat seedlings grown under stagnant conditions with or without ACC pre-treatments, ethylene was measured in roots and the expression of a gene encoding ACS (TaACS2; accession no. U42336, a key enzyme of ethylene biosynthesis) was measured in first seminal roots of wheat seedlings (Fig. 5). The content of ethylene in the roots of the ‘–ACC’ seedlings increased ~2.3-fold at 24h after initiation of growth under stagnant conditions, and then gradually decreased (Fig. 5A). The ethylene contents were ~500 times higher in the roots of the ‘+ACC’ seedlings than in those of the ‘–ACC’ seedlings immediately after pre-treatments (0h; Fig. 5B). The ethylene contents decreased in the roots of the ‘+ACC’ seedlings at 72h after initiation of growth under stagnant conditions, possibly because ACC was not included in the stagnant solution. However, the ethylene contents were still ~5 times higher in the ‘+ACC’ seedlings than in the ‘–ACC’ seedlings (Fig. 5B).

22The TaACS2 gene is predominantly expressed in roots of wheat seedlings. The TaACS2 gene expression was measured at two locations in the first seminal roots of wheat seedlings after transfer to stagnant conditions. At 0–30mm from the root tips, the TaACS2 expression immediately after ACC pre-treatment (0h) was comparable between the ‘+ACC’ and the ‘–ACC’ seedlings, but was significantly higher in the ‘+ACC’ seedlings at 24h and 48h (Fig. 5C). At 30–50mm from the root tips, the TaACS2 expression was significantly higher in the ‘+ACC’ seedlings at 24h and 48h, but was much higher in the ‘–ACC’ seedlings at 72h (Fig. 5D).

23To investigate expression levels of RBOH genes in first seminal roots of wheat seedlings, three TaRBOH genes (accession nos AK334304, AK334324, and AK335454) were selected (Supplementary Table S1 at JXB online). Expression levels of the TaRBOH (AK334304) and TaRBOH (AK334324) genes at 0–30mm and 30–50mm from the root tips were decreased in the ‘–ACC’ seedlings under stagnant conditions, whereas the levels were kept higher in the ‘+ACC’ seedlings than in the ‘–ACC’ seedlings (Fig. 6). On the other hand, the expression levels of the TaRBOH (AK335454) gene at 0–30mm from the root tips was gradually increased during growth of the ‘+ACC’ seedlings, but not of the ‘–ACC’ seedlings, under stagnant conditions (Fig. 6A). At 30–50mm from the root tips, TaRBOH (AK335454) expression was increased in both the ‘+ACC’ seedlings and the ‘–ACC’ seedlings under stagnant conditions, but its levels were slightly higher in the ‘+ACC’ seedlings than in the ‘–ACC’ seedlings (Fig. 6B). Among three TaRBOH genes, the expression level of the TaRBOH (AK334304) gene was highest in the first seminal roots immediately after ACC pre-treatment (0h), whereas TaRBOH (AK335454) expression was highest at 72h after initiation of growth under stagnant conditions (Fig. 6).

24To understand the effect of an NADPH oxidase inhibitor, DPI, on lysigenous aerenchyma formation in first seminal roots of wheat seedlings, the percentage of aerenchyma was investigated and compared among first seminal roots of seedlings treated under stagnant conditions with 2 d pre-treatment with different concentrations of DPI (0, 0.1, and 1 μM) together with 20 μM ACC. The lengths of the first seminal roots immediately after 0, 0.1, and 1 μM DPI pre-treatment (0h) were 80–100, 80–100, and 60–80mm, respectively. The ACC-induced aerenchyma formation at each position of the first seminal roots at 0h was significantly reduced by the DPI pre-treatment in a dose-dependent manner (Fig. 7A). At 72h after initiation of growth under stagnant conditions, the ACC-induced aerenchyma formation was also significantly reduced by the DPI pre-treatment in a dose-dependent manner (Fig. 7A).

25TTC reduction assay was performed using segments at 0–30mm and 30–50mm from the root tips. The value of TTC reduction of the first seminal roots was not affected by 0.1 μM DPI pre-treatment, but was reduced by 1 μM DPI pre-treatment (0h) (Fig. 7B). At 0–30mm from the root tips of the first seminal roots with 0.1 μM DPI pre-treatment, the value of TTC reduction was significantly reduced at 72h after initiation of growth under stagnant conditions, while the values without DPI pre-treatment was relatively the same between at 0h and 72h (Fig. 7B). At 30–50mm from the root tips, no significant changes in the values of TTC reduction were observed between at 0h and 72h (Fig. 7B).

26qRT–PCR was performed to investigate whether the DPI pre-treatment represses ACC-induced expression of the TaADH and the TaPDC genes in the first seminal roots of wheat seedlings. During the pre-treatment, the expression of the TaPDC (AK332508) and the TaPDC (BT009420) genes in first seminal roots of seedlings were comparable between pre-treatment with 20 μM ACC (the ‘+ACC’ seedlings) and pre-treatment with 20 μM ACC and 0.1 μM DPI (the ‘+ACC & +DPI’ seedlings) (Fig. 8A). However, the expression levels of the TaADH1, 2, and 3 genes were significantly reduced at 0–30mm from the root tips of the first seminal roots of the ‘+ACC & +DPI’ seedlings, compared with the ‘+ACC’ seedlings (Fig. 8B). At 30–50mm from the root tips, the TaADH3 gene expression was significantly reduced in the first seminal roots of the ‘+ACC & +DPI’ seedlings (0h; Fig. 8B). At 72h, significant reductions in TaPDC gene expression were observed at 30–50mm from the root tips (Fig. 8A). The expression of the TaADH2 gene was significantly reduced at 0–30mm from the root tips of the first seminal roots of the ‘+ACC & +DPI’ seedlings (Fig. 8B). At 30–50mm from the root tips, the expression of the TaADH1 and the TaADH2 genes was reduced in the first seminal roots of the ‘+ACC & +DPI’ seedlings (Fig. 8B). Although the expression level of the TaPDHE1α gene was reduced under stagnant conditions, reduction of the expression in the first seminal roots was significantly higher in the ‘+ACC & +DPI’ seedlings than in the ‘+ACC’ seedlings (Supplementary Fig. S4A, B at JXB online). The expression level of the TaACS2 gene was comparable between the ‘+ACC’ seedlings and the ‘+ACC & +DPI’ seedlings at 0h and 72h (Supplementary Fig. S4C, D), suggesting that ethylene contents in the roots of the ‘+ACC & +DPI’ seedlings were identical to those of the ‘+ACC’ seedlings. The RBOH expression is considered a reliable indicator of ROS production because treatment with H2O2 induced transcription of the RBOH gene in Arabidopsis and maize, and treatment with scavengers of H2O2 suppressed induction of RBOH expression in maize. The expression levels of the TaRBOH genes in the first seminal roots of the ‘+ACC & +DPI’ seedlings were significantly lower than those in the ‘+ACC’ seedlings (Supplementary Fig. S5), supporting that ROS production was severely depressed by DPI pre-treatment.

27To assess the effect of the DPI treatment on the adaptation to oxygen-deficient conditions in wheat seedlings, 5-day-old seedlings were pre-treated with 0.1 μM DPI together with 20 μM ACC for 2 d in aerated conditions. Subsequently, 7-day-old wheat seedlings were transferred to aerated or stagnant conditions (Supplementary Fig. S1 at JXB online). After 7 d (14 d old), growth of seedlings treated with each condition was measured (Table 1; Supplementary Fig. S6 at JXB online). Stagnant conditions reduced the shoot length of the ‘+ACC & +DPI’ seedlings by 22%, and decreased shoot dry weight by 16%, compared with the seedlings grown under aerated conditions (Table 1). Leaf number was reduced to three in the ‘+ACC & +DPI’ seedlings grown under stagnant conditions (Table 1). The longest adventitious root length and root dry weight in stagnant conditions were reduced by 31% and 46%, respectively, for the ‘+ACC & +DPI’ seedlings, compared with seedlings grown under aerated conditions (Table 1).

28Stagnant conditions severely reduced shoot length, leaf number, and shoot and root dry weights of wheat seedlings (Fig. 1A, B; Table 1), whereas all of the reductions of these growth parameters were alleviated by pre-treatment with the ethylene precursor ACC (Fig. 1A, B; Table 1). Exposure of several plants to hypoxic conditions greatly improved their anoxic stress tolerance. Because the effects of ACC are similar to the effects of hypoxic pre-treatments in several plants, ethylene may be a factor involved in the induction of the adaptive responses during hypoxic pre-treatments. Hormonal signalling pathways regulated by ethylene and ROS are involved in the adaptation of plants to abiotic stress. The cancelling of the effect of ACC by DPI (Table 1; Supplementary Fig. S6 at JXB online) suggests that ethylene-mediated ROS signalling plays a role in the adaptation of wheat seedlings to oxygen-deficient conditions. Leaf chlorosis has been used to evaluate the tolerance of plants to waterlogging because nutrient uptake and photosynthesis are severely affected by waterlogging (, c). The finding that ACC prevented the decrease of chlorophyll content of leaves grown under stagnant conditions (Fig. 1C, D; Table 1) is another indication that ACC enhanced the tolerance of wheat seedlings to oxygen-deficient conditions.

29The ACC pre-treatment significantly induced the expression levels of the TaPDC genes and the TaADH genes in the first seminal roots (Fig. 4), indicating that the induction of expression of the genes encoding ethanol fermentation enzymes is regulated by ethylene. Subsequent growth under stagnant conditions further induced the expression levels of the TaPDC genes and the TaADH genes, suggesting that ethylene itself was not sufficient for full induction of expressions of these genes. This result is supported by the study in Arabidopsis showing that the hypoxic induction of the ADH gene could not be completely suppressed by the treatment with an inhibitor of ethylene biosynthesis. Production of H2O2 via a DPI-sensitive NADPH oxidase is necessary for the induction of ADH gene expression as well as the enhancement of the ADH activity in Arabidopsis. Because the expression level of three homologues of NADPH oxidase in wheat was significantly higher in the first seminal roots of the ‘+ACC’ seedlings than in those of the ‘–ACC’ seedlings (Fig. 6A), RBOH-mediated ROS generation may be stimulated by ethylene in the first seminal roots. The co-treatment of ACC and DPI partly suppressed the ACC-induced expression of the TaADH genes in the first seminal roots at 0–30mm from the root tips (0h; Fig. 8B), suggesting that expression of the TaADH genes is regulated not only by ethylene but also by ROS. At 72h after initiation of growth under stagnant conditions, the expression of the TaADH2 gene was also significantly reduced in the first seminal roots at 0–30mm from the root tips of the ‘+ACC & +DPI’ seedlings when compared with the ‘+ACC’ seedlings (Fig. 8B). Since the TaADH2 gene expression was highest at 72h after initiation of the treatments (Fig. 8B), the protein levels and the activity of ADH in roots of the wheat seedlings may be stimulated by ethylene-mediated ROS signalling. The expression of the TaADH1 and TaADH2 genes and of the TaPDC genes in the first seminal roots at 30–50mm from the root tips was significantly reduced by the DPI pre-treatment at 72h after initiation of growth under stagnant conditions (Fig. 8). These results suggest that the expression levels of the genes encoding ethanol fermentation enzymes were regulated by both ethylene and ROS signalling.

30Aerenchyma formation in each position of the first seminal roots of wheat seedlings was enhanced by ACC pre-treatment (Fig. 3A), and ACC-induced aerenchyma formation was significantly inhibited by the DPI pre-treatment in a dose-dependent manner (Fig. 7A). During the aerenchyma formation in maize roots, expression of a gene encoding RBOH is induced in the cortical cells of the primary roots, and its induction is suppressed by pre-treatment with an inhibitor of ethylene perception. Moreover, aerenchyma formation of maize roots under waterlogged conditions was inhibited by the treatment with DPI. Together, ethylene-induced RBOH-mediated ROS generation may be commonly involved in lysigenous aerenchyma formation in roots of cereal crops.

31Although the accumulation of ethylene in the roots of the ‘–ACC’ seedlings peaked at 24h after initiation of growth under stagnant conditions (Fig. 5A), the TaACS2 expression in the first seminal roots of the ‘–ACC’ seedlings dropped to a low level at 24h and 48h after initiation of growth under stagnant conditions (Fig. 5C, D). These observations imply that the initial accumulation of ethylene mainly occurred through ‘physical entrapment’ of ethylene gas as a result of the low rate of gas diffusion under stagnant conditions. Under aerated conditions, adventitious roots of the ‘–ACC’ seedlings first emerged at 72h after the initiation of growth, while under stagnant conditions they first emerged at 48h (Supplementary Fig. S7 at JXB online). Therefore, initially, ethylene accumulated predominantly at the most basal part of the roots where adventitious roots emerged, raising the possibility that it contributes to the formation of the adventitious roots. In rice, adventitious root emergence under submergence is thought to be stimulated by ethylene accumulation through both increased ethylene biosynthesis and physical entrapment. In spite of the early induction of the ethylene accumulation in the roots of the ‘–ACC’ seedlings, the expression level of the TaACS2 gene was strongly induced at 72h after initiation of growth under stagnant conditions (Fig. 5C, D). The aerenchyma formation and the expression levels of the genes encoding ADH, PDC, and RBOH at 30–50mm from the root tips in the first seminal roots of the ‘–ACC’ seedlings were also increased at 72h after initiation of growth under stagnant conditions (Figs 3A, 4B, 6B). These results suggest that local accumulation of ethylene stimulated by ethylene biosynthesis is essential for the adaptive responses to oxygen-deficient conditions. If this is the case, the late induction of TaACS2 gene expression may be one of the reasons why wheat is less tolerant to oxygen-deficient conditions in waterlogged soils than wetland species. ACC pre-treatment remarkably enhanced the accumulation of ethylene in the roots of wheat seedlings even under the aerated conditions (at 0h; Fig. 5B), indicating that the roots of the ‘+ACC’ seedlings were ready to respond to oxygen-deficient conditions.

32Stagnant conditions severely reduced the expression level of the TaPDHE1α gene, but the reduction of the expression level was alleviated by the ACC pre-treatment in the first seminal roots (Fig. 4C, D). Interestingly, the expression level of the TaPDHE1α gene at 30–50mm from the root tips of the first seminal roots of the ‘+ACC’ seedlings started to increase at 48h after initiation of growth under stagnant conditions. The percentage aerenchyma formation plateaued at 48h after initiation of growth under stagnant conditions, supporting that the internal transport of oxygen to roots through the aerenchyma contributes to sustaining the level of aerobic respiration through the enhancement of TaPDHE1α expression. Oxygen is internally transported to roots through the aerenchyma. Moreover, aerenchymatous roots of maize showed higher values for ATP content, adenylate energy charge, and ATP/ADP ratios than non-aerenchymatous roots in anaerobic conditions. A relatively high level of aerobic respiration can be sustained for the roots with extensively formed aerenchyma, even in anoxic conditions. In the present study, it was also demonstrated that both the aerenchyma formation and the expression level of the TaPDHE1α gene under stagnant conditions were significantly reduced in the first seminal roots of the ‘+ACC & +DPI’ seedlings when compared with the ‘+ACC’ seedlings (Fig. 7A; Supplementary Fig. S4A, B at JXB online). These results further support the correlation between aerobic respiration and aerenchyma formation.

33In conclusion, it was found that the formation of lysigenous aerenchyma and the expression levels of the genes encoding ethanol fermentation enzymes were enhanced in the first seminal roots of the ACC-pre-treated seedlings that were further grown under stagnant conditions. Moreover, DPI pre-treatment mostly diminished the effect of ACC on the adaptive response to stagnant conditions. These results suggest that ethylene-mediated ROS signalling is involved in regulating the adaptive responses to oxygen-deficient conditions in waterlogged soil.

34Supplementary data are available at JXB online.

35Figure S1. Growth conditions to assess the effect of the ethylene precursor, ACC, and the NADPH oxidase inhibitor, DPI, on adaptation of wheat seedlings to oxygen-deficient conditions.

36Figure S2. Lateral root numbers and longest lateral root lengths of wheat seedlings in stagnant conditions with or without ACC pre-treatment.

37Figure S3. Time-course of aerenchyma formation and cell viability in shorter first seminal roots of wheat seedlings in stagnant conditions with or without ACC pre-treatment.

38Figure S4. qRT–PCR analyses of TaPDHE1α and TaACS2 genes in first seminal roots of wheat seedlings in stagnant conditions with or without pre-treatment with DPI together with ACC.

39Figure S5. qRT–PCR analyses of TaRBOH genes (AK334304, AK334324, and AK335454) in first seminal roots of wheat seedlings in stagnant conditions with or without pre-treatment with DPI together with ACC.

40Figure S6. Growth of wheat seedlings under aerated conditions and stagnant conditions with or without pre-treatment with DPI together with ACC.

41Figure S7. Time-course of emerged adventitious root numbers of wheat seedlings in aerated or stagnant conditions with or without ACC pre-treatment.