Seedlings of the following lines were used in this study: Arabidopsis thaliana ecotype Columbia-0 (Col-0) and the ethylene-insensitive mutants ein2-5 (Alonso et al., 1999) and ein3-1eil1-1 (Alonso et al., 2003) in the Col-0 background. All seeds were surface-sterilized by incubation in 70% ethanol containing 0.05% Triton X-100 (Solar Bio, Beijing, China) for 10 min, rinsed thoroughly with ethanol for 1 min and then washed with sterilized water. The sterilized seeds were sown on agar plates containing a MS basal salt mixture (Sigma–Aldrich, St. Louis, MO, USA), 1% sucrose and 0.8% agar (pH 5.7). The plates were maintained at 4°C in the dark for 48 h to enhance the germination process and then placed vertically at 23°C under a light intensity of 200 μml m^-2 s^-1 (16/8 h light cycle).

To investigate the effects of Cd stress on root growth, 4-day-old seedlings cultivated in control agar plates were transferred to freshly prepared medium supplemented by various concentration of CdCl₂ (25, 50, 75, and 100 μM) as indicated (Cao et al., 2009; Radeva et al., 2010) for 2 and 4 days. Preliminary experiments with seedling grown with various CdCl₂ concentrations showed that 75 μM CdCl₂ decreased root growth by about 40% and so this concentration was used to compare the effects of Cd stress between the wild type and the ethylene-insensitive mutants.

To investigate the effects of exogenous ACC on the accumulation of superoxide in the root tips of the Col-0 plants under Cd stress. Four-day-old seedlings were transferred to various pretreated agar plates with 75 μM CdCl₂ or 75 μM CdCl₂ plus various concentrations of ACC (0.01, 0.05, and 0.5 μM, a precursor of ethylene biosynthesis) with or without 5 μM AIB (an inhibitor of ethylene biosynthesis), as indicated, for 4 days.

All the experiments were repeated three times with three replicates of each treatment unless otherwise noted.

To investigate whether ethylene biosynthesis and signaling were involved in the plant response to Cd stress, the relative expression of genes encoding for ethylene biosynthesis, and of perception and signaling proteins were measured in seedlings that were treated with different concentrations of CdCl₂ for 4 days. Total RNA was isolated from 100 seedling roots using TRIzol solution (Invitrogen, Carlsbad, CA, USA). Two micrograms of total RNA was used for first-strand cDNA synthesis using RevertAid Reverse Transcriptase and Oligo d(T)primers (Takara, Dalian, China). The cDNA yield was measured according to the PCR signal generated from the internal standard, with the housekeeping gene Actin 2 used as the internal control. The primers that were used in the quantitative RT-PCR are listed in Table 1 (Liu et al., 2010; Lin et al., 2013b; Li et al., 2014).

Quantitative RT-PCR was performed using a RealMasterMix kit (Tiangen, Beijing, China) with 25 cycles as follows: 94°C for 30 s, 58°C for 30 s and 72°C for 30 s, followed by 72°C for 10 min gene expression quantifications was performed using the relative 2^-ΔΔCt method (Livak and Schmittgen, 2001). All experiments were performed in triplicate for each treatment.

To investigate the mechanism by which CdCl₂ affects the synthesis and distribution of ethylene, the relative expression of the ethylene reporter construct, EBS::GUS, in which the GUS reporter gene is driven by a synthetic EIN3-responsive promoter, was tested. GUS staining was as described by He et al. (2011). Four-day-old seedlings were transferred to agar plates supplemented with 75 μM CdCl₂ for 4 days, then were collected and washed with staining buffer without X-Gluc and stained with GUS staining buffer (50 mM sodium phosphate buffer, pH 7.0, 10 mM Na₂EDTA, 0.5 mM K₄[Fe(CN)₆]⋅3H2O, 0.5 mM K₃[Fe(CN)₆], 0.1% Triton X-100, and 1 mg/mL X-Gluc). Ethanol (70%) was used to terminate the staining reaction, and the seedlings were mounted on slides in 50 μL Hoyer’s solution (chloral hydrate: water: glycerol; 8:3:1; w/v/v) and examined and photographed with a DM 4000B stereomicroscope (Leica, Germany).

To further test the involvement of ethylene signaling in regulating the plant response to Cd stress, two ethylene-insensitive mutants (ein2-5 and ein3-1eil1-1) together with the wild type plant were subjected to Cd stress with or without ACC. The length of the whole primary root of the seedlings were measured either directly with a ruler or using NIS Elements software (Nikon, Japan) from digital images captured with a Nikon camera. The root tip length was also determined in the same way as the primary root. The number of the lateral roots was counted. The number of root hairs in a 2 mm section at the appropriate midpoint of the root was counted under a dissecting microscope. Root hairs from the photographs were measured with a ruler, and their length in μm was determined by comparison with an ocular micrometer photograph at the same magnification. All experiments for phenotypic analysis were performed at least three times, and the data represent one independent experiment.

The samples were rinsed twice with tap water then with de-ionized water before air drying is used to remove dirt and salt from root surface. High purity deionized water from a Millipore water purification system (Bedford, NY, USA) was used throughout our study. The root materials were dried at 80°C to a constant weight. The dried tissues were weighed and digested in HNO3 (100%) using the heat block (Shah et al., 2014; Olowu et al., 2015). The metal concentration was determined by inductively coupled plasma mass spectrometry (ICP-MS). Blanks (only HNO3) were analyzed for reference purposes. The cadmium standard solutions for ICP-MS (TraceCERT^®, Sigma–Aldrich) were produced according to the ISO Guide 34 in the analysis of Cd concentrations. External calibration was performed using a five-point analytical curve, prepared by diluting the individual cadmium standards with 5.0% (v/v) HNO3.

Bioaccumulation factor (Bf) was calculated to measure plant uptake of Cd:

Four-days-old wild-type seedlings cultivated on MS agar plates were transferred to agar plates supplemented with 75 μM CdCl₂ or 75 μM CdCl₂ plus various concentrations of ACC (0.01, 0.05, and 0.5 μM, a precursor of ethylene biosynthesis) with or without 5 μM AIB (an inhibitor of ethylene biosynthesis). Superoxide within the root tip was detected by nitroblue tetrazolium (NBT) staining which is used to detect O₂^- production, as well as other compounds (such as ascorbate) as described by Hernández et al. (2001). Seedlings were collected, and the roots were immersed in a 0.1% solution of NBT in 50 mM K–phosphate buffer (pH 6.4), containing 10 mM Na-azide in the absence of light. Seedlings were then transferred to distilled water to stop the reaction. The roots were observed and photographed using a DM 4000B Leica stereomicroscope equipped with a DC300F Nikon camera. The NIH ImageJ software was used to assess the mean staining intensity of the elongation and meristematic zones (0, white; 255, black) following Dunand et al. (2007). At least 20 individual roots were analyzed for each genotype and treatment, and one representative image was selected for the figure.

The determination of the NADPH dependent O₂^- generating activity in isolated plasma membrane vesicles was carried out according to Quartacci et al. (2001) and Van Gestelen et al. (1997), by measuring the rate of SOD inhibitable reduction of NBT using NADPH as electron donor. The reaction mixture consisted of 50 mM TRIS HCl buffer, pH 7.5, 0.25 M sucrose, 0.1 mM NBT, and 50–100 μg proteins. After 1 min preincubation the reaction started by the addition of 0.1 mM NADPH and the absorbance changes at 530 nm were followed for 5 min. Rates of O₂^- generation were calculated using an extinction coefficient of 12.8 mM^-1cm^-1.

To investigate whether ethylene reduced Cd stress-induced superoxide accumulation through the SOD pathway, 4-day-old seedlings were transferred to various agar plates for 4 days, and the activities of SOD and its three isoenzymes (Cu/Zn-SOD, Fe-SOD and Mn-SOD) were measured according to the method of Yu and Rengel (1999) with minor modifications. Frozen root samples (0.05 g) were weighed and homogenized on ice for 2 min in 5 ml of homogenizing solution containing 50 mm HEPES buffer and 0.1 mm Na₂EDTA (pH 7.6). The homogenate was centrifuged at 15000 g for 15 min at 4°C to produce the crude extract for SOD assays. SOD activity was assayed by monitoring the inhibition of the photochemical reduction of NBT. For the total SOD assay, a 5-ml reaction mixture containing 50 mM HEPES (pH 7.6), 0.1 mM EDTA, 50 mM Na₂CO₃ (pH 10.4), 13 mM methionine, 0.025% (w/v) Triton X-100, 75 μM NBT, 2 μM riboflavin and an appropriate aliquot of enzyme extract was utilized. The reaction mixtures were illuminated for 15 min at a light intensity of 350 μmol⋅m^-2⋅s^-1. One unit of SOD activity was defined as the amount of enzyme required to cause a 50% inhibition in the reduction of NBT, as monitored at 560 nm. The activities of different SOD forms were identified by adding KCN and/or H₂O₂ to the reaction mixture (Giannopolitis and Ries, 1977). KCN inhibits Cu/Zn-SOD but does not affect Mn-SOD or Fe-SOD, whereas H₂O₂ inactivates Cu/Zn-SOD and Fe-SOD but not Mn-SOD. In addition, peroxidases might interfere with the SOD assay in the presence of exogenous H₂O₂ (Yu et al., 1998). After the extensive preliminary testing of a range of concentrations, KCN (at final concentration of 3 mM) was added to the reaction mixture before the addition of H₂O₂ (5 mM final concentration) to eliminate the interference of peroxidase and catalase enzymes (Chen and Asada, 1989). Mn-SOD activity was determined in the presence of both 3 mM KCN and 5 mM H₂O₂. Fe-SOD activity was obtained by subtracting the Mn-SOD activity from the activity in the presence of 3 mM KCN, and Cu/Zn-SOD activity was calculated from the differences between the total activity and that of Mn-SOD and Fe-SOD. Identical reaction mixtures that had not been illuminated were used to correct for background absorbance.

To assess cell death in the root tips under Cd stress, the roots were immersed in 4 mg ml^-1 Trypan blue solution (Sigma–Aldrich, Saint Louis, MO, USA) for 15 min at room temperature and then washed with distilled water three times. The samples were then observed by light microscopy, and pictures were taken as detailed earlier.

For PI staining, the seedlings were immersed in the PI (Sigma–Aldrich, St. Louis, MO, USA) solution (final concentration of 1 μg ml^-1) for 10 min at room temperature in the dark and then washed with phosphate buffer solution (PBS) (pH 7.4) three times. The samples were then examined using a DM 4000B fluorescence microscope (Leica, Germany) with an excitation wavelength of 546 nm. Both experiments were repeated three times. For each treatment and genotype, at least 20 roots were analyzed for both stains, and one representative image was selected for the figure.

Cadmium-induced cell death in root tips can occur through either necrosis or PCD. To examine whether PCD is involved in the cell death at the root tips, we investigated the chromatin condensation and the internucleosomal fragmentation of the DNA by DAPI staining and the TUNEL assay, respectively. To detect cell apoptosis caused by Cd stress, an In Situ Cell Death Detection Kit, AP (Roche, Germany) was used. The TUNEL assay was performed according to the manufacturer’s instructions with a few modifications. In brief, whole seedlings were fixed in 4% paraformaldehyde in PBS (pH 7.4) for 20 min (20°C). After washing the samples for 30 min with PBS, the samples were incubated in permeabilization solution (0.1% Triton X-100, 0.1% sodium citrate) for 2 min (4°C), followed by two washes with PBS. Fifty microliters of TUNEL reaction mixture was then added to the sample, and it was incubated for 60 min (37°C) in a humidified atmosphere in the dark. After being rinsed three times with PBS, the samples were analyzed under a DM 4000B fluorescence microscope (Leica, Germany) with a 488 nm excitation line and a 530 nm emission filter.

DAPI staining was performed by fixing roots in a solution of ethanol and acetic acid (3:1) for 1 h. The roots were then washed twice with PBS (pH 7.4) and stained in PBS containing 1 μg/ml of DAPI for 20 min at room temperature. After two PBS washes, the samples were analyzed under a DM 4000B fluorescence microscope (Leica, Germany) with a UV light filter, and the images were captured immediately. Both experiments were repeated three times. For each treatment and genotype, at least 20 roots were analyzed for each experiment, and one representative image was selected for the figure.

For all experiments, the data were statistically analyzed using SPSS 17.0 (SPSS, Chicago, IL, USA). One-way analysis of variance with a Duncan post hoc test was used to test the differences at a 0.05 level for the primary root length and distance to the first root hair. The data presented here are the means with standard error (SE).

NADPH oxidase and SOD enzymes activities values were used for the calculation of abundance ratios between groups and for statistical evaluation by Student’s t-test (P = 0.05).

To investigate whether ethylene biosynthesis and signaling were involved in the plant response to Cd stress, quantitative RT-PCR was performed using a Real Master Mix kit (Tiangen, Beijing, China) to measure the relative expression of genes encoding for ethylene biosynthesis, and of perception and signaling proteins. The relative expression of ACS2, ACS6, ERF1, and EIN3 increased slightly under small concentrations of Cd but increased drastically after increasing Cd concentrations to 75 μM but decreased with further increase to 100 μM although expression still higher than the control (Figure 1A).

To investigate the mechanism by which CdCl2 affects the synthesis and distribution of ethylene, the relative expression of the ethylene reporter construct, EBS::GUS, in which the GUS reporter gene is driven by a synthetic EIN3-responsive promoter, was tested. Remarkably, less than 10% of root tips in unstressed seedlings exhibited GUS staining while nearly 80% of them in stressed seedlings were stained (Figures 1B,C). Taken these results together, suggested the involvement of ethylene signaling in plant responses to Cd stress.

The primary root lengths of the wild-type (Col-0) and the ein2-5 and ein3-1eil1-1 plants were severely inhibited by 75 μM CdCl₂ (Supplementary Figure S1).

However, Figure 2A showed that the primary roots lengths of both the ein2-5 and ein3-1eil1-1 plants were slightly longer than those of the wild-type (Col-0) plants after 2 days of Cd stress. This phenomenon was reversed after 4 days treatment as the primary roots of the ein2-5 and ein3-1eil1-1 plants were significantly shorter than the wild-type plants.

The root tip length (distance to the first root hair) was also determined after 4 days of Cd treatment (Figures 2B,C). Similar to the results for the primary root length, the length of the meristematic and elongation/transition zone was less inhibited in the ein2-5 and ein3-1eil1-1 mutants than in the wild-type plants after 2 days of Cd treatment. The length of the zones in the wild-type plant Col-0, ein2-5, and ein3-1eil1-1 plants was 1433.1, 1614.6, and 1741.7 μm, respectively. In contrast, after 4 days treatment, the root tip lengths of both the ein2-5 (1056.6 μm) and ein3-1eil1-1 (983.8 μm) plants were significantly shorter compared with the Col-0 (1393.1 μm) plants.

The lateral root number of the wild-type plants also significantly increased, whereas it was reduced in the ein2-5 and ein3-1eil1-1 mutants in response to Cd stress (Figure 3A). This result was expected since the primary root of wild-type plants was longer than that of the ein2-5 and ein3-1eil1-1 mutants under of Cd stress.

Moreover, the root hair density also significantly increased in the Col-0 plants, whereas it severely decreased in both the ein2-5 and ein3-1eil1-1 plants under Cd stress (Figure 3B). In addition, it is worth noting that very few root hairs emerged from mutants under Cd stress and also the length of these hairs decreased (Supplementary Figure S2).

These results show that the wild type plants improve their root system development under Cd stress in a better way than ein2-5 and ein3-1eil1-1 plants, this led us to suggest the involvement of ethylene in modulating the root system development in response to Cd stress.

Three-week-old A. thaliana seedlings were exposed to 0, 10, 20, and 40 μM CdCl2 for 8 days. (Figure 4) showed that a significant accumulation of Cd content in roots treated with Cd. In the Col-0 root, the Cd content was highest-level after exposure to 20 μM CdCl2. At the 20 μM CdCl2 treatment, the Cd contents of ein2-5 and ein3-1eil1-1 were even higher than the Cd content of Col-0. Bioaccumulation factors (Bfs) in the Col-0 got their maximum at 20 μM CdCl2. Furthermore, the Bfs of ein2-5 and ein3-1eil1-1 were even bigger than Col-0 under 20 μM CdCl2 treatment. Ein2-5 and ein3-1eil1-1 are both ethylene insensitive mutants. Moreover, application of ACC with different concentrations of Cd (50, 60, and 75 μM) decreased the Cd content in roots compared with roots treated with Cd alone (Supplementary Figure S3).

These results showed that ethylene could play a role in plant response to Cd stress supporting the above mentioned results and led us to wonder about the mechanism by which ethylene modulates root system in response to Cd stress.

To understand the mechanism of ethylene response to Cd stress, we determined the superoxide in the root tips of Col-0, ein2-5, and ein3-1eil1-1 under Cd stress. As the results in Figure 5 indicate, superoxide was predominantly localized at the meristematic and elongation zones. Treatment of the 4-days-old seedlings with 75 μM Cd for 2 days significantly increased the production of superoxide in all plants. However, the accumulation of superoxide in both the ein2-5 and ein3-1eil1-1 mutants was significantly greater than in the wild-type plants. Furthermore, after 4 days of Cd treatment, the concentration of superoxide in the root tips was still higher in the two ethylene-insensitive mutants than in the wild-type plants.

The altered response pattern of superoxide accumulation in ein2-5 and ein3-1eil1-1 suggested that ethylene signaling regulates superoxide accumulation in the Cd-stressed root tips.

To investigate whether regulating superoxide is the mechanism by which ethylene modulate root system under Cd stress, effect of exogenous ACC (a precursor of ethylene biosynthesis) on the accumulation of superoxide in the root tips of the Col-0 plants under Cd stress was determined. Generally, Cd stress-induced increases in superoxide accumulation were detected after 4 days of treatment in the Col-0 root tips as shown in Figure 6. Although the supplementation with ACC slightly increased superoxide accumulation, treatment with ACC together with Cd markedly reduced the production of superoxide compared with the Cd treatment alone in the root tips, especially with 0.01 μM concentration of ACC. Conversely, the ACC-induced suppression of the production of superoxide was reversed by AIB (an inhibitor of ethylene biosynthesis) application with ACC and Cd, confirming the results in Figure 5 and supporting the involvement of ethylene in regulation of superoxide accumulation in the Cd-stressed root tips.

It was well established that PMs NADPH oxidase transfers electrons from cytoplasmic NADPH to O₂ to form O₂^- thus we subsequently determined activity of NADPH in seedlings roots. Same results were also obtained in determining NADPH oxidase activities when exogenous ACC were applied to CdCl2 stressed seedlings and also with the addition of AIB (Figure 7).

The effect of ACC on the root system development in Col-0 plants under Cd stress was also determined. Application of exogenous ACC significantly increased lateral roots number and the length of root hairs. The effect of ACC was reversed by AIB that decreased the root hairs density and root hairs length (Supplementary Figure S4).

Taken these results together, we suggested that ethylene modulates root system development under Cd stress by regulating superoxide concentration.

To investigate whether ethylene reduces Cd stress-induced superoxide accumulation through the SOD pathway, the activities of SOD and its three isoenzymes (Cu/Zn-SOD, Fe-SOD, and Mn-SOD) were measured in Col-0, ein2-5, and ein3-1eil1-1 under Cd stress.

As the results in Figure 8 illustrate, the SOD activities were increased after Cd treatment in the Col-0 but decreased in the ein2-5 and ein3-1eil1-1 plants. In addition, the application of exogenous ACC significantly increased these activities compared with the Cd stress in both Col-0 and the ein3-1eil1-1 mutants. However, ACC displayed no effects on the SOD activities of the ein2-5 plants. Furthermore, the Cu/Zn-SOD and Fe-SOD activities significantly increased while the Mn-SOD activity did not change after 5 days of Cd treatment in the Col-0. Conversely, the activities of all three isoenzymes decreased in the ein2-5 and ein3-1eil1-1 roots. The application of ACC to the Cd-stressed seedlings increased the activities of all three isoenzymes in both the wild-type and ein3-1eil1-1 plants, whereas ACC only increased the activity of Fe-SOD in the ein2-5 mutant.

The increased SOD activities in the Col-0 compared with the ein2-5 and ein3-1eil1-1 plants under Cd stress suggests that ethylene reduces Cd stress-induced superoxide accumulation through the SOD pathway.

High concentrations of superoxide induce oxidative stress, which ultimately leads to cell death. We evaluated the progression of cell death in the root tips using PI and Trypan blue. As the results of the PI staining show, exposing the seedlings to Cd stress generally led to significant cell death in the primary root tips after treatment, especially from the end of the meristematic zone to the elongation zone. However, the PI-positive cells staining intensity were significantly increased in the ein2-5 and ein3-1eil1-1 mutant root tips compared with the wild-type plants after 2 days of Cd treatment. Furthermore, after 4 days of Cd treatment, the entire meristematic and most of the elongation zone of the stressed root tips were strongly stained in the ein2-5 and ein3-1eil1-1 mutants, whereas only the end of the elongation zone was affected in the wild-type plants. Staining analyses using Trypan blue showed a similar progression of cell death, confirming the above mentioned results (Figure 9).

As shown in Figure 10, only weak DAPI (weak fluorescence and round, homogenously stained nuclei) and TUNEL signals were detected in the untreated seedling roots. However, a marked increase in DAPI fluorescence and condensed and granular nuclear staining and TUNEL-positive signals were detected in the meristematic zone to the elongation zone from root tips after Cd treatment. However, higher intensity of DAPI fluorescence and TUNEL-positive signals were also detected in the wild-type root tips compared with the ein2-5 and ein3-1eil1-1 mutants.