Arabidopsis thaliana (Col-0 ecotype) and tobacco (Nicotiana tabacum var. Xanthi) seeds were surface-sterilized with 70% ethanol for 1 min and 1% hypochlorite for 5 min, and then rinsed five times with sterilized water. To determine the effects of GSH and buthionine sulfoximine (BSO) on seed germination and seedling growth, seeds were germinated and grown on half-strength MS medium (Murashige and Skoog, 1962) containing 50–200 μM GSH or BSO. The plants were grown at 22°C under long-day conditions (16 h-light/8 h-dark cycle).

To evaluate the effects of heavy metals on seed germination, seeds were sown on half-strength MS medium supplemented with 75 μM CdCl₂, 75 μM CuSO₄, 750 μM ZnSO₄, or 10–40 μM HgCl₂. To test the effects of Hg on seedling growth, the seeds were fully germinated on normal MS medium, 5-day-old seedlings were transferred to MS medium containing various concentrations of HgCl₂, and root length was measured. To investigate the effects of GSH on multiple types of heavy metal stress, Arabidopsis seeds were germinated in the presence of Cd, Cu, Zn, or Hg with or without 50 μM GSH.

Approximately 200 mg of 10-day-old seedlings treated with or without GSH in the presence of 10 or 20 μM Hg were extracted with 0.1% trichloroacetic acid (TCA) at 4°C, and the extract was mixed with 0.5 ml of 100 mM potassium phosphate buffer (pH 7.0) and 1 ml of 1 M KI. The reaction mixture was placed in the dark for 1 h, and the H₂O₂ content was determined by measuring absorbance at 410 nm. In situ accumulation of H₂O₂ and O₂^- was detected by histochemical staining with diaminobenzidine (DAB) and NBT, respectively, according to the procedure described by Romero-Puertas et al. (2004) with minor modifications. Briefly, 10-day-old Arabidopsis seedlings grown on 1/2 MS agar medium and 3-week-old Arabidopsis leaves were carefully removed, transferred to solution containing 20 μM Hg either with or without GSH, and further incubated for 1 day. To detect H₂O₂, the seedlings and leaves were immersed in DAB solution (1 mg mL^-1) and incubated at room temperature for 7 h under continuous light until brown spots developed, which are derived from the reaction of DAB with H₂O₂. After de-staining by soaking in 100% ethanol solution, the color of the leaves and roots was observed under a microscope. For O₂^- detection, the leaves were immersed in NBT solution (1 mg mL^-1) dissolved in 10 mM phosphate buffer (pH 7.8) at room temperature. The immersed leaves were illuminated for 2 h until dark spots appeared, which are characteristic of blue formazan precipitates.

Lipid peroxidation and chlorophyll content were measured in the same samples used for H₂O₂ detection as previously described (Kim et al., 2013). Lipid peroxidation was determined by measuring malondialdehyde (MDA) content using the thiobarbituric acid (TBA) method. Briefly, harvested seedlings were homogenized with 0.5% TBA in 20% TCA, and the mixture was incubated at 95°C and then cooled. Absorbance of the supernatant was measured at 532 and 600 nm. To measure chlorophyll content, the seedlings were extracted with 95% (v/v) ethanol, and absorbance of the extract was measured at 664 and 647 nm.

Fourteen-day-old Arabidopsis seedlings grown in hydroponic culture were treated with 10 or 20 μM Hg with or without GSH for 1 day and harvested for Hg content analysis. In addition, seeds were sown and grown for 14 days on half-strength MS medium containing 10–20 μM HgCl₂, 75 μM CdCl₂, 75 μM CuSO₄, or 750 μM ZnSO₄ with or without GSH, and the seedlings were harvested for metal content analysis. Harvested seedlings were thoroughly rinsed with H₂O, dried at 80°C, and used to measure metal content using an inductively coupled plasma optical emission spectrometer (ICP-OES). All experiments were repeated at least three times.

Data represent mean values with standard deviations of three independent replications, and statistical significance was analyzed by Student’s t-test (p < 0.05; SIGMAPLOT software; systat software Inc.).

Glutathione was reacted with HgCl₂, CdCl₂, CuCl₂, ZnCl₂, NiCl₂, or PbCl₂ (molar ratio, 1:2) for 1 h at room temperature. The mixture was then spotted on a TLC plate with silica gel G. After drying at room temperature, the TLC plate was developed with buthanol/acetic acid/water (3:1:1, v/v/v).

Samples containing GSH alone (4 × 10^-3 M) or GSH together with heavy metals, including Cd²⁺, Cu²⁺, Hg²⁺, and Zn²⁺, were prepared by adding 1 equivalent amount of cations to the GSH solution in D₂O. ¹H NMR spectra were recorded on a 300 MHz Varian Unity spectrometer using tetramethylsilane (TMS) as an internal standard.

To evaluate the effects of GSH on seed germination and seedling growth, Arabidopsis seeds were sown on half-strength MS medium supplemented with various concentrations of GSH, as well as BSO that inhibits a key enzyme necessary for GSH biosynthesis and decreases cellular GSH levels (Griffith and Meister, 1979; Howden et al., 1995a,b). Both GSH and BSO delayed seed germination and inhibited post-germination seedling growth in a concentration-dependent manner (Supplementary Figure S1). When seeds were germinated and grown with a low concentration of 50 μM GSH or BSO, the germination rate and primary root growth of Arabidopsis plants were comparable to those of untreated plants, whereas the germination rate and root growth of Arabidopsis plants were severely inhibited when treated with 100–200 μM GSH or BSO. Since application of 50 μM GSH or BSO did not affect seed germination and seedling growth of Arabidopsis, we used 50 μM GSH or BSO to evaluate the effects of GSH on heavy metal tolerance in subsequent experiments.

To assess whether exogenous GSH contributes to heavy metal tolerance, seed germination, and seedling growth of Arabidopsis plants were evaluated on MS medium containing various concentrations of heavy metals with or without GSH. The results showed that 75 μM Cd, 75 μM Cu, or 750 μM Zn markedly inhibited the germination rates and post-germination seedling growth of Arabidopsis; root length of 7-day-old Arabidopsis was approximately 2.0, 1.8, and 3.1 mm in the presence of 75 μM Cd, 75 μM Cu, or 750 μM Zn, respectively, and application of GSH or BSO did not influence the inhibitory effects of these heavy metals on seed germination and seedling growth (Figure 1). Moreover, exogenous GSH did not affect the seedling growth of Arabidopsis in the presence of other heavy metals, such as 100 μM Ni, 500 μM Pb, and 500 μM Co (Supplementary Figure S2). These results suggest that exogenous GSH does not affect Arabidopsis tolerance against all these tested heavy metals.

We next evaluated whether exogenous GSH confers tolerance to Hg, which is one of the most toxic metals and seriously damages plant growth (Chen and Yang, 2012). As expected, Hg severely retarded seed germination and post-germination seedling growth of Arabidopsis plants (Figure 2). Notably, exogenous GSH significantly increased Hg tolerance; application of 50 μM GSH markedly promoted seed germination and seedling growth of Arabidopsis in the presence of 20–40 μM Hg, while 50 μM GSH treatment completely abolished Hg toxicity at a low (10 μM) concentration (Figures 2A,B). By contrast, the seed germination and post-germination seedling growth of Arabidopsis plants were more strongly inhibited by the application of 50 μM BSO, an inhibitor of GSH biosynthesis, in the presence of Hg (Figure 2B). The positive effects of GSH on seed germination and post-germination growth in the presence of Hg were also observed in other plant species, including tobacco, camelina, rice, alfalfa, and rape seed (Supplementary Figure S3). These results indicate that exogenous GSH confers Hg tolerance in diverse plant species.

To further evaluate the influence of GSH on Hg tolerance, 5-day-old seedlings grown on normal MS medium were transferred to MS medium containing Hg with or without GSH or BSO, and root growth was examined. The results showed that exogenous GSH significantly alleviated Hg-induced toxicity in the presence of 10–20 μM Hg. In particular, GSH treatment completely abolished Hg toxicity at a low (10 μM) concentration; root length of Arabidopsis 5 days after 10 μM Hg treatment was approximately 1.8 cm, whereas root length of the plants under 10 μM Hg with GSH was approximately 3.0 cm which was comparable with that grown under normal conditions (Figures 2C,D). By contrast, application of BSO further increased Hg toxicity; root length of Arabidopsis under 10 μM Hg with 50 μM BSO was approximately 0.8 cm (Figures 2C,D). The contribution of GSH to Hg tolerance was also observed in other plant species, such as tobacco and pepper (Supplementary Figure S4). Collectively, these results indicate that exogenous GSH confers Hg tolerance in diverse plant species.

To determine whether GSH confers tolerance specifically to Hg in the presence of other heavy metals, Arabidopsis seeds were sown and grown on MS medium supplemented with various combinations of 20 μM Hg and other heavy metals, including 50 μM Cd, 50 μM Cu, or 750 μM Zn, with or without 50 μM GSH. The results showed that Arabidopsis root growth was more severely inhibited in the presence of both Hg and each additional heavy metal than in the presence of Hg or each heavy metal alone on day 7 (Figure 3). Clearly, GSH treatment significantly increased the root growth of Arabidopsis plants in the presence of both Hg and each of the other heavy metals (Figure 3 and Supplementary Figure S5). These positive effects of GSH on Hg tolerance in the presence of other heavy metals were more clearly visible on day 14 (Supplementary Figure S6). Taken together, these results indicate that GSH contributes to Hg tolerance in the presence of other heavy metals.

Heavy metals are well-known to generate ROS. Since GSH are responsible for reducing the toxic effects of ROS in plant tissues, we examined H₂O₂ levels after treatment of Hg with or without GSH. The levels of H₂O₂ were markedly increased under 10 or 20 μM Hg stress but were significantly diminished by GSH application (Figure 4A). The results were further supported by histochemical detection of H₂O₂ (Figure 4B). More intense brown formazan precipitates were detected in the leaves and roots of Arabidopsis after Hg treatment, whereas the brown signals were relatively weak after GSH treatment (Figure 4B). We also detected the superoxide anion (O₂^-) in mature leaves of Arabidopsis. More intense purple precipitates were observed in mature leaves of Arabidopsis after Hg treatment, whereas the purple signals decreased to control levels after GSH treatment (Figure 4B). These results indicate that Hg-produced H₂O₂ and O₂^- were efficiently scavenged by GSH treatment.

To investigate the physiological effects of exogenous GSH against Hg stress, chlorophyll content, and degree of lipid peroxidation were analyzed in Arabidopsis seedlings under Hg stress with or without GSH. When Arabidopsis seedlings were incubated in the presence of 10 or 20 μM Hg for 1 day, the color of the leaves changed from green to yellow, whereas the color of the leaves remained green under treatment of Hg plus GSH (Figure 5A). The chlorophyll content was decreased to 52 and 31% of the control under 15 and 30 μM Hg exposure, respectively, whereas GSH treatment increased the chlorophyll content up to 99 and 64% of the control in the presence of 10 and 20 μM Hg, respectively (Figure 5A). These results indicate that GSH can ameliorate Hg-induced damage on the cellular level. The MDA level is generally used to reflect the extent of membrane lipid peroxidation, which is a sensitive diagnostic index of oxidative damage (Cargnelutti et al., 2006). The levels of MDA were significantly increased in the presence of 10 or 20 μM Hg, whereas MDA levels were markedly decreased by GSH treatment (Figure 5B).

As it is evident that exogenous GSH alleviates Hg toxicity, the next important question is how GSH confers Hg tolerance in plants. To answer this question, the cellular concentrations of Hg as well as other heavy metals were determined in Arabidopsis seedlings grown in the presence of Hg with or without GSH. When 14-day-old Arabidopsis plants grown in hydroponic cultures were treated with Hg with or without GSH, Hg content inside the cells reduced down to approximately 64 and 55% levels at 10 or 20 μM Hg, respectively, after GSH treatment (Figure 6A). Furthermore, when 14-day-old Arabidopsis plants grown on MS agar were treated with Hg with or without GSH, Hg content reduced down to approximately 48 and 60% levels at 10 or 20 μM Hg, respectively, after GSH treatment (Figure 6A). On the other hand, exogenous GSH slightly decreased the levels of other heavy metals, such as Cd, Cu, and Zn (Figure 6B). These results suggest that GSH much more significantly inhibits the cellular accumulation of Hg than that of other heavy metals in plants.

Given the evidence that GSH inhibits Hg accumulation in plant cells, the next critical question is how GSH affects the accumulation of heavy metals in plants. To obtain some insight into this important point, we determined the binding affinity of GSH for Hg and other heavy metals using TLC and NMR analysis. First, TLC was used to investigate the migration of GSH-heavy metal complexes on a TLC plate. GSH itself migrated efficiently to the upper part of the plate, whereas Hg significantly suppressed the migration of GSH on the TLC plate (Figure 7A). By contrast, the migration of GSH was not affected by other heavy metals, such as Cd, Cu, Zn, Ni, or Pb (Figure 7A). The impediment of the migration of GSH in the presence of Hg suggests that GSH forms a stable complex with Hg.

To further determine the binding affinity of GSH for Hg and other heavy metals, we next conducted a ¹H NMR experiment. The ¹H NMR spectrum of GSH showed a triplet at δ 4.58 (H4), a singlet at δ 3.82 (H6), a triplet at δ 3.78 (H1), a doublet at δ 2.95 (H5), a triplet at δ 2.56 (H3), and a quartet at δ 2.16 (H2), which can be assigned to each proton in the GSH molecule (Figure 7B). We then measured the ¹H NMR spectra of GSH in the presence of 1 equivalent of HgCl₂, CdCl₂, CuCl₂, or ZnCl₂. When HgCl₂ was added to the GSH solution, the chemical shits of three protons, H1, H2, and H3, did not change, whereas a large downfield shift of three protons, H4, H5, and H6, was observed (Figure 7B). By contrast, the chemical shifts of the protons in GSH did not change in the presence of other heavy metals, such as Cd, Cu, and Zn, although the addition of Cu caused a small change in the chemical shifts and a broadening of the bands (Figure 7B). These results suggest that GSH binds more strongly to Hg than to other heavy metals.