Arsenic alters uptake and distribution of sulphur in Pteris vittata



Low-molecular-weight thiol (LMWT) synthesis has been reported to be directly induced by arsenic (As) in Pteris vittata, an As hyperaccumulator. Sulphur (S) is a critical component of LMWTs. Here, the effect of As treatment on the uptake and distribution of S in P. vittata was investigated. In P. vittata grown under low S conditions, the presence of As in the growth medium enhanced the uptake of SO42−, which was used for LMWT synthesis in fronds. In contrast, As application did not affect SO42− uptake in Nephrolepis exaltata, an As non-hyperaccumulator. Moreover, the isotope microscope system revealed that S absorbed with As accumulated locally in a vacuole-like organelle in epidermal cells, whereas S absorbed alone was distributed uniformly. These results suggest that S is involved in As transport and/or accumulation in P. vittata. X-ray absorption near-edge structure analysis revealed that the major As species in the fronds and roots of P. vittata were inorganic As(III) and As(V), respectively, and that As–LMWT complexes occurred as a minor species. Consequently, in case of As accumulation in P. vittata, S possibly acts as a temporary ligand for As in the form of LMWTs in intercellular and/or intracellular transport (e.g. vacuolar sequestration).


Arsenic (As) is one of the most toxic elements for living organisms, including plants. Arsenate [As(V)], the dominant inorganic species of As in many upland soils (Francesconi et al. 2002), enters the root cells via phosphate transporters because As(V) is a chemical analogue of phosphate (Caille et al. 2004; Kertulis et al. 2005). In higher plants, absorbed As(V) is rapidly reduced to arsenite [As(III)] by As(V) reductase (Bleeker et al. 2006) or non-enzymatically by glutathione (GSH) (Delnomdedieu et al. 1994). As(III) reacts with the thiol groups (−SH) of proteins in plant tissues, leading to the inhibition of cellular functions (Ullrich-Eberius, Sanz & Novacky 1989). It has also been suggested that As induces oxidative stress in plants (Requejo & Tena 2005). In general, plants avoid mineral toxicity by excluding mineral elements. However, several plant species with high internal tolerance to a toxic mineral element can accumulate huge amounts of the element, particularly in shoot tissues. They are often called ‘hyperaccumulators’. Pteris vittata L. (Chinese brake fern) is the first fern to be identified as an As hyperaccumulator (Ma et al. 2001). This species can accumulate more than 27 000 mg As kg−1 dry weight in its fronds (Wang et al. 2002). In addition to internal As detoxification mechanisms, effective xylem loading of As is necessary for As hyperaccumulation in shoots. In P. vittata, efficient reduction of As(V) to As(III) in roots is considered to contribute to effective xylem loading of As (Wang et al. 2002; Su et al. 2008). In rice, the silicate transporter Lsi2, belonging to the NIP subfamily of aquaporins, transports As(III) into the xylem (Ma et al. 2008). Wang et al. (2011) suggested that two different affinity systems (low and high) were involved in the xylem loading process for As in P. vittata, and the passive component played a more important role in the low affinity system.

It is well known that higher plants exposed to As show a substantial increase in low-molecular-weight thiols (LMWTs) such as cysteine (Cys), γ-glutamylcysteine (γEC), GSH and phytochelatin (PC, oligomers of GSH) (Grill, Winnacker & Zenk 1985; Srivastava et al. 2009). γEC synthesis is mediated by γ-glutamylcysteine synthetase (γECS) and is a rate-limiting factor in the LMWT synthesis pathway (Tripathi et al. 2007). The −SH group of these LMWTs, particularly of GSH and PC, binds to As(III), resulting in As inactivation in the plant. Furthermore, LMWTs contribute to alleviating oxidative stress induced by As (Shri et al. 2009). Although the role of LMWTs in As tolerance and/or hyperaccumulation in P. vittata has not been established, some researchers have suggested their potential involvement. Sakai et al. (2010) reported a similarity in the distribution of LMWTs (GSH and PC) and As in whole fronds, and that the expression of PvECS2, a putative gene encoding γECS, was remarkably up-regulated in fronds after As exposure. Moreover, Vetterlein et al. (2009) reported a significant positive correlation between As and total S, a key element of LMWTs, in the fronds of P. vittata. Based on these results, it appears that S nutrition is closely related to As hyperaccumulation in the shoots of P. vittata. In this study, we therefore comprehensively examined the interactive effects of the presence of As and S in the growth medium on As, sulphur (S) and LMWT accumulation in P. vittata to elucidate the role of S in As hyperaccumulation.

Materials and Methods

Experiment 1: Effect of As on accumulation and distribution of S

Plant materials

Plants of P. vittata were supplied by Fujita Corporation (Tokyo, Japan). Plants of Nephrolepis exaltata, a non-hyperaccumulator fern, were purchased from a garden centre. The plants were cultivated in peat moss prior to experimentation. Roots were carefully washed with tap water to remove adhering peat moss. Subsequently, the plants of uniform frond height (ca. 10 cm) were transferred to continuously aerated 40 L containers (15 plants per container) containing a nutrient solution for pre-culture in hydroponics. The standard nutrient solution contained 2.14 mm N (NH4NO3), 0.32 mm P (NaH2PO4·2H2O), 0.77 mm K (KCl), 1.25 mm Ca (CaCl2·2H2O), 0.82 mm Mg (MgCl2), 35.8 μm Fe (FeSO4·7H2O), 9.1 μm Mn (MnSO4·4H2O), 46.3 μm B (H3BO3), 3.1 μm Zn (ZnSO4·7H2O), 0.16 μm Cu (CuSO4·5H2O) and 0.05 μm Mo [(NH4)6Mo7O24·4H2O]; total SO4 = 0.05 mm. For pre-culture, S concentration was adjusted to 1.0 mm by adding Na2SO4. The pH of the solution was adjusted daily to 5.3–5.5 using 1 m NaOH or HCl and the solution was renewed once a week. The plants were grown in hydroponic culture in a greenhouse at Hokkaido University (13–15 h photoperiod and day/night temperature of 25–28/18–22 °C) for 2–3 weeks until new roots developed.


Plants of each fern species were transferred to continuously aerated 4 L pots (4 plants per pot) filled with four different treatment solutions consisting of the standard nutrient solution with defined concentrations of As (0 or 30 μm) and S [0.05 (low) or 1.0 (standard) mm)]. The treatments were −As/0.05S (0 μm As, 0.05 mm S), −As/1.0S (0 μm As, 1.0 mm S), +As/0.05S (30 μm As, 0.05 mm S) and +As/1.0S (30 μm As, 1.0 mm S). As was supplied as Na2HAsO4. The pH of the solution was adjusted daily to 5.3–5.5 and the solution was renewed once a week. The plants were harvested and rinsed with de-ionized water after 4 weeks of treatment and separated into fronds and roots. The basal portion of the fronds (ca. 1 cm) was not used in this study to avoid contamination of As in the culture solution. The samples were immediately frozen in liquid nitrogen, lyophilized, weighed and ground for the determination of As, S and LMWT concentrations. Each treatment was performed in triplicate.


To determine As concentration in each plant, lyophilized samples were digested with HNO3–H2O2, as described previously (Sakai et al. 2010). The As concentration was determined by inductively coupled plasma mass spectrometry (ICP-MS; ELAN DRC-e; Perkin Elmer, Waltham, MA, USA). Cys, γEC, GSH and PC2 (a dimer of GSH) concentrations in the plants were measured using a modification of the methods of Sneller et al. (2000) and Zhao et al. (2003), as described previously (Sakai et al. 2010). The midrib was removed for LMWT analysis because the frond midribs interfere with the extraction process and subsequently inhibit the detection of LMWTs (Sakai et al. 2010). S concentration in each plant was determined using a CNS analyzer (Vario Max; Elementar Analysensysteme GmbH, Hanau, Germany); samples (200 mg) were mixed with 150–200 mg of tungsten trioxide and analysed. Sulphate (SO42−) in the lyophilized samples was extracted with Milli-Q water and analysed by capillary electrophoresis (Quanta 4000 CE; Waters, Milford, MA, USA). The measurement conditions were as follows: electrolyte, 2.5% CIA-PACTM OFM Anion-BT; 5 mm Na2CrO4·4H2O; 17 μm H2SO4; capillary, fused silica (75 μm × 60 cm); detection, 254 nm; run voltage, 20 kV.

Experiment 2: Effect of γECS inhibition in roots on As accumulation in P. vittata


Plants were prepared as described in Experiment 1. Plants of P. vittata were transferred to 4 L pots filled with the standard nutrient solution with or without 250 μm L-buthionine-[S, R]-sulphoximine (BSO), a γECS inhibitor. After the 5 d BSO pre-treatment, the solutions were renewed and As treatment (0 and 50 μm arsenate) was applied. BSO treatment was maintained during As treatment. The treatments were as follows: −As/−BSO (0 μm As, 0 μm BSO), −As/+BSO (0 μm As, 250 μm BSO), +As/−BSO (50 μm As, 0 μm BSO) and +As/+BSO (50 μm As, 250 μm BSO). The plants were sampled 6 days after the onset of As treatment as described in Experiment 1. Fresh samples of fronds and roots were cut into small pieces. A portion was used for analysis of lipid peroxidation, and the rest was immediately frozen in liquid nitrogen and lyophilized.


Lipid peroxidation was estimated by measuring the amount of thiobarbituric acid reactive substances (TBARs) (Hodges et al. 1999; Singh et al. 2006). In brief, fresh plant tissue samples (0.5 g) were cut into small pieces, homogenized using mortar and pestle on ice with the addition of 2.5 mL of 5% (w/v) trichloroacetic acid, and centrifuged at 10 000 g for 15 min at room temperature. Equal volumes of the supernatant and either (1) −TBA solution comprising 20% (w/v) trichloroacetic acid or (2) +TBA solution containing the above plus 0.5% (w/v) thiobarbituric acid were added to a new tube, covered and incubated at 96 °C for 30 min. After incubation, the tubes were immediately placed on ice and centrifuged at 8000 g for 5 min. The absorbance of the resulting supernatant was recorded at 440, 532 and 600 nm. TBAR concentrations (nmol mL−1) were calculated as follows:

display math


math formula


display math

As and GSH concentrations were determined as described in Experiment 1.

Experiment 3: Effect of As on cellular distribution of S


Plants of P. vittata were prepared as described in Experiment 1. After pre-culture in the standard nutrient solution containing 1 mm SO4, the solution was replaced with a low S nutrient solution (0.05 mm S) and the plants were grown for 1 week. Subsequently, the plants were transferred to 200 mL conical flasks (1 plant per flask) filled with the treatment solutions containing 1 mm isotopically labelled 34SO4 (98% CP, as Na2SO4; Isotec, Miamisburg, OH, USA; the S nutrient source) with or without 30 μm As (as Na2AsO4). The plants were sampled 7 days after treatment initiation.

Isotope microscope analysis

The isotope microscope system is a mass spectrometer that is capable of permil-precision isotope ratio imaging with micro-scale resolution under high mass resolution (Yurimoto, Nagashima & Kunihiro 2003). Fresh pinna were cut into small pieces using a razor, embedded in Tissue-Tek OCT compound 4583 (Sakura Finetechnical, Tokyo, Japan) containing 0.1% Tween 20 and frozen in liquid nitrogen. The embedded samples were then cross-sectioned (20 μm) using a cryostat (Leica CM3050S; Leica Microsystems, Wetzlar, Germany). The section was transferred to the aluminum block, vacuum degasified, coated with gold (30 nm deep) and observed using the isotope microscope system (Hokkaido University, Hokkaido, Japan) (Yurimoto et al. 2003) and a scanning electron microscope (SEM) (S-800; Hitachi, Tokyo, Japan). For isotope microscope analysis, the observation area was homogeneously irradiated by a Cs+ primary ion beam (20 keV, 45 nA) by rastering (80 μm × 80 μm). Because molecular interference can be a problem when analysing S isotope ratios by secondary ion mass spectrometry, the mass interference contributions were cut by the exit slit to ensure sufficient separation of 31P1H and 16O16O from 32S (Yurimoto et al. 2003).

For SEM observations, an accelerated voltage of 10 kV was used and the emission current was maintained at 10 μA.

Experiment 4: As speciation in P. vittata by X-ray absorption near-edge structure (XANES) analysis

Sample preparation

Plants were prepared as described in Experiment 1. Plants of P. vittata were transferred to 4 L pots filled with the standard nutrient solution with or without 50 μm As and were grown for 20 days. They were then sampled as previously described, immediately frozen in liquid nitrogen, lyophilized and ground. As(III) (NaAsO2), As(V) (NaH2AsO4) and As(III)–GSH complex were prepared as As standards. The As(III)–GSH complex was obtained by lyophilizing a mixture of 3 mm GSH and 0.8 mm NaAsO2 (1:1, v/v) at pH 4. It should be noted here that the As(III)–GSH complex is simply an analogue of the As(III)–thiol complex that is often used as a standard for As(III)–thiol spectra (Lombi et al. 2009). As standards were diluted sevenfold with boron nitride. Standards and plant samples were pressed into pellets (1.5 mm in diameter, 1 mm thick) for XANES analysis.


As K-edge (11 867 eV) XANES spectra were acquired at BL01b1 of SPring-8 (Japan Synchrotron Radiation Research Institute, Harima, Japan). The electron storage ring operated at 8 GeV with a beam current of 100 mA. A double crystal Si(111) monochromator and two Rh-coated mirrors at a grazing incidence angle of 4.0 mrad (to remove higher harmonics) were used. The incoming beam was measured with a N2 85% Ar 15%-filled ion chamber. The XANES spectra of the plant samples were collected in the fluorescence detection mode using a 19-element Ge semiconductor detector, whereas those of the reference materials were collected in the transmission mode. Energy calibration was performed by the white line peak of NaAsO2 assigned at 11 865 eV. Composition of the As species in the solid phases of plants was evaluated by linear combination fitting (LCF) of the XANES spectra with model compounds. The model compounds for LCF included Na2AsO3 [As(III)], NaHAsO4 [As(V)] and As(III)–GSH complex solids. The fitting range was 11 850–11 880 eV. The REX 2000 ver. 2.5 program packages (Rigaku, Tokyo, Japan) were used to subtract the pre-edge background and normalize the spectra and LCFs. In LCFs, the fraction of each reference compound was set as an adjustable parameter, and optimization was achieved by minimizing the residual of the fit, defined as the normalized root-square difference between the data and the fit (R).


Experiment 1: effect of As on accumulation and distribution of S

The low S treatment and a 4 week exposure to 30 μm As did not significantly affect the dry weight of P. vittata or N. exaltata (Fig. 1a). Furthermore, no visible symptoms of toxicity were observed. In roots, As exposure increased As concentration, but the S nutrient level in the culture solution did not affect it significantly in either fern species (Fig. 1b). The As concentration in the fronds of N. exaltata increased only very slightly with As exposure (Fig. 1b). In contrast, As concentration in the fronds of P. vittata increased markedly with As exposure and was slightly higher under low S nutrient conditions (Fig. 1b). Despite the different S nutrient levels in the culture solution, no significant difference was observed in S concentration between the treatments in N. exaltata (Fig. 1c). In P. vittata, however, the S concentration tended to be high under high S conditions, except in fronds that received As treatment. In the fronds of P. vittata under low S conditions, As application significantly increased the S concentration. No significant difference was observed in the S concentration of fronds between +As/0.05S and +As/1.0S treatments.

Figure 1.

Effect of arsenic (As) and sulphur (S) treatments on dry weight (DW) (a), As concentration (b) and S concentration (c) in Pteris vittata and Nephrolepis exaltata. Bars represent ±SEs. Different letters indicate statistically significant differences between treatments (P < 0.05).

Changes in LMWT and SO42− levels in plants are expressed both as a concentration and as a percentage of the total S concentration (Table 1). In both species, Cys concentration was not affected by the treatment, except that it was higher in roots with +As/0.05S treatment. γEC and GSH concentrations in the fronds of P. vittata were increased by As application. PC2 accumulation was observed in the high As accumulation sites, that is, fronds and roots of P. vittata and roots of N. exaltata. SO42− concentration in N. exaltata was not affected by either As or S treatment. In P. vittata, the SO42− concentration increased in response to an increase in the S level of the growth medium, whereas no increase was observed in fronds in the presence of As.

Table 1. Concentration of total sulphur and each sulphur species (μmol g−1 DW) in frond and root of two fern species
 Total SCys-SγEC-SGSH-SPC2-SSO42−-S
  1. Values are the means of three replicates.
  2. The numbers in the brackets indicate the proportion of each sulphur species (each sulphur species/total sulphur).
  3. Different letters indicate significant differences at P < 0.05.
  4. γEC, γ-glutamylcysteine; GSH, glutathione; ND, not detected.
Pteris vittataFrond−As/0.05S50.4c0.209a (0.415a)ND0.456c (0.905b)ND2.80b (5.55b)
 −As/1.0S66.7b0.244a (0.365a)ND0.703b (1.054b)0.06b (0.10b)11.90a (17.82a)
 +As/0.05S80.1a0.317a (0.396a)0.072a (0.090a)1.081a (1.350a)16.28a (20.32a)4.12b (5.15b)
 +As/1.0S75.3ab0.336a (0.447a)0.051a (0.067a)1.084a (1.440a)10.39a (13.81a)5.78b (7.67b)
Root−As/0.05S20.7b0.162b (0.782b)ND0.130a (0.628a)0.19a (0.91a)5.55b (26.77b)
 −As/1.0S28.6ab0.203ab (0.711b)ND0.087a (0.304a)ND13.18a (46.16a)
 +As/0.05S21.9b0.305a (1.395a)ND0.091a (0.415a)0.78a (3.58a)6.33b (28.94b)
 +As/1.0S30.8a0.266ab (0.865b)ND0.118a (0.382a)1.97a (6.40a)10.15a (32.99a)
Nephrolepis exaltataFrond−As/0.05S102.2a0.035a (0.035bc)0.056a (0.055a)0.200a (0.195a)ND55.91a (54.68a)
 −As/1.0S88.1a0.025a (0.028bc)0.044a (0.050a)0.151a (0.171a)ND40.93a (46.49a)
 +As/0.05S79.8a0.033a (0.041ab)0.056a (0.070a)0.091a (0.114a)0.30a (0.38a)35.07a (43.98a)
 +As/1.0S101.4a0.022a (0.021c)0.040a (0.040a)0.181a (0.179a)0.27a (0.26a)50.13a (49.45a)
Root−As/0.05S26.7a0.009a (0.034a)ND0.012ab (0.044bc)ND1.55a (5.81c)
 −As/1.0S35.2a0.031a (0.088a)ND0.074a (0.210ab)ND6.14a (17.43bc)
 +As/0.05S37.2a0.003a (0.008a)ND0.004b (0.011bc)0.42a (1.12a)3.40a (9.14bc)
 +As/1.0S24.4a0.009a (0.039a)ND0.006b (0.023c)0.38a (1.57a)5.72a (23.46ab)

Experiment 2: effect of γECS inhibition in roots on As accumulation in P. vittata

Although BSO treatment decreased GSH concentration in roots, lipid peroxidation did not change in roots even in the presence of As (Fig. 2). In roots exposed to As, As concentration was not significantly affected by BSO treatment. In contrast, BSO treatment significantly decreased the As concentration in fronds with As exposure.

Figure 2.

Effect of γ-glutamylcysteine synthetase inhibition in roots on accumulation of As, glutathione and thiobarbituric acid reactive substances (TBARS) in Pteris vittata. Bars represent ± SEs. Different letters indicate statistically significant differences between treatments (P < 0.05). BSO, L-buthionine-[S, R]-sulphoximine; DW, dry weight; FW, fresh weight; GSH, glutathione.

Experiment 3: effect of As on the cellular distribution of S

Distribution of S isotopes in the fronds of P. vittata cultivated in nutrient solution containing 34S with or without As was investigated using the isotope microscope system. Because P. vittata was grown in nutrient solution with unlabelled S before As exposure, we could determine the distribution of 34S taken up during As exposure. Sulphur isotope imaging was performed on the red square area of the optical microscope image (Fig. 3). A high isotope ratio of 34S to 32S was observed locally in some regions of fronds treated with As (Fig. 3a,b). In contrast, no localization of the ratio was found in the −As treatment (Fig. 3c). Furthermore, the subcellular localization of the high 34S/32S isotope ratio zone was identified by SEM. An SEM image of a cross section of a frond of P. vittata was obtained from the same area in Fig. 3a. In the SEM image of this area, mesophyll and epidermal cells and vacuole-like organelles were observed (Fig. 4b). The composite image of the 34S/32S isotope ratio and the SEM images indicate that the high 34S/32S isotope ratio zone overlapped with the vacuole-like organelles (Fig. 4c).

Figure 3.

34S/32S isotope ratio images in a cross section of a frond of Pteris vittata treated with 34S-containing nutrient solution with or without As. A high isotope ratio of 34S/32S was locally observed in some regions of the fronds of P. vittata only when treated with As. The 34S/32S isotope ratio in nature is 0.044. (a, b) + As; (c) − As. The white scale bar corresponds to 10 μm.

Figure 4.

A composite image of the 34S/32S isotope ratio and scanning electron microscopy images in a cross section of a frond of Pteris vittata treated with As. (a) 34S/32S isotope ratio image obtained from Fig. 3a; (b) scanning electron microscope (SEM) image; (c) composite image.

Experiment 4: As speciation in P. vittata by XANES analysis

The XANES spectra of fronds and roots are shown and compared with those of the As standards [As(III), As(V) and As(III)–GSH complex] in Fig. 5. The ratio of As(III), As(V) and As(III)–GSH complex in the fronds was 81, 11 and 7%, respectively, whereas the ratio in the roots was 3, 92 and 5%, respectively (Fig. 5). This analysis revealed that the dominant form of As species was As(III) in the fronds and As(V) in the roots. The As(III)–thiol complex was only a minor species in both fronds (7%) and roots (5%).

Figure 5.

X-ray absorption near-edge structure spectra of As standards and Pteris vittata grown with As. (a) As(III)-GSH; (b) As(III); (c) As(V); (d) frond; and (e) root.


Many studies have investigated the effect of S nutrition on tolerance to and/or accumulation of heavy metals and As in plants (Srivastava & D'Souza 2010; Na & Salt 2011; Zhang et al. 2011) because LMWTs, which contain S as a key element, have roles in inactivating heavy metals and As by direct binding and/or in alleviating oxidative stress induced by these elements. As described in the Introduction, the major cause of As toxicity is protein dysfunction caused by As(III) binding to the −SH group. Because LMWTs contain a higher percentage of –SH groups, they can detoxify As(III) by binding to it, thus preventing As from binding to proteins. In the aquatic plant Hydrilla verticillata, on the other hand, an increase in the S level in the growth medium enhanced As uptake in addition to alleviation of As-induced oxidative stress (Srivastava & D'Souza 2009). Moreover, it was reported that S application effectively enhanced As uptake in P. vittata (Wei et al. 2010). In the present study, we could not determine the effect of plant S status on As uptake in P. vittata because S treatment did not affect S concentration in plants in the presence of As (Table 1). Under low S conditions, however, As application increased the S concentration in the fronds of P. vittata, suggesting that P. vittata requires S (enhances S uptake) to respond to As hyperaccumulation. In other words, S uptake in P. vittata may depend on the amount of As absorbed rather than on the S level in the growth medium.

We attempted to determine where the actively absorbed S is distributed in fronds under As exposure. We have demonstrated for the first time, using an isotope microscope system, that the 34S absorbed with As was preferentially and locally delivered into a specific organelle, most presumably a type of vacuole, in epidermal cells (Fig. 4). Under standard S conditions (1.0 mm S), As application increased the concentration of LMWTs, particularly of PC2, in the fronds of P. vittata, whereas SO4 concentration decreased (Table 1). Under low S conditions, moreover, As application increased S accumulation in fronds, which did not accumulate as SO4 but was used for LMWT synthesis (Table 1). An increase in LMWT concentrations accompanied by As accumulation has been reported in various plant species, including P. vittata (Mokgalaka-Matlala et al. 2009; Srivastava et al. 2009; Sakai et al. 2010). Although we have not performed the speciation of 34S in fronds directly, these results strongly suggest that the chemical form of 34S accumulated in the vacuole-like organelle is LMWT. Moreover, Lombi et al. (2002) reported that As was compartmentalized mainly in the epidermal cells, probably in the vacuoles. Thus, it is possible that both As and 34S, absorbed and translocated during As exposure, were co-accumulated in the vacuoles of epidermal cells as As–LMWT complexes. However, according to the present results of the XANES spectra, As(III)–LMWT is a minor species in fronds (Fig. 5).

Thus, we suppose that LMWTs play a role in As transport, including long-distance transport or transport between organs, cells and/or cell organelles, and that As–LMWT complexes finally dissociate when they have accumulated in the vacuole. Although it has been suggested that LMWTs are involved in long-distance transport of Cd from roots to shoots in Arabidopsis (Gong, Lee & Schroeder 2003), no similar results have been reported for As. Exposure to BSO, an γECS inhibitor, inhibits GSH synthesis in roots. In the present study, BSO treatment in the roots of P. vittata inhibited GSH synthesis in the roots and significantly decreased As concentration in the fronds (Fig. 2). Because inhibition of GSH synthesis did not affect lipid peroxidation in the roots (Fig. 2), this decrease in the As concentration in the fronds was not related to oxidative stress in the roots. Thus, the involvement of GSH and/or PC in long-distance transport of As in P. vittata seems plausible. However, As bound to LMWTs was only a minor species in roots (Fig. 5) and was almost undetectable in the xylem sap (Su et al. 2008) of P. vittata grown with As. Recently, Lei et al. (2013) also reported a decrease in As accumulation in the shoots of P. vittata with BSO treatment. While they assumed that this was caused by inhibition of GSH-mediated reduction in As(V) in the roots, they also reported that acceleration of LMWT synthesis in the roots enhanced As and –SH coordination, resulting in a significant increase in the tolerance of P. vittata to As.

Consequently, the hypothesis that LMWTs are involved in long-distance transport of As in P. vittata seems very unlikely. Another possibility for the involvement of LMWTs in As accumulation of P. vittata may be through temporary intercellular transport and/or intracellular transport (e.g. vacuolar sequestration) before xylem loading in roots and after unloading in shoots. As described in the Introduction, it has been suggested that the rapid reduction of As(V) to As(III) in roots contributes to effective xylem loading of As. Because As(III) strongly inhibits the protein functions in cells, it is easy to assume that the temporal formation of As(III)–LMWT complexes is essential for transporting As(III) from root cells to xylem, and from xylem to frond cells, without any cellular damage. The barely detectable levels of As bound to thiols in XANES analysis (Fig. 5) may support this hypothesis. To date, it has been reported that P. vittata has ACR3 that can transport As(III) into vacuoles, and the gene encoding ACR3 is induced by As in sporophyte roots and gametophytes (Indriolo et al. 2010). However, vacuolar membrane proteins transporting As(III)–LMWT complexes have not been reported. In Arabidopsis thaliana, Song et al. (2010) identified two ABCC-type transporters responsible for transporting As(III)–PC complexes into vacuoles. The results of the present study strongly suggest the existence of such a transporter in the fronds of P. vittata. Because As(III)–LMWT complexes are dissociated under strongly acidic conditions (Rey, Howarth & Pereira-Maia 2004), the vacuolar pH in P. vittata may be sufficiently acidic to induce dissociation of As(III)–LMWT complexes, resulting in the detection of inorganic uncomplexed As(III) in fronds (Fig. 5).


We thank M. Yasui (Electron Microscope Laboratory, Research Faculty of Agriculture, Hokkaido University) for guidance in taking SEM pictures.