Journal of Applied Botany and Food Quality 88, 234 - 240 (2015), DOI:10.5073/JABFQ.2015.088.034 1 Department of Environment, Ecology and Plant Sciences, Stockholm University, Stockholm, Sweden 2 Department of Land and Water Resources Engineering, Royal Institute of Technology, Stockholm, Sweden Influence of silicon on arsenic uptake and toxicity in lettuce Maria Greger1*, Claes Bergqvist1, Arifin Sandhi1, 2, Tommy Landberg1 (Received November 20, 2014) * Corresponding author Summary Lettuce grown in soil is found to contain high concentrations of arsenic (As). This paper investigates the uptake and speciation of As in lettuce as well as the influence of silicon (Si) on As uptake, since Si may decrease it. Lettuce plants were cultivated in nutrient solution containing arsenite or arsenate with or without silicate. The uptake and distribution of As between roots and shoots, As accu- mulation in cell walls, As speciation, and toxic effects on growth were analysed. Results indicate that arsenite was more toxic to let- tuce than was arsenate. Silicate decreased arsenate toxicity but had little effect on arsenite toxicity. In contrast, Si decreased arsenite uptake more than arsenate uptake. The concentration of arsenate was higher than that of arsenite in the plants independent of the As species added. When arsenate was added, the As concentration in shoots was half of that in the roots and this distribution did not change with Si addition. When arsenite was added, approximately 10 % of As was found in the shoots and 90 % in the roots; this pattern changed in the presence of Si, and As became evenly distributed in the plant. In both roots and shoots, approximately 40 % of the As was found in the cell wall fraction; when arsenite was added, the presence of Si increased this fraction to 47 %, but only in the shoots. The extraction efficiency when analysing the As species was lower in shoots than in roots, especially in the presence of arsenite and Si. The opposite was found for As concentration in pellets after ex- traction. This indicated variation in the binding strength of arsenite and arsenate between roots and shoots and between Si- and non-Si- treated plants. Introduction Arsenic (As) is ubiquitous in nature due to natural causes and an- thropogenic activity. Arsenic is toxic to humans and may be taken into the human body via polluted waters and food (Bundschuh et al., 2012). Crop plants accumulate As and lettuce (Lactuca sativa) accumulates high As levels in the edible parts compared with other crops (GreGer, 2006; BerGqvist et al., 2014). Arsenic exists in various species in nature, the inorganic species arsenite and arsenate being predominant (sadiq, 1997). In water or waterlogged systems, arsenite is the predominant As species, while arsenate is more common in terrestrial systems with higher redox potential (sadiq, 1997). Small amounts of methylated arsenic spe- cies such as methylarsonic acid (MMA) and dimethylarsinic acid (DMA) may also be found due to microbiological activity (Wood, 1974). Plants mainly contain the inorganic As species arsenate and arsenite (smith et al., 2008). Several plant species also contain methylated As species (BerGqvist and GreGer, 2012). The levels of methylated As are usually low (raaB et al., 2007), and some argue that methylated As is not formed in plants but is taken up from the surroundings (Lomax et al., 2012). Arsenic is generally considered toxic to plants, such as lettuce (sturchio et al., 2011). The toxicity, however, differs between As species. To most organisms, inorganic As species are regarded as more toxic than organic As species, and of the inorganic As species, arsenite is considered more toxic than arsenate (meharG and hartLey-Whitaker, 2002). Plants generally contain small amounts of As, especially in their aboveground parts, possibly due to restricted uptake and trans- location from roots to shoots (WanG et al., 2002). However, some plants, such as the radish (Raphanus sativus), when cultivated hydro- ponically, accumulate higher concentrations of As in the shoots than the roots (smith et al., 2008). Arsenite-sulphur compounds localized in the phloem could account for the high distribution to shoots in the radish (smith et al., 2008). Much of the As in plants is bound in the apoplasmic cell-wall fraction, and 30-60 % was found in the apoplasm of rice (Oryza sativa) and Cretan brake (Pteris cretica) (Bravin et al., 2008; FenG et al., 2011). Cellular uptake of arsenate is thought to occur through the high-affinity phosphate transporters in terrestrial plants (moreno-Jiménez et al., 2012), while cellular uptake of arsenite occurs through aquaporins used for glycerol and silicon (Bienert et al., 2008). The As accumulated in the cytoplasm is generally thought to be reduced to arsenite, bound to phytoche- latins, and transported further into the vacuole (moreno-Jiménez et al., 2012). Some plants additionally display arsenite cellular efflux ability when growing in an As-containing medium (Bienert et al., 2008). Silicon is one of the most abundant elements on earth, but most of it has low availability to plants. Although Si is beneficial for some plant species, such as rice, bamboo, and sugar cane, its essentiality to other plants has been questioned (epstein, 2009). Silicon was found to prevent the toxicity and decrease the uptake of toxic elements in plants (treder and ciesLinski, 2005). Silicon may also alleviate the negative effects of As on plants; for example, in rice, Si inhibits the uptake of As, possibly due to the competition between silicic acid and arsenite (BoGdan and schenk, 2008). Such effects, how- ever, are not seen in the As hyperaccumulator Chinese brake (Pteris vittata), whose uptake mechanisms and routes differ from those of non-accumulators (WanG et al., 2010). Lettuce is not a Si-accumu- lator and not an As hyperaccumulator. The influence of Si on As up- take and toxicity to lettuce is therefore difficult to predict from data on Si-accumulators such as rice and As accumulators such as Pteris vittata. Crops are cultivated on As rich soils, such as fertile alum shale soils, and lettuce is a popular vegetable to be cultivated. It is thus im- portant to elucidate the influence of Si as an agent to decrease As uptake in lettuce. The aim of the study was therefore to investigate the effect of Si on the toxicity, accumulation, and speciation of As in lettuce (Lactuca sativa) when As was added as arsenate and arsenite. The hypothesis was that Si would decrease the arsenic uptake by, accumulation in, and toxicity to lettuce. The effect would likely be greatest when arsenite was added. Arsenite and silicate are thought to have the same uptake site on cell membranes; because Si decreas- es the cellular entrance of arsenite, a higher As level would be found in the apoplasm. More As would therefore accumulate in roots in relation to shoots with the addition of Si. By means of these effects, As toxicity will likely decrease with Si addition. Silicon effect on arsenic in lettuce 235 Material and methods Plant material Seeds of lettuce (Lactuca sativa L. cv Amerikanischer brauner) were germinated in vermiculite for approximately 10 days until the plants were 1 cm tall. The plants were then transferred to 25 %-strength nutrient medium; 100 % medium contained the following concen- trations of nutrients: 10 mM KNO3, 1.5 mM KH2PO4, 3.0 mM Ca(NO3)2, 2.0 mM MgSO4, 0.5 mM (NH4)2 HPO4, 25 μM FeCl3, 18.8 μM NaEDTA, 10 μM H3BO3, 16 μM MnCl2, 0.16 μM CuSO4, 0.35 μM ZnSO4, and 0.21 μM Na2MoO4. The pH was 6.4. When the plants were 10 cm tall, they were transferred to 50 %-strength nutri- ent medium, three plants per 1-L black pot. The nutrient solution was changed every week. Two weeks later, the plants were used in the toxicity tests and uptake experiments. Plants were grown in a climate chamber with a 16 h light / 8 h dark regime at 23 °C / 19 °C and a photon flux density of 270 μmol m–2 s–1 during the light period. The humidity was 80 %. Toxicity test The plants were weighed and two plants were transferred to each 1-L pot containing 50 %-strength nutrient solution. The solution was supplemented with various concentrations (i.e., 0, 0.1, 1, 10, 50, 100, 500, 1000, or 5000 μM) of arsenite (NaAsO2) or arsenate (Na2HAsO4 × 7H2O), with or without 1 mM K2SiO3 added. The silicate concentration was chosen from a pilot study were 0-5 mM K2SiO3 and 0-10000 μM arsenate or arsenite was used under the same condition as described above. After five days treatment plants were divided into roots and shoots and weighed, dried at 80 °C for 48 hrs, and weighed again. The pH of the solution at the end of the five days of treatment was measured to be approximately 6.0 without and 6.5 with Si; four replicates were used. Arsenic uptake Two plants were transferred to each 1-L pot containing 50 %-strength nutrient solution. After 24 h, the solution was replaced with a new nutrient medium containing 10 μM arsenite or arsenate in combina- tion with 0 or 1 mM K2SiO3 added and treated during four days. The plants were then harvested and total As was measured in the whole tissue and in the cell wall fraction of the roots and shoots. In addition, the As speciation was analysed in the tissue fraction. This experiment was performed in four replicates. Cell walls were prepared according to Lozano-rodriGuez et al. (1997) as follows. Plant material was first homogenized in liquid nitrogen using a mortar, and then ultra-mixed using a Polytron PT2000 homogenizer (Kinematica AG, Luzern, Switzerland) in ex- traction buffer containing 500 mM sucrose, 50 mM HEPES, 1 mM sodium dithionite, 5 mM ascorbic acid, and adjusted to pH 7.5. The sample was centrifuged at 500 g for 10 min. The pellet was used for As analysis of cell walls. Short term arsenic uptake Two times 1 ml samples were taken from 1-L pots containing 50 %-strength nutrient solution in a combination of 1 μM arsenate or arsenite with or without 1 mM K2SiO3 added. One plant was placed in each1-L pot and 2 times 1 ml samples were taken after 10, 20, 30, 60 min, 2, 4, 6 and 24 hours. The total content of As in the samples was analysed in the samples using an atomic absorption spectropho- tometer (SpectrAA 55B, Varian Inc., Palo Alto, CA) with the vapour generation technique (VGA-77, Varian Inc.). Sodium borohydride (3 %; Merck, Whitehouse Station, NJ), sodium hydroxide (2.5 %; EKA Chemicals), and hydrochloric acid (6M; VWR International, Radnor, PA) were used for hydride generation. Standards of As were added to the samples to eliminate matrix interaction effects. The de- tection limit was 7 μg As L–1. The total As uptake was calculated based on plant fresh weight and the As content in the solution sam- ples. Five replicates were used. Analysis of arsenic Cell walls and whole tissue of roots and shoots were dried at 80 °C for 48 hrs. The material was then wet digested in concentrated HNO3:HClO4, (7:3 v:v). The total content of As in the wet digested plant material, and solution from the short term uptake study, was analysed using an atomic absorption spectrophotometer. Arsenic was extracted from plant material to analyse As speciation according to a modified mir et al. (2007) method. Approximate- ly 0.5 g DW of air-dried plant material was put in 50-mL Falcon tubes and then ultra-mixed using a Polytron PT2000 homogenizer (Kinematica AG) at maximum speed for approximately 10 s in 10 mL of MeOH:H2O (1:1). The tubes were then placed in a soni- cator (Transenic Digital S; ELMA, Singen, Germany) for 10 min followed by centrifugation at 3000 × g for 10 min. The superna- tant was extracted into a new 50-mL Falcon tube. To each 50-mL Falcon tube, 10 mL of MeOH:H2O (1:1) was added to the pellets for resuspension and the extraction procedure was repeated. In total, this extraction procedure was performed twice, rendering 20 mL of MeOH:H2O (1:1) solution. The same procedure was then repeated two additional times with 0.1% HCl, rendering 20 mL of HCl so- lution. The volume of the extracted MeOH:H2O (1:1) solution was reduced at 60 °C and resuspended in deionized H2O to 3 mL to re- move MeOH and to concentrate the low levels of organic As before analysis. In the case of HCl, the extracted solution was also concen- trated to 3 mL. Arsenic species in samples were separated using ion chromatogra- phy (IC) with a PRP X-100 (250 × 4.6 mm) anion exchange col- umn (Hamilton, Reno, NV). Before injection of 100 μL of solution, samples were filtered through a 0.22-μm filter. The eluent was am- monium-phosphate buffer (pH 5.8) with a flow rate of 1 mL min–1. The atomic absorption spectrophotometer (SpectAA 55B, Varian Inc.) vapour generation technique (VGA-77, Varian Inc.) was used for peak detection. Sodium borohydride (3 %; Merck), sodium hy- droxide (2.5 %; EKA Chemicals), and hydrochloric acid (6M; VWR International) were used for hydride generation. Peak amounts of As were quantified by means of external calibration with standard solutions of arsenate, arsenite, MMA, and DMA using the following chemicals: sodium arsenate dibasic heptahydrate (Sigma-Aldrich, St. Louis, MO) for arsenate, sodium-meta-arsenite (Merck) for ar- senite, sodium methylarsonate (Sigma-Aldrich) for MMA, and cacodylic acid (Sigma Aldrich) for DMA. SpectrAA Worksheet Oriented AA Software, Version 5.1 (Varian Inc.) was used to detect the peaks, and a formula devised by BurrieL-marti et al. (1968) was used to calculate peak area and implemented in Microsoft Office Excel. Analysis of silicon Roots and shoots of plants treated five days in nutrient medium added 0 and 1 mM K2SiO3 combined with 0 and 10 μM arsenite or arsenate, originating from the toxicity experiment, were dried at 80 °C for 48 hrs. The material was then wet digested in concentrated HNO3:HClO4, (7:3 v:v). The total content of Si in the wet digested plant material was analysed using atomic absorption spectropho- tometty (Varian SpectrAA 55B) with furnace. Calculations and statistics The dose-response and sensitivity of various As and silicon combi- nations in the external medium were evaluated using the modified Weibull function (1) according to tayLor et al. (1991). This was 236 M. Greger, C. Bergqvist, A. Sandhi, T. Landberg done for the increase in the fresh weight of the whole plants over five days. y = a + b exp[– (x/c) d ] (1) where y is the dependent variable, yield; x is the independent vari- able, the concentration of metal in the growth solution; a is the abso- lute minimum growth; b is the magnitude of the uninhibited growth response above the absolute minimum growth; and c and d indicate the shape of the curve, with parameter c altering the scaling on the x-axis and parameter d affecting the skewness of the dose-response (tayLor et al., 1991, 1992). The modified Weibull function provides direct estimates of some biological parameters of toxicity. These parameters are maximum unit toxicity (UTmax), defined as the maxi- mum reduction in growth per unit of concentration of arsenic (2), and the empirical toxicity threshold (TT95b), which is the concentra- tion of arsenic that results in a yield reduction of 5% (3) (tayLor et al., 1991). When d was <1, the UTmax was calculated as dy/dx at TT95b (tayLor et al., 1991): UTmax = bd/cd exp[–(d –1)/d] c(d –1) [(d –1)/d][(d –1)/d] (2) TT95b = c[(– ln 0.95)(1/d)] (3) When comparing two values, a higher TT95b and a lower UTmax indicate that a plant is less sensitive to arsenic than is a plant with a lower TT95b and a higher UTmax. The concentration of As that induced a 50 % reduction in growth increase was also calculated (i.e., EC50 values). The Student t-test and ANOVA were used to detect differences be- tween the treatments and linear regression was used to find signifi- cant differences between lines. Significance levels at p ≤ 0.05 were calculated using the statistical program JMP (version 10.0, 2012; SAS Institute Inc., Cary, NC). Results Arsenic affected the growth of lettuce to different extents depending on the As species added. Arsenite produced a more severe effect and inhibited the plant-growth increase at lower concentrations than did arsenate (Fig. 1, Tab. 1). The lower TT95b and EC50 and higher UTmax values for arsenite than for arsenate indicate this. Plants died when treated with concentrations higher than 50 μM arsenite but survived arsenate at concentrations as high as 1000 μM (not shown). The dry weight:fresh weight (DW:FW) ratio increased with increasing As addition, as fresh weight was more affected than was dry weight (Tab. 2). The shoot:root ratio (based on fresh weight) did not change with arsenate addition, but decreased slightly with high arsenite additions (Tab. 2). Addition of Si decreased the effect of arsenate on growth more than that of arsenite, and the presence of Si increased TT95b and EC50 but decreased UTmax (Tab. 1). The Si effect was especially pronounced in the As concentration range of 10 -100 μM (Fig. 1, Tab. 2). Although Si influenced the effect of As on fresh and dry weight, it did not influence the change in DW:FW ratio or shoot:root ratio caused by As (Tab. 2). Lettuce accumulated more As in roots than in shoots (Tab. 3), inde- pendent of whether As was added as arsenite or arsenate. The [As] shoot:[As]root ratio was lower when As was added as arsenite than as arsenate. When adding Si to the arsenite treatment, however, the As concentration increased in shoots and decreased in roots compared with the treatment without Si. No such Si effect was found in the arsenate case. The same concentrations found in plant tissue were also found in the cell wall fraction. Approximately 40 % of As in the plant tissue was found in the cell walls in both roots and shoots. However, when Si was added to the arsenite treatment, a larger per- centage, 47 %, of As was bound to the cell walls in shoots. The Si concentration in lettuce roots increased about 30 % when arsenite or arsenate was added but not Si (Fig. 2). Similar effect was found in shoots, but only when arsenite was added. In the presence of 1 mM Si, however, the Si concentration in both roots and shoots decreased upon addition of arsenite or arsenate. This decrease was larger in shoots than in roots and As species added did not influence the magnitude. The short-term uptake of As is shown in Fig. 3. In the first 30 min, when apoplasmic uptake occurs, treatment with arsenite + Si re- sulted in a lower uptake rate and the curve was saturated at a lower internal As concentration than in the other treatments. In the later part of the curve, showing symplasmic uptake, less As was taken up from the arsenite solution than the arsenate solution; furthermore, when Si was added, less As was taken up from either the arsenate or arsenite solution. The effects of various treatments on As speciation in plant tissue are presented in Tab. 4. In all cases, arsenite and arsenate were found in Tab. 1: Assessing response of fresh weight increase of lettuce to various combinations of arsenite or arsenate with potassium silicate using the modified Weibull function (see Calculations and statistics). Treatment Weibull parameter TT95b EC50 Utmax a b c d R2 (μM As g-1 FW) (μM As g-1 FW) (%) Arsenate 0.505 1.095 67 1.01 0.455 3.53 46.5 0.016 Arsenate + Si 0.170 1.095 615 1.01 0.685 32.50 428.0 0.002 Arsenite – 0.406 1.953 42 1.01 0.585 2.25 29.5 0.044 Arsenite + Si – 0.368 1.743 66 1.01 0.684 3.47 45.8 0.025 Fig. 1: Fresh weight decrease in relation to control of lettuce over five days of treatment with arsenite or arsenate in the presence or absence of 1 mM potassium silicate using modified Weibull function; ±SE is included in the figure, n = 4. Silicon effect on arsenic in lettuce 237 Ta b. 2 : D ry w ei gh t ( D W ), dr y w ei gh t:f re sh w ei gh t r at io (D W :F W ) o f w ho le p la nt s, a nd s ho ot :w ho le p la nt ra tio b as ed o n fr es h w ei gh t o f l et tu ce a ft er tr ea tm en t f or fi ve d ay s w ith v ar io us c on ce nt ra tio ns o f a rs en ite or a rs en at e w ith o r w ith ou t 1 m M p ot as si um s ili ca te ; n = 4 , ± SE . C ha ng e in w ei gh t w ith A s ad di tio n, % D W :F W R oo t:S ho ot A s, A rs en at e A rs en at e A rs en ite A rs en ite A rs en at e A rs en at e A rs en ite A rs en ite A rs en at e A rs en at e A rs en ite A rs en ite µM + Si + Si + S i + Si + Si + Si D W FW D W FW D W FW D W FW 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 5. 58 ± 0 .6 9 5. 89 ± 0 .4 6 5. 58 ± 0 .6 9 5. 89 ± 0 .4 6 82 .9 ± 1 .2 4 82 .0 ± 1 .0 1 82 .9 ± 1 .2 4 82 .0 ± 1 .0 1 0. 1 93 80 10 4 90 95 76 99 91 5. 15 ± 0 .3 2 7. 15 ± 0 .8 2 5. 98 ± 0 .3 3 5. 38 ± 0 .1 2 79 .9 ± 0 .3 5 79 .2 ± 1 .7 0 79 .1 ± 1 .9 7 82 .6 ± 0 .2 0 1 95 83 98 82 94 75 10 8 11 8 6. 26 ± 0 .8 3 6. 71 ± 0 .8 3 5. 82 ± 0 .3 4 5. 91 ± 0 .0 6 77 .9 ± 1 .8 6 81 .0 ± 1 .3 5 82 .5 ± 0 .7 9 83 .5 ± 0 .8 6 10 88 54 10 3 10 7 98 77 96 57 6. 26 ± 0 .3 2 5. 82 ± 0 .7 3 6. 57 ± 0 .4 9 8. 34 ± 0 .1 6 77 .9 ± 1 .8 4 80 .5 ± 1 .2 7 82 .2 ± 1 .3 7 85 .6 ± 0 .9 6 50 90 57 10 5 91 88 43 91 42 6. 77 ± 0 .4 5 7. 32 ± 0 .2 4 8. 11 ± 0 .3 2 8. 47 ± 0 .5 8 82 .9 ± 1 .2 2 78 .1 ± 0 .9 2 81 .9 ± 1 .7 2 84 .0 ± 0 .2 9 10 0 92 58 98 62 81 24 91 45 6. 85 ± 0 .3 2 8. 46 ± 0 .5 6 8. 40 ± 0 .2 2 8. 32 ± 1 .3 8 78 .9 ± 1 .5 9 79 .1 ± 1 .2 7 81 .9 ± 1 .4 6 83 .2 ± 1 .3 1 50 0 91 51 10 3 80 85 26 82 20 7. 74 ± 0 .9 0 8. 08 ± 0 .4 7 11 .1 1 ± 0. 71 8. 02 ± 0 .4 0 83 .4 ± 0 .4 8 82 .7 ± 1 .8 6 78 .7 ± 1 .2 5 81 .0 ± 1 .1 3 10 00 95 52 93 47 84 27 90 31 9. 02 ± 1 .1 8 8. 33 ± 0 .6 4 10 .2 2 ± 0. 49 10 .8 8 ± 1. 32 79 .3 ± 2 .4 1 79 .9 ± 1 .1 6 76 .0 ± 0 .7 6 77 .8 ± 0 .6 9 50 00 95 59 89 38 79 17 80 12 8. 15 ± 0 .6 3 8. 23 ± 0 .0 2 9. 59 ± 1 .1 8 9. 85 ± 2 .4 6 82 .1 ± 1 .4 3 76 .1 ± 2 .3 9 73 .5 ± 3 .6 9 74 .0 ± 2 .9 8 both plant parts. Independent of the As species added, arsenate pre- dominated in both roots and shoots. The arsenite:arsenate concentra- tion ratio was higher when arsenite was added, and this ratio was higher in shoots than in roots. The addition of silicon decreased the arsenate and arsenite concentrations in both roots and shoots, either significantly or as a tendency. Adding silicon did not change the ratio between arsenite and arsenate in either roots or shoots. Monomethylarsonic acid was detected in both the absence and presence of Si, but only when arsenate was added (Tab. 4). Silicon addition decreased the MMA concentration in both plant parts. To determine whether the MMA concentration in plants was due to bacterial activity in nutrient medium containing arsenate, steri- lized and unsterilized media without plants were analysed after four days. The results indicated that the unsterilized solution contained 0.03 ± 0.006 (SE) μM MMA, while the sterilized solution contained no MMA (not shown). Arsenite or arsenate concentrations did not differ between the sterilized and unsterilized solutions and no DMA was detected. No DMA was found in the plants after the various treatments (Tab. 4). The extraction efficiency for the As species analysis of lettuce tis- sue was higher in roots (≥ 60 %) than in shoots (31- 45 %; Tab. 4). The addition of Si reduced the extraction efficiency in shoots but significantly only in the shoots of arsenite-treated plants, where the efficiency decreased from 45 % to 13 %. During extraction for As species analysis, a pellet fraction was formed as a by-product. When Si was added, a higher percentage of As was found in this pellet fraction for the shoots of arsenite-treated plants. Discussion This study demonstrated that silicon influences the toxicity, accumu- lation, and speciation of As in lettuce (Lactuca sativa). As suggested by the hypothesis, the uptake of As, as both arsenite and arsenate, decreased when Si was present (Fig. 3). Arsenite, ar- senate, and silicate may interact with each other at uptake into the root tissue. The interaction between arsenite and silicate was likely strongly antagonistic in the cell wall sites, since the apoplasmic up- take of As, as shown in the first 30 min in Fig. 3, was significantly affected by Si only when As was added as arsenite. The ionic interac- tions in the cell walls resulting in the decreased arsenite accumulation in the apoplasmic compartments could have been due to Si-induced secondary cell wall modifications (yamamoto et al., 2012), for ex- ample, of As binding functional groups (vithanaGe et al., 2012), reducing the binding affinity for arsenite. Low cell wall binding may increase the efflux of arsenite from roots, as suggested by Bienert et al. (2008). In addition, lack of binding in root apoplasm may in- crease As translocation to shoots, increasing the As concentration in the shoots and reducing it in roots (Tab. 3). At the same time As addition decrease the translocation of Si to the shoot; the decrease of Si concentration of shoots more than that of roots (Fig. 2), showing that interaction between Si and As occurs in both directions. The effect of Si on the cellular uptake of As (shown after 30 min in Fig. 3) was, on the other hand, similar for arsenite and arsenate. Sili- cate might therefore interact with arsenite at aquaglyceroporins, i.e., the site of both arsenite and silicon uptake (meharG and Jardine, 2003; Bienert et al., 2008), and with arsenate at the high-affinity phosphate transporters (Guo et al., 2007; moreno-Jiménez et al., 2012). After being taken up, As distribution in the plants depended on the As species added and on whether or not Si was present. Most As was found in roots, especially when added as arsenite (Tab. 3). It is known that arsenite is taken up by the root cells, where it is detoxi- fied by being bound to phytochelatins, while arsenate may be trans- located further to the shoots (meharG and hartLey-Whitaker, 2002). For this reason, more As may be found in the shoots when 238 M. Greger, C. Bergqvist, A. Sandhi, T. Landberg in the arsenate concentration, nor in the arsenite concentration, in shoots in the presence of Si (Tab. 4). Adding Si did not result in any changes in the ratio between arsenite and arsenate in either roots or shoots compared with the non-silicon treatments. This suggests that enzymes responsible for arsenate/arsenite metabolism (moreno- Jiménez et al., 2012) were unaffected by Si. The change in As concentration in shoots, but no change in arsenite or arsenate, might therefore depend on the increase of some other form of As in the shoots, for example, MMA or DMA. Dimethyl- arsinic acid (DMA) was, however, not found at all in lettuce (Tab. 4). Methylarsonic acid (MMA) was only found in roots and shoots when As was added as arsenate but not as arsenite, and the MMA concentration decreased in the presence of Si (Tab. 4). The MMA originated from the unsterilized nutrient solution (not shown) and was unlikely produced in the plant itself. Therefore, the formation of DMA or MMA could not contribute to the increased As concentra- tion in shoots after Si addition. One possibility of Si increases As concentration in shoot at arsenite addition, but no change in arsenite and arsenate, is that the addition of Si may decrease the ability of As to be extracted from the plant material. This is seen by the lower extraction efficiency in the arsen- ite + Si-treated lettuce (12 %) than in the non-arsenite + Si-treated lettuce (44 %; Tab. 4). Thus, As would be more bound to the shoot material in the presence than in the absence of Si. When analysing the different As species, the extraction efficiency was lower in shoots than in roots, meaning that the binding of As to the tissue likely dif- fers between roots and shoots. Pellets remained from the extraction Fig. 2: Silicon concentration in root and shoot of lettuce after cultivation for five days in 0 and 1 mM potassium silicate with or without 10 μM arsenite (AsIII) or arsenate (AsV); n = 9, ±SE. * indicate significant difference from control. Fig. 3: Uptake of As by lettuce over 24 h from nutrient medium containing 1 μM arsenite or arsenate with or without 1 mM potassium silicate; n = 5, ±SE. Tab. 3: Arsenic concentration in whole tissue and cell walls of roots and shoots of lettuce after treatment with 10 μM arsenite or arsenate with or without 1 mM potassium silicate for four days; n = 4, ±SE. Treatment As concentration, μg gDW-1 As in cell wall fraction, % Whole tissue Cell wall Root Shoot Root Shoot Root Shoot Arsenate 1260 ± 106a 508 ± 54c 1255 ± 53s 511 ± 37u 40.7 ± 6.04 35.6 ± 2.06 Arsenate + Si 1082 ± 220a 404 ± 28c 1089 ± 199s 456 ± 47u 37.0 ± 2.76 38.9 ± 1.45 Arsenite 1131 ± 250a 179 ± 38d 1017 ± 211s 196 ± 35v 34.4 ± 1.41 38.8 ± 3.87 Arsenite + Si 560 ± 51b 488 ± 43c 547 ± 66t 581 ± 32u 38.2 ± 2.36 47.2 ± 3.57 Different letters indicate significant differences between values in each column. added as arsenate than as arsenite, which was the case in this study (Tab. 3). Silicon affected the distribution of As between roots and shoots in lettuce when added as arsenite, but not as arsenate (Tab. 3). Much more As was translocated to the shoots in the presence than in the absence of Si, resulting in similar concentrations in shoots and roots in the case of arsenite (Tab. 3). Silicate interaction might have in- hibited the cellular uptake via aquaporines of arsenite consigning arsenite to instead be translocated to the shoot. Silicate might have promoted the transformation of arsenite to arsenate in roots, which was thereafter translocated to shoots. However, there was no change Silicon effect on arsenic in lettuce 239 Tab. 4: Concentration of various As species (μg g–1 DW) in roots and shoots of lettuce after treatment with 10 μM arsenite or arsenate with or without 1 mM potassium silicate for four days. Indicated here is As in pellets left after extraction for As species in relation to total As in tissue as well as the extraction efficiency of arsenite species; n = 4, ±SE, nd = not detected. Treatment Arsenate Arsenite MMA DMA Extraction As in pellet efficiency, % in relation to total As in tissue, % Roots Arsenate 1226 ± 50a 40.9 ± 1.34a 50.8 ± 2.38a nd 58.0 ± 4.5a 6.9 ± 1.5ab Arsenate + Si 1109 ± 21a 29.9 ± 3.01b 43.7 ± 2.05a nd 65.7 ± 10.4ab 22.1 ± 10.2abc Arsenite 1145 ± 74a 56.8 ± 0.43c nd nd 70.3 ± 19.5ab 8.0 ± 1.4ab Arsenite + Si 928 ± 61b 42.9 ± 4.98a nd nd 96.2 ± 6.7b 4.8 ± 1.7b Shoots Arsenate 180 ± 15a 43.0 ± 4.09a 49.7 ± 1.38a nd 31.6 ± 4.6ab 45.6 ± 8.0c Arsenate + Si 92 ± 11b 21.3 ± 2.73b 27.9 ± 4.13b nd 20.1 ± 2.9bc 31.4 ± 7.6c Arsenite 70 ± 3b 44.1 ± 0.28a nd nd 44.6 ± 10.4a 29.8 ± 10.7cb Arsenite + Si 61 ± 8b 40.0 ± 3.30a nd nd 12.6 ± 1.2c 83.9 ± 12.1d Different letters indicate significant differences between values in the same column. procedures before the analysis of arsenite and arsenate, and the As concentration in the pellets in relation to the total concentration in the tissue was higher in shoots than in roots (Tab. 4). In addition, in the shoots of arsenite-treated plants, the share of As in the pellet fraction in relation to the As in the tissue was even higher when Si was present (Tab. 4). The same trend was found for As in the cell wall fraction (Tab. 3). One suggested explanation of this is that the pellet material originated largely from the cell walls and that Si promoted a tighter binding of As to the cell walls, possibly by modifying functional groups in the walls responsible for As binding (vithanaGe et al., 2012), thereby increasing the As concentration in the pellets. In this study, we demonstrated that As was toxic to lettuce at ele- vated concentrations, and that arsenite was more toxic than arsenate (Fig. 1, Tab. 1 and 2). The latter has been demonstrated for other plant species as well (meharG and hartLey-Whitaker, 2002). Less arsenite than arsenate was distributed to the shoots, likely to prevent the toxic effects of arsenite on the photosynthetic appara- tus, which has also been seen in cucumber (Cucumis sativus; uroic et al., 2012). One would think from the uptake data, indicating that Si influenced the uptake of arsenite more than that of arsenate (Fig. 3), that Si would have influenced the toxicity of arsenite more than that of arse- nate. However, the opposite was found. Silicate decreased the toxic effects of both arsenite and arsenate, but most efficiently decreased the toxicity caused by arsenate (Fig. 1, Tab. 1 and 2). One reason why Si generally decreased the arsenate and arsenite effect on growth was that Si decreased the net As uptake (Fig. 3), as also demonstrated by BoGdan and schenk (2008). In addition, Si may increase anti- oxidant activities in plants, alleviating the negative effects of reac- tive oxygen species (Liu et al., 2009). Silicon would decrease the toxicity of arsenate more than that of arsenite because, when arse- nate was added, both arsenite and arsenate concentrations decreased in both roots and shoots, while with arsenite addition, the concentra- tion of both species decreased only in roots in the presence of Si (Tab. 3). In the latter case, therefore, the effect on photosynthesis would be the same as without added Si. In addition, Si might de- crease the reactivity of the arsenate molecule. We conclude that silicon can be used to decrease As uptake by lettuce, and thus, reducing toxic effects on the plant and, thus, in lettuce-crops. Lowered uptake also reduces the arsenic health-risk in lettuce as human food. During lettuce production, arsenate will be the major problem, while in solution culture, arsenite could be an additional problem due to the lower redox potential than in ter- restrial conditions. In addition, cultivation in soil might include Si effects on the binding of As to soil colloids, which would in turn influence the As uptake (seyFFerth and FendorF, 2012). Lettuce has a relative high uptake of arsenic compared to many other crop plants (mcBride, 2013) but there are no international accepted limit for arsenic in food. There are suggestions eg. 200 mg/kg and 15 μg/ kg b.w (EFSA, 2009). In field, lettuce can reach these levels and sili- con would therefore be a useful tool to reduce As in food crops. Acknowledgements This work was funded by the Knut and Alice Wallenberg foundation and the Swedish Research Council. We would like to acknowledge Lillvor Wikander for assisting with the experiments. References BerGqvist, c., GreGer, M., 2012: Arsenic accumulation and speciation in plants from different habitats. Appl. Geochem. 27, 615-622. BerGqvist, c., herBert, r., persson, i., GreGer, M., 2014: Plants influ- ence on arsenic availability and speciation in the rhizosphere, roots and shoots of three different vegetables. Environ. Pollut. 184, 540-546. Bienert, G.p., thorsen, m., schüssLer, m.d., niLsson, h.r., WaGner, a., tamás, m.J., Jahn, T.P., 2008: A subgroup of plant aquaporins fa- cilitate the bi-directional diffusion of As(OH)3 and Sb(OH)3 across mem- branes. BMC Biol. 6, article no. 26. BoGdan, k., schenk, M.K., 2008: Arsenic in rice (Oryza sativa L.) related to dynamics of arsenic and silicic acid in paddy soils. Environ. Sci. Tech- nol. 42, 7885-7890. Bravin, m.n., travassac, F., Le FLoch, m., hinsinGer, p., Garnier, J.M., 2008: Oxygen input controls the spatial and temporal dynamics of ar- senic at the surface of a flooded paddy soil and in the rhizosphere of low- land rice (Oryza sativa L.): a microcosm study. Plant Soil 312, 207-218. Bundschuh, J., nath, B., Bhattacharya, p., Liu, c.W., armienta, m.a., moreno López, m.v., Lopez, d.L., Jean, J.s., corneJo, L., Lauer macedo, L.F., FiLho, A.T., 2012: Arsenic in the human food chain: the Latin American perspective. Sci. Total Environ. 429, 92-106. 240 M. Greger, C. Bergqvist, A. Sandhi, T. Landberg BurrieL-marti, F., condaL-Bosch, L., Gassiot-matas, M., 1968: Chrono- metrical method for calculation of chromatographic peak areas. Chroma- tographia 1, 507-509. EFSA, 2009: Scientific Opinion on Arsenic in Food. EFSA Panel on Con- taminants in the Food Chain (CONTAM). EFSA J. 7(10), 1351. epstein, E., 2009. Silicon: Its manifold roles in plants. Ann. Appl. Biol. 155, 155-160. FenG, r., Wei, c., tu, s., tanG., s., Wu, F., 2011: Simultaneous hyperaccu- mulation of arsenic and antimony in Cretan brake fern: Evidence of plant uptake and subcellular distributions. Microchem. J. 97, 38-43. GreGer, M., 2006: Metallupptag i växter odlade i rödfyr- och alunskiffer- jord. (in Swedish) - Report Västra Götalands Länstyrelse. Guo, W., zhu, y.G., Liu, W.J., LianG, y.c., GenG, c.n., WanG, S.G., 2007: Is the effect of silicon on rice uptake of arsenate (AsV) related to inter- nal silicon concentrations, iron plaque and phosphate nutrition? Environ. Pollut. 148, 251-257. Liu, J., Lin, s., xu, p., WanG, x., Bai, J., 2009: Effects of exogenous sili- con on the activities of antioxidant enzymes and lipid peroxidation in chilling-stressed cucumber leaves. Agr. Sci. China 8, 1075-1086. Lomax, c., Liu, W.-J., Wu, L., xue, k., xionG, J., zhou, J., mcGrath, s.p., meharG, a.a., miLLer, a.J., zhao, F.-J., 2012: Methylated arsenic species in plants originate from soil microorganisms. New Phytol. 193, 665-672. Lozano-rodriGuez, e., hernandes, L.e, Boany, p., carpena-ruiz, R.O., 1997: Distribution of cadmium in shoot and root tissue of maize and pea plants: physiological disturbances. J. Exp. Bot. 48, 123-128. meharG, a.a., hartLey-Whitaker, J., 2002: Arsenic uptake and meta- bolism in arsenic resistant and nonresistant plant species. New Phytol. 154, 29-43. mcBride, M.B., 2013: Arsenic and Lead Uptake by Vegetable Crops Grown on Historically Contaminated Orchard Soils. Appl. Environ. Soil Science 2013, 1-8. meharG, a.a., Jardine, L., 2003: Arsenite transport into paddy rice (Oryza sativa) roots. New Phytol. 157, 39-44. mir, k.a., rutter, a., koch, i., smith, p., reimer, k.J., poLand, J.S., 2007: Extraction and speciation of arsenic in plants grown on arsenic contaminated soils. Talanta 72, 1507-1518. moreno-Jiménez, e., esteBan, e., peñaLosa, J.M., 2012: The fate of ar- senic in soil-plant systems. Rev. Environ. Contam. Toxicol. 215, 1-37. raaB, a., Ferreira, k., meharG, a.a., FeLdmann, J., 2007: Can arsenic- phytochelatin complex formation be used as an indicator for toxicity in Helianthus annuus? J. Exp. Bot. 58, 1333-1338. sadiq, M., 1997: Arsenic chemistry in soils: an overwiev of thermodynamic predictions and field observations. Water Air Soil Poll. 93, 117-136. seyFFerth, a.L., FendorF, S., 2012: Silicate mineral impacts on the uptake and storage of arsenic and plant nutrients in rice (Oryza sativa L.). Envi- ron. Sci. Tech. 46, 13176-13183. smith, p.G., koch, i., reimer, K.J., 2008: Uptake, transport and transforma- tion of arsenate in radishes (Raphanus sativus). Sci. Total Environ. 390, 188-197. sturchio, e., Boccia, p., meconi, c., zaneLLato, m., marconi, s., Beni, c., aromoLo, r., ciampa, a., diana., G., vaLentini, M., 2011: Effects of arsenic on soil-plant systems. Chem. Ecol. Supplement 27, 67-78. tayLor, G.J., stadt, k.J., daLe, M.R.T., 1991: Modelling the phytotoxicity of aluminium, cadmium , copper, manganese, nickel and zinc using the Weibull frequency distribution. Can. J. Bot. 69, 359-367. tayLor, G.J., stadt, k.J., daLe, M.R.T., 1992: Modelling the interactive effects of aluminium, cadmium, copper, manganese, nickel and zinc stress using the Weibull frequency distribution. Environ. Exp. Bot. 32, 281-293. treder, W., ciesLinski, G., 2005: Effect of silicon application on cadmium uptake and distribution in strawberry plants grown on contaminated soils. J. Plant Nutr. 28, 917-929. uroic, m.k., saLaün, p., raaB, a., FeLdmann, J., 2012: Arsenate impact on the metabolite profile, production, and arsenic loading of xylem sap in cucumbers (Cucumis sativus L.). Front. Phys. 3, art. 55, 1-23. vithanaGe, m., daBroWska, B.B., mukherJee, a.B. sandhi, a., Bhatta- charya, P., 2012: Arsenic uptake by plants and possible phytoremedia- tion applications: a brief overview. Environ. Chem. Lett. 10, 217-224. WanG, J., zhao, F.J., meharG, a.a., raaB, a., FeLdmann, J., mcGrath, S.P., 2002: Mechanisms of arsenic hyperaccumulation in Pteris vittata. Uptake kinetics, interactions with phosphate, and arsenic speciation. Plant Physiol. 130, 1552-1561. WanG, x., ma, L.q., rathinasaBapathic, B., Liu, y., zenG, G., 2010: Up- take and translocation of arsenite and arsenate by Pteris vittata L.: Ef- fects of silicon, boron and mercury. Environ. Exp. Bot. 68, 222-229. Wood, J.M., 1974: Biological cycles for toxic elements in the environment. Science 183, 1049-1052. yamamoto, t., nakamura, a., iWai, h., ishii, t., ma, J.F., yokoyama, r., nishitani, k., satoh, s., FurukaWa, J., 2012: Effect of silicon de- ficiency on secondary cell wall synthesis in rice leaf. J. Plant Res. 125, 771-779. Address of the corresponding author: E-mail: maria.greger@su.se © The Author(s) 2015. This is an Open Access article distributed under the terms of the Creative Commons Attribution Share-Alike License (http://creative- commons.org/licenses/by-sa/4.0/).