P

Mmh

Ra

b

c

a

ARRAA

KHTSMT

I

eHeeefps

h0

Journal of Plant Physiology 183 (2015) 32–40

Contents lists available at ScienceDirect

Journal of Plant Physiology

journa l h om epage: www.elsev ier .com/ locate / jp lph

hysiology

olybdenum accumulation, tolerance andolybdenum–selenium–sulfur interactions in Astragalus selenium

yperaccumulator and nonaccumulator species

achael Ann DeTara, Élan R. Alfordb,c, Elizabeth A.H. Pilon-Smitsa,∗

Biology Department, Colorado State University, Fort Collins, CO 80523, USAGraduate Degree Program in Ecology, Colorado State University, Fort Collins, CO 80523, USADepartment of Forest and Rangeland Stewardship, Colorado State University, Fort Collins, CO 80523, USA

r t i c l e i n f o

rticle history:eceived 25 April 2015eceived in revised form 26 May 2015ccepted 27 May 2015vailable online 3 June 2015

eywords:yperaccumulatorransporterseleniumolybdenum

olerance

a b s t r a c t

Some species hyperaccumulate selenium (Se) upwards of 0.1% of dry weight. This study addressedwhether Se hyperaccumulators also accumulate and tolerate more molybdenum (Mo). A field surveyrevealed on average 2-fold higher Mo levels in three hyperaccumulator Astragali compared to threenonaccumulator Astragali, which were not significantly different. Next, a controlled study was performedwhere hyperaccumulators Astragalus racemosus and Astragalus bisulcatus were compared with nonaccu-mulators Astragalus drummondii and Astragalus convallarius for Mo accumulation and tolerance, aloneor in the presence of Se. When grown on agar media with 0, 12, 24 or 48 mg L−1 molybdate and/or 0,1.6 or 3.2 mg L−1 selenate, all species decreased in biomass with increasing Mo supply. Selenium did notimpact biomass at the supplied levels. All Astragali accumulated Mo upwards of 0.1% of dry weight. Sele-nium levels were up to 0.08% in Astragalus racemosus and 0.04% Se in the other species. Overall, therewas no correlation between Se hyperaccumulation and Mo accumulation capacity. However, the hyper-accumulators and nonaccumulators differed in some respects. While none of the species had a highertissue Mo to sulfur (S) ratio than the growth medium, nonaccumulators had a higher Mo/S ratio thanhyperaccumulators. Also, while molybdate and selenate reduced S accumulation in nonaccumulators, itdid not in hyperaccumulators. Furthermore, A. racemosus had a higher Se/S ratio than its medium, while

the other species did not. Additionally, Mo and Se treatment affected S levels in nonaccumulators, butnot in hyperaccumulators. In conclusion, there is no evidence of a link between Se and Mo accumulationand tolerance in Astragalus. Sulfate transporters in hyperaccumulating Astragali appear to have highersulfate specificity over other oxyanions, compared to nonaccumulators, and A. racemosus may have atransporter with enhanced selenate specificity relative to sulfate or molybdate.

© 2015 Elsevier GmbH. All rights reserved.

ntroduction

The element selenium (Se) is toxic at high levels, but also anssential micronutrient for many organisms (Terry et al., 2000).owever, Se is not an essential nutrient for plants (Valdez Barillast al., 2011). While many animals, bacteria and some algae havenzymes that require Se as a component of certain antioxidantnzymes, vascular plants do not have a confirmed physiological use

or Se (Valdez Barillas et al., 2011). However, many plants show aositive growth response to low levels of Se, which is why it is con-idered a beneficial element (Pilon-Smits et al., 2009). At optimal

∗ Corresponding author. Tel.: +1 970 491 4991.E-mail address: epsmits@lamar.colostate.edu (E.A.H. Pilon-Smits).

ttp://dx.doi.org/10.1016/j.jplph.2015.05.009176-1617/© 2015 Elsevier GmbH. All rights reserved.

levels, Se exposure has been linked to the reduction of various reac-tive oxygen species (ROS) by the upregulation of antioxidants andantioxidant enzymes (Feng et al., 2013). However, at high levels,Se oxyanions can also be a source of oxidative stress (Van Hoewyk,2013). Since Se is chemically similar to the macronutrient sulfur(S), plants are thought to take up and assimilate Se inadvertently,instead of S. The non-specific incorporation of Se into proteins maylead to toxicity because of reduced or nonexistent protein function(Terry et al., 2000).

Interestingly, certain plant species have the ability to accumu-late and tolerate extreme concentrations of Se in their tissues,

upwards of 0.1% of DW (Beath et al., 1939). Selenium hyperaccu-mulation occurs in several genera, including the legume Astragalus(Beath et al., 1939). The hyperaccumulation likely serves an ecolog-ical function, since high Se levels can protect plants from a variety

lant P

o(assaic2

lzbNanSb

atMsttioepSoid(

actto1c2Mtrs

c(pic(nwMs1tac2

dm

R.A. DeTar et al. / Journal of P

f herbivores including aphids, caterpillars, spider mites and thripsEl Mehdawi and Pilon-Smits, 2012). The Se tolerance in hyper-ccumulators is thought to be based in part on their capacity toequester Se in the form of the non-protein Se amino acids methyl-elenocysteine and selenocystathionine (Freeman et al., 2010). Inddition, Se hyperaccumulators may have higher levels of antiox-dants such as glutathione, which may prevent oxidative stressaused by inorganic forms of Se (Freeman et al., 2010; Van Hoewyk,013).

Apart from Se, other toxic elements that can be hyperaccumu-ated by plants include arsenic, cadmium, nickel, manganese andinc (Baker et al., 2000). While many hyperaccumulators have onlyeen shown to hyperaccumulate one element, some species likeoccaea caerulescens have been shown to accumulate and toler-te multiple metals (Lochlainn et al., 2011). To our knowledge,o research has been published that investigated whether or note hyperaccumulators can also accumulate other toxic elementsesides Se.

In view of the extreme capacity of Se hyperaccumulators toccumulate Se, it may be hypothesized that Se hyperaccumula-ors could also hyperaccumulate molybdenum (Mo), since Se and

o are both chemically similar to S (Marschner, 1995). Selenate,ulfate and molybdate are the oxyanion forms plants generallyake up from soil, and all can make use of the same sulfateransporters (Shinmachi et al., 2010). For example, the high affin-ty sulfate transporter SULTR1;2 in the root plasma membranef Arabidopsis thaliana can mediate selenate uptake (Shibagakit al., 2002). Similarly, molybdate has been shown to enter thelant through Stylosanthes hamata high-affinity sulfate transporterHST1 (Fitzpatrick et al., 2008; Tejada-Jiménez et al., 2013). More-ver, Mo and Se reach higher concentrations in S starved plants thann S replete plants, which further provides evidence that molyb-ate and selenate can enter the plant through sulfate transportersShinmachi et al., 2010).

Many sulfate assimilation enzymes can also use Se analogss a substrate, leading to Se incorporation into organic seleno-ompounds (Sors et al., 2005). Molybdenum appears to movehrough the first step of the S assimilation pathway, i.e. the reac-ion with ATP, mediated by ATP sulfurylase; however, the productf this reaction is thought to be unstable (Wilson and Bandurski,958). Due to its similarity to Se, Mo may also interact withertain Se binding proteins (SBP) such as SPB1 (Schiavon et al.,012). These overlaps in uptake and assimilation of Se, S, ando form the basis for the hypothesis that Se hyperaccumula-

ion capacity may also facilitate Mo accumulation. There are noeports of plant Mo hyperaccumulation so far, from field or labtudies.

Molybdenum is a micronutrient for plants, needed as a metalofactor in certain enzymes involved in nitrogen metabolismZimmer and Mendel, 1999; Tejada-Jiménez et al., 2013). Thelant requirement for Mo is extremely low (0.1 mg kg−1 DW crit-

cal deficiency level, Marschner, 1995). Nevertheless, many plantsan tolerate tissue Mo concentrations of up to 1000 mg kg−1 DWMarschner, 1995). Such high levels have not been reported inature. For example, field tissue levels of Mo in the legume alfalfaere reported to be 0.23 mg kg−1 (Gupta, 1991). The highest planto levels are found in nodules of leguminous species, where Mo

erves as a cofactor for rhizobacterial nitrogenase (Marschner,995). In addition to being taken up non-specifically via sulfateransporters, Mo can be taken up specifically through MOT1, a highffinity molybdate transporter reported to be present in both theell membrane and the mitochondrial membrane (Tomatsu et al.,

007; Tejada-Jiménez et al., 2013).

The objective of the study described in this paper was toetermine whether or not Se hyperaccumulators Astragalus race-osus and Astragalus bisulcatus accumulate and tolerate more

hysiology 183 (2015) 32–40 33

Mo than nonaccumulators Astragalus drummondii and Astra-galus convallarius. The reasoning behind the hypothesis that Sehyperaccumulators can also accumulate more Mo than nonaccu-mulators is that they have upregulated sulfate/selenate/molybdatetransporters as a Se hyperaccumulation mechanism (Schiavonet al., 2015). Thus, hyperaccumulators could potentially accu-mulate high concentrations of molybdenum through thesesulfate/selenate/molybdate transporters if there were high con-centrations of Mo in the growth medium. In addition to havinghigher Mo accumulation capacity, Se hyperaccumulators may behypothesized to have enhanced Mo tolerance. The reasoning forthis hypothesis is that some of the Se tolerance mechanisms, suchas the enhanced antioxidant levels may also alleviate Mo stress(Freeman et al., 2010; Van Hoewyk, 2013).

An alternative hypothesis is that Se hyperaccumulator plantswould not accumulate more Mo than nonaccumulators, with thereasoning that they have evolved specialized selenate transportersthat have a greater specificity for Se than for other oxyanions. Ifso, hyperaccumulators would be more likely to enrich themselveswith selenate relative to sulfate and molybdate. This hypothesis issupported by the finding that many Se hyperaccumulators have ahigher Se/S ratio relative to their growth medium and to other plantspecies (White et al., 2007). Selenium hyperaccumulators couldalso be hypothesized not to tolerate significantly more Mo thannonaccumulators on the reasoning that some mechanisms of Setolerance (e.g. the methylation of selenocysteine) are specific to Se,rather than general stress tolerance mechanisms. The experimentdescribed in this paper tests these opposing hypotheses. Further-more, we investigated the interactions between Mo, Se, and S inthese species, in an attempt to better understand the mechanismsinvolved in the uptake of these elements, and the possible role ofSe in Mo tolerance.

Materials and methods

Field survey

Leaf material was collected from six Astragalus species growingin their natural habitats in Colorado, USA as described by Alfordet al. (2012). The selenium data for these samples have been pub-lished (Alford et al., 2012). Three Se hyperaccumulator species(Astragalus bisulcatus (Hook.) A. Gray (two-grooved milkvetch),Astragalus racemosus Pursh (cream milk vetch), Astragalus prae-longus Sheldon (stinking milkvetch) and three non-accumulators(Astragalus convallarius Greene (lesser rushy milkvetch), Astragalusmissouriensis Nutt, (Missouri milkvetch), Astragalus argophyllusNutt. (silverleaf milkvetch)) were sampled. There were on aver-age six bioreplicates per species. The leaf material was analyzed forMo concentration as described below.

Growth chamber study

Seeds of hyperaccumulators A. racemosus and A. bisulcatus, andof nonaccumulators A. convallarius and A. drummondii Douglasex. Hook. (Drummond’s milkvetch) were obtained from WesternNative Seed (Coaldale, CO, USA). The seeds were scarified withsandpaper, then surface-sterilized in 70% ethanol for 2 min, fol-lowed by 15% bleach for 20 min. The seeds were then rinsed fivetimes in sterile water, and stratified at 4 ◦C for five days. The seedswere then transplanted onto sterile wet tissue paper in Petri dishes,sealed and allowed to germinate in the dark. Solid medium was pre-

pared in 100 mL aliquots in Phytatray containers (Sigma, St Louis,MO). The medium (pH 5.8) consisted of 1/2 strength Murashigeand Skoog (MS) salts (Murashige and Skoog, 1962), 1% sucroseand phytoagar (Sigma, St Louis, MO), spiked with various Se and

3 lant Physiology 183 (2015) 32–40

Ma(1c

fPgauhwwictter

wirBwwWpnaslA(

t(pod

R

Mn

AMtane

Tm

semtms

Fig. 1. Leaf Mo levels in Astragalus nonaccumulators A. convallarius, A. missouriensis,A. argophyllus and hyperaccumulators A. racemosus, A. bisulcatus, A. praelongus in thefield. The left part of the graph shows means of individual species, while the rightside shows the means of all hyperaccumulators and nonaccumulators. Error barsrepresent standard error of the mean, and letters above bars denote statistically

4 R.A. DeTar et al. / Journal of P

o concentrations. Selenium was supplied in the form of Na2SeO4nd Mo as Na2MoO4. The treatments were: 1.6 parts per millionmg L−1) Se, 3.2 mg L−1 Se, 16 mg L−1 Mo, 32 mg L−1 Mo, 48 Mo,.6 mg L−1 Se + 32 mg L−1 Se, 1.6 mg L−1 Se + 48 mg L−1 Mo and aontrol with no added Se or Mo.

Into each Phytatray, four to five emerging seedlings were trans-erred from sterile paper. Per plant species and treatment, threehytatray boxes were planted. The Phytatrays were placed in arowth chamber with a 14 h light/10 h dark photoperiod at 23 ◦Cnd 200 �E light intensity. The plants were grown for seven weeksntil harvest. The number of live plants per treatment and species atarvest was on average seven, with a range from 3 to 13. Plants thatere dead at the time of harvest were not sampled; the death ratesere similar for the four species. Because of the difficulty of obtain-

ng sufficient root material, only shoot elemental analysis could bearried out. There was an average of six bioreplicates for elemen-al analysis. For two A. convallarius treatments, i.e. 48 mg L−1 Moreatment, and 48 mg L−1 Mo + 1.6 mg L−1 Se treatment, no shootlemental data could be collected for lack of sufficient plant mate-ial.

At harvest, the plants were gently removed from the agar. Theyere rinsed in tap water, divided into roots and shoots, and placed

n tinfoil packets for elemental analysis as described below. Theoots and shoots were dried in a 50 ◦C oven until constant weight.iomass production was determined from total root and shoot dryeight. For elemental analysis, 20–100 mg of dried shoot materialas weighed and placed in a 25 mm × 200 mm glass digestion tube.here needed, two plants were pooled in order to obtain enough

lant material for elemental analysis. One milliliter of concentrateditric acid was added to each tube, and the tubes were covered with

funnel and heated at 60 ◦C for 2 h and then at 130 ◦C for 6 h. Theamples were then diluted to 10 mL with distilled water and ana-yzed for elemental concentrations via Inductively Coupled Plasmatomic Emission Spectroscopy (ICP-AES) according to Zarcinas et al.

1987).The data on biomass and elemental composition were statis-

ically analyzed using the program JMP-IN from the SAS instituteCary, NC, USA), by a one way ANOVA followed by Tukey Kramerost hoc analysis when replicates followed a normal distribution,r by non-parametric Wilcoxon/Kruskal Wallis test when replicatesid not follow a normal distribution ( = 0.05).

esults

olybdenum levels in Se hyperaccumulators andonaccumulators in the field

When three hyperaccumulator and three nonaccumulatorstragalus species were sampled in the field and analyzed for leafo concentration, there were no significant differences between

he species (Fig. 1, left side). However, the leaf Mo levels were onverage almost twice as high in hyperaccumulators as compared toonaccumulators (p = 0.067, Fig. 1). None of the average Mo levelsxceeded 15 mg kg−1 DW.

he effects of selenium and molybdenum on plant growth on agaredium

To test Se hyperaccumulator and nonaccumulator Astragaluspecies for their Mo (and Se) accumulation and tolerance prop-rties under more controlled conditions, they were grown on agar

edium spiked with different levels of Mo and Se. Molybdenum

reatment decreased the mean shoot and root biomass for A. race-osus, A. bisulcatus, A. convallarius and A. drummondii, but the

pecies differed with respect to the extent of growth inhibition by

significantly different means ( = 0.05).

Mo (Figs. 2 and 3). The mean shoot and root biomass of Se hyper-accumulator A. racemosus treated with 16 mg L−1 Mo decreasedby 43% (Fig. 2A, p < 0.05) and 27% (Fig. 3A, p < 0.05), respectively,compared to the mean control shoot and root biomass. Increas-ing the external Mo concentration to 32 or 48 mg L−1 did notfurther reduce the mean biomass of roots or shoots in A. race-mosus. The other Se hyperaccumulator species, A. bisulcatus, hada 16% increase (NS), a 17% decrease (NS), and a 38% (p < 0.05)decrease in shoot biomass when treated with 16, 32, and 48 mg L−1

Mo respectively (Fig. 2B). This species had a 77% increase, a 30%decrease, and a 50% decrease in root biomass compared to thecontrol when treated with 16, 32, and 48 mg L−1 Mo, respectively(NS, Fig. 3B).

For nonaccumulator A. drummondii, the mean shoot biomasswas reduced by 44% (p < 0.05), 60% (p < 0.05), and 56% (p < 0.05),when treated with 16, 32, and 48 mg L−1 Mo, respectively (Fig. 2C).The mean root biomass decreased by 35% (NS), 80% (p < 0.05), and77% (p < 0.05), respectively (Fig. 3C). The mean shoot biomass ofA. convallarius was not affected by treatment with 16 mg L−1 Mo(Fig. 2D, NS). Treatment with 32 mg L−1 Mo and 48 mg L−1 Modecreased the shoot biomass by 40% (NS) and 55% (p < 0.05), respec-tively. Root biomass of A. convallarius was reduced by 30%, 18%, and59% when treated with 16 mg L−1 Mo, 32 mg L−1 Mo, and 48 mg L−1

Mo, respectively (NS, Fig. 3D).Selenium treatment did not significantly affect root or shoot

biomass in any of the species at any treatment level (Suppl. Figs.1 and 2). The effect of Se on shoot and root biomass in the presenceof Mo varied among the species, with no relation to hyperaccu-mulation (Suppl. Figs. 3 and 4). For example, A. racemosus treatedwith 48 mg L−1 Mo and 1.6 mg L−1 Se had a significantly greaterroot biomass than those treated only with Mo (Suppl. Fig. 4A,p < 0.05). Yet, A. racemosus treated with 1.6 mg L−1 Se in additionto 32 mg L−1 Mo had similar root and shoot biomass compared toplants treated with 32 mg L−1 Mo alone (Suppl. Figs. 3A and 4A).For A. bisulcatus, additional treatment with Se did not significantlyaffect biomass production as compared to treatment with Mo alone(Suppl. Figs. 3B and 4B). For nonaccumulator A. drummondii, treat-ment with 1.6 mg L−1 Se in addition to 48 mg L−1 Mo lowered shootand root biomass around 2-fold as compared to plants treated onlywith 48 mg L−1 Mo, but this was only significant for the shoot;in a background of 32 mg L−1 Mo, addition of Se did not affectbiomass (Suppl. Figs. 3C and 4C). Addition of Se did not significantly

affect biomass production of A. convallarius as compared to Moalone.

R.A. DeTar et al. / Journal of Plant Physiology 183 (2015) 32–40 35

Fig. 2. The effect of Mo on shoot biomass of hyperaccumulator (A, B) and nonaccumulator (C, D) Astragalus species grown on agar medium. Panel A: A. racemosus; Panel B: A.bisulcatus; Panel C: A. drummondii; Panel D: A. convallarius. Error bars represent standard error of the mean, and letters above bars denote statistically significantly differentmeans ( = 0.05).

Fig. 3. The effect of Mo on root biomass of hyperaccumulator (A, B) and nonaccumulator (C, D) Astragalus species grown on agar medium. Panel A: A. racemosus; Panel B: A.bisulcatus; Panel C: A. drummondii; Panel D: A. convallarius. Error bars represent standard error of the mean, and letters above bars denote statistically significantly differentmeans ( = 0.05).

36 R.A. DeTar et al. / Journal of Plant P

Fig. 4. The shoot concentration of Mo and Se (mg kg−1 DW) relative to the suppliedcee

M

ft3(satAet

fmeastt

I

tblh(M

oncentration of Mo and Se in the agar media (mg L−1). Significant ( = 0.05) differ-nces between species within each treatment are denoted by different letters. Therror bars represent the standard error of the mean.

olybdenum and selenium accumulation

When treated with different concentrations of molybdate, allour species accumulated significantly higher concentrations of Mohan their controls (p < 0.05). Shoot Mo levels ranged from 1000 to000 mg kg−1 DW, with no apparent link to Se hyperaccumulationFig. 4A, darker vs. lighter bars). Among the four species, the averagehoot Mo levels were highest for nonaccumulator A. drummondii atll treatments; this was significantly different from other speciesested at 48 mg L−1 Mo and from A. racemosus at 16 mg L−1 Mo.mong the other three species there were no significant differencesxcept that A. racemosus contained less Mo than A. bisulcatus whenreated with 16 or 48 mg L−1Mo (Fig. 4A).

The Se hyperaccumulator A. racemosus consistently had twoold higher Se concentration in its shoots compared to nonaccu-

ulators A. convallarius and A. drummondii (Fig. 4B). Contrary toxpectation, the known Se hyperaccumulator A. bisulcatus did notccumulate significantly more Se than the nonaccumulators in thisystem (Fig. 4B). Note that plants not supplied with Se still con-ained this element in their tissue, likely because Se was present inhe seed.

nteractions between Mo, Se and S

Effects of Mo on S: Molybdate treatment significantly reducedhe shoot S concentration in the nonaccumulators (Fig. 5C and D),ut did not have a significant effect on S levels in hyperaccumu-

ators (Fig. 5A and B), with one exception. The S concentration inyperaccumulator A. racemosus was 22% reduced by 32 mg L−1 Mo,p < 0.05, Fig. 5A). Otherwise there were no significant effects of

o on S concentrations in A racemosus or A. bisulcatus. The shoot S

hysiology 183 (2015) 32–40

concentration in nonaccumulator A. drummondii was significantlyreduced compared to the control at all treatment levels, by 28%,34%, and 34% when treated with 16 mg L−1 Mo, 32 mg L−1 Mo,and 48 mg L−1 Mo, respectively (Fig. 5C). The S concentration innonaccumulator A. convallarius was also significantly reducedcompared to the control at all Mo treatment levels, by 39% and 56%when treated with 16 mg L−1 Mo and 32 mg L−1 Mo, respectively(Fig. 5D). There was not sufficient biomass to obtain 3 bioreplicatesof A. convallarius treated with 48 mg L−1 Mo.

Effects of Se on S: Selenium treatment reduced shoot S con-centrations in nonaccumulators A. drummondii and A. convallarius(with one exception), and did not affect shoot S concentrations inhyperaccumulators A. racemosus and A. bisulcatus (Fig. 6). In nonac-cumulator A. drummondii the S levels for treatments of 1.6 mg L−1

and 3.2 mg L−1 Se were reduced by 25% and 26%, respectively, com-pared to the control (p < 0.05, Fig. 6C). Treatment with 1.6 mg L−1 Sereduced shoot S concentrations in A. convallarius by 34% (p < 0.05)and increased S concentrations by12% (NS) in plants treated with3.2 mg L−1 Se (Fig. 6D).

Interactions between Mo and Se: Selenium treatment did not havea consistent effect on tissue Mo levels in any of the species (resultsnot shown), which was perhaps to be expected because the con-centration of Mo supplied was at least ten fold higher than that ofSe. The addition of Mo affected Se accumulation differently in thefour Astragalus species. Selenium levels in the hyperaccumulatorspecies were not significantly affected by adding Mo (Fig. 7A andB). Nonaccumulator A. drummondii treated with 1.6 mg L−1 Se and48 mg L−1 Mo had a 39% reduction in internal Se concentration com-pared to plants treated only with 1.6 mg L−1 Se (p < 0.05, Fig. 7C).Mean Se concentration in A. drummondii treated with 32 mg L−1 Moand 1.6 mg L−1 Se was 12% lower compared to plants only treatedwith selenate, and did not differ significantly from the control,nor from the 1.6 mg L−1 Se + 48 mg L−1 Mo treatment (Fig. 7C). InA. convallarius there was no significant effect of Mo treatment onshoot Se concentration (Fig. 7D). There were not enough biorepli-cates of A. convallarius treated with 48 mg L−1 Mo or 48 mg L−1 Moand 1.6 mg L−1 Se to perform statistical analysis. In conclusion, theeffect of Mo on Se accumulation was not different between speciesclassified as Se hyperaccumulators or nonaccumulators.

Tissue concentration ratios of Mo/S, Se/S, and Mo/Se: The Mo/Sratio of the growth medium influenced the internal Mo/S ratio inall four species: With each increase in Mo/S ratio in the growthmedium, there was a statistically significant increase in the internalMo/S ratio in all four species (Fig. 8A). All four species exhib-ited lower internal Mo/S ratios than were present in the growthmedia (Fig. 8A, dashed line). Hyperaccumulator A. racemosus had a75–85% lower mean Mo/S ratio than the growth medium, followedby hyperaccumulator A. bisulcatus, nonaccumulator A. convallariusand nonaccumulator A. drummondii, which all had a 50–60% lowerMo/S ratio than the medium (Fig. 8A). The internal Mo/S ratio wasconsistently lower in the hyperaccumulators than in the nonaccu-mulators. A. drummondii and A. convallarius had 1.5- to 4 fold higherMo/S ratios than A. racemosus and A. bisulcatus. However, this trendwas not statistically significant at all points (Fig. 8A).

The Se hyperaccumulator A. racemosus consistently had Se/Sratios that were 160% of the ratios in the growth medium; in con-trast, the other three species tended to have Se/S ratios that were80% to 110% of the Se/S ratios in the growth media (Fig. 8B). It isimportant to note that all four species exceeded the Se/S ratio inthe control (0) treatment, likely because Se was present in the seed.For each external Se/S ratio, A. racemosus had a significantly higher(two-fold) mean internal Se/S ratio than A. drummondii. The Se/S

ratio in A. racemosus was also consistently about 1.5 times higherthan A. convallarius and A. bisulcatus (Fig. 8B).

The Mo/Se ratio in the shoot of all four species was less thanthat in the growth medium (Fig. 8C). The nonaccumulators tended

R.A. DeTar et al. / Journal of Plant Physiology 183 (2015) 32–40 37

F the stab ondii

tMtiM

Fb

ig. 5. The effect of Mo treatment on shoot S concentration. The error bars represent

etween treatments. Panel A: A. racemosus; Panel B: A. bisulcatus; Panel C: A. drumm

o have higher Mo/Se ratios than the hyperaccumulators. The

o/Se ratio in nonaccumulator A. drummondii was very similar to

he medium, with ratios about 13% lower than the Mo/Se ration the medium. The other nonaccumulator, A. convallarius, had

o/Se ratios that were about 30% lower than the Mo/Se ratio in

ig. 6. The effect of Se treatment on shoot S concentration. The error bars represent the staetween treatments. Panel A: A. racemosus; Panel B: A. bisulcatus; Panel C: A. drummondii

ndard error of the mean. Different lettering denotes statistical differences ( = 0.05); Panel D: A. convallarius.

the medium. The hyperaccumulator A. bisulcatus had Mo/Se ratios

that were 35–55% lower than the medium. Hyperaccumulator A.racemosus had 80% lower Mo/Se ratios than the medium at all treat-ments, and also consistently (up to 5 fold) lower Mo/Se ratios thanall other species (Fig. 8C). In all four species, the mean internal

ndard error of the mean. Different lettering denotes statistical differences ( = 0.05); Panel D: A. convallarius.

38 R.A. DeTar et al. / Journal of Plant Physiology 183 (2015) 32–40

F e barB

Milmi

Fet

ig. 7. The effect of Mo supply on shoot Se concentration. Different letters above th: A. bisulcatus; Panel C: A. drummondii; Panel D: A. convallarius.

o/Se ratio increased as the Mo/Se ratio in the growth medium

ncreased. However, the mean internal Mo/Se ratio in A. racemosuseveled off after an external Mo/Se ratio of 20, while in A. drum-

ondii and A. bisulcatus it continued to increase as external ratiosncreased.

ig. 8. The molar ratios of Mo/S (Panel A), Se/S (Panel B) and Mo/Se (Panel C) in the plants

xpected if internal ratio would be the same as external ratio. Different letters denote stahe standard error of the mean.

s denote statistically significant differences ( = 0.05). Panel A: A. racemosus; Panel

Discussion

The primary objective of this study was to investigate whetherSe hyperaccumulator species from the genus Astragalus couldaccumulate higher levels of Mo compared to closely related

as a function of those in the agar medium. The dashed lines represent the plant ratiotistically ( = 0.05) significant differences between points. The error bars represent

lant P

nssfmrsDMwMihnsbnds

artTatwdShm

afachpaahmtnSthat(tNlpnihcH

ahbmoa

R.A. DeTar et al. / Journal of P

onaccumulator species. In a field survey, Se hyperaccumulatorpecies accumulated almost twice as much Mo as nonaccumulatorpecies (NS, but p < 0.1), but the Mo levels were in the normal rangeor all species. When two hyperaccumulator and two nonaccu-

ulator species were compared under controlled conditions withespect to their ability to accumulate Mo it was found that all fourpecies accumulated high levels of Mo (more than 1000 mg kg−1

W) from their growth media. Also, under controlled conditions,o accumulation was not correlated with Se accumulation. Thus,hile Se hyperaccumulators showed a trend toward having highero levels than nonaccumulators in the field, the controlled exper-

ment showed no evidence for enhanced Mo accumulation by Seyperaccumulators. Furthermore, Se hyperaccumulators showedo evidence of Mo hypertolerance or preferential Mo uptake overimilar oxyanions. Interestingly, S and Se uptake were less affectedy competing oxyanions in Se hyperaccumulators compared toonaccumulators, perhaps indicating the existence of indepen-ent transporters for each of these elements in hyperaccumulatorpecies.

One hypothesis tested here, that Se hyperaccumulators wouldccumulate more Mo than nonaccumulators, was based on theeasoning that Se hyperaccumulators have up-regulated sulfateransporters that also mediate selenate and molybdate transport.he alternative hypothesis was that hyperaccumulators would notccumulate more Mo than nonaccumulators, with the reasoninghat hyperaccumulators may contain selenate-specific transportersith enhanced selenate specificity relative to sulfate and molyb-ate. The results support the second hypothesis. Molybdenum ande accumulation had no correlation, and thus the mechanism of Seyperaccumulation does not appear to also confer Mo hyperaccu-ulation.There were some differences between Se hyperaccumulators

nd nonaccumulators in Se, S and Mo interactions, indicating dif-erences in uptake mechanisms. Selenium, S, and Mo appeared toffect each other’s uptake less in hyperaccumulators than in nonac-umulators (Figs. 5–8). This trend suggests that hyperaccumulatorsave evolved independent high-specificity transporters to betterrevent competition between sulfate and selenate ions. This mech-nism would allow the plants to accumulate high levels of Se whilevoiding S deficiency. Evidence that the sulfate transporters of Seyperaccumulators are more S-specific than those from nonaccu-ulators is that the presence of Se in the growth medium lessened

he internal concentration of S in nonaccumulators, but had no sig-ificant effect on hyperaccumulators (Fig. 6). Also, Mo inhibited

accumulation to a greater degree in the nonaccumulators thanhe hyperaccumulators (Fig. 5). As a result, nonaccumulators hadigher Mo/S ratios (Fig. 8). Evidence for Se-specific transportersre that the hyperaccumulator A. racemosus had a higher Se/S ratiohan the medium (Fig. 8). Surprisingly in view of earlier reportse.g. White et al., 2007), A. bisulcatus did not have a higher Se/S ratiohan the medium, nor did it accumulate as much Se as A. racemosus.onaccumulators had higher Mo/Se ratios than the hyperaccumu-

ators, suggesting that hyperaccumulators have a relatively higherreference for Se over Mo than the nonaccumulators. However,one of the species showed evidence of preferentially accumulat-

ng Mo over either of the other two oxyanions. Other researchersave similarly found evidence for the hypothesis that Se hyperac-umulators have a selenate-specific transporter (White et al., 2007;arris et al., 2014).

Another objective of this study was to compare Mo toler-nce between Se hyperaccumulators and nonaccumulators. Oneypothesis tested here was that the hyperaccumulators would

e more tolerant to Mo stress on the reasoning that molecularechanisms for tolerating Se stress (such as Se binding proteins

r the up-regulation of antioxidants) may confer some toler-nce to Mo. Judged from the percentage growth inhibition, the

hysiology 183 (2015) 32–40 39

hyperaccumulators were indeed somewhat more Mo resistant thanthe nonaccumulators (Figs. 2 and 3). In addition to shared molec-ular mechanisms for Se and Mo tolerance, a possible explanationis that Se hyperaccumulators were better able to exclude toxic Molevels from their cells. Indeed, the Mo/S ratios (Fig. 8) and S accu-mulation data (Figs. 4 and 5) indicate that Se hyperaccumulatorsare better able to prevent molybdate from non-specifically enter-ing their roots via sulfate transporters, which may allow them toavoid Mo toxicity. Selenium did not appear to ameliorate Mo tox-icity in either Se hyperaccumulators or nonaccumulators. Thus, theresults do not suggest Se has a physiological role in Mo tolerance.

In future studies it will be interesting to further investigate themechanisms underlying the apparently enhanced sulfate- and sele-nate uptake specificity of Se hyperaccumulators. If transporterswith enhanced sulfate or selenate specificity can be identified, andthe associated genes cloned, these may have applications in cropproduction, phytoremediation and biofortification. Expression of asulfate-specific transporter could be used to bioengineer crops lesssusceptible to S starvation, particularly in the presence of compet-ing oxyanions. Genes encoding selenate-specific transporters couldbe used to bioengineer Se-enriched plants for biofortification orphytoremediation, particularly in high-sulfate environments.

Acknowledgements

The authors thank the American Society of Plant Biologistsfor providing a Summer Undergraduate Research Fellowship toRachael DeTar, which supported this study.

Appendix A. Supplementary data

Supplementary data associated with this article can be found,in the online version, at http://dx.doi.org/10.1016/j.jplph.2015.05.009

References

Alford ER, Pilon-Smits EAH, Marcus MA, Fakra SC, Paschke MW. No evidence for acost of tolerance: selenium hyperaccumulation by Astragalus does not inhibitroot nodule symbiosis. Am J Bot 2012;99:1930–41.

Baker AJM, McGrath SP, Reeves RD, Smith JAC. Metal hyperaccumulator plants: areview of the ecology and physiology of a biological resource for phytoremedia-tion of metal polluted soils. In: Terry N, Banuelos GS, editors. Phytoremediationof contaminated soil and water. Boca Raton, FL, USA: CRC Press; 2000. p. 85–107.

Beath OA, Gilbert CS, Eppson HF. The use of indicator plants in locating seleniferousareas in Western United States. Am J Bot 1939;26:257–69.

El Mehdawi AF, Pilon-Smits EAH. Ecological aspects of plant selenium hyperaccu-mulation. Plant Biol 2012;14:1–10.

Feng R, Wei C, Tu S. The roles of selenium in protecting plants against abiotic stresses.Environ Exp Bot 2013;87:58–68.

Fitzpatrick K, Tyerman S, Kaiser B. Molybdate transport through the plant sulfatetransporter SHST1. FEBS Lett 2008;582:1508–13.

Freeman JL, Tamaoki M, Stushnoff C, Quinn CF, Cappa JJ, Devonshire J, et al. Molecularmechanisms of selenium tolerance and hyperaccumulation in Stanleya pinnata.Plant Physiol 2010;153:1630–52.

Gupta UC. Boron, molybdenum and selenium status in different plant parts in foragelegumes and vegetable crops. J Plant Nutr 1991;14:613–21.

Harris J, Schneberg KA, Pilon-Smits EAH. Sulfur–selenium–molybdenum inter-actions distinguish selenium hyperaccumulator Stanleya pinnata from non-hyperaccumulator Brassica juncea (Brassicaceae). Planta 2014;239:479–91.

Lochlainn SO, Bowen HC, Fray RG, Hammond JP, King GJ, White PJ, et al. Tandemquadruplication of HMA4 in the zinc (Zn) and cadmium (Cd) hyperaccumulatorNoccaea caerulescens. PLoS ONE 2011;6:e17814.

Marschner H. Mineral nutrition of higher plants. San Diego, CA, USA: Academic Press;1995. p. 889.

Murashige T, Skoog F. A revised medium for rapid growth and bioassays with tobaccotissue culture. Physiol Plant 1962;15:437–97.

Pilon-Smits EAH, Quinn CF, Tapken W, Malagoli M, Schiavon M. Physiological func-tions of beneficial elements. Curr Opin Plant Biol 2009;12:267–74.

Schiavon M, Pilon-Smits EAH, Wirtz M, Hell R, Malagoli M. Selenate and molybdatealter sulfate transport and assimilation in Brassica juncea L. Czern.: implicationsfor phytoremediation. Environ Exp Bot 2012;75:41–51.

Schiavon M, Pilon M, Malagoli M, Pilon-Smits EAH. Exploring the importance ofsulfate transporters and ATP sulfurylases for selenium hyperaccumulation—a

4 lant P

S

S

S

T

T

T

0 R.A. DeTar et al. / Journal of P

comparison of Stanleya pinnata and Brassica juncea (Brassicaceae). Front PlantSci 2015;6(2):13.

hibagaki N, Rose A, McDermott JP, Fujiwara T, Hayashi H, Tadakatsu Y, et al. Selenateresistant mutants of Arabidopsis thaliana identify Sultr1;2, a sulfate transporterrequired for efficient transport of sulfate into roots. Plant J 2002;29:475–86.

hinmachi F, Buchner P, Stroud J, Parmar S, Zhao F, McGrath S, et al. Influence of sulfurdeficiency on the expression of specific sulfate transporters and the distributionof sulfur, selenium, and molybdenum in wheat. Plant Physiol 2010;153:327–36.

ors TG, Ellis DR, Salt DE. Selenium uptake, translocation, assimilation and metabolicfate in plants. Photosynth Res 2005;86:373–89.

ejada-Jiménez M, Chamizo-Ampudia A, Galván A, Fernández E, Llamas Á. Molyb-

denum metabolism in plants. Metallomics 2013;5:1191–203.

erry N, Zayed AM, de Souza MP, Tarun AS. Selenium in higher plants. Annu RevPlant Physiol Plant Mol Biol 2000;51:401–32.

omatsu H, Takano J, Takahashi H, Watanabe-Takahashi A, Shibagaki N, Fuji-wara T. An Arabidopsis thaliana high-affinity molybdate transporter required

hysiology 183 (2015) 32–40

for efficient uptake of molybdate from soil. Proc Natl Acad Sci U S A2007;104:18807–12.

Valdez Barillas JR, Quinn C, Pilon-Smits EA. Selenium accumulation inplants—phytotechnological applications and ecological implications. Int JPhytoremediation 2011;13:166–78.

Van Hoewyk D. A tale of two toxicities: malformed selenoproteins and oxida-tive stress both contribute to selenium stress in plants. Ann Bot 2013;112:965–72.

White PJ, Bowen HC, Marshall B, Broadley MR. Extraordinarily high leaf selenium tosulfur ratios define ‘Se-accumulator’ plants. Ann Bot 2007;100:111–8.

Wilson LG, Bandurski RS. Enzymatic reactions involving sulfate, sulfite, selenate and

molybdate. J Biol Chem 1958;233:975–81.

Zarcinas BAB, Cartwright BAB, Spouncer LR. Nitric acid digestion and multi elementanalysis of plant material by inductively coupled plasma spectrometry. CommunSoil Sci Plant Anal 1987;18:131–46.

Zimmer W, Mendel R. Molybdenum metabolism in plants. Plant Biol 1999;1:160–8.