Selenium accumulation by plants

Abstract

Background Selenium (Se) is an essential mineral element for animals and humans, which they acquire largely from plants. The Se concentration in edible plants is determined by the Se phytoavailability in soils. Selenium is not an essential element for plants, but excessive Se can be toxic. Thus, soil Se phytoavailability determines the ecology of plants. Most plants cannot grow on seleniferous soils. Most plants that grow on seleniferous soils accumulate <100 mg Se kg^–1 dry matter and cannot tolerate greater tissue Se concentrations. However, some plant species have evolved tolerance to Se, and commonly accumulate tissue Se concentrations >100 mg Se kg^–1 dry matter. These plants are considered to be Se accumulators. Some species can even accumulate Se concentrations of 1000–15 000 mg Se kg^–1dry matter and are called Se hyperaccumulators.

Scope This article provides an overview of Se uptake, translocation and metabolism in plants and highlights the possible genetic basis of differences in these between and within plant species. The review focuses initially on adaptations allowing plants to tolerate large Se concentrations in their tissues and the evolutionary origin of species that hyperaccumulate Se. It then describes the variation in tissue Se concentrations between and within angiosperm species and identifies genes encoding enzymes limiting the rates of incorporation of Se into organic compounds and chromosomal loci that might enable the development of crops with greater Se concentrations in their edible portions. Finally, it discusses transgenic approaches enabling plants to tolerate greater Se concentrations in the rhizosphere and in their tissues.

Conclusions The trait of Se hyperaccumulation has evolved several times in separate angiosperm clades. The ability to tolerate large tissue Se concentrations is primarily related to the ability to divert Se away from the accumulation of selenocysteine and selenomethionine, which might be incorporated into non-functional proteins, through the synthesis of less toxic Se metabilites. There is potential to breed or select crops with greater Se concentrations in their edible tissues, which might be used to increase dietary Se intakes of animals and humans.

INTRODUCTION: SELENIUM IN SOILS, PLANTS AND ANIMALS

Selenium (Se) is an essential mineral element for both human and animal nutrition (White and Brown, 2010). In humans, Se deficiency is associated with hypothyroidism, cardiovascular disease, a weakened immune system, male infertility, cognitive decline and increased incidence of various cancers (Fairweather-Tait et al., 2011; Rayman, 2012; Fordyce, 2013). The Institute of Medicine (USA) has proposed a recommended dietary allowance of 55 μg Se d^–1 for adult humans (Institute of Medicine, 2000). Unfortunately, it is estimated that the diets of as many as 1 billion people might lack sufficient Se for their well-being (Combs, 2001; Fairweather-Tait et al., 2011; Joy et al., 2014; Stoffaneller and Morse, 2015). Since much of the Se in human diets is derived, either directly or indirectly, from edible plants, the lack of Se in human diets is generally attributed to crop production on soils with low Se content or Se phytoavailability (Broadley et al., 2006; White and Broadley, 2009; Chilimba et al., 2011; Fairweather-Tait et al., 2011; Rayman, 2012; Fordyce, 2013; Joy et al., 2015).

Excessive dietary Se intakes can also be harmful to humans and animals (Fairweather-Tait et al., 2011; Rayman, 2012; Fordyce, 2013). The symptoms of mild selenosis in humans include dermatitis, cracking of nails, hair loss and garlicky breath (due to exhalation of dimethylselenide), while severe selenosis can cause acute respiratory distress, myocardial infarction and renal failure. The Institute of Medicine (USA) has suggested a tolerable upper intake of 400 μg Se d^–1 for adults (Institute of Medicine, 2000). The symptoms of selenosis in animals, which occur when they consume feed with >1–5 mg Se kg^–1 dry matter (DM), include garlicky breath, hair loss, hoof deformation (in cattle), abnormal posture, lack of vitality, slow growth, anorexia, diarrhoea, reduced reproductive performance, fetal deformities and respiratory failure (Dhillon and Dhillon, 2003; Fordyce, 2013). Plants growing on seleniferous soils have tissue Se concentrations sufficient to cause selenosis in animals (Rosenfeld and Beath, 1964; Brown and Shrift, 1982; Dhillon and Dhillon, 2003; Fordyce, 2013).

Selenium concentrations in plants are directly related to Se phytoavailability in the soil, as witnessed by the larger Se concentrations in (1) plants growing in natural soils with greater Se phytoavailability (Rosenfeld and Beath, 1964; Brown and Shrift, 1982; Ihnat, 1989); (2) plants growing in soils anthropogenically contaminated with Se (Fang and Wu, 2004; Wu, 2004); (3) produce grown on agricultural soils with greater Se phytoavailability (Ihnat, 1989; Broadley et al., 2006; Williams et al., 2009; Lee et al., 2011; Garrett et al., 2013; Joy et al., 2015); and (4) produce to which soil or foliar Se fertilizers have been applied (Broadley et al., 2006; White and Broadley, 2009; Chilimba et al., 2012; Fordyce, 2013; Alfthan et al., 2015). Indeed, the application of inorganic Se fertilizers has been particularly effective in increasing Se concentrations in edible crops, increasing the Se content of diets and improving the Se status, and health, of both animals and humans (White and Broadley, 2009; Alfthan et al., 2015).

The concentration and chemical forms of Se in natural soils are determined primarily by geology (Dhillon and Dhillon, 2003; Broadley et al., 2006; White et al., 2007b; Fordyce, 2013; Pilbeam et al., 2015). Selenium concentrations in most soils lie in the range 0·01–2·0 mg Se kg^–1, but soils associated with particular geological features can reach concentrations of 1200 mg Se kg^–1 (Dhillon and Dhillon, 2003; Fordyce, 2013; Pilbeam et al., 2015). The Se concentrations in the latter soils are toxic to many plants and they support a unique flora (Rosenfield and Beath, 1964; Brown and Shrift, 1982). Seleniferous soils are widespread in the Great Plains of the USA, Canada, South America, Australia, India, China and Russia (Dhillon and Dhillon, 2003; Fordyce, 2013; Pilbeam et al., 2015).

Selenate (SeO₄^2–) is the main water-soluble form of Se in oxic soils (pH + pe > 15), which include most cultivated soils, whereas selenite (SeO₃^2–) predominates in anaerobic soils with a neutral to acidic pH (pH + pe = 7·5–15), such as paddy soils (Mikkelsen et al., 1989; White et al., 2007b; Fordyce, 2013; Pilbeam et al., 2015). Selenide (Se^2–) species are stable only under low redox conditions (pH + pe < 7·5) and are rarely present in cultivated soils. Selenate is relatively mobile in the soil solution, but selenite is strongly absorbed by iron and aluminium oxides/hydroxides and, to a lesser extent, by clays and organic matter (Fordyce, 2013; Pilbeam et al., 2015). Thus, the addition of selenate to soils facilitates immediate Se accumulation by plants, while selenite provides a longer lasting Se fertilizer (Broadley et al., 2006; Fordyce, 2013; Pilbeam et al., 2015).

Selenium is not considered to be an essential element for flowering plants (angiosperms), although it is considered to be a beneficial element since it can stimulate growth, confer tolerance to environmental factors inducing oxidative stress, and provide resistance to pathogens and herbivory (Quinn et al., 2007; Pilon-Smits et al., 2009; White and Brown, 2010; El Mehdawi and Pilon-Smits, 2012; Feng et al., 2013). Angiosperm species have been divided into three ecological types according to their ability to accumulate Se in their tissues (Rosenfeld and Beath, 1964; Brown and Shrift, 1982; White et al., 2007a). These types are designated non-accumulator, Se-indicator and Se-accumulator species. Most angiosperm species are non-accumulator species. These species cannot tolerate tissue Se concentrations >10–100 μg Se g^–1 DM and cannot colonize seleniferous soils (Rosenfeld and Beath, 1964; White et al., 2004; Dhillon and Dhillon, 2009; Fordyce, 2013). In contrast, Se-indicator species are able to tolerate tissue Se concentrations approaching 1 mg Se g^–1 DM and colonize both non-seleniferous and seleniferous soils (Rosenfeld and Beath, 1964; Moreno Rodriguez et al., 2005). Tissue Se concentration in Se-indicator plants is directly related to Se phytoavailability in the soil and, therefore ‘indicates’ soil Se phytoavailability (cf. Baker, 1981). The distribution of Se-accumulator species is generally restricted to seleniferous soils, where their leaf Se concentrations can exceed 1 mg Se g^–1 DM (Table 1; Rosenfeld and Beath, 1964; Brown and Shrift, 1982). These species include several members of the Asteraceae, Brassicaceae and Fabaceae, which accommodate large Se concentrations in leaf trichomes and epidermal cells (Freeman et al., 2006, 2010; El Mehdawi and Pilon-Smits, 2012). Several members of the Lecythidaceae family [e.g. Brazil nut (Bertholletia excelsa Humb. and Bonpl.), paradise nut (Lecythis zabucajo Aubl.), coco de mono (Lecythis ollaria Loefl.) and monkeypot nut (Lecythis minor Jacq., syn. Lecythis elliptica Kunth.)] are also renowned for accumulating large Se concentrations in their fruit and seed (Chang et al., 1995; Hammel et al., 1996; Dernovics et al., 2007). Selenium concentrations can reach 512 μg g^–1 f. wt, which is equivalent to about 530 μg g^–1 DM, in Brazil nuts (Chang et al., 1995), 5–12 mg g^–1 DM in seeds of coco de mono (Hammel et al., 1996; Ferri et al., 2004) and 4–6 mg g^–1 in monkeypot nuts (Dernovics et al., 2007; Németh et al., 2013). It is thought that the ability to accumulate Se arose by convergent evolution of appropriate Se transport and biochemical pathways in disparate angiosperm clades during geological periods when seleniferous soils were more widespread than they are today (Brown and Shrift, 1982; White et al., 2007a; Cappa and Pilon-Smits, 2014). Species are defined as ‘Se-hyperaccumulators’ if their leaves contain >1 mg Se g^–1 DM when sampled from the natural environment (Reeves and Baker, 2000; Terry et al., 2000), although there is debate as to whether this threshold should be lowered to 100 μg Se g^–1 DM (Reeves and Baker, 2000; van der Ent et al., 2013). Thus, species that hyperaccumulate Se are an extreme sub-set of Se-accumulator species.

Table 1.

Angiosperm species credited with the appellation of Se-(hyper)accumulator which is formally defined as a species for which plants with shoot Se concentrations >1000 mg Se kg^–1 dry matter have been sampled from a natural environment

Species	Authority	Synonyms	Location	Se concentration (mg Se kg–1 DM)	Reference
Asteraceae (Asterales)
Dieteria canescens	(Pursh) Nutt.	Machaeranthera ramosa	Midwest USA	1600	Beath et al. (1939a)
Grindelia squarrosa	(Pursh) Dunal		Lower Brule Reservation, SD, USA	930	Lakin and Byers (1941)
Gutierrezia microcephala	(DC.) A.Gray		Thompson, UT, USA	1287	Beath (1943)
Oonopsis foliosa	Greene	Haplopappus fremontii var. fremontii	Lascar, CO, USA	3630	Beath et al. (1939b)
Oonopsis wardii	(A.Gray) Greene	O. condensata, Haplopappus fremontii var. wardii	Albany County, WY, USA	9120	Byers (1935)
Symphyotrichum ascendens	(Lindl.) G.L.Nesom		Soda Springs, ID, USA	4455	Pfister et al. (2013)
Symphyotrichum ericoides	(L.) G.L.Nesom	Aster ericoides	Pine Ridge, Fort Collins, CO, USA	1378	El Mehdawi et al. (2015)
Symphyotrichum lateriflorum	(L.) Á.Löve & D.Löve	Aster multiflorus	SD, USA	1800	Moxton et al. (1939)
Xylorhiza glabriuscula	Nutt.	X. villosa, Machaeranthera glabriuscula, Aster parryi	Huerfano County, CO, USA	1750	Byers et al. (1938)
Xylorhiza parryi	Greene	Machaeranthera parryi	Albany County, WY, USA	5390	Byers (1935)
Xylorhiza venusta	(M.E.Jones) A.Heller	Machaeranthera venusta	Midwest USA	3486	Rosenfeld and Beath (1964)
Fabaceae (Fabales)
Acacia cana	Maiden		NW Queensland, Australia	1121	McCray and Hurwood (1963)
Astragalus albulus	Wooton & Standl.		La Ventana, NM, USA	530	Beath et al. (1941), listed by Rosenfeld and Beath (1964)
Astragalus asclepiadoides	M.E.Jones				Listed by Brown and Shrift (1982)
Astragalus beathii	C.L.Porter		Cameron, AZ, USA	3135	Beath et al. (1940)
Astragalus beckwithii var. purpureus	M.E.Jones	A. artemisarium	Clark County, NE, USA	970	Lakin and Byers (1941)
Astragalus bisulcatus	(Hook.) A.Gray	A. bisulcatus var. bisulcatus, A. diholcus, A. scobinatulus	Pine Ridge, Fort Collins, CO, USA	13 685	Sura-de Jong et al. (2015)
Astragalus bisulcatus var. haydenianus	(A.Gray) Barneby	A. haydenianus	Cuba, NM, USA	2377	Beath et al. (1941)
Astragalus bisulcatus var. nevadensis	(M.E.Jones) Barneby				Listed by Brown and Shrift (1982)
Astragalus canadensis	L.	A. carolianus	Las Vegas, NE, USA	1110	Byers et al. (1938)
Astragalus crotalariae	A.Gray	A. limatus	Truckhaven, CA, USA	2175	Beath et al. (1941)
Astragalus eastwoodiae	M.E.Jones	A. preusii var. eastwoodiae	Utah, USA	1664	Beath (1943)
Astragalus flavus	Torr. & A.Gray	A. flavus var. flavus, A. convertiflorus var. flaviflorus, A. flaviflorus	Aztec, NM, USA	1361	Beath et al. (1941)
Astragalus flavus var. argillosus	(M.E.Jones) Barneby	A. argillosus	Greenriver, UT, USA	631	Beath et al. (1941), listed by Brown and Shrift (1982)
Astragalus flavus var. candicans	A.Gray	A. convertiflorus	Thompson, UT, USA	1322	Beath (1943)
Astragalus grayi	S.Watson		Carbon County, WY, USA	4450	Byers (1935)
Astragalus linifolius	(Osterh.) Osterh.				Listed by Brown and Shrift (1982)
Astragalus mokiacensis	A.Gray				Listed by Brown and Shrift (1982)
Astragalus moencoppensis	M.E.Jones				Listed by Brown and Shrift (1982)
Astragalus nelsonianus	Barneby	A. pectinatus var. platyphyllus			Listed by Brown and Shrift (1982)
Astragalus oocalycis	M.E.Jones				Listed by Brown and Shrift (1982)
Astragalus osterhoutii	M.E.Jones		Kremmling, CO, USA	2678	Beath et al. (1940)
Astragalus pattersonii	A.Gray		Thompson, UT, USA	8512	Beath (1943)
Astragalus pectinatus	(Hook.) G.Don		Teton County, MT, USA	5170	Williams et al. (1940)
Astragalus praelongus	E.Sheld.	A. pattersoni var. praelongus, A. recedens	Leupp, AZ, USA	4835	Beath et al. (1941)
Astragalus praelongus var. ellisiae	(Rydb.) B.L.Turner	A. ellisiae	Valmont, NM, USA.	656	Beath et al. (1941), listed by Brown and Shrift (1982)
Astragalus praelongus var. longchopus					Listed by Brown and Shrift (1982)
Astragalus preussii	A.Gray	A. preussii var. preusii, A. preussii var. latus	Thompson, UT, USA	4188	Beath (1943)
Astragalus preussii var. laxiflorus	A.Gray				Listed by Brown and Shrift (1982)
Astragalus racemosus	Pursh.	A. racemosus var. racemosus, A. racemosus var. treleasei	WY, USA	14 920	Knight and Beath (1937)
Astragalus racemosus var. longisetus	M.E.Jones				Listed by Brown and Shrift (1982)
Astragalus rafaelensis	M.E.Jones		Jensen, TX, USAA	716	Beath et al. (1941), listed by Brown and Shrift (1982)
Astragalus sabulosus	M.E.Jones		Thompson, UT, USA	2210	Beath et al. (1941)
Astragalus saurinus	Barneby				Listed by Brown and Shrift (1982)
Astragalus toanus	M.E.Jones		ID, USA	990	Lakin and Byers (1948)
Astragalus urceolatus	(Greene ex Rydb.) Greene ex Ch. Porter				Listed by Beath et al. (1940)
Astragalus woodruffi	M.E.Jones				Listed by Brown and Shrift (1982)
Neptunia amplexicaulis	Domin		Richmond, Queensland, Australia	4334	Knott and McCray (1959)
Brassicaceae (Brassicales)
Cardamine hupingshanensis	K.M.Liu, L.B.Chen, H.F.Bai & L.H.Liu		Yutangba, Enshi, China	1965	Yuan et al. (2013)
Stanleya bipinnata	Greene	S. pinnata var. gibberosa, S. pinnata var. bipinnata	Laramie, WY, USA	2490	Beath et al. (1940)
Stanleya pinnata	(Pursh) Britton	S. pinnata var. pinnata	Pine Ridge, Fort Collins, CO, USA	>4000	Galeas et al. (2007)
Stanleya pinnata var. integrifolia	(E. James) Rollins	S. integrifolia	Vernal, UT, USA	977	Beath et al. (1941)
Amaranthaceae (Caryophyllales)
Atriplex confertifolia	(Torr. & Frém.) S.Watson		Thompson, UT, USA	1734	Beath (1943)
Atriplex nuttallii	S.Watson		WY, USA	930	Beath et al. (1937)
Rubiaceae (Gentianales)
Coelospermum decipiens	Baill.	Morinda reticulata	Cape York Peninsula, Queensland, Australia	1141	Knott and McCray (1959)
Orobanchaceae (Lamiales)
Castilleja angustifolia var. dubia	A.Nelson	C. chromosa	Lysite, WY, USA	3460	Beath et al. (1941)

For each species the largest tissue Se concentration known to the author, and location of the plant that was analysed, are listed.

Species binomials, authorities and synonyms were consistent with The Plant List (http://www.theplantlist.org/) in July 2015.

SELENIUM UPTAKE, TRANSLOCATION AND METABOLISM IN PLANTS

Plant roots can take up Se as selenate (SeO₄^2–), selenite (SeO₃^2–; HSeO₃^–; H₂SeO₃) or organoselenium compounds, such as selenocysteine (SeCys) and selenomethionine (SeMet), but are unable to take up colloidal elemental Se or metal selenides (White and Broadley, 2009). Selenate uptake by root cells from the rhizosphere is catalysed by high-affinity sulphate transporters (HASTs) homologous to the arabidopsis (Arabidopsis thaliana [L.] Heynh.) AtSULTR1;1 and AtSULTR1;2 transporters (Terry et al., 2000; White et al., 2004, 2007b; Sors et al., 2005b; Shinmachi et al., 2010; Gigolashvili and Kopriva, 2014). In arabidopsis, AtSULTR1;1 contributes little to selenate uptake in S-replete plants, but its relative contribution is increased greatly when plants have insufficient S for growth (El Kassis et al., 2007; White et al., 2007b). Phosphate transporters, such as rice OsPT2, catalyse the uptake of HSeO₃^– (Zhang et al., 2014), and homologues of the rice aquaporin channel OsNIP2;1 catalyse the uptake of H₂SeO₃ (Zhao et al., 2010; Pommerrenig et al., 2015). Transporters that catalyse the uptake and movement of cysteine and methionine within the plant might transport SeCys and SeMet (Tegeder, 2012).

The arabidopsis genome contains at least 12 genes encoding sulphate transporters, which are divided into four distinct groups that encode proteins with contrasting physiological functions (Gigolashvili and Kopriva, 2014). An equivalent number of genes encoding sulphate transporters are likely to be present in the genomes of other angiosperms, including species that hyperaccumulate Se (Buchner et al., 2004, 2010; Shinmachi et al., 2010; Cabannes et al., 2011; Takahashi et al., 2012; Gigolashvili and Kopriva, 2014). The expression of genes encoding SULTR1;1 and SULTR1;2 generally increases in roots of non-accumulator and Se-indicator species when their growth is restricted by S supply (El Kassis et al., 2007; Rouached et al., 2008; Shinmachi et al., 2010; Schiavon et al., 2015), or when tissue Se concentrations rise (Takahashi et al., 2000; Van Hoewyk et al., 2005; Zhang et al., 2006a; Rouached et al., 2008; Hsu et al., 2011; Inostroza-Blancheteau et al., 2013). Roots of Se-hyperaccumulator species have constitutively high expression of these genes, which might account for their large selenate uptake capacity (Freeman et al., 2010; Cabannes et al., 2011; Schiavon et al., 2015). The increased expression of genes encoding HASTs, particularly SULTR1;1, results in greater uptake capacity for both sulphate and selenate, and accounts for the greater tissue Se concentrations in S-starved plants compared with S-replete plants (Terry et al., 2000; White et al., 2004, 2007b; Hsu et al., 2011). Sulphur-replete arabidopsis mutants lacking SULTR1;2, but not those lacking other sulphate transporters, take up less selenate and exibit greater tolerance to Se in the rhizosphere than wild-type plants (Shibagaki et al., 2002; El Kassis et al., 2007; Barberon et al., 2008). Similarly, the expression of OsPT2 increases in roots of plants lacking sufficient phosphorus and results in a greater capacity for selenite uptake (Zhang et al., 2014), and rice mutants lacking OsPT2 take up significantly less selenite than wild-type plants (Zhang et al., 2014).

To account for the characteristically greater Se/S quotient in shoots of Se-hyperaccumulator plants than in shoots of other plants growing under the same conditions (Rosenfeld and Beath, 1964; Bell et al., 1992; Feist and Parker, 2001; Galeas et al., 2007; White et al., 2007b; Freeman et al., 2010; Cappa et al., 2014; Harris et al., 2014; DeTar et al., 2015; Schiavon et al., 2015), it has been proposed that the complement of HASTs present in the plasma membranes of root cells differs in its selenate/sulphate selectivity between Se-hyperaccumulator and non-accumulator plants (White et al., 2004, 2007a). Specifically, it is hypothesized that the dominant HASTs in the plasma membrane of roots of Se-hyperaccumulator plants are selective for selenate, whereas those in other angiosperms are selective for sulphate. Interestingly, Cabannes et al. (2011) reported that the amino acid sequence of the SULTR1 transporters cloned from all the Astragalus species they studied (the Se-hyperaccumulator species A. bisulcatus [Hook.] A. Gray, A. crotalariae A. Gray and A. racemosus Pursh., and the non-Se-hyperaccumulator species A. glycyphyllos L. and A. drummondii Hook.) differed from that of other angiosperms. In particular, they identified an alanine residue in the SULTR1 cloned from the Astragalus species that corresponded to a conserved glycine residue in all other transporters of the eukaryotic sulphate permease (SulP) family in a position that might determine the selectivity of this transporter. Harris et al. (2014) observed that increasing sulphate concentration in the rhizosphere reduced leaf molybdenum (Mo) concentration in the Se-hyperaccumulator species Stanleya pinnata (Pursh) Britton but not in the Se-indicator plant Indian mustard [Brassica juncea (L.) Czern.] which, they suggested, might reflect different specificities of the complement of selenate/sulphate/molybdate transporters in Se-hyperaccumulator species and those of other angiosperms. Conversely, increasing the molybdate concentration in the rhizosphere had no effect on shoot S concentration in the Se-hyperaccumulator species Astragalus bisulcatus and A. racemosus, but reduced shoot S concentration in congeneric non-hyperaccumulator species (DeTar et al., 2015).

Selenite is rapidly converted to organoselenium compounds in the root, whereas selenate is delivered immediately to the xylem (White et al., 2004; Ximénez-Embún et al., 2004; Li et al., 2008). Sulphate transporters homologous to arabidopsis AtSULTR2;1, AtSULTR2;2 and AtSULTR3;5 have been implicated in the long-distance transport of selenate in the xylem (Takahashi et al., 2000; Gigolashvili and Kopriva, 2014). Selenium is also transported, to a very limited extent, as SeMet and selenomethionine Se-oxide (SeOMet) in the xylem (Li et al., 2008). In arabidopsis, the low-affinity sulphate transporters AtSULTR2;1 and AtSULTR2;2 are thought to catalyse selenate uptake into cells within the stele, whereas AtSULTR3;5 appears to modulate the activity of AtSULTR2;1, but does not catalyse transport itself (Kataoka et al., 2004a). The expression of AtSULTR2;1, AtSULTR2;2 and their homologues in other plants is induced both by S starvation and by increasing Se availability (Takahashi et al., 2000; Buchner et al., 2004, 2010; Van Hoewyk et al., 2005; Gigolashvili and Kopriva, 2014). Interestingly, the expression of SULTR2 genes in roots of S-replete plants of Se-hyperaccumulating Astragalus species is greater than in S-replete plants of non-Se-hyperaccumulator Astragalus species and S-starved plants of other non-Se-hyperaccumulator species (Cabannes et al., 2011). This might account for the constitutively large Se fluxes from the root to the shoot in Astragalus species that hyperaccumulate Se. In addition, the amino acid sequences of SULTR2 and SULTR3;4 from the Se-hyperaccumulator species A. racemosus and A. bisulcatus differ from those of the congeneric non-Se-hyperaccumulator species A. drummondii (Cabannes et al., 2011). Stanleya pinnata also exhibits a high constitutive expression of SpSULTR2;1 (Schiavon et al., 2015).

Selenate is assimilated into organoselenium compounds in plastids (White et al., 2007b; Pilon-Smits and LeDuc, 2009; Pilon-Smits, 2012). The sulphate transporter AtSULTR3;1 is localized in the chloroplast membrane (Cao et al., 2013) and might catalyse selenate transport into plastids. Selenate is first activated by adenosine triphosphate sulphurylase (ATPS) to form adenosine 5'-phosphoselenate (APSe), which is then reduced to selenite by adenosine 5'-phosphosulphate reductase (APR) using reduced glutathione (GSH) as the electron donor. There are four genes encoding ATPS and three genes encoding APR in the arabidopsis genome, and equivalent numbers in the genomes of other plant species (Schiavon et al., 2015). In non-accumulator and Se-indicator species, the expression of genes encoding ATPS (APS) decreases as S supply is reduced, whereas in Se-hyperaccumulator species, such as S. pinnata, they appear to be constitutively expressed (Freeman et al., 2010; Schiavon et al., 2015). Intriguingly, the expression of APS and several SULTR genes appears to be co-regulated through the expression of micro RNA (miRNA), such as miRNA395 (Paul et al., 2015). The conversion of selenate to selenite appears to be the rate-limiting step in the assimilation of Se into organic compounds (Pilon-Smits et al., 2009). Overexpressing genes encoding ATPS or APR in transgenic plants leads to the accumulation of organic Se in their leaves (Pilon-Smits et al., 1999b; Van Huysen et al., 2004; Bañuelos et al., 2005b; Sors et al., 2005a). Selenite is reduced to selenide enzymatically by sulphite reductase (Pilon-Smits, 2012) or non-enzymatically by reduced glutathione (Terry et al., 2000).

The synthesis of SeCys from serine and selenide is catalysed by cysteine synthase, an enzyme complex containing both serine acetyl transferase (SAT) and O-acetylserine (thiol) lyase (OAS-TL) subunits (Birringer et al., 2002; Sors et al., 2005b; White et al., 2007b; Ogra and Anan, 2012; Pilon-Smits, 2012). Many genes encoding enzymes in the primary S/Se assimilation pathway are upregulated when plant Se supply is increased, and often exhibit constitutively high expression in Se-hyperaccumulator species (Van Hoewyk et al., 2005, 2008b; Freeman et al., 2010). Selenomethionine is synthesized from SeCys and O-phosphohomoserine (OPHS) through the sequential actions of cystathionine γ-synthase (CγS), which produces selenocystathionine (SeCysta), cystathionine β-lyase (CBL), which produces selenohomocysteine (SeHCys), and methionine synthase (MTR). Selenocysteine is the most abundant form of Se in unselenized garlic (Allium sativum L.; Cai et al., 1995), and SeMet is often the most abundant form of Se in edible seeds and cereal grains (Smrkolj et al., 2005, 2006, 2007; Broadley et al., 2006; Kápolna et al., 2007; Rayman et al., 2008; Thavarajah et al., 2008; Zhu et al., 2009; Seppänen et al., 2010; Hart et al., 2011; Fairweather-Tait et al., 2011; Shao et al., 2014), in seeds of Lecythidaceae (Vonderheide et al., 2002; Dumont et al., 2006; Ferri et al., 2004; Németh et al., 2013; da Silva et al., 2013) and in potato (Solanum tuberosum L.) tubers (Gionfriddo et al., 2012). Selenocystathionine appears to be the most abundant form of Se in the non-Se-hyperaccumulator species Stanleya albescens M.E. Jones, and is also present at high concentrations in tissues of several Se-hyperaccumulator species (Birringer et al., 2002; Ferri et al., 2004; Freeman et al., 2006, 2010; Németh et al., 2013). It is also the main Se compound in cladodes and fruit of selenized prickly pear (Opuntia ficus-indica [L.] Mill.; Bañuelos et al., 2011). Interestingly, most of the Se in roots and shoots of the Se-hyperaccumulator species Cardamine hupingshanensis KM Liu et al. is found as selenocystine (SeCys₂; Yuan et al., 2013), which is also abundant in fruits of Lecythidaceae (Dumont et al., 2006; da Silva et al., 2013), and Se biofortification of some plants, such as Japanese pungent radish (Raphanus sativus L.), results in the formation of selenohomolanthionine from SeHCys (Ogra et al., 2007). Selenized brassicas, such as broccoli, cauliflower (Brassica oleracea L.) and black mustard (Brassica nigra [L.] K.Koch), can also contain large concentrations of seleno-glucosinolates and their Se-aglycons (Matich et al., 2012, 2015; Ouerdane et al., 2013), and selenosugars, possibly of cell wall origin, have also been reported in appreciable concentrations in selenized plants (Aureli et al., 2012).

Selenium toxicity has been attributed to the non-specific replacement of cysteine and methionine in proteins by SeCys and SeMet (Brown and Shrift, 1982; Van Hoewyk, 2013). The magnitude of this appears to be related to the tissue Se/S quotient, rather than the Se content alone (White et al., 2004; El Kassis et al., 2007). In particular, the replacement of cysteine with SeCys prevents the formation of disulphide bridges, which are essential for protein structure and function, and the replacement of cysteine with SeCys in the active site of enzymes impairs catalytic activity (Brown and Shrift, 1982; Van Hoewyk, 2013). Thus, the conversion of SeCys and SeMet to non-toxic or volatile Se metabolites can increase plant Se tolerance (Sors et al., 2005b; White et al., 2007b; Pilon-Smits and LeDuc, 2009; Van Hoewyk, 2013). Selenocysteine methyltransferase (SMT) catalyses the methylation of SeCys to Se-methylselenocysteine (SeMSeCys), and S-adenosyl-methionine:methionine methyl transferase (MMT) catalyses the methylation of SeMet to Se-methylselenomethionine (SeMSeMet; Sors et al., 2005b; White et al., 2007b; Pilon-Smits and LeDuc, 2009; Van Hoewyk, 2013). Genes encoding functional SMT are not thought to exist in plants with little Se tolerance, such as arabidopsis (Lyi et al., 2005; Van Hoewyk, 2013; Zhao et al., 2015), and there is only a single gene encoding MMT in the arabidopsis genome (Tagmount et al., 2002). The expression of BoSMT increases upon exposure of broccoli to selenate and correlates with the accumulation of SeMSeCys (Lyi et al., 2005), while differences among Astragalus and Stanleya species in their ability to accumulate Se appear to be directly correlated with SMT activity (Sors et al., 2005a, 2009; Freeman et al., 2010). The AbSMT gene appears to be expressed constitutively in Astragalus bisulcatus (Pickering et al., 2003). SeMSeCys is the most abundant form of Se in roots and shoots of Se-hyperaccumulator species, such as A. bisulcatus and Stanleya pinnata (Birringer et al., 2002; Pickering et al., 2003; Sors et al., 2005a; Freeman et al., 2006, 2010; Lindblom et al., 2013; Alford et al., 2014), in allium (chive, garlic, leek, onion) and brassica (broccoli, Brussels sprouts, cabbage, cauliflower, Chinese cabbage, kale) crops fertilized with either selenate or selenite (Birringer et al., 2002; Sugihara et al., 2004; Rayman et al., 2008; Zhu et al., 2009; Fairweather-Tait et al., 2011; Kápolna et al., 2012; Ávila et al., 2014; Thosaikham et al., 2014), and in leaves of other vegetable crops fertilized with selenite (Sugihara et al., 2004; Mazej et al., 2008). It is also present in large concentrations in tubers of selenized potato (Gionfriddo et al., 2012) and seeds of selenized legumes (Smrkolj et al., 2007; Shao et al., 2014). Selenocysteine can also be converted to alanine and elemental Se by a SeCyslyase (cpNifS) located in the chloroplast (van Hoewyk et al., 2008a; Pilon-Smits and Leduc, 2009). Although elemental Se is not commonly observed in leaves, significant amouts of elemental Se have been found in stems, nodules and roots of Se-hyperaccumulator plants grown in the presence of appropriate endosymbiotic bacteria and fungi (Valdez Barillas et al., 2012; Lindblom et al., 2013; Sura-de Jong et al., 2015). It is also noteworthy that plant genomes contain genes encoding putative Se-binding proteins (SBPs) that might contribute to Se tolerance in plant tissues (Agalou et al., 2005; Dutilleul et al., 2008). In the arabidopsis genome, there are three genes encoding SBPs. The expression of AtSBP1, and its homologues in other plants, is upregulated in response to S starvation (Hugouvieux et al., 2009; Byrne et al., 2010).

Both SeMSeCys and SeMSeMet can be conjugated with glutamate to form γ-glutamyl-SeMSeCys (γ-Glu-SeMSeCys) or γ-glutamyl-SeMSeMet (γ-Glu-SeMSeMet), or converted to dimethyldiselenide (DMSe) or dimethylselenide (DMDSe) and volatilized (Sors et al., 2005b; White et al., 2007b; Pilon-Smits and LeDuc, 2009; Ogra and Anan, 2012; Van Hoewyk, 2013). SeMSeMet can also be converted to dimethylselenonium propionate and thence to DMSe (Grant et al., 2004). Many Se-hyperaccumulator species, such as A. bisulcatus (Freeman et al., 2006; Alford et al., 2014), and allium crops (garlic, leek, onion) grown on Se-rich soils accumulate significant concentrations of γ-glutamyl-SeMeSeCys (Sugihara et al., 2004; Ogra et al., 2005; Broadley et al., 2006; White et al., 2007b; Rayman et al., 2008; Fairweather-Tait et al., 2011; Kápolna et al., 2012). In A. bisulcatus, the formation of γ-glutamyl-SeMeSeCys appears to be promoted by rhizobial symbiosis, which has been attributed to a greater supply of glutamate in nodulated plants (Alford et al., 2014). The Se compound γ-glutamyl-Secystathionine has also been reported in some Se-hyperaccumulator plants (e.g. monkeypot nuts; Dernovics et al., 2007). In general, Se is volatilized as DMSe in non-hyperaccumulator species and as DMDSe in Se-hyperaccumulator species (Pilon-Smits and LeDuc, 2009). There is considerable variation among angiosperms in their ability to volatilize Se (Terry et al., 1992; Pilon-Smits et al., 1999a; de Souza et al., 2000), and the production of these volatiles appears to be determined by the conversion of SeCys to SeMet, and transgenic plants overexpressing CγS volatilize more Se than untransformed plants (Pilon-Smits and LeDuc, 2009).

Selenium concentrations tend to be greatest in the younger leaves of plants and generally increase to a maximum during seedling growth, then decline before, or upon, flowering, when Se is translocated from leaves to reproductive organs (Rosenfeld and Beath, 1964; Turakainen et al., 2004; Galeas et al., 2007; White et al., 2007b; Cappa et al., 2014; Harris et al., 2014). This is consistent with transcriptional analyses suggesting that Se/S assimilation occurs predominantly in younger leaves and especially the first leaves a plant produces (White et al., 2007b). Selenium is readily redistributed in the phloem as both selenate and the organoselenium compounds SeMet and SeMSeCys (Carey et al., 2012). In arabidopsis, the HAST AtSULTR1;3 is thought to catalyse selenate uptake into the phloem and the expression of AtSULTR1;3 is increased in S-deficient plants (Yoshimoto et al., 2003).

Most plant cells can accumulate selenate in their vacuoles. When non-accumulator plants are fertilized with selenate, much of this is translocated to the shoot and sequestered in the vacuoles of cells within the vasculature and leaf meophyll (Ximénez-Embún et al., 2004; Mazej et al., 2008). Sulphate transporters homologous to AtSULTR4;1 and AtSULTR4;2 are present in the tonoplast of plant cells and are thought to catalyse the efflux of selenate from the vacuole (Kataoka et al., 2004b; Gigolashvili and Kopriva, 2014). The expression of AtSULTR4;1 and AtSULTR4;2 increases both upon S starvation and when plants are exposed to Se (Van Hoewyk et al., 2005; Gigolashvili and Kopriva, 2014). The expression of both SULTR4;1 and SULTR4;2 is greater in shoots of the Se-hyperaccumulator species Stanleya pinnata than in the congeneric Se-indicator species S. albescens when grown in the presence of selenite (Freeman et al., 2010). Increased expression of TaSULTR4;1 has been linked to greater grain Se concentrations in S-starved wheat than in S-replete wheat (Shinmachi et al., 2010). The expression of a number of genes encoding ABC transporters is increased in roots and leaves of perennial ryegrass (Lolium perenne L.) upon exposure to Se, and it has been suggested that some of these might be involved in the transport of Se compounds within the plant (Byrne et al., 2010), although there is presently no direct evidence to support this hypothesis.

THE EVOLUTION OF SELENIUM HYPERACCUMULATION

There can be considerable variation in shoot Se concentration among angiosperm species growing in the same environment (Rosenfeld and Beath, 1964; Brown and Shrift, 1982; Ihnat, 1989; White et al., 2004, 2007a; Bitterli et al., 2010). However, little of this variation can be attributed to systematic differences between angiosperm orders, and it is thought to reflect species-specific adaptations (White et al., 2004; Watanabe et al., 2007). In general, Se concentration in leaf tissues declines in the order Se-accumulator > Se-indicator > non-accumulator species. Differences in Se accumulation between species are most pronounced within genera containing Se-accumulator or Se-indicator plants, such as Astragalus and Stanleya (White et al., 2004).

When grown in the same environment, Se concentrations in leaves of Se-hyperaccumulating species are significantly greater than those of other angiosperms (Rosenfeld and Beath, 1964; Brown and Shrift, 1982; White et al., 2007b), suggesting that these species might have distinct physiological adaptations enabling this trait. Since Se-hyperaccumulating species occur in several unrelated families (Table 1; Fig. 1A), it is thought that the traits of Se tolerance and accumulation arose by convergent evolution of appropriate biochemical pathways in several angiosperm clades (Brown and Shrift, 1982; White et al., 2004; Cappa and Pilon-Smits, 2014). The ability to accumulate Se appears to have evolved independently in the core eudicot families Amaranthaceae (Caryophyllales), Asteraceae (Asterales), Brassicaceae (Brassicales), Fabaceae (Fabales), Orobanchaceae (Lamiales) and Rubiaceae (Gentianales). The Fabaceae contains the greatest number of species known to hyperaccumulate Se. The ability to hyperaccumulate Se appears to have evolved several times within the Asteraceae, Brassicaceae and Fabaceae (Table 1). Indeed, it even appears to have evolved several times among North American Astragalus (Fabaceae): in the Homaloboid Phalanx within the seleniferous Homalobi, for which it can be used as a taxonomic character (Barneby, 1964), and the Preussiani (Fig. 1B), and also within the Piptoloboid and Ceridothrix Phalanxes. The evolution of Se hyperaccumulation in Stanleya (Brassicaceae) has also been studied in some detail (Fig. 1C; Cappa et al., 2014, 2015). Cappa et al. (2015) have observed that Se hyperaccumulation is restricted to the S. bipinnata/pinnata clade and is likely to have evolved once and then been lost in various ecotypes, such as those described as S. pinnata var. inyoensis and S. pinnata var. texana. Cappa et al. (2014) reported that S. pinnata ecotypes differed markedly in their ability to hyperaccumulate Se and observed that the trait was restricted to populations on the east side of the continental divide. They suggested that Se hyperaccumulation could have evolved in eastern USA and either (a) the Rocky Mountains formed a geographical barrier for gene flow to the west; (b) a reproductive barrier prevented gene flow because of ploidy differences among populations in the east and west; or (c) there is a greater cost to Se hyperaccumulation in the west. Evidence suggests that the ability to tolerate large tissue Se concentrations evolved earlier than the trait of Se hyperaccumulation in Stanleya and might have been a necessary predisposition enabling Se hyperaccumulation (Cappa et al., 2015).

Fig. 1.

(A) Distribution of proposed Se-hyperaccumulating species among angiosperm orders. Phylogenetic relationships between the angiosperm orders are reproduced from the Angiosperm Phylogeny Group (2009). The number of Se-hyperaccumulating genera and Se-hyperaccumulating species in each order are given in parentheses based on data presented in Table 1. (B) Distribution of proposed Se-hyperaccumulating taxa among sections of the Homaloboid astragali of North America. Taxonomic relationships are derived from Barneby (1964). The number of Se-hyperaccumulating taxa and Se-hyperaccumulating species in each section are given in parentheses based on data presented in Table 1. (C) Distribution of proposed Se-hyperaccumulating taxa among Brassicaceae indicating a single origin of Se hyperaccumulation (filled circles) in the Stanleya pinnata/bipinnata clade (Cappa et al., 2015).

Several hypotheses have been proposed for the evolution of Se hyperaccumulation in plants. First, there is a clear evolutionary advantage in being one of a few stress-tolerant plant species able to colonize seleniferous soils (Brown and Shrift, 1982). This character might have occurred through the evolution of mechanisms for Se exclusion by roots, tissue Se tolerance or Se volatilization. However, although Se exclusion by roots allows non-accumulator plants to survive greater rhizosphere Se concentrations, it does not confer the ability to colonize seleniferous soils (Rosenfeld and Beath, 1964; Brown and Shrift, 1982). In contrast, the ability to accumulate Se in non-toxic forms, and to remove Se by volatilization, are characteristics shared by many Se-indicator and Se-accumulator plants that colonize seleniferous soils, and there is considerable variation in the expression of these between and among plant species (Terry et al., 2000; Bañuelos et al., 2005a; White et al. 2007b; Pilon-Smits and LeDuc, 2009). Since the colonization of seleniferous soils by angiosperm species appears to require the ability to tolerate Se in their tissues, it is unsurprising that biochemical pathways that restrict the incorporation of selenoamino acids into proteins through the production of non-toxic Se metabolites appear to have evolved before those of Se hyperaccumation (Cappa et al., 2015). The accumulation of Se in plant tissues protects them against pathogens and herbivores (Quinn et al., 2007; Pilon-Smits et al., 2009; El Mehdawi and Pilon-Smits, 2012), and it has been proposed that this might be the primary ecological driver for the evolution of Se hyperaccumulation (Quinn et al., 2007; El Mehdawi and Pilon-Smits, 2012). It is also possible that the deposition of leaf litter with large Se concentrations around Se-hyperaccumulator plants could prevent competition by species with less tolerance of Se in the rhizosphere (El Mehdawi et al., 2011).

In addition to their exceptional ability to accumulate Se, Se-hyperaccumulator species have several other characteristics that appear to distinguish them from Se-indicator and non-accumulator species. When compared with other angiosperms, Se-hyperaccumulator species (1) constitutively express genes encoding sulphate transporters (Cabannes et al., 2011; Schiavon et al. 2015); (2) have significantly greater leaf Se/S quotients (Bell et al., 1992; Feist and Parker, 2001; Galeas et al., 2007; White et al., 2007a; Freeman et al., 2010; Cappa et al., 2014; Harris et al., 2014; DeTar et al., 2015); (3) exhibit reduced Mo accumulation with increasing rhizosphere sulphate or selenate concentrations (Harris et al., 2014); (4) restrict the incorporation of selenoamino acids into proteins through greater expression of appropriate genes (see ‘Selenium Uptake, Translocation and Metabolism in Plants’); and (5) accumulate Se in leaf trichomes and epidermal cells (Freeman et al., 2006, 2010; El Mehdawi and Pilon-Smits, 2012). These traits have been proposed as additional diagnostic characteristics for species that hyperaccumulate Se (White et al., 2007a; El Mehdawi and Pilon-Smits, 2012; Harris et al., 2014).

VARIATION IN SELENIUM ACCUMULATION WITHIN PLANT SPECIES

In addition to the considerable variation between species in their ability to accumulate Se in their tissues, there is often significant variation among genotypes of a particular species in this character. It has been observed, for example, that ecotypes of the Se-hyperaccumulator species Stanleya pinnata (Feist and Parker, 2001; Cappa et al., 2014) and Symphyotrichum ericoides (L.) G.L.Nesom (El Mehdawi et al., 2015) differ significantly in their leaf Se concentrations when grown in the same environment. Ecotypes collected from seleniferous soils generally have greater leaf Se concentrations and leaf Se/S quotients than ecotypes collected from soils with less Se in common garden experiments (Feist and Parker, 2001; Cappa et al., 2014; El Mehdawi et al., 2015). Significant genetic variation in shoot Se concentration has also been reported among tall fescue (Festuca arundinacea Schreb.) genotypes (McQuinn et al., 1991; Wu, 1998), and it would appear that there is a negative correlation between shoot Se concentration and shoot yield in this plant species (Wu, 1998).

Arabidopsis accessions differ both in their tolerance of Se in the rhizosphere and in their shoot Se concentration when grown in the same environment (Zhang et al., 2006a, b, 2007; Tamaoki et al., 2008; Chao et al., 2014). However, there appears to be no correlation between tolerance of Se in the rhizosphere and relative shoot Se concentration among arabidopsis accessions (Zhang et al., 2007). Analyses of crosses between arabidopsis accessions suggest that a single major gene controls selenite tolerance in this species, but that at least three chromosomal quantitative trait loci (QTLs) control selenate tolerance (Zhang et al., 2006a, b, 2007). Selenite tolerance in arabidopsis has been correlated with concentrations of non-protein thiols (e.g. cysteine, glutathione, phytochelatins) in roots, and tolerance to both selenate and selenite has been correlated with shoot SeCys and SeCys₂ concentrations (Zhang et al., 2006a). In addition, it has been noted that the shoot S concentration of a selenite-tolerant accession of arabidopsis (Col-0) was greater than that of a selenite-sensitive accession (Ws-2) when they were exposed to selenite (Tamaoki et al., 2008). This has been attributed to greater expression of genes encoding SULTR2;2, SURTR3;1 and SULTR3;5, together with several genes involved in S assimilation, in the selenite-tolerant accession than in the selenite-sensitive accession, which is consistent with the hypothesis that upregulation of the S transport and assimilation pathways is one mechanism to increase selenite tolerance (Tamaoki et al., 2008). Chao et al. (2014) failed to identify any QTLs affecting leaf Se concentration in arabidopsis when they applied genome-wide association mapping techniques to a diverse set of 349 accessions, despite most of the variation in leaf Se concentration being accounted for by genotype (heritability 0·68) in their experiments. However, AtAPR2 was inferred to influence leaf Se accumulation in a population of arabidopsis derived from an accession with a large leaf Se concentration (Hodonín) and the Col-0 accession using extreme array mapping (Chao et al., 2014). The influence of AtAPR2 on leaf Se accumulation was further confirmed by phenotyping mutants lacking AtAPR2 and accessions with contrasting AtAPR2 activities (Chao et al., 2014). A single amino acid substitution apparently led to the loss of function of AtAPR2 and Se accumulation in leaves in the Hodonín accession (Chao et al., 2014).

There also appears to be sufficient genetic variation to breed for crops that can accumulate more Se in their edible tissues (White and Broadley, 2009). Genetic variation in grain Se concentration has been reported for a number of cereals (Table 1). Although several studies have suggested little genetic variation in grain Se concentration among bread wheat (Triticum aestivum L.) genotypes (Table 2; Lyons et al., 2005a; Zhao et al., 2009; Lee et al., 2011; Nelson et al., 2011), other studies have reported significant genetic variation in this trait (Garvin et al., 2006; Murphy et al., 2008; Rodríguez et al., 2011; Pu et al., 2014). It is evident that the expression of this trait in bread wheat is strongly dependent upon weather conditions, crop husbandry and Se fertilization (Lyons et al., 2005a; Garvin et al., 2006; Zhao et al., 2009; Lee et al., 2011; Nelson et al., 2011). In addition, there appears to be a negative relationship between grain Se concentration and grain yield among genotypes of bread wheat (Zhao et al., 2007; Fan et al., 2008; Murphy et al., 2008), although this is not always observed (Lyons et al., 2005a; Zhao et al., 2009). No genetic variation has been observed to date in the distribution of Se within wheat grain (Lyons et al., 2005b). Significant genetic variation in grain Se concentration has been observed in other cereals including durum wheat (Triticum turgidum L.; Rodríguez et al., 2011), barley (Hordeum vulgare L.; Ilbas et al., 2012; Mangan et al., 2015), wild barley (Hordeum spontaneum K.Koch; Yan et al., 2011), oat (Avena sativa L.; Eurola et al., 2004) and rice (Orzya sativa L.; Zhang et al., 2006c; Norton et al., 2010, 2012).

Table 2.

Examples of the variation in selenium concentrations in edible tissues among genotypes of common crops grown under the same conditions

Crop	Plant species	Tissue	Trial	Details	Selenium (mg Se kg–1 DM)	Genotypes	Reference
Wheat	Triticum aestivum L.	Grain	Field trial, Sonora, Mexico		0·045 (0·010–0·130)	n = 100	Lyons et al. (2005a)
Wheat	Triticum aestivum L.	Grain	Field trial, Sonora, Mexico		0·076 (37-120)	n = 40	Lyons et al. (2005a)
Wheat	Triticum aestivum L.	Grain	Field trial, Manhattan, KS, USA		0·045 (0·039–0·055) ns	n = 14	Garvin et al. (2006)
Wheat	Triticum aestivum L.	Grain	Field trial, Hutchinson, KS, USA		0·36 (0·28–0·48)***	n = 14	Garvin et al. (2006)
Wheat	Triticum aestivum L.	Grain	Field trial, Pullman, WA, USA		0·015 (0·002–0·030)***	n = 63	Murphy et al. (2008)
Wheat	Triticum aestivum L.	Grain	Field trial, Martonvásár, Hungary		0·099 (0·033–0·238)	n = 150	Zhao et al. (2009)
Wheat	Triticum aestivum L.	Grain	Six field trials, Europe		0·070 (0·032–0·091) ns	n = 26	Zhao et al. (2009)
Spring wheat	Triticum aestivum L.	Grain	Two field trials, SD, USA		0·832 (0·730–0·940) ns	n = 10	Lee et al. (2011)
Winter wheat	Triticum aestivum L.	Grain	Two field trials, SD, USA		0·418 (0·370–0·460) ns	n = 8	Lee et al. (2011)
Wheat	Triticum aestivum L.	Grain	Three field trials, Edmonton, Canada	Conventional agronomy	0·023 (0·020–0·028) ns	n = 5	Nelson et al. (2011)
Wheat	Triticum aestivum L.	Grain	Three field trials, Edmonton, Canada	Organic cultivation	0·131 (0·115–0·151) ns	n = 5	Nelson et al. (2011)
Wheat	Triticum aestivum L.	Grain	Field trial, Canary Islands		0·072 (0·032–0·130)*	n = 11	Rodríguez et al. (2011)
Wheat	Triticum aestivum L.	Grain	Hydroponics	10 μM Na2SeO4	0·232 (0·190–0·300)	n = 20	Souza et al. (2014)
Durum wheat	Triticum turgidum L.	Grain	Field Trial, Martonvásár, Hungary		0·081 (0·039–0·146)	n = 10	Zhao et al. (2009)
Durum wheat	Triticum turgidum L.	Grain	Field trial, Canary Islands		0·079 (0·031–0·154)*	n = 8	Rodríguez et al. (2011)
Einkorn wheat	Triticum monococcum L.	Grain	Field Trial, Martonvásár, Hungary		0·279 (0·179–0·440)	n = 5	Zhao et al. (2009)
Emmer wheat	Triticum dicoccon (Schrank) Schübl.	Grain	Field Trial, S. Italy		0·028 (0·018–0·035)	n = 10	Piergiovanni et al. (1997)
Emmer wheat	Triticum dicoccon (Schrank) Schübl.	Grain	Field Trial, Martonvásár, Hungary		0·229 (0·151–0·326)	n = 5	Zhao et al. (2009)
Spelt wheat	Triticum spelta L.	Grain	Field Trial, S. Italy		0·039 (0·019–0·058)	n = 10	Piergiovanni et al. (1997)
Spelt wheat	Triticum spelta L.	Grain	Field trial, Martonvásár, Hungary		0·209 (0·125–0·244)	n = 5	Zhao et al. (2009)
Barley	Hordeum vulgare L.	Grain	Field trials on loess soil, China		0·045 (0·000–0·144)	n = 10	Yan et al. (2011)
Barley	Hordeum vulgare L.	Grain	Field trial, Moldova	No Se fertilizer	0·111 (0·084–0·142) ns	n = 3	Ilbas et al. (2012)
Barley	Hordeum vulgare L.	Grain	Field trial, Moldova	12·5 g ha–1 Na2SeO3	0·218 (0·154–0·252)*	n = 3	Ilbas et al. (2012)
Wild barley	Hordeum spontaneum K.Koch	Grain	Field trial on loess soil, China		0·045 (0·000–0·387)	n = 92	Yan et al. (2011)
Oat	Avena sativa L.	Grain	Field trials, 8–10 sites × 3 years, Finland	Se fertilizer	0·110 (0·110–0·140)***	n = 6	Eurola et al. (2004)
Oat	Avena sativa L.	Grain	Six field trials, ND, USA		0·380 (0·310 - 0·412) ns	n = 18	Doehlert et al. (2013)
Rice	Oryza sativa L.	Brown grain	Field trials, Wuxi City, China		0·057 (0·029–0·103)*	n = 151	Zhang et al. (2006c)
Common bean	Phaseolus vulgaris L.	Seed	Twelve field trials, Saskachewan, Canada		0·430 (0·381–0·500) ns	n = 9	Ray et al. (2014)
Common bean	Phaseolus vulgaris L.	Seed	Glasshouse soil	No Se fertilizer	0·052 (0·046–0·081) ns	n = 4	Smrkolj et al. (2007)
Common bean	Phaseolus vulgaris L.	Seed	Glasshouse soil	Foliar Se fertilizer	2·081 (1·892–2·379) ns	n = 4	Smrkolj et al. (2007)
Field pea	Pisum sativum L.	Seed	Twelve field trials, Saskachewan, Canada		0·457 (0·373–0·519) ns	n = 17	Thavarajah et al. (2010)
Chickpea	Cicer arietinum L.	Seed	Field trials, ND, USA		0·333 (0·153–0·563)*	n = 10	Thavarajah and Thavarajah (2012)
Chickpea	Cicer arietinum L.	Seed	Ten field trials, Saskachewan, Canada		0·732 (0·629–0·846)**	n = 8	Ray et al. (2014)
Lentil	Lens culinaris Medik.	Seed	Sixteen field trials, Saskachewan, Canada		0·523 (0·425–0·672)	n = 19	Thavarajah et al. (2008)
Lentil	Lens culinaris Medik.	Seed	Field trial, Tel Hadya, Syria (2008)		0·020 (0·016–0·024)*	n = 10	Thavarajah et al. (2011)
Lentil	Lens culinaris Medik.	Seed	Field trial, Tel Hadya, Syria (2009)		0·025 (0·016–0·044)*	n = 17	Thavarajah et al. (2011)
Lentil	Lens culinaris Medik.	Seed	Field trial, Surkhet, Nepal		0·439 (0·277–0·577) ns	n = 17	Thavarajah et al. (2011)
Lentil	Lens culinaris Medik.	Seed	Field trial, Nawalpur, Nepal		0·064 (0·036–0·117)*	n = 17	Thavarajah et al. (2011)
Lentil	Lens culinaris Medik.	Seed	Field trial, Annacuur, Morocco		0·015 (0·008–0·023)*	n = 20	Thavarajah et al. (2011)
Lentil	Lens culinaris Medik.	Seed	Field trial, Sidi ElAidi, Morocco		0·042 (0·025–0·063)*	n = 16	Thavarajah et al. (2011)
Lentil	Lens culinaris Medik.	Seed	Field trial, Pullman, WA, USA		0. 026 (0·09–0·028) ns	n = 72	Thavarajah et al. (2011)
Lentil	Lens culinaris Medik.	Seed	Field trial, Horsham, Australia		0·255 (0·172–0·287) ns	n = 17	Thavarajah et al. (2011)
Lentil	Lens culinaris Medik.	Seed	Field trial, Melton, Australia		0·041 (0·027–0·063)*	n = 17	Thavarajah et al. (2011)
Lentil	Lens culinaris Medik.	Seed	Field trial, Diyarbakir, Turkey (2008)		0·049 (0·035–0·065)*	n = 11	Thavarajah et al. (2011)
Lentil	Lens culinaris Medik.	Seed	Field trial, Sanliufa, Turkey (2009)		0·045 (0·030–0·067)*	n = 10	Thavarajah et al. (2011)
Lentil	Lens culinaris Medik.	Seed	Four field trials, Saskachewan, Canada		1·180 (0·970–1·637)**	n = 18	Ray et al. (2014)
Mung bean	Vigna radiata (L.) R.Wilczek	Seed	Field, ICRISAT, India (2012)		0·502 (0·210–0·910)**	n = 20	Nair et al. (2015)
Soybean	Glycine max (L.) Merr.	Seed	Field	0·111 mg Se kg–1 soil; Se fertilizer	6·923 (6·355–7·491)**	n = 2	Yang et al. (2003)
Indianmustard	Brassica juncea (L.) Czern.	Leaves	Hydroponics	2 mg L–1 Na2SeO4	707 (501–1092)*	n = 9	Bañuelos et al. (1997)
Indian mustard	Brassica juncea (L.) Czern.	Leaves	Field trial	Se-laden soil	543 (407–769)*	n = 9	Bañuelos et al. (1997)
Rapid-cycling Brassica	Brassica oleracea L.	Leaves	Hydroponics	E1: 2 mg L–1 Na2SeO4	604 (120–988)	n = 219	Kopsell and Randle (2001)
Rapid-cycling Brassica	Brassica oleracea L.	Leaves	Hydroponics	E2: 2 mg L–1 Na2SeO4	341 (152–531)**	n = 190	Kopsell and Randle (2001)
Broccoli	Brassica oleracea L. (Italica Group)	Floret	Greenhouse soil	Non-saline irrigation	0·5 (0·4–0·7) ns	n = 4	Bañuelos et al. (2003)
Broccoli	Brassica oleracea L. (Italica Group)	Leaves	Greenhouse soil	Non-saline irrigation, no Se	0·3 (0·2–0·5) ns	n = 4	Bañuelos et al. (2003)
Broccoli	Brassica oleracea L. (Italica Group)	Floret	Greenhouse soil	Non-saline irrigation, 250 μg Se L–1	48 (43-51)*	n = 4	Bañuelos et al. (2003)
Broccoli	Brassica oleracea L. (Italica Group)	Leaves	Greenhouse soil	Non-saline irrigation, 250 μg Se L–1	29 (26-31)*	n = 4	Bañuelos et al. (2003)
Broccoli (hybrid)	Brassica oleracea L. (Italica Group)	Floret	Three field trials, SC, USA		0·068 (0·053–0·085)**	n = 20	Farnham et al. (2007)
Broccoli (inbreds)	Brassica oleracea L. (Italica Group)	Floret	Three field trials, SC, USA		0·063 (0·049–0·080) ns	n = 15	Farnham et al. (2007)
Broccoli	Brassica oleracea L. (Italica Group)	Leaves	Hydroponics	20 μM Na2SeO4	1100 (801–1798)	n = 38	Ramos et al. (2011b)
Onion	Allium cepa L.	Bulb	Hydroponics	2 mg L–1 Na2SeO4	0·085 (0·060–0·113)***	n = 16	Kopsell and Randle (1997)
Lettuce	Lactuca sativa L.	Leaves	Hydroponics	15 μM Na2SeO4	5·28 (2·76–7·20)	n = 30	Ramos et al. (2011a)
Lettuce	Lactuca sativa L.	Leaves	Hydroponics	15 μM Na2SeO3	2·87 (1·67–5·33)	n = 30	Ramos et al. (2011a)
Chicory	Cichorium intybus L.	Leaves	Aeroponics	7 mg Se L–1 (as Na2SeO4)	391 (167–480)*	n = 4	Mazej et al. (2008)
Tomato	Solanum lycopersicum L.	Fruit	Five glasshouses, Almería, Spain		0·188 (0·015–0·363)*	n = 8	Guil-Guerrero and Rebolloso-Fuentes (2009)
Pepper	Capsicum annuum L.	Fruit	Five glasshouses, Almería, Spain		0·110 (0·047–0·200)*	n = 10	Guil-Guerrero et al. (2006)
Potato	Solanum tuberosum L.	Tubers	Field, CO, USA		1·551 (0·014–5·816)*	n = 8	Perla et al. (2012)

Data show the mean and, in parentheses, the minimum and maximum Se concentrations for n genotypes

Significant differences are indicated as *P < 0·05, **P < 0·01 and ***P < 0·001.

Chromosomal loci (QTLs) influencing grain Se concentration have been identified in wheat (Yang et al., 2013; Pu et al., 2014). Pu et al. (2014) identified four QTLs affecting grain Se concentration on chromosomes 3D at 218 cM, 4A at 91 cM, 5B at 169 cM and 7D at 215 cM in a cross between a synthetic wheat (SHW-L1) and Chuanmei 32, and a single QTL on chromosome 4D at 100 cM in a cross between Chuanmai 42 and Chuannong 16. Yang et al. (2013) identified four QTLs affecting grain Se concentration in a genetic mapping population derived from a cross between wild emmer wheat (Triticum dicoccoides [Körn. ex Asch. and Graebn.] Schweinf.) and a tetraploid durum wheat. These occurred on chromosomes 5B, 6A and 6B. Chromosomal loci affecting Se concentrations of leaves and grain of paddy rice have also been reported in a genetc mapping population derived from an indica (Bala) and a japonica (Azucena) variety (Norton et al., 2010, 2012). Several QTLs were found to affect grain Se concentration in this population, although the magnitude of their effects differed between environments. Chromosomal loci affecting grain Se concentration in rice were located on chromosome 1 (27·4 and 246·4 cM), chromosome 3 (80·6 cM), chromosome 6 (12·0, 20·3 and 103·5 cM), chromosome 7 (149·8 cM), chromosome 8 (16·9 cM), chromosome 9 (61·5 cM), chromosome 10 (66·1 cM) and chromosome 11 (105·9 cM). Two of these QTLs (on C3 and C7) also influenced leaf Se concentration, suggesting that Se accumulation in leaves, and its subsequent remobilization to developing grain, could be important in determining grain Se concentrations (Norton et al., 2010). None of the causal genes underpinning QTLs affecting Se accumulation in gain of wheat or rice is currently known.

Genetic variation in seed Se concentration has been reported among genotypes of several legume species (Table 2), although data from field trials indicate that genetic effects on seed Se concentration are generally small when compared with environmental effects (Thavarajah et al., 2010; Garrett et al., 2013; Ray et al., 2014). When grown at several sites in Saskatchewan (Canada), genetic effects on seed Se concentration of common bean (Phaseolus vulgaris L.) or field pea (Pisum sativum L.) were not significant (P > 0·1), although genotype × environment interactions did affect seed Se concentrations in common bean (Thavarajah et al., 2010; Garrett et al., 2013; Ray et al., 2014). This is consistent with studies performed in the glasshouse on common bean (Smrkolj et al., 2007). Similarly, no single nucleotide polymorphism (SNP) markers could be associated with variation in seed Se concentration among 94 pea genotypes grown in the field in Saskatchewan (Diapari et al., 2015). In contrast, genotypic variation was found to affect seed Se concentrations in both chickpea (Cicer arietinum L.) and lentil (Lens culinaris Medik.) grown in Saskatchewan (Thavarajah et al., 2008; Thavarajah and Thavarajah, 2012; Ray et al., 2014; Rahman et al., 2015). Significant genetic variation in seed Se concentration of lentil has also been observed in field trials conducted in other countries including Morocco, Turkey, Syria, Nepal, Australia and the USA (Thavarajah et al., 2011). Significant genetic variation in seed Se concentration has also been observed among genotypes of mung bean (Vigna radiata [L.] R.Wilczek; Nair et al., 2015) and soybean (Glycine max [L.] Merr.; Yang et al., 2003), and two QTLs have been identified, one on chromosome 8 and another on chromosome 18, that explain about 21 % of the variation in seed Se concentration in a recombinant inbred population of soybean derived from a cross between Williams82 and DSR-173 (Ramamurthy et al., 2014). Interestingly, the QTL on chromosome 8 includes GmSULTR2;1 (Ramamurthy et al., 2014).

Despite large environmental effects, significant genetic effects on Se concentration have been observed for onion bulbs (Allium cepa L.; Kopsell and Randle, 1997), leaves of rapid-cycling Brassica oleracea (Kopsell and Randle, 2001), broccoli florets (B. oleracea L. Italica Group; Bañuelos et al., 2003; Farnham et al., 2007; Ramos et al., 2011b), sprouts of cauliflower (B. oleracea L. Botrytis Group), kale (B. oleracea L. Acephala Group) and Chinese cabbage (Brassica rapa L.; Ávila et al., 2014), shoots of Indian mustard (Bañuelos et al., 1997), leaves of chicory (Cichorium intybus L.; Mazej et al., 2007) and leaves of lettuce (Lactuca sativa L.; Ramos et al., 2011a). In lettuce, the ability of genotypes to accumulate Se supplied as selenate was positively correlated with the expression of LsSULTR1;1, LsAPS1 and LsAPR1 (Ramos et al., 2011a). Significant genetic variation has also been observed in tomato (Solanum lycopersicum L.) fruit (Guil-Guerrero and Rebolloso-Fuentes, 2009), pepper (Capsicum annuum L.) fruit (Guil-Guerrero et al., 2006) and potato tubers (Perla et al., 2012).

TRANSGENIC APPROACHES TO INCREASE SELENIUM ACCUMULATION

Transgenic plants have been generated with greater Se tolerance, Se accumulation or Se volatilization than their non-transgenic counterparts (Table 3; Terry et al., 2000; Pilon-Smits and LeDuc, 2009; Pilon-Smits, 2012). These have been created for a variety of purposes. They have been used to provide fundamental knowledge of the transport proteins involved in the uptake and movement of Se in plants and to gain insight into the biochemical pathways and, in particular, the rate-limiting steps and control of Se metabolism in plants. The manipulation of Se transport and biochemistry can benefit crop production either directly, by allowing the development of crops with greater Se tolerance that can grow on soils with high soil Se concentrations, or indirectly, through the remediation of agricultural land with high soil Se concentrations using plants that can remove more Se from soils either by accumulating more Se in harvested tissues or by volatilizing more Se to the atmosphere. It can also benefit crop quality through Se biofortification of produce, not only by enabling greater Se concentrations to be accumulated in edible produce but also by synthesizing the most beneficial Se compounds for human and animal health.

Table 3.

Phenotypes of transgenic plants overexpressing genes involved in selenium uptake and metabolism

Overexpressed gene	Enzyme	Plant	Tolerance of Se in the rhizosphere			Leaf Se concentration			Reference
			Field soil	Selenite	Selenate	Total	Organic	Volatilization
ShST1	Sulphate transporter	Indian mustard			No effect	Increase, selenate supply			de Souza et al. (2000)
AtATPS1	ATP-sulphurylase	Arabidopsis			Decrease	Decrease	Increase		Sors et al. (2005a)
AtATPS1	ATP-sulphurylase	Indian mustard	No effect		Increase	Increase	increase	No effect	Pilon-Smits et al. (1999b);
									Van Huysen et al. (2004);
									Bañuelos et al. (2005b)
AtATPS1	ATP-sulphurylase	Tobacco			No effect	No effect, selenate supply			McKenzie et al. (2009)
PaAPR	Adenosine 5′-phosphosulphate reductase	Arabidopsis			Increase	Decrease	Increase		Sors et al. (2005a)
AtATPS1 + PaAPR	ATP-sulphurylase + adenosine 5′-phosphosulphate reductase	Arabidopsis				Decrease	Increase		Sors et al. (2005a)
Escherichia coli gshII	Glutathione synthetase	Indian mustard	Increase			Increase			Bañuelos et al. (2005b)
Escherichia coli gshI	γ-Glutamyl-cysteine synthetase	Indian mustard	No effect			Increase			Bañuelos et al. (2005b)
TgSAT (m)	Serine acetyltransferase	Arabidopsis			No effect	Decrease	Increase		Sors et al. (2005a)
SoCS (cp)	Cysteine synthase	Indian mustard			No effect	No effect			de Souza et al. (2000)
AtCGS1 (cp)	Cystathionine-γ-synthase	Indian mustard	No effect	Increase	No effect	Decrease, selenite supply		Increase	Van Huysen et al. (2003, 2004)
						No effect, selenate supply
AbSMT1	SeCys methyltransferase	Arabidopsis		Increase	Increase	Increase	Increase in SeMSeCys and GluMSeCys	Increase	LeDuc et al. (2004);
									Ellis et al. (2004)
AbSMT	SeCys methyltransferase	Indian mustard	No effect	Increase	Increase	Increase	Increase in SeMSeCys	Increase in laboratory, no effect in field	LeDuc et al. (2004);
									Bañuelos et al. (2007);
									Kubachka et al. (2007)
AbSMTA	SeCys methyltransferase	Tobacco			No effect	Increase	Increase in SeMSeCys and GluMSeCys	Increase	Matich et al. (2009);
									McKenzie et al. (2009)
AtATPS1 + AbSMT	ATP-sulphurylase + SeCys methyltransferase	Indian mustard		Increase	Increase	Increase	Increase in SeMSeCys and GluMSeCys		LeDuc et al. (2006)
									Kubachka et al. (2007)
BoATPS1 + AbSMTA	ATP-sulphurylase + SeCys methyltransferase	Tobacco			No effect	Increase	Increase in SeMSeCys and GluMSeCys	Increase	Matich et al. (2009);
									McKenzie et al. (2009)
AtCpNifS	Cysteine lyase	Arabidopsis		No effect	Increase	Increase	Decrease in protein		Van Hoewyk et al. (2005)
Mus musculus SL (cp)	Selenocysteine lyase	Indian mustard	Increase	Decrease	Decrease	Increase	Decrease in protein	No effect	Garifullina et al. (2003);
									Bañuelos et al. (2007)
Mus musculus SL (cyt)	Selenocysteine lyase	Arabidopsis		Increase	Increase	Increase	Decrease in protein		Pilon et al. (2003)
Mus musculus SL (cp)	Selenocysteine lyase	Arabidopsis		Decrease	Decrease	Increase	Decrease in protein		Pilon et al. (2003)
AtSBP1	Selenium-binding protein	Arabidopsis		Increase					Agalou et al. (2005)

The phenotypes listed are tolerance of Se in the soil under field conditions, tolerance of selenite and selenate in the rhizosphere determined in laboratory experiments, effects on total Se concentration and organic Se compounds in leaves, and Se volatilization.

Much of the research using transgenic plants has been directed towards the remediation of land with high soil Se concentrations (Terry et al., 2000; Pilon-Smits and LeDuc, 2009; Zhu et al., 2009; Pilon-Smits, 2012). This research has focused on (a) increasing plant tolerance of high soil Se concentrations; (b) increasing Se transport to the shoot; (c) increasing Se accumulation in shoot tissues; and (d) increasing Se volatilization. Overexpressing genes encoding transporters for selenate, selenite or selenoamino acids in the plasma membrane of particular cells can increase the capacity for Se uptake and transport within the plant. However, unless this is accompanied by an ability to tolerate greater tissue Se concentrations or volatilize more Se, it is unlikely to allow greater tolerance of Se in the rhizosphere or phytoremediation potential.

In non-accumulator plants, the conversion of selenate to selenite within plastids appears to be the rate-limiting step in the assimilation of Se into organic compounds (Pilon-Smits et al., 2009). Overexpression of AtATPS1, PaAPR or both AtATPS1 and PaAPR in arabidopsis results in greater concentrations of organic Se in leaves, but a decrease in total leaf Se concentration (Table 3; Sors et al., 2005a) and, although overexpression of PaAPR results in greater tolerance of selenate in the rhizosphere in arabidopsis, the overexpression of AtATPS1 does not (Sors et al., 2005a). In contrast, the overexpression of AtATPS1 in Indian mustard, a Se-indicator plant, results in greater concentrations of Se and organic Se in leaves and greater tolerance of selenate in the rhizosphere (Pilon-Smits et al., 1999b; Van Huysen et al., 2004; Bañuelos et al., 2005b). The overexpression of genes involved in glutathione synthesis, such as glutathione synthase and γ-glutamyl-cysteine synthase, also appears to increase Se concentrations in leaves and Se tolerance of Indian mustard grown on seleniferous soils (Bañuelos et al., 2005b), whereas overexpression of cystathione-γ-synthase results in greater tolerance of selenite in the rhizosphere, reduced leaf Se concentrations and greater Se volatilization (Van Huysen et al., 2003, 2004).

The ability to tolerate Se in plant tissues and, thereby, to accumulate greater Se concentrations can be increased by the overexpression of genes encoding SMT, particularly if combined with overexpressing ATPS (Table 3). The overexpression of SMT, with or without the overexpression of ATPS, results in greater tolerance of selenite, and sometimes also selenate, in the rhizosphere, greater total Se, SeMSeCys and γ-glutamyl-SeMSeCys concentrations in leaves, and greater Se volatilization in transgenic plants compared with untransformed controls (Ellis et al., 2004; LeDuc et al., 2004, 2006; Bañuelos et al., 2007; Kubachka et al., 2007; Matich et al., 2009; McKenzie et al., 2009). The overexpression of genes encoding SeCyslyases has had variable effects on the tolerance of transgenic plants to selenate and selenite in the rhizosphere, but has consistently resulted in greater leaf Se concentrations and less Se incorporation into proteins in transgenic plants exposed to selenite or selenite than in untransformed plants (Garifullina et al., 2003; Pilon et al., 2003; Van Hoewyk et al., 2005; Bañuelos et al., 2007). Finally, the overexpression of AtSBP1 has been shown to increase selenite tolerance in transgenic arabidopsis (Agalou et al., 2005).

CONCLUSIONS AND PERSPECTIVES

Selenium is an essential mineral element for the well-being of animals and a beneficial element for plants. However, excess Se can be toxic to both animals and plants. There is considerable interest in understanding how plants acquire and accumulate Se, not only to facilitate appropriate dietary Se intakes for animal and humans, which often requires Se biofortification of edible crops, but also to remediate land contaminated anthropogenically by excess Se and to appreciate the ecology of native plants inhabiting seleniferous soils. Recently, researchers have begun to identify the genetic factors influencing Se acquisition and accumulation by plants. Initially, this work focused on elucidating the genes encoding enzymes involved in Se uptake, metabolism and distribution within the plant. Application of this knowledge has allowed the genetic manipulation of Se metabolism to increase Se accumulation in harvested tissues and Se volatilization to the atmosphere, benefitting both biofortification and phytoremediation strategies. It has also informed our appreciation of the possible mechanisms driving the evolution of species that hyperaccumulate Se in their tissues. Appreciable variation in Se concentrations in analogous tissues has been attributed to genetic factors both between and within plant species. Considerable effort is currently being invested in identifying chromosomal loci (QTLs) underlying these differences, which will enable the selection and breeding of crops with greater ability to acquire and accumulate Se in appropriate chemical forms in their edible tissues. Although our knowledge of the genetics of Se accumulation in plants appears rudimentary at present, it will increase rapidly as the modern toolbox of molecular techniques are applied. It is laudable that this effort will be built on the solid foundations of plant physiology and biochemistry.