植物硫代谢

Review ArticleAbstract:Sulfur metabolite levels and sulfur metabolism have been studied in a significant number of herbaceous and woody plant species.However,only a limited number of datasets are comparable and can be used to identify similarities and differ-ences between these two groups of plants.From these data,it appears that large differences in sulfur metabolite levels,as well as the genetic organization of sulfate assimilation and me-tabolism do not exist between herbaceous plants and trees.The general response of sulfur metabolism to internal and/or exter-nal stimuli,such as oxidative stress,seems to be conserved be-tween the two groups of plants.Thus,it can be expected that, generally,the molecular mechanisms of regulation of sulfur me-tabolism will also be similar.However,significant differences have been found in fine tuning of the regulation of sulfur me-tabolism and in developmental regulation of sulfur metabolite levels.It seems that the homeostasis of sulfur metabolism in trees is more robust than in herbaceous plants and a greater change in conditions is necessary to initiate a response in trees. This view is consistent with the requirement for highly flexible defence strategies in woody plant species as a consequence of longevity.In addition,seasonal growth of perennial plants ex-erts changes in sulfur metabolite levels and regulation that cur-rently are not understood.In this review,similarities and differ-ences in sulfur metabolite levels,sulfur assimilation and its reg-ulation are characterized and future areas of research are iden-tified.Key words:Adenosine5′-phosphosulfate reductase(APR),cell differentiation,compartmentation,cysteine,genetic engineer-ing,glutathione,methionine,plant development,protein, ozone,oxidative stress,sulfate,sulfate assimilation,transgenic poplar.IntroductionTrees differ from herbaceous plants not only in size and lon-gevity,but also in numerous features of nutrient acquisition and use in growth and development.Nutrient uptake of trees is largely uncoupled from seasonal growth of vegetative and generative organs.In many tree species,reproductive tissues are developed when uptake processes are strongly limited by low temperatures in the soil(Kozlowski,1971).Other tree spe-cies,like European beech,develop their leaves entirely in early spring,when unfavourable conditions for nutrient uptake from the soil prevail.Also,root growth of trees may take place under such conditions in spring and autumn(Kozlowski, 1971).As a consequence,developmental processes in trees that rely on the availability of a significant amount of nu-trients,depend on storage and mobilization processes rather than actual nutrient availability and acquisition(Millard et al.,2006),with the primary site of storage differing between conifers(needles)and deciduous trees(stem and roots). Exploitation by trees of nutrients in the soil is strongly im-proved by association with mycorrhizal fungi,most important-ly with ectomycorrhizal fungi.This symbiosis enables the oc-cupation of nutrient-poor environments by trees(Vitousek and Howarth,1991),in particular,growth on soils low in nitro-gen and/or phosphorus(Smith and Read,1997),and also im-proves sulfur nutrition of trees(Rennenberg,1999).In many temperate forest trees,ectomycorrhizal association maintains the preference for ammonium compared to nitrate uptake that is also found in non-mycorrhizal roots of trees(Kreuzwieser et al.,1997a;Geßler et al.,1998).However,organic nitrogen com-pounds liberated from leaf litter,decaying roots and microbial biomass are also thought to contribute considerably to nitro-gen nutrition of trees(Chalot and Brunn,1998).Also,organic sulfur can be taken up by the roots of trees(Seegmüller and Rennenberg,2002),but the significance of organic sulfur up-take for sulfur nutrition has not been elucidated.It appears that growth and development of trees is connected with general differences in acquisition and metabolism of nu-trients,compared to herbaceous plants.This review focuses on such differences in the metabolism of sulfur,the fifth most abundant element in plants,that is not only involved in pro-tein structure and function,redox regulation,and chloroplast membrane structure,but also in biotic and abiotic stress de-Sulfur Metabolism in Plants:Are Trees Different?H.Rennenberg1,C.Herschbach1,K.Haberer1,and S.Kopriva1,21Institute of Forest Botany and Tree Physiology,Chair of Tree Physiology,University of Freiburg,Georges-Köhler-Allee053/054,79110Freiburg,Germany2Present address:John Innes Centre,Norwich,Norfolk,NR47UH,UKReceived:January15,2007;Accepted:March23,2007Plant Biol.9(2007):620–637©Georg Thieme Verlag KG Stuttgart·New York DOI10.1055/s-2007-965248ISSN1435-8603Dedicated to Prof.Dr.L.Bergmann at the occasion of his80th birthday.620fence(Benning,1998;Rausch and Wachter,2005;Ströher and Dietz,2006).Sulfur Metabolite Levels in Herbaceous andPerennial PlantsThe most abundant sulfur metabolites in plants,i.e.,sulfate, cysteine,and glutathione,have been determined in a large number of plant species(Tables1–3).The herbaceous species analyzed include plants that are rich in sulfur due to the syn-thesis of secondary sulfur metabolites,such as Arabidopsis and Brassica that produce significant amounts of glucosino-lates(Fahey et al.,2001;Wittstock and Halkier,2002).How-ever,most herbaceous species and all tree species investigated do not possess such secondary sulfur metabolites.The range in which sulfate was found in plants varied between less than1 and more than30μmol g–1fresh weight in roots and less than 1and more than100μmol g–1fresh weight in leaf tissues.For a given species,root and leaf sulfate levels are comparable,and general differences between herbaceous and woody species were not found.High levels of sulfate were found in roots and leaves of Brassica oleracea,Populus sp.,and Quercus ilex as well as in needles of Pinus sylvestris(Table1).Similar observations were made for the glutathione levels of leaves and roots that, for most plant species,varied between50and400nmol g–1 fresh weight in roots and25and800nmol g–1fresh weight inTable1Sulfate levels in lateral roots,mature leaves,xylem,and phloem of herbaceous plants,conifers,and deciduous treesLateral roots(μmol g–1tissue fw)Mature leaves(μmol g–1tissue fw)Xylem(μmol l–1)Phloem(nmolμmol–1sucrose)Arabidopsis thaliana 2.5–1017Brassica oleracea8–161113–4311Brassica napus0.0005–0.00096 Allium cepa 5.3–7.216 2.8–3.216Nicotiana tabacum6–7.315Spinacea oleracea0.4–1.1100.6–1.310Lemna minor–1–2013Solanum tuberosum 2.3–2.514 1.4–2.014Fagus sylvatica2–513–6110–3702Populus sp.15–33310–203100–300310–807 Quercus robur280–7009Quercus ilex12–1888–148Picea abies0.3–3.1420–2805Pinus sylvestris40–110121Kreuzwieser et al.,1997b;2Rennenberg et al.,1994a;3Herschbach et al.,2000; 4Schupp and Rennenberg,1992;5Köstner et al.,1998;6Lappartient and Tour-aine,1996;7Herschbach,unpublished;8Schulte et al.,2002;9Rennenberg,1999;10Poortinga and De Kok,2000;11De Kok et al.,1997;12Polle et al.,1994; 13Kopriva et al.,2002;14Hopkins et al.,2005;15calculated from Matityahu et al., 2006;16Durenkamp and De Kok,2004;17Loudet et al.,2007.Table2Cysteine levels in lateral roots,mature leaves,xylem,and phloem of herbaceous plants,conifers,and treesLateral roots (nmol g–1tissue fw)Mature leaves(nmol g–1tissue fw)Xylem(μmol l–1)Phloem(nmolμmol–1sucrose)Arabidopsis thaliana30–501010–4010Brassica oleracea40175–3017Brassica napus20–80890–110220.005–0.0098 Nicotiana tabacum15–2758–125Spinacea oleracea40–501520–6015Lemna minor–ca.218Solanum tuberosum6–8195–1919,20Fagus sylvatica2–1218–2410.05–1.22,3Populus sp.7–1142–1040–4.04,130.2–1.59 Quercus pubescens n.d.5–1511Quercus ilex4–81230–4112Quercus robur 1.5–2.614Picea abies10–9060.25–1.257,21Pinus sylvestris60–100161Kreuzwieser et al.,1997b;2Rennenberg et al.,1994;3Schupp et al.,1991; 4Herschbach et al.,2000;5Herschbach and Rennenberg,1994;6Schupp and Rennenberg,1992;7Köstner et al.,1998;8Lappartient and Touraine,1996; 9Herschbach,unpublished;10Vauclare et al.,2002;11Schulte et al.,1997;12Schulte et al.,2002;13Schneider et al.,1994b;14Rennenberg,1999;15Poor-tinga and De Kok,2000;16Polle et al.,1994;17Westerman et al.,2000;18Kopriva et al.,2002;19Hopkins et al.,2005;20Harms et al.,2000;21Blaschke et al.,1996; 22Ruiz and Blumwald,2002.Sulfur Metabolism in Trees Plant Biology9(2007)621leaves.Exceptionally high levels of glutathione were found in the roots of Brassica sp.(550to1800nmol g–1fresh weight) und high levels of glutathione in leaves were measured in Pi-nus sylvestris(1700nmol g–1fresh weight).Root and leaf levels of glutathione were similar in most herbaceous plants,where-as most tree species contained significantly higher glutathione levels in the leaves than in roots.However,clear-cut differ-ences in glutathione levels between herbaceous and woody plants were not found for roots or for leaves(Table3).In most herbaceous species analyzed,cysteine levels of roots were higher(15to50nmol g–1fresh weight)than in woody plants (2–11nmol g–1fresh weight).Cysteine levels of leaves varied between2and100nmol g–1fresh weight,with peak values measured in conifer needles.Clear-cut differences between herbaceous and woody plants were not observed;for a given species,leaf and root levels of cysteine were similar in most cases(Table2).In the comparison of sulfur metabolite levels in Tables1–3, only fully expanded,mature leaves and needles were consid-ered.However,sulfur metabolite levels show a strong varia-tion with season and lifetime.Mature conifer needles,for ex-ample,had highest cysteine and glutathione levels during winter(Schupp and Rennenberg,1992).The high glutathione levels in one-year-old conifer needles not only protect against oxidative damage,mediated in winter by high-light intensities at low temperatures(Esterbauer and Grill,1978),but also function as a storage pool of reduced sulfur for the new needle generation in the following year(Schupp et al.,1992;Schnei-der et al.,1994a;Blaschke et al.,1996).Glutathione from one-year-old needles must,therefore,be transported in spring ei-ther in the phloem or in the xylem to the developing needle generation(Schupp et al.,1992;Schneider et al.,1994a).As a consequence,the already high glutathione level at budburst is maintained during initial needle development(Schupp and Rennenberg,1992).High glutathione levels are also found in young,developing leaves of tobacco(Herschbach and Rennen-berg,1994)and poplar(Herschbach,unpublished results)and these high levels may also originate from high glutathione lev-els present in buds and/or from glutathione transported to the developing leaves.The latter is unlikely for poplar,since sul-fate rather than reduced sulfur is allocated from mature to developing leaves in this species(Hartmann et al.,2000).In beech,seasonally high glutathione levels were not only found in young,developing leaves in spring,but also during early stages of leaf senescence in autumn(Luwe,1996;Herschbach, unpublished results).Apparently,glutathione synthesis and export also contribute to the removal of nutrients from senes-cing beech leaves and transport to storage tissues in living bark and wood.There is little information about sulfur composition and levels in saps of the long-distance transport pathways in xylem and phloem.Sulfur levels of xylem saps from several herbaceous plants were investigated(for examples see:Teyker et al., 1991;Sato et al.,1998;do Amarante et al.,2006),but these studies did not distinguish reduced sulfur and sulfate.Data from perennial plants indicate that sulfate is the dominant sul-fur compound in xylem sap and that glutathione and cysteine levels are at least one order of magnitude lower(Tables1–3). Whether glutathione or cysteine constitutes the most abun-dant reduced sulfur compound in xylem sap depends on the species analyzed.In beech,cysteine is the most abundant re-duced sulfur compound(Rennenberg et al.,1994),whereas in Quercus ilex,Q.robur(Rennenberg,1999;Schulte et al.,2002), and Picea abies(Köstner et al.,1998)glutathione dominates the reduced sulfur pool of the xylem sap.Peak values of both sulfate and reduced sulfur compounds were found during spring,before bud break in deciduous trees(Schupp et al., 1991;Rennenberg et al.,1994;Schneider et al.,1994b),and during development of the new needle generation in conifer-ous trees(Köstner et al.,1998).The increasing cysteine concen-Table3Glutathione levels in lateral roots,mature leaves,xylem,and phloem of herbaceous plants,conifers,and deciduous treesLateral roots (nmol g–1tissue fw)Mature leaves(nmol g–1tissue fw)Xylem(μmol l–1)Phloem(nmolμmol–1sucrose)Arabidopsis thaliana130–28014,15220–37014,15Brassica oleracea600–70011,21500–60021Brassica napus900–180012700–800260.16–0.3512 Allium cepa220–38024260–60024Nicotiana tabacum200–3005200–3005Spinacea oleracea50–3006,1025–2506,10Lemna minor–ca.50025Solanum tuberosum250–40022175–45022,23Fagus sylvatica50–2001250–8001,200.02–0.42,3Populus sp.100–3501,4400–7001,43–124,190.1–3.713 Quercus pubescens n.d.175–30016Quercus ilex50–10018300–47518Quercus robur100–2001150–28015–801Picea abies200–45070.05–49,17Pinus sylvestris200–170081Herschbach and Rennenberg,2001;2Rennenberg et al.,1994a;3Schupp et al., 1991;4Herschbach et al.,2000;5Herschbach and Rennenberg,1994;6Hersch-bach et al.,1995;7Schupp and Rennenberg,1992;8Taulavuori et al.,1999; 9Köstner et al.,1998;10Poortinga and De Kok,2000;11Westerman et al.,2001; 12Lappartient and Touraine,1996;13Herschbach,unpublished;14Vauclare et al.,2002;15Howden et al.,1995a,b;16Schulte et al.,1997;17Blaschke et al.,1996; 18Schulte et al.,2002;19Schneider et al.,1994b;20Luwe,1996;21Westerman et al.,2000;22Hopkins et al.,2005;23Harms et al.,2000;24Durenkamp and De Kok;2004,25Kopriva et al.,2002;26Ruiz and Blumwald,2002.Plant Biology9(2007)H.Rennenberg et al. 622tration in xylem sap collected in April with increasing height in beech trunks may also be a consequence of mobilization of cysteine from storage pools(Rennenberg et al.,1994).These findings clearly indicate that,in spring,sulfur is remobilized from storage tissues and loaded into the xylem.Feeding ex-periments with2-year-old beech seedlings proved this.35S-sulfur from sulfate fed to a leaf accumulated in the stem in living bark and wood tissues as soluble and insoluble sulfur during winter.In spring,the stored radiolabelled sulfur was mobilized before bud break and accumulated in the newly de-veloped leaves(Herschbach and Rennenberg,1994).Sulfur compounds in the phloem have been preferentially an-alyzed in perennial plants.Few data are available from herba-ceous plants but,due to different experimental approaches, data are often not directly comparable.In phloem exudates from wheat,S-methylmethionine was the predominant re-duced sulfur compound,followed by glutathione(Bourgis et al.,1999),but sulfate was not measured in these samples.In other herbaceous plants,sulfate levels of phloem exudates were negligible compared to reduced sulfur compounds(Lap-partient and Touraine,1996;Kuzuhara et al.,2000).In con-trast,the phloem sap of poplar was dominated by sulfate and reduced sulfur was present in considerably lower amounts, with glutathione as the most abundant reduced sulfur com-pound(Herschbach et al.,1998,2000;Herschbach and Ren-nenberg,2001).Whether relatively high sulfate levels and a high contribution of glutathione to the reduced sulfur fraction are general features for phloem exudates of woody plants re-mains to be elucidated.The reduced sulfur level in phloem exudates from poplar was found to be higher than from Bras-sica(Tables2,3;Lappartient and Touraine,1996).However, analyses of phloem exudates generally lack completeness.No single study in herbaceous or perennial plants has analyzed the whole set of sulfur compounds identified in phloem exudates,i.e.,sulfate,cysteine,glutathione,methionine and S-methylmethionine.Cellular and Subcellular Compartmentation of Glutathione Most investigations have analyzed sulfur levels at the level of organs and have not considered cell-specific differences or compartmentation within the cells.Subcelluar compartmen-tation of glutathione has been discussed for many years(Ren-nenberg,1982),and different experimental approaches have revealed conflicting findings(Noctor et al.,2002).From isolat-ed chloroplasts,it was calculated that the plastidic glutathione fraction ranged from5to65%of total cellular glutathione and could amount to several mM(Noctor et al.,2002).However, the different methods used for chloroplast isolation may have resulted in different amounts of glutathione retained in this organelle.Recently,glutathione was localized in cells of Cucur-bita pepo by immunogold cytochemistry(Müller et al.,2004). According to this study,the main proportion of glutathione was in mitochondria.Since glutathione levels in mitochondria have never been studied by other approaches,further studies are required to test these results.A new technique for glutathione determination at the cellular level uses confocal laser scanning microscopy after in situ la-belling with monochlorobimane(Sánchez-Fernández et al., 1997;Gutiérrez-Alcaláet al.,2000;Fricker et al.,2000;Meyer and Fricker,2000,2002;Meyer et al.,2001;Hartmann et al.,2003;Gómez et al.,2004a).However,it is important to note that only the cytosolic glutathione fraction is marked by in situ labelling with monochlorobimane.This has been demonstrat-ed with poplar leaves(Hartmann et al.,2003),where the plas-tidic glutathione pool did not undergo short-term exchange with the cytosolic GSH pool.From these first in situ measure-ments of glutathione in photosynthetically active tissues,a cy-tosolic glutathione concentration of0.2to0.3mM was calcu-lated for poplar leaves(Hartmann et al.,2003).This value is one order of magnitude lower than previous reports for herba-ceous plants(reviewed in Noctor et al.,2002).Noctor et al. (2002)calculated a cytosolic glutathione concentration of1to 2mM in wheat leaves,and Fricker et al.(2000)found cytosolic glutathione concentrations of2to3mM in most cell types of Arabidopsis roots.Since the glutathione levels determined by HPLC in wheat(Noctor et al.,2002),Arabidopsis(Table3)and poplar are similar,10to20times higher glutathione levels may be assumed for plastids or mitochondria of poplar than for wheat or Arabidopsis cells.Certainly further studies are re-quired to test this assumption and to address the question why glutathione is not readily transported from these organelles into the cytosol in poplar cells,even with such a high concen-tration gradient.Glutathione Levels in Plant Tissues Interact withCell Differentiation and Plant DevelopmentGlutathione levels in leaves depend on the age of the leaf.In poplar,the highest glutathione level was found in the apex and decreased with increasing leaf age(Herschbach,unpub-lished results).A high glutathione level in the apex or in young leaves was also observed for tobacco leaves(Herschbach and Rennenberg,1994)and for spruce(Schupp and Rennenberg, 1992),Scots pine needles,and bilberry leaves(Taulavuori et al.,1999).In the monocot Lolium perenne,the glutathione level was highest in the leaf base,the growing zone of these leaves, and decreased with increasing distance from the base of the leaf(Piquery et al.,2002).Apparently,glutathione levels are high in growing and developing leaves or leaf sections.In pop-lar,sulfate assimilation and glutathione synthesis in develop-ing leaves,but not GSH transport to these leaves,is responsible for the high glutathione level(Hartmann et al.,2000;Hersch-bach,2003).Whether these observations are a special feature of poplar or of more general significance remains to the eluci-dated.Not only cells of different age,but also different cell types con-tain different glutathione levels.In epidermal cells of Arabi-dopsis leaves glutathione levels of trichome cells were higher than of trichome basement cells or the epidermal cells them-selves(Gutiérrez-Alcaláet al.,2000).Other leaf cell types from Arabidopsis were not investigated.In poplar,a uniform distri-bution of glutathione was found in transverse leaf sections. Epidermal,palisade,spongy mesophyll,and guard cells had similar cytosolic glutathione levels(Hartmann et al.,2003). Since comparable data from the leaves of other perennial or herbaceous species are not available,more general conclusions on the distribution of glutathione between different types of leaf cells cannot be established.Since the activity of the key enzyme of sulfate assimilation, adenosine5′-phosphosulfate reductase(APR),in Zea mays roots decreased with increasing distance from the root tip(Ko-Sulfur Metabolism in Trees Plant Biology9(2007)623priva et al.,2001),decreasing glutathione levels from the root tip to older root sections may be assumed.Indeed,in situ label-ling of glutathione and detection of glutathione-bimane conju-gates by confocal laser scanning microscopy(CLSM)showed that glutathione decreased along Arabidopsis roots with in-creasing distance from the root tip(Sánchez-Fernández et al., 1997;Espunya et al.,2006).Also,trichoblasts and atrichoblasts of Arabidopsis roots were shown by CLSM to possess different levels of glutathione,with the latter being significantly bright-er after labelling with monochlorobimane(Meyer and Fricker, 2000).As meristematic cells showed very low fluorescence from glutathione-bimane derivatives,it seems that these cells contain only low glutathione levels(Sánchez-Fernández et al., 1997;Meyer and Fricker,2000).In addition,Vernoux et al. (2000)found that glutathione accumulation is induced during the G1to S transition in root cell development.More recently, transcriptional profiling showed a high abundance of the gshII gene in the cells of the quiescent centre of Arabidopsis roots (Nawy et al.,2005).Apparently,a high rate of GSH synthesis is necessary to control cell division.This assumption is sup-ported by findings with the Arabidopsis mutant rml1that con-tains a mutation in the gene forγ-glutamylcysteine synthetase, the first enzyme of glutathione biosynthesis(Vernoux et al., 2000).This mutant was unable to establish an active post-em-bryonic meristem in the root apex,but feeding with gluta-thione partially restored the observed phenotype.On the other hand,a complete disruption ofγ-glutamylcysteine synthetase (γ-ECS)led to embryo lethality(Cairns et al.,2006).In this con-text,it is interesting that glutathione levels in the cytosol and the nucleus are higher in roots compared to leaves(Müller et al.,2004).However,Espunya et al.(2006)showed that external glutathione supplied to wild-type Arabidopsis roots reduced root growth.Since Sánchez-Fernández et al.(1997)demon-strated that feeding with glutathione resulted in an increased root hair,but a reduced trichoblast,length,it appears that a balanced glutathione level is necessary for optimal root growth.Such a regulatory mechanism may adapt root growth to a changing environment(Sánchez-Fernández et al.,1997). Poplar roots with secondary growth show higher glutathione levels than fine roots(Herschbach,unpublished results),but this may be the consequence of GSH storage in ray and pith cells(Hartmann et al.,2000).However,detailed analyses of GSH levels in fine roots are not available for poplar or other tree species,so that a comparison between herbaceous and perennial plants is not possible.Glutathione levels also play a crucial role in tracheary element development.In cultured mesophyll cells of Zinnia elegans,ox-idized glutathione(GSSG)rather than reduced glutathione (GSH)initiates cell differentiation(Henmi et al.,2001).Also, in wild-type Arabidopsis,tracheary development was promot-ed by application of GSSG(Henmi et al.,2005).When gluta-thione reductase was over-expressed in Zinnia cells and Arabi-dopsis,tracheary development was delayed in a similar way as observed with GSH treatment.Apparently,tracheary element development is subject to redox control by glutathione.Al-though over-expression of glutathione reductase in poplar did not result in a visible phenotype(Foyer et al.,1995),nothing is known about vessel development in these transgenic plants. Flowering also seems to be dependent on glutathione levels (Ogawa et al.,2001,2004).The cad2-1mutant of Arabidopsis with reduced activity ofγ-ECS(Cobbett et al.,1998)exhibits delayed flowering.This effect can be counteracted by GSH treatment(Ogawa et al.,2001).Together,these results clearly indicate that glutathione levels and the redox state of gluta-thione are involved in the control of cell and tissue develop-ment in various plant organs.A lack of data for perennial plants,however,prevents any conclusions about differences between herbaceous and woody species.Consequences of Genetic Engineering forGlutathione LevelsGlutathione levels have been altered in plants using several manipulation strategies(Table4).The first transgenic poplars with increased glutathione levels were produced by over-ex-pressing the gene for bacterial glutathione reductase(GR,gor; Foyer et al.,1995).The level of glutathione was not affected when the gene product was targeted to the cytosol,but foliar glutathione levels doubled when GR was targeted to plastids. These transgenic poplars showed increased resistance to pho-toinhibition and had a higher capacity to withstand oxidative stress mediated by methylviologen treatment(Foyer et al., 1995).Poplar plants over-expressing bacterialγ-glutamylcys-teine synthetase(γ-ECS,gsh1),but not poplar plants over-expressing bacterial glutathione synthetase(GSHS,gsh2), showed enhanced glutathione levels,irrespective of whether the proteins were targeted to the cytosol or to the plastids (Strohm et al.,1995;Arisi et al.,1997;Noctor et al.,1998).Sim-ilar results were obtained by transformation of other plant species(Table4).Arabidopsis plants over-expressing theγ-ECS gene from Arabidopsis(Table4;Xiang et al.,2001)showed in-creased glutathione levels in the transformants,indicating that the origin ofγ-ECS is not a decisive factor for the increase in glutathione level.Reduced or negligible glutathione levels were found in Arabidopsis mutants inγ-ECS gene cad2,rml2, and rax1(Cobbett et al.,1998;Vernoux et al.,2000;Ball et al., 2004).These data clearly identifyγ-ECS activity as a key factor in glutathione synthesis.Tobacco plants over-expressing the bacterial gsh1or gsh2had increased GSH levels,which were significantly greater when both genes were over-expressed simultaneously(Table4; Creissen et al.,1999).Targeting of the gsh1to plastids,but not to the cytosol,caused symptoms of oxidative stress in trans-genic tobacco plants(Creissen et al.,1999).Symptoms of oxi-dative stress were not observed in other plant species when gsh1was targeted to the cytosol(Noctor et al.,1998;Xiang et al.,2001),even after prolonged growth(Herschbach,unpub-lished results).Contrary to results from tobacco,Brassica seed-lings(Zhu et al.,1999a)and young poplar plants(Noctor et al., 1998)targeting gsh1to plastids did not cause a visible pheno-type.But,after prolonged growth(4to5months),transgenic poplar over-expressing gsh1in the plastids had similar symp-toms of injury to tobacco plants(Creissen et al.,1999),such as reduced growth,premature senescence,and,occasionally,the development of side branches(Herschbach,unpublished re-sults).However,transgenic tobacco over-expressing maize gsh1targeted to chloroplasts did not showed any visible symp-toms of oxidative stress(Gómez et al.,2004b).Whether these differences are a result of plant versus bacterial gsh1over-ex-pression,or a result of different expression levels of the trans-gene remains to be elucidated.Plant Biology9(2007)H.Rennenberg et al. 624。