Specific Roles of a-andg-Tocopherol in Abiotic Stress · Specific Roles of a-andg-Tocopherol in Abiotic Stress Responses of Transgenic Tobacco1[W][OA] Ali-Reza Abbasi, Mohamad Hajirezaei,
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Specific Roles of a- and g-Tocopherol in Abiotic StressResponses of Transgenic Tobacco1[W][OA]
Ali-Reza Abbasi, Mohamad Hajirezaei, Daniel Hofius2, Uwe Sonnewald, and Lars M. Voll*
Friedrich-Alexander-Universitat Erlangen-Nurnberg, Lehrstuhl fur Biochemie, 91058 Erlangen,Germany (A.-R.A., U.S., L.V.); and Institut fur Pflanzengenetik und Kulturpflanzenforschung,D–06466 Gatersleben, Germany (M.H., D.H.)
Tocopherols are lipophilic antioxidants that are synthesized exclusively in photosynthetic organisms. In most higher plants,a- and g-tocopherol are predominant with their ratio being under spatial and temporal control. While a-tocopherolaccumulates predominantly in photosynthetic tissue, seeds are rich in g-tocopherol. To date, little is known about the specificroles of a- and g-tocopherol in different plant tissues. To study the impact of tocopherol composition and content on stresstolerance, transgenic tobacco (Nicotiana tabacum) plants constitutively silenced for homogentisate phytyltransferase (HPT) andg-tocopherol methyltransferase (g-TMT) activity were created. Silencing of HPT lead to an up to 98% reduction of totaltocopherol accumulation compared to wild type. Knockdown of g-TMT resulted in an up to 95% reduction of a-tocopherol inleaves of the transgenics, which was almost quantitatively compensated for by an increase in g-tocopherol. The response ofHPT and g-TMT transgenics to salt and sorbitol stress and methyl viologen treatments in comparison to wild type was studied.Each stress condition imposes oxidative stress along with additional challenges like perturbing ion homeostasis, desiccation, ordisturbing photochemistry, respectively. Decreased total tocopherol content increased the sensitivity of HPT:RNAi transgenicstoward all tested stress conditions, whereas g-TMT-silenced plants showed an improved performance when challenged withsorbitol or methyl viologen. However, salt tolerance of g-TMT transgenics was strongly decreased. Membrane damage ing-TMT transgenic plants was reduced after sorbitol and methyl viologen-mediated stress, as evident by less lipid peroxidationand/or electrolyte leakage. Therefore, our results suggest specific roles for a- and g-tocopherol in vivo.
Under natural conditions, plants are exposed to avariety of biotic and abiotic stresses, including path-ogens, adverse temperature, drought, salt, and highlight. Under these stress conditions, reactive oxygenspecies (ROS) derived from molecular oxygen canaccumulate in leaves, resulting in the oxidation ofcellular components, including proteins, chlorophyll,and lipids. To cope with oxidative stress plants haveevolved two general protective mechanisms, enzy-matic and nonenzymatic detoxification, of which thelatter involves vitamin E (Dat et al., 2000; Alscher andHeath, 2002).
Eight structurally related tocochromanols are com-monly addressed as vitamin E, namely tocopherols
and tocotrienols, which can be distinguished by thesaturation of their prenyl moiety and by the methyl-ation degree of their chromonal ring. Subcellulardistribution and enzymatic steps of vitamin E biosyn-thesis have been characterized in detail (for recentreview, see DellaPenna and Pogson, 2006) and a num-ber of molecular approaches to manipulate vitamin Esynthesis have been published (summarized in Herbers,2003; DellaPenna and Last, 2006; Dormann, 2007).
Vitamin E synthesis is restricted to photosyntheticorganisms including plants, algae, and some cyano-bacteria. The first committed steps in tocopherol andtocotrienol biosynthesis are the prenylation of homogen-tisic acid by either phytyl diphosphate or geranylgeranyldiphosphate yielding 2-methyl-6-phytyl-benzoquinoneor methyl geranylgeranyl benzoquinone, respectively.These reactions are catalyzed by homogentisate phy-tyltransferase (HPT) and homogentisate geranylger-anyl transferase. The two subsequent steps catalyzedby tocopherol cyclase (TC) and g-tocopherol methyl-transferase (g-TMT) are in common for tocopherol andtocotrienol biosynthesis (see Fig. 1). With the recentcharacterization of phytol and phytyl phosphate ki-nase from Arabidopsis (Arabidopsis thaliana; Ischebecket al., 2006; Valentin et al., 2006), it could be demon-strated that the prenyl moiety for g-tocopherol bio-synthesis is derived from free phytol in seeds,substantiating the assumption that phytol gets recy-cled during chlorophyll breakdown (Peisker et al.,1989; Rise et al., 1989; Dormann, 2007).
1 This work was supported by the University of Teheran andthe Ministry of Science, Research, and Technology of the I.R. Iran (toA.-R.A.).
2 Present address: Department of Molecular Biology, CopenhagenBiocenter, University of Copenhagen, Ole Maaloes Vej 5, 2200Copenhagen N, Denmark.
The author responsible for distribution of materials integral to thefindings presented in this article in accordance with the policydescribed in the Instructions for Authors (www.plantphysiol.org) is:Lars M. Voll ([email protected]).
[W] The online version of this article contains Web-only data.[OA] Open Access articles can be viewed online without a sub-
1720 Plant Physiology, April 2007, Vol. 143, pp. 1720–1738, www.plantphysiol.org � 2007 American Society of Plant Biologists www.plantphysiol.orgon April 1, 2020 - Published by Downloaded from
In plants, tocopherol composition differs between dif-ferent species and between different tissues within onespecies (summarized in Grusak and DellaPenna, 1999).Usually, leaves commonly accumulate a-tocopherol,whereas seeds are rich in g-tocopherol. b- andd-tocopherol are not very abundant in most plantspecies. Therefore, a-tocopherol has been proposed toparticipate in the detoxification of ROS together withthe hydrophilic antioxidants glutathione and ascorbate(Foyer and Noctor, 2003). Based on inhibitor studies ithas been shown that tocopherol acts as singlet oxygenscavenger in PSII of Chlamydomonas reinhardtii (Trebstet al., 2002; Kruk et al., 2005). The inhibitory effect couldbe overcome by addition of membrane-permeableshort chain a- and g-tocopherol derivatives, suggestingthat a-tocopherol can be substituted by g-tocopherol inleaves. A similar conclusion was drawn from the bio-chemical analysis of the Arabidopsis g-TMT-deficientmutant vte4-1, which accumulates comparable amountsof g-tocopherol instead of a-tocopherol (Bergmulleret al., 2003). Photosynthetic parameters in vte4-1 werefound to be indistinguishable from wild type undera variety of stress conditions (Bergmuller et al., 2003).In contrast to leaves, g-tocopherol is predominant inseeds, where total tocopherol contents are 10 to 100times higher than in leaves (Table I). In oil-storing seedslike Arabidopsis, g-tocopherol has been shown to pro-tect polyunsaturated fatty acids (PUFAs) from oxida-tion, thereby increasing seed longevity (Sattler et al.,2004). This suggests that g-tocopherol is involved indesiccation tolerance of seeds.
Excess salt and sorbitol both induce hyperosmoticand oxidative stress, albeit to varying degrees. Conse-quently, molecular events in response to salt and
osmotic stress are convergent, involving abscisic acid(ABA), ROS, and Ca21 signaling that activate variousdownstream signaling cascades, like mitogen-activatedprotein kinases, Suc nonfermenting-related kinase, orcalcineurin B-like Ca21-dependent signaling (for re-cent reviews, see Wang et al., 2003; Boudsocq andLauriere, 2005). The interplay of the foliar antioxidantnetwork and cellular signaling is pivotal for stresssensing. Leaf ascorbate content was shown to modu-late the response to ABA (Pastori et al., 2003), whileABA was shown to stimulate antioxidant capacity andtransient ROS production in parallel (Zhang et al., 2006,
Figure 1. The tocopherol, tocotrienol, and plas-toquinone biosynthetic pathway in plants. Thisfigure represents the enzymatic reactions andintermediates that are involved in the biosynthesisof tocopherol, tocotrienol, and plastoquinone.Arabidopsis mutants of the corresponding path-way genes are given in parentheses behindthe enzymes. DMPBQ, 2,3-dimethyl-5-phytyl-1,4-benzoquinone; HPP, p-hydroxyphenylpyruvate;MSBQ, 2-methyl-6-solanesyl-1,4-benzoquinone;phytyl-DP, phytyl-diphosphate; solanesyl-DP,solanesyl diphosphate; HPPD, HPP dioxygen-ase; HST, homogentisate solanesyltransferase;MPBQMT (vte3), MPBQ/MSBQ methyltransfer-ase; PK (vte5), phytol kinase; PPK, phytyl phos-phate kinase.
Table I. Tocopherol content and composition in tobacco tissues
Tocopherols were extracted from the indicated tobacco tissues with100% methanol as described in ‘‘Materials and Methods.’’ a-T and g-Tare a- and g-tocopherol, respectively; b-tocopherol was not detectableand d-tocopherol represented less than 5% of the tocopherol pool andwas omitted from the representation for clarity reasons. Consequently,the percentage distribution of a- versus g-tocopherol is given in thetable, being mean and SD of four independent samples.
and refs. therein). As proposed by Foyer and Noctor(2005), the balance of the antioxidant and ROSresponses finally determines whether the cells re-spond to water stress by acclimation and hardeningor with programmed cell death.
Nevertheless, excess salt and sorbitol are thought tohave different impact on cellular metabolism. After itsuptake into the symplast, sodium will be sequesteredin the vacuole by Na1/H1 antiporters (Fukuda et al.,1999; Gaxiola et al., 1999; Shi et al., 2000) to preventpoisoning of enzyme activities in the other cellularcompartments (Serrano et al., 1999). As this processutilizes the pH gradient across the tonoplast, highsalinity also impairs cellular pH homeostasis (Serranoet al., 1999). In contrast, sorbitol accumulates as com-patible solute in Rosaceae (e.g. apple [Malus domestica];Li and Li, 2005; Sircelj et al., 2005) and prokaryotes andpossesses a lower toxicity than NaCl. Sorbitol ac-cumulation of up to 3 mmol g fresh weight21 wastolerated in transgenic tobacco (Nicotiana tabacum;Sheveleva et al., 1998). Excess sorbitol more specificallyeffectuates desiccation of the cellular compartments,inevitably leading to increased ROS production at thephotosystem reaction centers when the hydration sta-tus in chloroplasts drops. Consequently, increasingROS quenching capacity by overexpression of super-oxide dismutase (Gupta et al., 1993) or glutathi-one S-transferase and glutathione peroxidase (Roxaset al., 1997) had the potential to alleviate oxidativestress.
In this study, we have constitutively silenced HPTand g-TMT posttranscriptionally in tobacco plantswith a dsRNAi approach. Assessing the response ofHPT and g-TMT knockdown lines to salt stress, sorbi-tol stress, and methyl viologen enabled us to comparethe effects of a general tocopherol deficiency (inHPT:RNAi) versus a replacement of a-tocopherolwith g-tocopherol (in g-TMT:RNAi) in oxidative stressscenarios. Silencing of both HPT and g-TMT resultedin an elevated susceptibility to salt stress. In contrast,the substitution of a- for g-tocopherol in g-TMT trans-genics resulted in an elevated osmotolerance whenplantlets were subjected to sorbitol. This was charac-terized by substantially increased membrane integritycompared to the wild type, as determined by ionleakage and lipid peroxidation. The g-TMT:RNAiplants exhibited a similar benefit when ROS produc-tion in the thylakoids was increased by methyl vio-logen treatment. Our results provide evidence thatg-tocopherol is more potent than a-tocopherol in me-diating osmoprotection in vivo.
RESULTS
The Ratio of a- to g-Tocopherol Is Tissue Dependentin Tobacco
To investigate the plasticity of tocopherol biosyn-thesis in tobacco, we examined the composition of thetocopherol pool in a variety of vegetative and repro-
ductive organs of wild-type Samsun NN (Table I). Thetocopherol pool in seedlings, leaves, roots, and stigmaconsisted more than 90% of a-tocopherol, female flowerorgans contained between 70% and 80% a-tocopherol,sepals and petals contained roughly equal amounts ofa- and g-tocopherol, while stamina (80% g-tocopherol)were dominated by g-tocopherol. Both total tocoph-erol pool size and the fraction of g-tocopherol in-creased progressively with seed maturation (data notshown). The tocochromanol pool in mature tobaccoseeds is dominated by g-tocopherol and g-tocotrienolwith negligible amounts of a-tocopherol (e.g. Falket al., 2003).
In summary, a-tocopherol accumulates in vegetativeorgans of tobacco, while reproductive tissues are en-riched in g-tocopherol, especially when they are des-iccation tolerant like stamina or seeds.
dsRNAi-Mediated Silencing of HPT Results inTocopherol Deficiency in Tobacco Plants
To manipulate tocopherol metabolism in transgenictobacco, potato (Solanum tuberosum) cDNA fragmentsof StHPT and Stg-TMT (see next paragraph) wereintroduced into intron-spliced hairpin RNA (RNAi)silencing constructs driven by the constitutive cauli-flower mosaic virus 35S promoter (Figs. 2A and 3A).Posttranscriptional silencing of tobacco HPT mediatedby pBin-StHPT-RNAi (Fig. 2A) resulted in 68 primarytransformants in the T0 generation, which were se-lected for kanamycin resistance. All positive lines weretransferred to the greenhouse and screened for to-copherol deficiency by HPLC analysis. Tocopherolcontent in the transgenic lines ranged from morethan 130% to less than 2% of wild type (data notshown), demonstrating efficient silencing of tobaccoHPT by the heterologous potato dsRNAi construct.HPT catalyzes the committed step in tocopherol bio-synthesis, the condensation of the precursors homo-gentisic acid and phytyl diphosphate (see Fig. 1).Therefore, silencing of HPT affected all tocopherolsto the same extent, leaving the ratio between individ-ual tocopherols constant (data not shown).
Five transgenic lines with the strongest tocopheroldeficiency lacking more than 97% of total tocopherol(StHPT:RNAi-28, -49, -35, -44, -2), one line with 70%reduction (StHPT:RNAi-39) and one with almost 40%reduction (StHPT:RNAi-65) were selected for furtheranalysis (Fig. 2B). Among the selected lines,StHPT:RNAi-28 showed the strongest depletion oftotal tocopherol with less than 2% of wild-type levels.Transgenics with a total tocopherol content of less than5% of wild type, e.g. StHPT:RNAi-28, -44, and -49,exhibited a significant growth reduction of about 30%when cultivated in the greenhouse (Table II), whilegrowth of HPT:RNAi plants with a residual tocoph-erol content of above 5% were indistinguishable fromthe wild type. For simplicity, only results for thestrongest line StHPT:RNAi-28 will be communicatedbelow.
Abbasi et al.
1722 Plant Physiol. Vol. 143, 2007 www.plantphysiol.orgon April 1, 2020 - Published by Downloaded from
Results in an Accumulation of g- Instead ofa-Tocopherol in Leaves
Transforming the pBin-Stg-TMT-RNAi construct(Fig. 3A) for constitutive silencing of g-TMT intotobacco yielded 82 kanamycin-resistant T0 trans-formants, in which the tocopherol composition wassurveyed by HPLC. Silencing of g-TMT lead to dimin-ished a-tocopherol contents in leaves of the trans-genics. The T0 transformant Stg-TMT:RNAi-55 showedthe strongest a-tocopherol-deficient phenotype withless than 5% of a-tocopherol wild-type levels (Fig. 3B).Stg-TMT:RNAi-82 and -22 were also strongly affected,accumulating 7% to 10% of wild-type a-tocopherol,respectively.
The reduction in a-tocopherol was mostly paralleledby an increase in g-tocopherol in the g-TMT trans-formants (Fig. 3C), so that strong silencing of g-TMTcorrelated with a decrease of the a- to g-tocopherolratio in leaves of the transgenics (Fig. 3E). The inter-mediate lines Stg-TMT:RNAi-1, -21, and -13 displayedequal contents of a- and g-tocopherol, while the threelines with the most dramatic change in the ratio ofa- to g-tocopherol (to far below 1), namely Stg-TMT:RNAi-55, -82, and -22 (see Fig. 3E), were selectedfor closer study. For comparison, the wild type dis-played an a- to g-tocopherol ratio of about 15 (Fig. 3E).
In the lines Stg-TMT:RNAi-22, -55, and -82, totaltocopherol contents remained rather unchanged inrelation to the wild type (Fig. 3D), resulting in a netsubstitution of a- for g-tocopherol in leaves. The stron-gest line Stg-TMT:RNAi-55 was characterized in moredetail along with StHPT:RNAi-28, as described below.
Silencing of g-TMT Results in Increased Biomass
Production under Osmotic Stress, while Silencing ofBoth HPT and g-TMT Decreases Tolerance to Salt Stress
In contrast to recent reports on the functional char-acterization of plants with altered tocopherol compo-sition (Porfirova et al., 2002; Bergmuller et al., 2003;Hofius et al., 2004; Havaux et al., 2005; Kanwischeret al., 2005; Maeda et al., 2006), we aimed at examiningthe oxidative and osmotic stress tolerance instead offocusing on high light and cold stress. Wild-typetobacco and the transgenics were grown on Murashigeand Skoog medium supplemented with increasingamounts of 0 to 400 mM sorbitol or sodium chloride inaxenic culture; plant growth (this section), metabolitecontents, and physiological parameters (other para-graphs) were scored after 4 weeks of either stress.
Biomass production of wild-type plants negativelycorrelated with increasing sorbitol or NaCl concentra-tions and was diminished by 90% and 65% on 400 mM
sorbitol and 400 mM sodium chloride, respectively(Tables III and IV, column 1). Vegetative growth ofHPT and g-TMT transgenics was not significantly dif-ferent from wild type on control plates and on platessupplemented with up to 200 mM sorbitol or sodiumchloride, except for a 25% growth retardation of HPTknockdown plants on 200 mM NaCl (Tables III and IV;see Figs. 4, A–C and 5, A–C for the correspondingphenotypes). Sorbitol and sodium chloride of equal ormore than 300 mM provoked distinct phenotypes forHPT:RNAi and g-TMT:RNAi tobacco (Figs. 4, A–C and5, A–C for lines StHPT:RNAi-28 and Stg-TMT:RNAi-55;see Supplemental Figs. S1 and S2 for the performanceof lines StHPT:RNAi-49 and Stg-TMT:RNAi-82 chal-lenged with sorbitol and NaCl, respectively).
Silencing of HPT did not alter the response towardsorbitol significantly (Table III), while the HPT trans-genics were more susceptible to salt stress than thewild type, exhibiting only 63% of wild-type freshweight on 400 mM NaCl (Table IV). Leaves ofHPT:RNAi plants were almost entirely chlorotic on300 mM sodium chloride, while the wild type showedchlorosis on 400 mM NaCl only (compare Fig. 5, A andC; see Supplemental Fig. S2).
As in HPT:RNAi tobacco, silencing of g-TMT alsoaggravated the sensitivity of the transgenics towardsalt stress: this effect was less dramatic than for theHPT knockdown plants at 300 mM NaCl, but similar at400 mM sodium chloride (Table IV; Fig. 5, A andB; Supplemental Fig. S2). Surprisingly, silencing ofg-TMT led to a substantial growth benefit of more than2-fold over the wild type in osmotic stress of 300 and400 mM sorbitol (Table III). Despite a decrease in
Figure 2. RNAi-mediated silencing of HPT results in tocopherol defi-ciency in transgenic tobacco plants. A, Representation of the T-DNA fromthebinary intron-spliced hairpinRNA (RNAi)expression constructused fortransformation of tobacco plants. The 650-bp StHPT PCR fragment fromthe potato EST (accession no. BI919738) was cloned between the cauli-flower mosaic virus 35S promoter and the ocs terminator of the Bin19-derived vector in sense and antisense orientation, separated by intron 1 ofpotato GA 20-oxidase (200 bp) using the introduced BamHI and SalIrestriction sites. B, The total tocopherol content (a-, g-, and d-tocopherol)in fully expanded leaves of HPT:RNAi transformants and wild type wasquantified fluorometrically by HPLC and normalized to fresh weight.
Tocopherol Composition and Stress Tolerance
Plant Physiol. Vol. 143, 2007 1723 www.plantphysiol.orgon April 1, 2020 - Published by Downloaded from
growth of 75% on 400 mM sorbitol compared to thecontrol plates without osmolyte, g-TMT transgenicsremained virescent at 400 mM sorbitol, even whilewild-type and HPT:RNAi plants exhibited advancedchlorosis (Fig. 4, A–C; Supplemental Fig. S1).
Foliar Tocopherol Pools Exhibit Distinct Responses to
Sorbitol and Salt Stress in the Tobacco Transgenics
It is well documented that the amount of tocopherolincreases during stress (e.g. Collakova and DellaPenna,2003; Havaux et al., 2005; Kanwischer et al., 2005;Maeda et al., 2006). Therefore, we examined how the
disturbances in tocopherol biosynthesis in HPT:RNAiand g-TMT:RNAi affected tocopherol accumulation inthe studied stress conditions. To this end, tocopherolcontents in stress-acclimated leaves were determinedafter 4 weeks of osmotic and salt stress.
The wild-type Samsun NN displayed a more than4-fold increase in total tocopherol from around 70ng cm2 in controls to about 400 ng cm2 at 400 mM
sorbitol or NaCl (Figs. 4D and 5D). The total tocoph-erol content correlated directly with the concentrationof sorbitol and sodium chloride in the medium, witha-tocopherol contributing to more than 90% of thetocopherol pool in any condition in the wild type.
Figure 3. RNAi-mediated silencingof g-TMT lead to accumulation ofg-tocopherol and a deficiency ina-tocopherol in transgenic tobaccoplants. A, Representation of the T-DNAfrom the binary intron-spliced hairpinRNA (RNAi) expression construct usedfor transformation of tobacco plants.The 625-bp Stg-TMT PCR fragmentfrom the potato EST (accession no.BQ116842) was cloned between thecauliflower mosaic virus 35S promoterand the ocs terminator of the Bin19-derived vector in sense and antisenseorientation, separated by intron 1 ofpotato GA 20-oxidase (200 bp) usingthe introduced BamHI and SalI restric-tion sites. B to D, Contents of tocoph-erols in fully expanded leaves ofg-TMT:RNAi transformants and wildtype were determined by HPLC. Thetransgenic plant lines are orderedfrom left to right with increasing a- tog-tocopherol ratio. B, a-Tocopherolcontent. C, g-Tocopherol content. D,Total tocopherol content (sum of a-,g-, and d-tocopherol). E, The ratio of a- tog-tocopherol in wild-type and trans-genic plants. The diagrams depict dataof single measurements for individualsselected for further study in the T0 gen-eration screening. The data were laterreproduced with all tested individualsof the lineage progenies.
Table II. Growth performance of the three strongest HPT:RNAi tobacco lines (HPT) in comparison tothe wild type (WT)
The parameters were recorded at the end of the vegetative phase and are means and SD of fiveindividuals of one representative population.
The total tocopherol content in unchallengedg-TMT:RNAi plants was comparable to wild type,between 60 to 80 to ng cm2 (see Figs. 4D and 5D),except that g- substituted a-tocopherol in the g-TMTtransgenics. Likewise, the a-tocopherol-deficient g-TMT:RNAi transgenic tobacco plants accumulated similaramounts of g-tocopherol instead of a-tocopherol dur-ing osmotic and salt stress (Figs. 4D and 5D), whichwas up to 350 ng cm2 g-tocopherol. Compared tounstressed wild type, this would determine a 70-foldincrease in g-tocopherol. In conclusion, however, to-copherol biosynthesis got induced to a similar extentin g-TMT:RNAi and in wild-type tobacco, resulting inthe accumulation of g- and a-tocopherol, respectively.
In contrast, silencing of HPT abrogated the capacityfor tocopherol biosynthesis in the transgenics, so thatthe applied stress conditions did not result in signif-icant tocopherol abundance. Nevertheless, total to-copherol content was induced from 1.6 (i.e. 2% ofunstressed wild type) to 9.8 ng cm2 (i.e. 12% ofunstressed and 1% of stressed wild type, respectively)after 4 weeks of either osmotic or salt stress in the HPTtransgenics (Figs. 4D and 5D).
Pigment Analysis of Tobacco Transgenics and Wild-TypePlants Subjected to Sorbitol and Salt Stress
Tocopherols are involved in protecting lipids andthe photosynthetic apparatus against oxidative dam-
age together with the antioxidants ascorbate, glutathi-one, and the photosystem-associated xanthophyllcycle (Porfirova et al., 2002; Sattler et al., 2004; Foyerand Noctor, 2005; Havaux et al., 2005; Kanwischeret al., 2005). Therefore, we studied how chloro-phyll and carotenoid contents were affected in thetransgenic tobacco plants under osmotic and saltstress.
In control and low stress conditions (100 and 200 mM
sorbitol or salt) there were no significant differences inchlorophyll and carotenoid content of wild type andboth transgenic plants (Fig. 6, A, C, E, and F). Incontrast, chromophore contents in wild type andtransgenics were diminished whenever a chloroticphenotype appeared, paralleling the observations de-scribed in Figures 4, A to C and 5, A to C andSupplemental Figures S1 and S2. Compared to thecontrols, chlorophyll and carotenoid contents weresubstantially decreased by about 25% each inHPT:RNAi and g-TMT:RNAi leaves at sodium chlo-ride concentrations of 300 mM and higher, while thewild type showed a 30% reduction in pigments onlyon 400 mM NaCl (Fig. 6).
A decline in chromophore contents was also deter-mined for wild type and HPT transgenics on 400 mM
sorbitol (Fig. 6), while chlorophyll and carotenoidpools in g-TMT:RNAi plants remained constant irre-spective of the sorbitol concentration in the medium(Fig. 6, D and E). Interestingly, the chlorophyll a/b
Table IV. Biomass of wild-type and transgenic lines grown on medium supplemented with differentconcentrations of NaCl
Wild-type (WT), g-TMT:RNAi-55, and HPT:RNAi-28 tobacco plants were treated with the sorbitol andNaCl concentrations indicated. Four weeks after the onset of sorbitol and NaCl stress, biomass productionfrom stressed plants was determined. The depicted data are the means and SDs of four independentmeasurements. *, P , 0.10; **, P , 0.05; and ***, P , 0.01. Significantly different values for identicaltreatments were classified into groups a and b, as indicated by the respective letter.
NaCl Wild Type g-TMT:RNAi-55 HPT:RNAi-28
mM mg Fresh Weight mg Fresh Weight % Wild Type mg Fresh Weight % Wild Type
Table III. Biomass of wild-type and transgenic lines grown on medium supplemented with differentconcentrations of sorbitol
Wild-type (WT), g-TMT:RNAi-55, and HPT:RNAi-28 tobacco plants were treated with the sorbitol andNaCl concentrations indicated. Four weeks after the onset of sorbitol and NaCl stress, biomass productionfrom stressed plants was determined. The depicted data are the means and SDs of four independentmeasurements. **, P , 0.05; ***, P , 0.01. Significantly different values for identical treatments wereclassified into groups a and b, as indicated by the respective letter.
Sorbitol Wild Type g-TMT:RNAi-55 HPT:RNAi-28
mM mg Fresh Weight mg Fresh Weight % Wild Type mg Fresh Weight % Wild Type
ratio decreased from 2.4 on control plates to approx-imately 1 for both wild type and HPT:RNAi trans-genics on 400 mM sorbitol, due to a strong decline inchlorophyll a, but not b (Fig. 6F). This suggests that thephotosystem core and accessory complexes becomemore damaged under osmotic stress than antennacomplexes, which are rich in chlorophyll b.
Altering Tocopherol Composition and Quality by
Silencing HPT and g-TMT Does Not Lead to Strong
Effects on the Ascorbate Pool in Response to Salt and
Osmotic Stress
The ascorbate pool size is reportedly increased inresponse to oxidative stress (Noctor and Foyer, 1998)
Figure 4. Sorbitol-induced osmotic stress in wild-type tobacco and tobacco lines silenced for HPT:RNAi and g-TMT:RNAi. Two-week-old seedlings of Samsun NN, wild type (A), g-TMT:RNAi-55 (B), and HPT:RNAi-28 (C) tobacco seedlings were subjected toMurashige and Skoog medium containing different amounts of sorbitol (0, 200, 300, and 400 mM, from left to right). Pictures offive representative individuals were taken after 4 weeks of sorbitol stress. D, Tocopherol composition in leaves of the wild-typeand transgenic tobacco plants shown in A to C after 4 weeks of sorbitol stress of the concentration indicated above the columns.The depicted results are from a single representative experiment with four replicate samples 6 SD.
Abbasi et al.
1726 Plant Physiol. Vol. 143, 2007 www.plantphysiol.orgon April 1, 2020 - Published by Downloaded from
and there is strong experimental evidence that ascor-bate constitutes the interface of the lipophilic tocopherolsto the hydrophilic antioxidant network represented bythe ascorbate-glutathione cycle (as summarized inMunne-Bosch and Alegre, 2002; Foyer and Noctor,2005). However, it is unclear whether the ascor-
bate pool gets induced in the tocopherol-deficientArabidopsis vte1 mutant, as contrasting results havebeen obtained (Havaux et al., 2005; Kanwischer et al.,2005).
After 4 weeks of treatment, total ascorbate con-tents increased in the wild-type Samsun NN with
Figure 5. Salt-induced stress in HPT:RNAi and g-TMT:RNAi transgenic and wild-type tobacco plants. Samsun NN, wild type (A),g-TMT:RNAi-55 (B), and HPT:RNAi-28 (C) tobacco seedlings were subjected to Murashige and Skoog medium containingdifferent amounts of NaCl. The phenotype of five representative individuals after 4 weeks of salt stress is shown for everytreatment, with 0, 200, 300, and 400 mM NaCl given in columns from left to right. D, Tocopherol composition in leaves of thewild-type and transgenic tobacco plants shown in A to C after 4 weeks of salt stress of the concentration indicated above thecolumns. The depicted results are from a single representative experiment with four replicate samples 6 SD.
Tocopherol Composition and Stress Tolerance
Plant Physiol. Vol. 143, 2007 1727 www.plantphysiol.orgon April 1, 2020 - Published by Downloaded from
increasing amounts of sorbitol and salt (Fig. 6, Band E). Salt treatment effectuated a more pronouncedincrease of the ascorbate pool than sorbitol at iden-tical concentrations (compare Figs. 7, B and E),e.g. 11.7 nmol cm22 ascorbate on 400 mM NaCl
compared to 7.4 nmol cm22 ascorbate on 400 mM
sorbitol.Unexpectedly, the ascorbate pool size in the HPT
transgenics was similar to the wild type even in severestress conditions (Fig. 6, B and E), although the progress
Figure 6. Total tocopherol, ascorbate, dedydroascorbate, chlorophyll, and carotenoid contents in wild-type and transgenictobacco leaves subjected to NaCl and sorbitol stress. Tobacco seedlings of wild-type Samsun NN (left bracket), g-TMT:RNAi-55(middle bracket), and HPT:RNAi-28 (right bracket) were subjected to oxidative and osmotic stress by supplementing 0 to 400 mM
salt (A–C) or sorbitol (D–F) in 100 mM increments to the Murashige and Skoog culture medium. The numbers below the diagramsindicate the following conditions: 1, 0 mM; 2, 100 mM; 3, 200 mM; 4, 300 mM; 5, 400 mM of sodium chloride (A–C) or sorbitol(D–F). A and D, Total carotenoid (black bars) and total tocopherol contents (gray bars). B and E, Dehydroascorbate (black bars)and total ascorbate contents (gray bars). C and F, Chlorophyll contents (chlorophyll a, black bars; chlorophyll b, gray bars) wereassessed in leaves after 4 weeks of stress. Total tocopherol was measured by HPLC (A and D), while chlorophyll (C and F),carotenoid (A and D), total ascorbate, and dehydroascorbate (DHA) were determined spectrophotometrically. The data shownare means from five replicate samples 6 SD from one representative experiment. For ascorbate, significantly different values foridentical treatments were classified into groups a and b, as indicated by the respective letter above the bar.
Abbasi et al.
1728 Plant Physiol. Vol. 143, 2007 www.plantphysiol.orgon April 1, 2020 - Published by Downloaded from
of chlorosis in the transgenics was aggravated in saltstress (Fig. 5; Supplemental Fig. S2).
Silencing of g-TMT resulted in 30% less accumula-tion of ascorbate than in the wild type during salt andosmotic stress of equal or more than 200 mM sodiumchloride or sorbitol, respectively (Fig. 6). Consistently,g-TMT transgenics produced more biomass and ex-hibited less chlorosis on sorbitol than the wild type(Table IV; Fig. 4; Supplemental Fig. S1). In contrast,knockdown of g-TMT resulted in elevated susceptibil-ity toward salt (Table III; Fig. 5; Supplemental Fig. S2),while no substantial effect on the ascorbate pool wasevident in salt-stressed g-TMT transgenics comparedto wild type.
In general, the oxidation state of the ascorbate poolwas comparable for the three genotypes in all testedconditions (Fig. 6, B and E).
Lipid Peroxidation and Membrane Damage in Response
to Oxidative Stress Are Increased by Silencing HPT andDecreased when g-TMT Is Silenced
The observed increase in the pools of the antioxi-dants ascorbate and especially tocopherol under se-vere sorbitol and salt stress (Fig. 6) suggests that wildtype and transgenics suffer from oxidative stress inboth conditions. Malondialdehyde (MDA) occurs as asoluble side product during lipid peroxidation and hasproven to be a reliable marker for oxidative stress(Muller-Moule et al., 2003; Baroli et al., 2004). Excesslipid oxidation will eventually result in membranedamage. The extent of membrane damage can bequantified by ion leakage from the symplast (Rizhskyet al., 2002). Unfortunately, determining ion leakagefrom salt-stressed plants is prone to artifacts due to theelevated ion concentrations, so that only data forsorbitol stress were recorded.
Four weeks after the onset of salt or sorbitol stress,lipid peroxidation in wild type and both transgenicswas correlated in a near-linear manner with increasingsalt and sorbitol stress (Fig. 7). Lipid peroxidation inthe g-TMT transgenics was significantly lowered com-pared to wild type in either salt or sorbitol of equal ormore than 200 mM (Fig. 7). This is in good accordancewith the decreased accumulation of ascorbate observedin g-TMT:RNAi leaves in both salt and sorbitol stress(Fig. 6, B and E). Together, this points toward a di-minished oxidative stress in challenged g-TMT:RNAileaves, even though chlorosis is increased under saltstress compared to wild type (Fig. 5).
However, ion leakage from g-TMT:RNAi and wild-type leaves was similar over a broad range of stressconditions (Table V). On 400 mM sorbitol, ion leakageremained comparable to lower sorbitol concentrationsin g-TMT knockdowns, while the wild type displayeda 2-fold increase compared to the transgenics (Table V).
In contrast to the g-TMT transgenics, MDA contentwas significantly elevated in HPT-silenced plants
Figure 7. Lipid peroxidation in leaves from HPT:RNAi andg-TMT:RNAi and wild-type tobacco plants after 4 weeks of osmoticand salt stress. Wild-type Samsun NN (WT, left bracket, black bars),g-TMT:RNAi-55 (middle bracket, gray bars), and HPT:RNAi-28 (rightbracket, gray bars) tobacco plants were treated with varying amountsbetween 0 and 400 mM of sodium chloride (top section) or sorbitol(bottom section). The numbers below the diagrams indicate thefollowing conditions: 1, 0 mM; 2, 100 mM; 3, 200 mM; 4, 300 mM; 5,400 mM of NaCl (A) or sorbitol (B). Lipid peroxidation was assayedfrom leaf samples after 4 weeks of stress by determining the amount ofMDA. Data shown is the mean and SD of five independent mea-surements.
Table V. Membrane damage is induced by osmotic stress in wild-typeand transgenic tobacco plants
Wild-type (WT), g-TMT:RNAi-55, and HPT:RNAi-28 tobacco plantswere cultivated on increasing concentrations of sorbitol as indicated onthe left. Four weeks after the onset of osmotic stress, membrane damagewas assayed from stressed plants by determining ion leakage from leafdiscs. Data shown are the means and SDs of four independent mea-surements. *, P , 0.10; **, P , 0.05; and ***, P , 0.01. Significantlydifferent values for identical treatments were classified into groups aand b, as indicated by the respective letter.
above 100 mM salt or sorbitol compared to the wildtype (Fig. 7). At 400 mM of either stressor, HPT:RNAitobacco accumulates 1.5-fold more MDA than the wildtype (Fig. 7). In addition, ion leakage was alreadyincreased by 25% in HPT:RNAi leaves in the absenceof sorbitol or sodium chloride supplements on controlplates (Table V). Ion leakage in the HPT transgenicsreached similar values than in the wild type when 400mM sorbitol was supplied.
To evaluate whether a substitution of a- forg-tocopherol in g-TMT:RNAi plants improved the re-silience of the transgenics toward oxidative stress in aless artificial system, we sprayed leaves of soil-grownplants with different amounts of methyl viologen (Fig.8). With the methyl viologen treatments, we aimed tospecifically impose oxidative stress on chloroplasts,mimicking the scenario in chloroplasts of sorbitol-stressed leaves. Ion leakage from wild-type leavescorrelated with the amount of applied methyl violo-gen, rising from 20% in the controls to 85% when 50 mM
methyl viologen had been sprayed (Fig. 8). With theexception of the untreated control leaves, ion leakagein HPT knockdown lines was similar to the wild type.As in the in vitro experiment, untreated HPT:RNAileaves displayed elevated ion leakage in the controlsrelative to the wild type (compare Fig. 8 and Table V).
Also paralleling the observations from sorbitol-stressed plantlets, leaves of g-TMT transgenics exhib-ited a 35% lower ion leakage than the wild type in theconcentration range of 5 to 20 mM methyl viologen (Fig.8). When 50 mM methyl viologen was applied, the ion
leakage from g-TMT:RNAi leaves was not signifi-cantly different from the other two genotypes. How-ever, membrane damage got already saturated in thewild-type and HPT:RNAi plants after spraying 20 mM
methyl viologen (Fig. 8), indicating that 50 mM methylviologen might be too harsh to be overcome by theantioxidant system.
Taken together, our results indicate that a lack oftocopherols in HPT:RNAi decreases membrane integ-rity in the absence of stress and exacerbates lipidoxidation in mild stress conditions already, while lipidperoxidation is diminished and membrane integrity iselevated in the g-TMT transgenics during oxidativestress caused by either sorbitol or methyl viologen.
Silencing of HPT and g-TMT Alters Sugar and AminoAcid Metabolism in Stress Conditions
Severely tocopherol-deficient maize (Zea mays) andpotato plants are reportedly compromised in sugarexport and accumulate soluble sugars and starch,paralleled by the formation of callose plugs at theplasmodesmata (Provencher et al., 2001; Hofius et al.,2004). The Arabidopsis HPT mutant vte2 displayedthis phenomenology only upon chilling (Maeda et al.,2006). Carbon and nitrogen metabolism interdependon each other (see Fritz et al., 2006 for a very recentreport and the refs. cited therein). In an oversimplifiedview, carbon availability triggers nitrogen metabolism(e.g. Henkes et al., 2001; Matt et al., 2001; Gibon et al.,2004), and vice versa, nitrogen availability modulatescarbon flow in the leaf (e.g. Scheible et al., 1997; Geigeret al., 1999). Therefore, we assessed how sugar andamino acid metabolism are affected in the HPT andg-TMT-silenced tobacco plants during salt and sorbitolstress.
Salt and sorbitol stress provoked clearly distinctresponses in sugar metabolism in the wild type, al-though 300 mM of both salt and sorbitol provoked anaccumulation of starch and soluble sugars (Fig. 9, A–J).Compared to unsupplemented plates, starch accumu-lated 3- and 20-fold in wild type on 300 mM sorbitol andsalt, respectively, with salt being about 5 times morepotent than sorbitol (compare Fig. 9, D–J); less than 300mM of either stressor did not result in a considerableaccumulation of starch. The accumulation of solublesugars in the wild type was found to correlate directlywith salt and sorbitol concentrations. The monosaccha-rides Glc and Fru accounted for the increase in totalsoluble sugar content with increasing sorbitol stress(Fig. 9, G–I), while Suc accumulation correlated withincreasing salt stress (Fig. 9, A–C). Glc and Fru contentswere finally increased about 20-fold on 300 mM sorbitolcompared to the controls, while Suc was found to ac-cumulate more than 10-fold on 300 mM sodium chloride.
The g-TMT transgenics displayed comparablesteady-state pools of soluble sugars as the wild type,with the remarkable exception that Glc and Fru accu-mulation were 2-fold lower in sorbitol stress of 200 mM
or higher compared to wild type (Fig. 9, G and H).
Figure 8. Ion leakage in wild-type and transgenic tobacco leaves aftermethyl viologen treatment. Fully expanded source leaves of 8-week-oldtobacco plants of wild-type Samsun NN (left bracket, black bars),g-TMT:RNAi-55 (middle bracket, gray bars), and HPT:RNAi-28 (rightbracket, gray bars) were sprayed with 1 mL of the indicated methylviologen solution on 2 d prior to the harvest of leaf disc samples. Thenumbers given behind the genotype acronym at the bottom indicate thefollowing conditions: 0, 0 mM; 5, 5 mM; 20, 20 mM; 50, 50 mM methylviologen. Ion leakage is given in percent of maximum leakage afterboiling. The data shown are means from five replicate samples 6 SD
from one representative experiment.
Abbasi et al.
1730 Plant Physiol. Vol. 143, 2007 www.plantphysiol.orgon April 1, 2020 - Published by Downloaded from
Abrogating HPT activity in HPT:RNAi abolished thecorrelation of sugar or starch accumulation with thedegree of salt or sorbitol stress. Sugar and starch poolsin the HPT transgenics did not respond to salt stress atall (Fig. 9D); soluble sugar contents in HPT:RNAileaves were already substantially elevated compared
to wild type or g-TMT:RNAi lines in low NaCl (Fig. 9,A–C), with starch contents in the HPT transgenicsbeing as high as in wild type or g-TMT transgenicsgrown on 300 mM sorbitol (Fig. 9, D and J). In sorbitolstress, up to 200 mM sorbitol, starch, and soluble sugarcontents in HPT transgenics remained comparable to
Figure 9. Carbohydrate and amino acidcontents in leaves from HPT:RNAi andg-TMT:RNAi and wild-type tobacco plantsafter 4 weeks of salt and sorbitol stress.Wild-type Samsun NN (WT, left bracket),g-TMT:RNAi-55 (middle bracket), andHPT:RNAi-28 (right bracket) tobaccoplants were treated with varying amountsbetween 0 and 300 mM of sodium chloride(left column) or sorbitol (right column).The numbers below the diagrams indicatethe following conditions: 1, 0 mM; 2, 100mM; 3, 200 mM; 4, 300 mM of either NaCl(left) or sorbitol (right). The contents of thesoluble sugars Glc (A and G), Fru (B andH), and Suc (C and I), of starch (D and J),the compatible solute Pro (E and K), andtotal amino acids (F and L) were assayedfrom leaf samples after 4 weeks of stress.The depicted data are mean and SD of fiveindependent measurements from one rep-resentative experiment.
Tocopherol Composition and Stress Tolerance
Plant Physiol. Vol. 143, 2007 1731 www.plantphysiol.orgon April 1, 2020 - Published by Downloaded from
untreated wild type on control plates (Fig. 9, G–J). Incontrast, HPT:RNAi leaves accumulated 30% moreSuc, 75% more Glc, and 100% more Fru than wild typeor g-TMT on 300 mM sorbitol, while starch contentsrose to similar levels as in the two other genotypes. Incontrast to maize sxd1, Arabidopsis vte2, and TC-silenced potato plants, we failed to observe a sugarexport block in the tocopherol depleted HPT trans-genics in any of the tested conditions, which wouldinvolve much stronger accumulation of starch andSuc. Attempts to probe plasmodesmatal obstruction inHPT transgenics remained inconclusive (data notshown).
Characteristic changes in the composition of the freeamino acid pool coincided with the accumulation ofstarch or soluble sugars (Supplemental Fig. S3). Thecompatible solute Pro accumulated 15- and 20-fold on300 mM sorbitol and NaCl compared to the unsupple-mented controls in all genotypes (Fig. 9, K and E,respectively), as did its precursor Glu in salt stress (seeSupplemental Fig. S3). Proline accounted for 40% to50% of the free amino acid pool when 300 mM of eithersalt or sorbitol were supplied (compare values in Fig.9, E and F, or Fig. 9, K and L). It is noteworthy that a9-fold accumulation of Pro could be observed when aslittle as 100 mM sodium chloride were supplied, while200 mM sorbitol led to a 4-fold Pro accumulation only(compare Fig. 9, E and K).
Apart from this indicative amino acid, we also ob-served correlations between the accumulation of Asn,Gln, His, most branched-chain, aromatic amino acidsand hexose contents with increasing sorbitol stress(Supplemental Fig. S3). Almost the same subset ofamino acids accumulated 4- to 9-fold in the wild typeon 400 mM NaCl, while the transgenics did not accu-mulate these amino acids (Supplemental Fig. S3). Ingeneral, the transgenics did not show pronounced dif-ferences in free amino acid contents to the wild type insorbitol stress, with g-TMT:RNAi plants being indis-tinguishable from the wild type (Supplemental Fig. S3).
DISCUSSION
Salt and Sorbitol Stress Have Different Targets inCellular Metabolism and Induce CharacteristicPhysiological Responses
The rationale behind employing salt and sorbitolstress in our study was to elicit specific stress patterns.Both severe salt and sorbitol stress finally result inoxidative stress caused by hyperosmolarity (as re-viewed by Wang et al., 2003). However, salt andsorbitol stress will primarily target different subcellu-lar compartments. Excess symplastic salt concentra-tions endanger protein integrity and therefore, have tobe sequestered in the vacuole or extruded from thecells, concomitantly disturbing pH and ion homeosta-sis across the tonoplast and the plasma membrane(Serrano and Rodriguez-Navarro, 2001). Increased salt
tolerance was achieved by generating transgenicplants overexpressing the plasma membrane Na1/H1 antiporter SOS1 (Shi et al., 2003) or the vacuolarNa1/H1 antiporter AtNHX1 (Apse et al., 1999;Zhang et al., 2001; Zhang and Blumwald, 2001), dem-onstrating that the sequestration capacity for sodiumis limiting in high salinity. Our data on carbohydrateand amino acid metabolism provide evidence that theassumed mode of sodium chloride toxicity also holdstrue for our experimental setup: in transgenics and thewild type, we observed a 9-fold accumulation of Pro(and its precursor Glu) in 100 mM salt stress (Fig. 9;Supplemental Fig. S3), while Pro biosynthesis, evi-denced by steady-state contents of itself and its pre-cursor Glu, was not induced in 100 mM sorbitol stress,indicating that the constraint to protect macromole-cules with the accumulation of compatible solutes ismore specific for salinity. Furthermore, a high salinityof 400 mM NaCl led to a dramatic accumulation ofstarch and Suc as well as Asp and pyruvate-derivedamino acids in the wild type (Fig. 9; Supplemental Fig.S3), indicating a severely disturbed central cellularmetabolism. Likewise, this was not observed in sorbi-tol stress. The correlation of Suc and starch accumu-lation with increasing salt stress also argues for aprogressing sugar export block from source leaves inresponse to poisoning by salinity, as commonly ob-served under various stress conditions (e.g. Voll et al.,2003; Sam et al., 2004; Maeda et al., 2006).
In contrast to sodium chloride, sorbitol is a compat-ible solute, which does not interfere with enzymaticactivities in plasmatic compartments. Sorbitol wouldmore specifically cause desiccation of the cells in addi-tion to hyperosmotic stress. In the second line, reducedwater availability would finally increase oxidativestress in the chloroplasts by stimulating the generationof ROS and, concomitantly, lipid peroxy radicals inilluminated thylakoids. Paralleling the increased tol-erance toward sorbitol stress, the g-TMT transgenicswere less susceptible than wild type toward the elec-tron donor methyl viologen that specifically inducesoxidative stress in the thylakoids by stimulating thegeneration of superoxide anions at PSI reaction centers(Fig. 8). Taken together, this indicates that g-tocopherolin g-TMT:RNAi leaves is capable of mitigating ROS-induced stress in chloroplasts better than a-tocopherolin wild-type leaves. Increasing ROS quenching capac-ity by, for example, overexpression of superoxide dis-mutase (Gupta et al., 1993) or glutathione S-transferaseand glutathione peroxidase (Roxas et al., 1997), hasalready been shown to possess the potential to allevi-ate oxidative stress.
In support of our model of sorbitol action, we notedan enhanced hexose accumulation in wild type andg-TMT:RNAi under sorbitol stress compared to saltstress (Fig. 9), indicating a rearrangement of carbohy-drates in favor of osmotically competent monosaccha-rides like Glc and Fru. The compatible solute Proaccumulated in elevated sorbitol stress of more than300 mM only (Fig. 9). As concluded from the metabolic
Abbasi et al.
1732 Plant Physiol. Vol. 143, 2007 www.plantphysiol.orgon April 1, 2020 - Published by Downloaded from
analysis, the specific accumulation of soluble sugarsindicates that desiccation predominates over sorbitoltoxicity and oxidative stress in sorbitol treatments. Anaccumulation of hexoses during drought (i.e. desicca-tion) stress has already been described for tobacco (e.g.Karakas et al., 1997) and was also observed for barley(Hordeum vulgare) leaves (Villadsen et al., 2005) andsweet potato (Ipomoea batatas) suspension cells (Wanget al., 1999) subjected to drought and osmotic stress.Mostly, drought stress was accompanied by a decreasein starch biosynthesis in favor of soluble sugars(Karakas et al., 1997; Villadsen et al., 2005).
Tocopherol Depletion by Silencing HPT Decreases Stress
Tolerance in Transgenic Tobacco
HPTcatalyzes the committed prenyl transferase stepin tocopherol biosynthesis at the branch point oftocopherol and tocotrienol biosynthesis (Fig. 1). Byposttranscriptional silencing of HPT activity, we couldobtain transgenic tobacco with less than 2% of wild-type foliar tocopherol content. Although tobacco seedsare capable to synthesize tocotrienols (Falk et al., 2003),wild-type tobacco leaves are devoid of tocotrienols(Cahoon et al., 2003; Rippert et al., 2004). Even insevere stress, we could only detect traces of tocochro-manols in HPT:RNAi leaves (data not shown), so thatit can be ruled out that the depletion of tocopherols inHPT knockdown tobacco is functionally compensatedby tocotrienols.
As expected, HPT:RNAi leaves displayed an in-creased sensitivity toward salt and sorbitol stress,although membrane damage and lipid peroxidationwere already elevated in unstressed HPT transgenics(Table V). Sorbitol stress, despite some minor changesin carbon and nitrogen metabolism, rather specificallyaffected lipid peroxidation in HPT transgenics com-pared to the two other genotypes (Fig. 7). In contrast,the extent of lipid peroxidation, pigment breakdown,and the changes in carbon and nitrogen metabolismwere aggravated in the HPT transgenics compared tothe wild type in salt stress (Tables III–V; Figs. 6, 7, and9; Supplemental Fig. S3). Aside from a direct impact ofelevated lipid peroxidation, the increased susceptibil-ity of HPT:RNAi tobacco toward oxidative stresscaused by salt and (to a lesser extent) sorbitol mightbe facilitated by enhanced ROS signaling in the trans-genics early in the acclimation process. Transcriptomeanalysis of unchallenged Arabidopsis vte2 seedlingsrevealed that the increased lipid peroxidation in thisHPT mutant triggered ROS-regulated genes, as duringpathogen challenge or oxidative stress-like ozone (Sattleret al., 2006). Interestingly, the induced genes in vte2 seed-lings were not positively regulated by methyl jasmonate(Sattler et al., 2006), the synthesis of which is elevatedwhen lipid peroxidation occurs. In contrast, ArabidopsisTC mutants (vte1) challenged with low temperatureand high light exhibited a transient increase in methyljasmonate (Munne-Bosch et al., 2007). However, thediverging results obtained for vte1 and vte2 could be
explained by the different age of the examined plants orby the accumulation of the tocopherol precursor 2,3-dimethyl-5-phytyl-1,4-benzoquinone in vte1 (Sattleret al., 2003), which is absent in vte2.
Despite the evidence that Arabidopsis vte2 mutantsexhibit an enhanced ROS response (Sattler et al., 2006),the soluble antioxidants ascorbate and glutathionewere not substantially induced in the Arabidopsistocopherol-deficient vte1 and vte2 mutants (Havauxet al., 2005; Kanwischer et al., 2005). We also found thatthe ascorbate pool was not significantly increased intocopherol-deficient HPT knockdown tobacco com-pared to wild type (Fig. 6). But as tocopherols connectthe lipophilic xanthophyll cycle to the soluble antiox-idant network governed by the ascorbate-glutathionecycle (Foyer and Noctor, 2003), high light sensitivity ofvte1 was greatly enhanced when other antioxidant sys-tems like the zeaxanthin or the glutathione-ascorbatecycle were compromised concomitantly (Havaux et al.,2005; Kanwischer et al., 2005). This demonstrates com-pensatory capacity in the foliar antioxidant network,which might also account for the comparable responseof the ascorbate pool in HPT transgenics and wild typeduring acclimation to the imposed oxidative and os-motic stress.
The Substitution of a- for g-Tocopherol in
g-TMT-Silenced Tobacco Increases Osmotolerance
We achieved an almost quantitative substitution ofa- for g-tocopherol by silencing g-TMT in tobacco withonly 5% of foliar a-tocopherol left compared to wildtype. To date, mutants with decreased g-TMT activityhave been isolated from Arabidopsis (vte4), sunflower(Helianthus annuus), and Synechocystis (slr0089) and allof them accumulate g-tocopherol instead of a-tocopherol(Shintani and DellaPenna, 1998; Bergmuller et al.,2003; Hass et al., 2006). Interestingly, the total tocoph-erol content in sunflower and Arabidopsis mutantleaves remained similar to the corresponding wildtype (Bergmuller et al., 2003; Hass et al., 2006), whilethe Synechocystis mutant lacked a-tocopherol with-out any compensation in g-tocopherol (Shintani andDellaPenna, 1998; Sakuragi et al., 2006). Among theseg-TMT mutants, only the Arabidopsis vte4 mutant wastested for abiotic stress tolerance and did not exhibitan altered stress response toward heat, cold, and highlight compared to the wild type (Bergmuller et al.,2003). In contrast, we imposed sorbitol, methyl vio-logen, and sodium chloride to probe whether replac-ing a-tocopherol with g-tocopherol affects oxidativestress tolerance in tobacco.
Silencing g-TMT in tobacco, thereby exchanginga- for g-tocopherol, resulted in an elevated suscepti-bility toward salt, but a diminished susceptibilitytoward osmotic stress and methyl viologen comparedto the wild type (Tables III and IV; Figs. 4, 5, and 8).Ascorbate pool size and lipid peroxidation were re-duced in the g-TMT transgenics compared to the wildtype in salt and sorbitol treatments (Fig. 7), indicating
Tocopherol Composition and Stress Tolerance
Plant Physiol. Vol. 143, 2007 1733 www.plantphysiol.orgon April 1, 2020 - Published by Downloaded from
less oxidative stress in the transgenics in both stressconditions. As judged from carotenoid and chloro-phyll contents (Fig. 6), the photosynthetic apparatus ing-TMT transgenics remained intact in sorbitol, but notin NaCl stress. While the chlorophyll a to b ratiocollapsed in wild-type leaves on 400 mM sorbitol,leaves of g-TMT:RNAi plants retained the same ratioon 400 mM sorbitol as on control plates.
Our results shed new light on the in vivo function ofg-tocopherol, allowing two major conclusions. First,g-tocopherol is more potent than a-tocopherol in con-ferring desiccation tolerance in vivo. This is presum-ably mediated by the higher in vivo lipid antioxidantactivity of g-tocopherol (that is present in the g-TMTtransgenics) compared to a-tocopherol (abundant inwild-type leaves). Consequently, lipid peroxidation isdiminished, membrane damage is decreased, andpigment loss is reduced in the g-TMT transgenicscompared to wild type when oxidative stress arisesduring desiccation. This assumption is supported bytwo observations: (1) g-TMT:RNAi plants exhibitedless membrane damage after targeted oxidative stresswas imposed on thylakoids (and the PUFAs containedtherein) by stimulating ROS production with methylviologen (Fig. 8); and (2) by the fact that g-tocopherolis the naturally predominant tocopherol derivativein most oil-storing seeds (Fig. 1) and that loss ofg-tocopherol was shown to result in elevated PUFAoxidation and diminished seed longevity in Arabi-dopsis (Sattler et al., 2004). Likewise, we also observeda substantially decreased germination efficiency in thetocopherol depleted HPT transgenics (data notshown). Thus, the presence of g-tocopherol is notonly pivotal for seed desiccation tolerance, but canalso increase desiccation tolerance in leaves.
In oil-storing seeds, storage lipids are located in oilbodies that reside in the cytosol. It was reported thattocochromanols are tightly associated with oil bodiesfrom sunflower and oat (Avena sativa) seeds (Fisk et al.,2006; White et al., 2006). As excess lipid peroxidationdoes not occur in, for example, Arabidopsis wild typebut in g-tocopherol-deficient seeds (Sattler et al., 2004),it can be assumed that tocochromanol contents in oilbodies are sufficient to efficiently protect seed PUFAsfrom oxidation. For leaves, it has been demonstratedthat tocopherols are abundant in plastoglobules andevidence is emerging that tocopherol biosynthesis alsooccurs in plastoglobules (Austin et al., 2006; Vidi et al.,2006). Plastoglobules are thylakoid protrusions com-posed of lipophilic constituents like triacylglycerols,quinones, chlorophyll, carotenoids, and also monoga-lactosyldiacylglycerol and digalactosyldiacylglycerol(Ghosh et al., 1994; Austin et al., 2006, and refs. therein)that increase in number during senescence and inoxidative stress (Steinmuller and Tevini, 1985; Munne-Bosch and Alegre, 2004) when chlorophyll turnoveris high. Free phytol from chlorophyll breakdownmight directly be salvaged for tocopherol biosynthesis(Ischebeck et al., 2006; Dormann, 2007), which is re-quired for antioxidant protection in these conditions.
As a second conclusion, g-tocopherol cannot sub-stitute a-tocopherol to ensure a better survival in saltstress, although markers for oxidative stress weredecreased in the g-TMT transgenics compared to wildtype. Consequently, a-tocopherol appears to better me-diate protection of macromolecules from denaturationby salt than g-tocopherol. Alternatively, g-tocopherolmight not be able to influence cellular signaling likea-tocopherol does.
In this study, we have discovered that g-tocopherolexerts a specific function in osmoprotection in vivo,providing evidence that a- and g-tocopherol may notbe functionally equivalent in many respects. Our re-sults raise a wealth of new questions on the roles ofindividual tocopherol derivatives. Future aims willencompass attempts to discover the specific in vivofunctions of a- and g-tocopherol in ROS-mediated sig-naling and their specific antioxidant activities, usingthe tobacco lines silenced for HPT and g-TMT as a tool.
MATERIALS AND METHODS
Plant Material and Growth Conditions
Tobacco (Nicotiana tabacum cv Samsun NN) plants were obtained from
Vereinigte Saatzuchten and maintained in tissue culture under a 16-h-light/
8-h-dark period (50 mE m22 s21 light, 21�C) at 50% relative humidity on
Murashige and Skoog medium (Sigma) containing 2% (w/v) Suc.
Seeds of transgenic plants were sown on Murashige and Skoog medium
containing kanamycin and after 14 d, the resistant seedlings were either
transferred to Murashige and Skoog medium containing NaCl and sorbitol of
the concentrations indicated in the figure legends (0, 100, 200, 300, 400, and 500
mM) or the kanamycin-resistant seedlings were transferred on soil and further
cultivated in the greenhouse in a 16-h/8-h light/dark cycle at 24�C, a relative
humidity of 50%, and a photon flux density of 150 mM m22 s21.
For methyl viologen treatments, the above selection procedure was mod-
ified. Seeds of transgenic lines were sown on Murashige and Skoog medium
without kanamycin and 14 d after the transfer to soil, one mature leaf from
each plant was harvested to determine its tocopherol content. Plantlets
matching the tocopherol profile of the parental line were propagated and
methyl viologen treatments were started when the plants were 8 weeks old.
Two hours after the beginning of the light phase, 1 mL of 0 to 50 mM methyl
viologen spray was applied on one fully expanded source leaf per plant on
two consecutive days. Leaf discs from sprayed leaves were harvested 6 h after
the second methyl viologen application and ion leakage was determined as
described below.
Cloning of Partial HPT and g-TMT cDNA Sequencesfrom Potato, RNAi Vector Construction,
and Tobacco Transformation
To obtain heterologous HPT and g-TMT sequences for RNAi constructs, we
identified partial expressed sequence tag (EST) clones for potato (Solanum
tuberosum) HPT (accession no. BI919738) and potato g-TMT (accession no.
BQ116842) by homology search with tBLASTn (Altschul et al., 1990). A 650-bp
fragment of the StHPT open reading frame containing nucleotides 51 to 701 of
the potato HPT EST clone (accession no. BI919738) was amplified by reverse
transcription-PCR using primers AA7 (5#-ggatccCTGATTTAGAAAT-
CAAAAATGGAATC-3#) and AA8 (5#-gtcgacCCAACGATCCATCCAAGC-
CAAAAAC-3#), which introduced terminal BamHI and SalI recognition
sites, respectively, into the sequence. To amplify a 625-bp fragment comprising
nucleotides 31 to 659 of the Stg-TMT open reading frame from the partial
potato EST clone (accession no. BQ116842) by reverse transcription-PCR,
primers AA9 (5#-ggatccGTTAAGAATCCTCTGCGAACAATAAA-3#) and AA10
(5#-gtcgacGGATTAGGGGATAAGGTTTCATTTCA-3#) were applied, also in-
troducing terminal BamHI and SalI recognition sites.
Abbasi et al.
1734 Plant Physiol. Vol. 143, 2007 www.plantphysiol.orgon April 1, 2020 - Published by Downloaded from