2010 engineering tocopherol biosynthetic pathway in arabidopsis leaves
Plant Science 178 (2010) 312–320 Contents lists available at ScienceDirect Plant Science journal homepage: www.elsevier.com/locate/plantsciEngineering tocopherol biosynthetic pathway in Arabidopsis leaves and itseffect on antioxidant metabolismYin Li a, Yin Zhou a, Zinan Wang a, Xiaofen Sun a, Kexuan Tang a,b,*a State Key Laboratory of Genetic Engineering, School of Life Sciences, Fudan-SJTU-Nottingham Plant Biotechnology R&D Center, Fudan University, Shanghai 200433, PR Chinab Plant Biotechnology Research Center, School of Agriculture and Biology, Fudan-SJTU-Nottingham Plant Biotechnology R&D Center,Shanghai Jiao Tong University, Shanghai 200030, PR ChinaA R T I C L E I N F O A B S T R A C TArticle history: With genetic manipulation, ﬁve genes (HPPD, VTE2, VTE3, VTE1, and VTE4), which encode enzymesReceived 24 November 2009 involved in tocopherol biosynthesis, were over-expressed in model plant Arabidopsis thaliana, eitherReceived in revised form 17 January 2010 alone or in couple combinations (VTE2 + VTE4 and VTE3 + VTE4), to value and compare the roles ofAccepted 19 January 2010 enzymes played in tocopherol biosynthetic pathway under the same genetic background. Our resultsAvailable online 28 January 2010 suggested that, elevated expression level of biosynthetic pathway gene affected either total tocopherol content or composition, it is recommended to choose two or more enzymes with different functions forKeywords: genetic manipulation. It was also found that metabolic engineering of tocopherol biosynthetic pathwayArabidopsis thalianaBiosynthetic pathway affected endogenous ascorbate and glutathione pools in leaves. Further study suggested that expressionHalliwell–Asada cycle levels of genes encoding enzymes of Halliwell–Asada cycle were up-regulated, such as APX, DHAR andOver-expression MDAR. These ﬁndings provide hints on the relationship of lipid-soluble antioxidant vitamin E and water-Tocopherol soluble antioxidants vitamin C and glutathione, which will help to perfect theory in plant physiology andVitamin E give practical instruction for metabolic engineering. ß 2010 Elsevier Ireland Ltd. All rights reserved.1. Introduction adults (http://ods.od.nih.gov/factsheets/vitamine.asp). Higher dai- ly vitamin E doses have been needed for cancer reduction, immune Vitamin E is an essential nutrient of daily diet for humans and response, and cardiovascular beneﬁts . Among the family, a-animals. Naturally occurring vitamin E exists in eight chemical tocopherol is believed to have the highest vitamin E activity toforms (a-, b-, g-, and d-tocopherol and a-, b-, g-, and d- meet human requirements, and it is preferentially retained andtocotrienol) that have varying levels of biological activity. The distributed throughout the body . Naturally synthesized a-nutritional values of vitamin E were afﬁrmed in 1922 . The tocopherol, which is a single (R, R, R) stereoisomer, has moreNational Institutes of Health (NIH) currently suggests a recom- activity than chemically synthesized a-tocopherol .mended daily allowance (RDA) of 15–19 mg a-tocopherol for Tocopherols can be synthesized only in photoautotrophy organisms, including plants and other oxygenic, photosynthetic organisms. Although the tocopherol biosynthetic pathway hadAbbreviations: APX, ascorbic acid peroxidase; AsA, ascorbic acid (vitamin C); DHA, been elucidated from 1979 , the genetic analysis of the pathwaydehydroascorbic acid; DHAR, dehydroascorbic acid reductase; GDPME, GDP-D- and key enzymes had only commenced since 1990s, with themannose-3, 5-epimerase; GDPMPPase, GDP-D-mannose pyrophosphorylase; GSH, approaches of genetic and genomics-based methodologies in theglutathione; GSSG, oxidized glutathione; HGA, homogentisic acid; HPLC, highperformance liquid chromatography; HPPD, p-hydroxyphenylpyruvic acid dioxy- model organisms Arabidopsis thaliana and Synechocystis sp.genase; L-GalDH, L-galactose dehydrogenase; L-GalPPase, L-galactose 1-P phospha- PCC6803.tase; L-GLDH, L-galactono-g-lactone dehydrogenase; MDAR, monodehydroascorbic Tocopherol biosynthesis mainly takes place in plastids of higheracid; MEP, methylerythritol phosphate; NC, non-transgenic control; PDP, phytyl- plants. The tocopherol biosynthetic pathway utilizes two com-diphosphate; PQ8, plastochromanol-8; PUFAs, protect polyunsaturated fatty acidchains; RDA, recommended daily allowance; ROS, reactive oxygen species; VTC2, pounds from different metabolic pathways as precursors, whichGDP-L-galactose phosphorylase; VTE1, tocopherol cyclase; VTE2, homogentisate include homogentisic acid (HGA), derived from cytosolic shikimatephytyltransferase; VTE3, 2-methyl-6-phytylplastoquinol methyltransferases; metabolic pathway for head group and phytyldiphosphate (PDP)VTE4, g-tocopherol methyltransferase.  from the plastidic methylerythritol phosphate (MEP) pathway * Corresponding author at: State Key Laboratory of Genetic Engineering, School of for tail synthesis [7,8] (Fig. 1). There are at least ﬁve enzymesLife Sciences, Fudan-SJTU-Nottingham Plant Biotechnology R&D Center, FudanUniversity, 220 Handan Road, Shanghai 200433, PR China. involved in the biosynthesis of tocopherols, excluding the bypass E-mail addresses: firstname.lastname@example.org, email@example.com (K. Tang). pathway of phytyl-tail synthesis and utilization (Table 1). HGA is0168-9452/$ – see front matter ß 2010 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.plantsci.2010.01.004
Y. Li et al. / Plant Science 178 (2010) 312–320 313 this work, genes encoding these enzymes were cloned from the model plant A. thaliana and constitutively over-expressed, alone or in combination, in order to achieve this assessment. In metabolic engineering, when target nutritional product is increasing, other related nutritional products in the same bioreactor might be affected, in some cases they would decrease. Many studies reported the antioxidant properties of tocopherols, such as photo protection and reduction of lipid peroxidation by reducing lipid peroxyl radicals to their corresponding hydroper- oxides [16,17,18]. As lipid-soluble antioxidant, tocopherols locate mainly on membranes of many cellular compartments, and will be oxidized to tocopherol radical . There are two important water-soluble antioxidants in plant cells—ascorbic acid (vitamin C, AsA) and glutathione (GSH), and they can scavenge reactive oxygen species by Halliwell–Asada cycle during normal metabolism and particularly during stress . Especially vitamin C is not only an important antioxidant in plant physiology, but also a metabolicFig. 1. Simpliﬁed tocopherol biosynthetic pathway from shikimate and MEP product with important nutritional and physiological values forpathways. humans and animals. Former works indicated there might be relationship existing among tocopherol, AsA and glutathioneproduced from the tyrosine aromatic amino acid catabolite p- contents [21,22]. It was reported  that deﬁciency in onehydroxyphenylpyruvate (HPP) by the cytosolic enzyme 4-hydro- antioxidant in tocopherol, AsA or glutathione led to increasedxyphenylpyruvate dioxygenase (HPPD) . Condensation of HGA oxidative stress and the concomitant increase in alternativeand PDP is catalyzed by homogentisate phytyltransferases (VTE2) antioxidants. In some other cases, such as sunﬂower cell lines,. The product of this reaction, 2-methyl-6-phytylbenzoquinol high content of tocopherols leads to higher content levels of(MPBQ), is the ﬁrst phytylquinol intermediate in the pathway and ascorbate and glutathione pools . However, in the study whichcan be methylated to 2,3-dimethyl-6-phytyl-1, 4-benzoquinol observed the plant responses to oxidant stress in the presence of(DMPBQ) by MPBQ methyltransferase (VTE3) . Both MPBQ and Cu or Cd , the correlation between high tocopherol and lowDMPBQ are substrates for tocopherol cyclase (VTE1) to yield the ascorbate/glutathione levels was not seen.ﬁrst tocopherols of the pathway, d-tocopherol and g-tocopherol, In this study AsA and glutathione contents in transgenic linesrespectively . Both d- and g-tocopherol can be methylated by over-expressing tocopherol biosynthetic pathway genes wereg-tocopherol methyltransferase (VTE4) to yield b- and a- analyzed in order to assess whether accumulation of vitamin Etocopherol, respectively . affected other antioxidants in plant. Furthermore, some studies As a member of plant secondary metabolites, vitamin E has were done in order to explain the change of vitamin C andvarious biological and pharmaceutical functions to humans as well glutathione in transgenic lines.as to plants. Crops and vegetables are the best source for naturalvitamin E. Nevertheless, vitamin E is of low content, and the 2. Materials and methodscomposition of the eight forms needs to optimize. Recently,metabolic engineering has been widely applied in order to achieve 2.1. Plant materials and growth conditionshigher yields of speciﬁc metabolites. The in-depth understandingof biosynthetic pathways, along with the increasing number of Seeds of wild-type Arabidopsis (Columbia ecotype) werecloned genes involved in biosynthesis, enable the exploration of sterilized with chlorox and spread on Murashige and Skoogmetabolic engineering as a potential effective approach to increase (1962) plates, then treated at 4 8C for 3–5 days and induced forthe yield of speciﬁc metabolites by enhancing rate-limiting steps germination in 24 h-white-light for 6 days. Then the seedlingsor by blocking competitive pathways . Increasing vitamin E were transferred into soil at 20 8C under 16-h photoperiod of lightcontent and a-tocopherol composition in vegetables and crops has at 120 mmol mÀ2 sÀ1.been an important aim for metabolic engineering. Strategies as toover-expressing different genes involved in the pathway have been 2.2. cDNA generation and vector constructionemployed. Although signiﬁcant work has been done , it is stillhard to assess the relative importance of each enzyme in For construction of plant expression vectors, myc tag was usedtocopherol biosynthetic pathway, due to different genetic back- as screening labels and sub-cloned into XhoI and PstI sites ofground and various manipulation used in former studies. Valuation pBluescript SK+ vector (pBS; Stratagene) to form the vectorof the contribution of different enzymes under the same genetic pBSmyc (a gift from Prof. Hongquan Yang, SIPPE, CAS). Total RNAbackground will be essential to provide effective strategies for was isolated from leaves of Arabidopsis thaliana (Columbialarge-scale commercial production of biosynthetic tocopherol. In ecotype) by using TRIzol reagent (GIBCO/BRL). The cDNAs ofTable 1Enzymes, basic functions, loci and genes encoding tocopherol biosynthetic enzymes in Arabidopsis thaliana. Pathway enzyme Basic function Arabidopsis Gene Locus p-Hydroxyphenylpyruvic acid dioxygenase (HPPD) Head group synthesis PDS1 At1g06570 Homogentisate prenyltransferase (HPT) Prenylation of HGA VTE2 At2g18950 2-Methyl-6-phytylbenzoquinone methyltransferase (MPBQ MT) Methylation of MPBQ and MGGBQ VTE3 At3g63410 Tocopherol cyclase (TC) Cyclization VTE1 At4g32770 g-Tocopherol methyltransferase (g-TMT) Methylation of d- and g-tocopherol VTE4 At1g64970
314 Y. Li et al. / Plant Science 178 (2010) 312–320Fig. 2. Schematic representation of the transformation vector construct used in this study. (A) Schematic representation of the single-gene transformation vector. (B)Schematic representation of dual-gene transformation vector. LB, left border; RB, right border; 35S, cauliﬂower mosaic virus 35S promoter; 35S polyA, cauliﬂower mosaicvirus 35S polyA terminator; Nos, nopaline synthase gene terminator.At-HPPD, VTE2, VTE3, VTE1, and VTE4 (Genbank accession Nos. screened on Murashige–Skoog (MS) basal medium supplementedAY072329, AY089963, AB054257, NM119430 and AF104220) were with 50 mg mLÀ1 hygromycin. More than ﬁfty independent linesprepared by using TaKaRa One Step RT-PCR kit (Code No.: expressing transgene were transferred to soil and screened with 1%DRR024A, TaKaRa, Japan). The ampliﬁed cDNA fragments of At- herbicide (glufosinate-ammonium). Screening of T1 and T2 seedsHPPD, VTE2, VTE3, and VTE1 with an introduced HindIII site at the were performed with herbicide in order to obtain transgenic linesstart codon (ATG) and an introduced BamHI site before the stop which accorded with Mendel’s Law. Polymerase chain reactioncodon were digested, and sub-cloned into HindIII–BamHI site of (PCR) and protein gel blot analysis were also performed onvector pBSmyc to obtain the expression cassette with recombina- independent transgenic lines (random chosen) of each generation,tion sequence of target gene and myc. The ampliﬁed fragment of in order to further conﬁrm and check the homozygosis and geneticVTE4 with an introduced XhoI site at the start codon and an stability. Homozygous seeds could be obtained in T4 generation.introduced BamHI site before the stop codon was digested, and The screening of co-expression transgenic plants was performedsub-cloned into XhoI–BamHI site of pBSmyc. The reading frame of with PCR and western blot analysis, while herbicide could not bepBS-target gene–myc was conﬁrmed to be correct by sequencing used because anti-herbicide gene was not contained in the binaryfrom both strands. The recombinant sequences of target gene–myc vectors.were introduced into pET28a vector and over-expressed in E. coli inorder to detect biological activity of fusing proteins in vitro. The 2.4. Molecular analysis of transgenic linesrecombinant sequences of target gene–myc were sub-cloned intothe botany expression vector PHB (a gift from Prof. Hongquan Genomic DNA was isolated from the transgenic plants and NCYang, SIPPE, CAS) , located downstream of double 35S CaMV plant with a CTAB method . The presence of the transgene waspromoters and upstream of rbcS polyA terminator (Fig. 2A). detected by PCR. In order to explore the function of two enzymes in plant Protein gel blot analysis was performed as described previouslytocopherol biosynthetic pathway, three of them were chosen for with minor modiﬁcations . Fifty mg of total protein, deter-dual expression vector construction. Two co-expression combina- mined by using the DC Protein Assay Kit (Bio-Rad, Hercules, CA),tions were designed: VTE2 + VTE4 and VTE3 + VTE4. The target was fractionated on 12% SDS-PAGE mini-gel and blotted onto agenes were also recombinant with myc. In order to put two genes in nitrocellulose membrane (PerkinElmer, 0.45 mm). The blots wereone vector, the expression cassette of vector pBI121 (Clontech) was probed with the primary antibody c-Myc (9E10) (sc-40, mousedigested with HindIII and EcoRI, and then sub-cloned into the monoclonal IgG1, Santa Cruz Biotechnology) [diluted in PBSTHindIII–EcoRI site of pCAMBIA1304 vector to construct the co- (80 mmol LÀ1 Na2HPO4, 20 mmol LÀ1 NaH2PO4, 100 mmol LÀ1expression vector. VTE4–myc fraction was sub-cloned into XbaI– NaCl, and 0.1% Tween 20)], washed with PBST three times, reactedSacI site, while VTE2–myc and VTE3–myc fragment were cloned into with goat anti-mouse IgG-AP (sc-2047, Santa Cruz Biotechnology),BglII–BstEII site, respectively (Fig. 2B). washed, and exposed with alkali phosphatase for 3 min. [The primers used for cloning were listed in Supplement Table2]. 2.5. Real-time PCR2.3. Transformation and screening of transgenic plants Total RNA was isolated and genomic DNA was removed by treatment with RQ1-RNase free DNase (Promega, Madison, WI). The binary vector PHB contains hygromycin resistant gene and One microgram of total RNA was reverse transcribed to generateanti-herbicide gene (BAR) inside the T-DNA for the selection of cDNA in 20 mL reactions for each sample using a Toyobo Rever Tra-transformants. The PHB:target gene constructs were transformed Plus-Kit according to manufacturer’s recommendations (Toyobo,into Agrobacterium tumefaciens C58C1 (pGV3101; rifampicin Osaka, Japan). An aliquot of cDNA corresponding to 10 pg – 10 ngresistance) and the ﬂoral-dip method was used for Arabidopsis total RNA was used in each real-time PCR assay (SYBR1 ExScripttransformation . The plants harboring empty plant expression RT-PCR Kit, TaKaRa, Shiga, Japan). Ubiquitin RNA was used tovector were used as non-transgenic control (NC). T0 seeds were normalize RNA concentrations. Standard curves were constructed
Y. Li et al. / Plant Science 178 (2010) 312–320 315for each gene and were used to calculate the corresponding mRNAconcentrations. [The primers used in the real-time PCR assays were listed inSupplement Table 3].2.6. Analyses of vitamin E, vitamin C, and glutathione Fifty independent plants (T4 generation) were chosen from thesame transgene line for vitamin E, vitamin C, and glutathioneanalyses. Data of one transgene line were collected for statistics. For each plant, the leaves of 3–4 weeks were harvested. Freeze-dried material (150 mg) was ground in liquid nitrogen, and thenextracted with 4 mL of n-hexane in dim light and in the presence ofargon to prevent the oxidation of vitamin E. After centrifugation at4000 Â g for 10 min, the clear supernatant was taken and the pelletwas re-extracted twice with 2 mL n-hexane. All the resulting Fig. 3. Phenotype of Arabidopsis transgenic lines. (A–C) Transgenic seedlingssupernatants were pooled, evaporated to dryness under nitrogen, transferred in soil for 3 days, 10 days, and 17 days. (D and E) Transgenic lines anddissolved in 750 mL of methanol, and stored at À80 8C until being non-transgenic control in full blooms. There are no signiﬁcant differences betweenanalyzed. transgenic lines and NC. The methanol extracts were resolved on a Phenomenex C18reverse-phase column (Calesil ODS-100, 5 mm, 4.6 mmI.D. Â 250 mm length) at 30 8C with a methanol:isopropanol not be detected (Fig. 4C). PCR and western blot could help to(95:5, v/v) mobile phase for 30 min at the ﬂow rate of 1 mL minÀ1 conﬁrm the plants harboring target transgene for each transfor-1 to equilibrate the column. Vitamin E was detected by mation event.ﬂuorescence with excitation at 292 nm and emission at 325 nm.Tocopherols and tocotrienols were identiﬁed by retention time and 3.2. Transcript changes of tocopherol biosynthetic pathway genes inquantiﬁed relative to dilution series of standards. Quantiﬁcation of transgenic linesvitamin E was carried out by measuring peak areas usingchromatograph data system D-2000 Elite software (Hitachi, Japan). To access gene expression levels of transgenic plants, mRNA Analyses of vitamin C were performed based on the method levels of tocopherol biosynthetic pathway genes were measured bydescribed by Kato and Esaka . The vitamin C was analyzed inleaves derived from the same plant which had tocopherol analysesperformed. Total ascorbate pool (AsA + DHA) was measured byreducing DHA to AsA with dithiothreitol and the reductive ratiocalculated as [AsA/(AsA + DHA)] Â 100. HPLC analysis of AsA wasperformed using a C18 reversed-phase column (Calesil ODS-100,5 mm, 4.6 mm I.D. Â 250 mm length) at 30 8C with a 0.1% oxalicacid:methanol (95:5, v/v) mobile phase for 30 min at the ﬂow rateof 1 mL minÀ1 to equilibrate the column. Samples (20 mL) wereinjected and analyzed with an UV detector at 254 nm. AsA wereidentiﬁed by retention time and quantiﬁed relative to dilutionseries of AsA standards. Glutathione was detected as described by Grifﬁth .Glutathione was extracted from frozen Arabidopsis leaves with6% (v/v) TCA. The extract was neutralized to pH 8.0 with 1 mol LÀ1K2CO3. Oxidized glutathione (GSSG) was reduced to 2Â GSH withdithiothreitol. Total glutathione was determined using a cyclingassay based on the reaction with 2-nitrobenzoic acid and detectedby ﬂuorescence with excitation at 380 nm and emission at 470 nm.The reductive ratio calculated as [GSH/(GSH + GSSG)] Â 100.3. Results3.1. Arabidopsis transformation and molecular analysis of transgenicplants Different constructs with single (HPPD, VTE2, VTE3, VTE1, andVTE4) and dual (VTE2 + VTE4 and VTE3 + VTE4) genes were Fig. 4. Molecular detection of transgenic lines. (A) PCR analyses for the presence ofintroduced into Arabidopsis. There were no developmental myc and hygromycin (hyg) fragments in single transgenic lines. M, DL-2000 Markerphenotypes in transgenic lines (Fig. 3). The hygromycin and (100–2000 bp); PC, vector used for transgenic (positive control); NC, non-herbicide resistant plants of single-gene transgenic lines were transgenic control; HPPD, VTE2, VTE3, VTE1, and VTE4 showed representativescreened by PCR with hygromycin and myc gene primers (Fig. 4A). plants of corresponding transgenic lines. (B) PCR analyses for the presence ofFor dual-gene transgenic lines, PCR was performed to analyze the transgenes in dual transgenic lines. VTE2 + VTE4 and VTE3 + VTE4 showed representative plants of corresponding dual transgenic lines. (C) Western blotspresence of the transgene (Fig. 4B). The transgenic lines over- of transgenic lines. HPPD, VTE2, VTE3, VTE1, and VTE4 showed fusing proteins withexpressing target protein could be screened by speciﬁc myc-tag myc tag. VTE2 + VTE4, VTE2–myc and VTE4–myc fusing proteins; VTE3 + VTE4,antibody, with which the corresponding endogenous protein could VTE3–myc and VTE4–myc fusing proteins.
316 Y. Li et al. / Plant Science 178 (2010) 312–320Fig. 5. Gene expression level of tocopherol biosynthetic pathway key enzymes intransgenic plant leaves. Y axis, relative mRNA expression; NC, non-transgeniccontrol; HPPD, VTE2, VTE3, VTE1, and VTE4 showed representative plants ofcorresponding transgenic lines; VTE2 + VTE4 and VTE3 + VTE4 showedrepresentative plants of corresponding dual transgenic lines.real-time PCR in NC and transgenic lines. The expression levels oftarget gene were elevated in both single- and dual-gene transgeniclines (Fig. 5), suggesting that exogenous induction elevatedexpression level of target gene in transgene lines. Further studyshowed that the changes of target gene over-expression hadlimited effect on the transcription of other tocopherol biosyntheticpathway genes in transgenic lines (P > 0.05).3.3. Genetic engineering of tocopherol biosynthetic pathway results inaccumulation of vitamin E and shift in a-tocopherol composition inleaves To assess the contributions of different enzymes in tocopherolbiosynthetic pathway, total vitamin E content and a-tocopherolcomposition were analyzed. Total vitamin E content was(14.28 Æ 2.26) pmol mgÀ1(FW) (n = 50, ÆSD, same below) and a-tocopherol was about (87.5 Æ 7.1) % of all vitamin E forms in leaves ofNC. In this study, homologous over-expression of HPPD could increaseat least 43% total vitamin E content in leaves with a-tocopherolcomposition increased modestly (88.1 Æ 7.5)%. However, accumulatedtocotrienols were not observed in this study. Over-expression of VTE2would increase total vitamin E in Arabidopsis leaves. Among ﬁftyindependent transgenic lines, total vitamin E increased to(29.71 Æ 4.31) pmol mgÀ1(FW) in average, and the best produced44.71 pmol mgÀ1(FW) vitamin E with only 85.7% a-tocopherolcomposition. The function of over-expression of VTE3 did not havedistinct function in vitamin E content (14.68 Æ 1.64) pmol mgÀ1 (FW),while a-tocopherol composition was elevated to 91.8% in the besttransgenic line. Over-expression of VTE1 or VTE4 alone signiﬁcantlyaffected tocopherol composition. Over-expression of VTE1 mainlyincreased the proportion of g-tocopherol and its isoforms, while a- Fig. 6. Average vitamin E, vitamin C and glutathione contents in leaves oftocopherol content did not show remarkable increase compared with representative transgenic Arabidopsis lines. (A) Vitamin E contents and a- tocopherol composition of transgenic lines and NC, measured in pmol mgÀ1NC. In transgenic line which expressed highest level of VTE1, the a- (FW). The numbers above the bars showed a-tocopherol composition, calculated bytocopherol only possessed 59.2% in total vitamin E, and the total [a-tocopherol/(a-tocopherol + g-tocopherol)] Â 100. (B) Dehydroascorbate (DHA)vitamin E increased to (24.51 Æ 2.69) pmol mgÀ1 (FW). Over-expres- and ascorbic acid (AsA) contents in the same NC and transgenic lines, measured insion of VTE4 elevated both content and composition [(96.8 Æ 3.1) % in mmol gÀ1 (FW). The numbers above the bars showed the reductive ratio of vitaminaverage] of a-tocopherol in Arabidopsis leaves, and increased the total C, calculated by [AsA/(AsA + DHA)] Â 100. (C) Glutathione (GSH) contents in the same NC and transgenic lines, measured in mmol gÀ1 (FW). The numbers above thevitamin E content to (16.96 Æ 2.87) pmol mgÀ1 (FW) (average of 50 bars showed the reductive ratio of reduced glutathione, calculated by [GSH/independent transgenic lines). The best transgenic line had total (GSH + GSSG)] Â 100.vitamin E content to (18.36 Æ 0.95) pmol mgÀ1 (FW), and a-tocopherolcomposition was 97.1 (Fig. 6A).
Y. Li et al. / Plant Science 178 (2010) 312–320 317 The dual-gene transgenic lines showed multiple functions ofboth enzymes. The VTE2 + VTE4 dual transgenic lines increasedtotal vitamin E content to (64.55 Æ 3.21) pmol mgÀ1 (FW) for thebest performing events, which was higher than VTE2 or VTE4 singleover-expression lines. In addition, the VTE3 + VTE4 dual transgenicline had a-tocopherol composition increased to 97.9% in the besttransgenic line by the expense of other tocopherol forms, and the totalvitamin E content was about (16.89 Æ 1.58) pmol mgÀ1 (FW)(Fig. 6A).3.4. Tocopherol accumulation may affect ascorbate pool and itsreductive ratio in Arabidopsis leaves To assess whether elevated vitamin E had inﬂuence on vitaminC, bioactive compounds in leaves were measured in both NC andsingle-gene over-expressing lines, including total vitamin Econtent, a-tocopherol composition, vitamin C content, andreductive ratio of AsA. As shown in Fig. 6B, the endogenous ascorbate pool wasobserved decrease in the vitamin E up-regulated plants, especiallyin the a-tocopherol increased lines. AsA content in HPPD over-expressing lines [(2.98 Æ 0.20) mmol LÀ1 (FW)] did not have obviousdifference with non-transgenic line [(3.07 Æ 0.19) mmol LÀ1 (FW)].AsA content decreased to (0.94 Æ 0.08) mmol LÀ1 (FW) in the bestVTE2 over-expressing line, whose vitamin E level increased by 3.7-fold compared with the NC. In dual transgenic line VTE2 + VTE4, whichhad remarkable increase in total vitamin E content (4.5-fold), vitaminC was observed decrease to (0.84 Æ 0.07) mmol LÀ1 (FW). It was alsofound that the level of AsA changed in VTE3, VTE1, and VTE4 over-expressing lines and VTE3 + VTE4 dual transgenic line, which had a-tocopherol composition changed. For VTE3 lines, AsA decreased to(1.65 Æ 0.13) mmol LÀ1 (FW). VTE1 and VTE4 over-expressing linesshowed similar decreasing trends in AsA, which had decreased to(0.90 Æ 0.13) mmol LÀ1 (FW) and (0.66 Æ 0.10) mmol LÀ1 (FW), re- Fig. 7. Gene expression level of biosynthetic pathway gene and vitamin C content. VTE1 expression lines were taken for example. (A) VTE1 expression level in VTE1spectively. In VTE3 + VTE4 dual transgenic line, vitamin C was transgenic lines. (B) Vitamin C content in corresponding transgenic lines.observed decrease to (0.65 Æ 0.10) mmol LÀ1 (FW). Our data pre-sented here indicated that, ascorbate pool might be inﬂuenced by thechange in vitamin E pool (Fig. 6B). Correlation analyses were showed C content. In HPPD transgenic line, the glutathione content wasnegative correlation between vitamin E content and AsA content (0.60 Æ 0.03) mmol LÀ1 (FW), which showed no obvious difference(r = À0.38, P = 0.0053, n = 50). between NC line [(0.63 Æ 0.04) mmol LÀ1 (FW)]. The VTE3 transgenic The reductive ratio of AsA was also observed in this study line had modest decrease in glutathione content(Fig. 6B). In the single transgenic lines VTE2, VTE1, and VTE4, whose [(0.44 Æ 0.03) mmol LÀ1 (FW)]. In the VTE2, VTE1, and VTE4 transgenicAsA level strongly decreased, the reductive ratio of AsA greatly lines, the glutathione contents were (0.31 Æ 0.03) mmol LÀ1 (FW),increased (87.3%, 90.7%, 91.9%, and 95.5% for NC, VTE2, VTE1, and (0.34 Æ 0.02) mmol LÀ1 (FW), and (0.28 Æ 0.02) mmol LÀ1 (FW), re-VTE4). The increase in composition of reductive ratio was also spectively. In dual transgenic lines VTE2 + VTE4 and VTE3 + VTE4,obvious in the transgenic lines VTE3 90.8%. On the contrary, no there were also existing similar trends, and the glutathione contentssigniﬁcant change in reductive ratio was observed in HPPD over- were (0.25 Æ 0.02) mmol LÀ1 (FW) and (0.24 Æ 0.03) mmol LÀ1 (FW)expressing lines 87.6%. In dual transgenic lines VTE2 + VTE4 and (Fig. 6C). Correlation analyses were showed negative correlationVTE3 + VTE4, the reductive ratio of AsA increased to 93.3% and between vitamin E content and glutathione content (r = À0.46,93.5%, respectively. P = 0.0044, n = 50). The content of AsA and glutathione had positive It was also observed that the lines which had high expression correlation (r = 0.98, P = 0.020, n = 50).level of tocopherol biosynthesis gene would show low AsA content. Considering the reductive glutathione ratio, it seemed that allTake VTE1 transgene line for example. It seemed that the lines with transgenic lines had increase compared with NC (70.7%). Forhigh VTE1 expression level (Fig. 7A, also with high total tocopherol single-gene transgenic line, the ratio was 73.2%, 76.3%, 72.5%,content) would have low AsA content (Fig. 7B). Correlation 74.5%, and 77.4% for HPPD, VTE2, VTE3, VTE1, and VTE4 lines. Inanalyses were showed signiﬁcantly negative correlation between dual-gene transgenic lines VTE2 + VTE4 and VTE3 + VTE4, the ratiotarget gene expression level and AsA content (r = À0.91, was 78.7% and 78.4%, respectively (Fig. 6C).P = 0.00065, n = 50). 3.6. Halliwell–Asada cycle is activated in transgenic lines3.5. Tocopherol accumulation and ascorbate decrease may affectglutathione pool and its reductive ratio in Arabidopsis leaves Decrease in ascorbate pool might either be inﬂuenced by a decrease in biosynthesis of AsA or an increase in AsA catabolism. In Because vitamin C was related with glutathione in plant order to test this hypothesis, the expression levels of enzymes inphysiology, the content of glutathione and reduced glutathione AsA metabolism were studied. In plant cells, AsA related metaboliccomposition were observed in the same line. It was observed that network can be divided into two categories: biosynthetic pathwaythe change of glutathione content had similar trends with vitamin and Halliwell–Asada cycle. AsA biosynthetic pathway includes
318 Y. Li et al. / Plant Science 178 (2010) 312–320Fig. 8. Gene expression level of enzymes of AsA metabolic pathways and Halliwell–Asada cycle in single transgenic lines. Y axis, relative mRNA expression level. (A–C) DHAR,MDAR, and APX expression level in transgenic lines. (D) Tocopherol for reactive oxygen species scavenging and its recycle pathway with Halliwell–Asada cycle. APX, DHAR,and MDAR are important enzymes of Halliwell–Asada cycle. AsA can reduce tocopherol radical to tocopherol with the help of APX, and itself is oxidized to MDA. With the helpof DHAR and MDAR, MDA can be recycled to AsA. AsA is the protector of tocopherol in plant cells. Data represent the means Æ SD of 50 individual plants.enzymes GDPME, GDPMPPase, VTC2, L-GalPPase, L-GalDH, and L- leaves of tobacco . In this study, homologous over-expressionGLDH. Halliwell–Asada cycle includes enzyme APX, DHAR, and of HPPD could increase 43% total vitamin E content in leaves. ThisMDAR. In order to investigate which step or steps of AsA result was consistent with Tsegaye et al. , who over-expressedmetabolism is regulated, the expression of key enzymes of the homozygous HPPD in Arabidopsis resulted in up to a 37% increase inpathways mentioned above was therefore measured by real-time leaf tocopherol levels. There might be two reasons for slightPCR. increase in vitamin E content. Firstly, the effect of HPPD gene over- Expression levels of AsA biosynthetic pathway related genes did expression may be counteracted by the limited expression level ofnot show signiﬁcant differences between NC and transgenic lines. downstream genes in tocopherol biosynthetic pathway, and theIt was suggested that the decrease of ascorbate pool could not be transgenic plant may not have enough capacity to phytylate all theattributed to the down-regulation of AsA biosynthetic pathway. In increased HGA. Secondly, the increased HGA may be used ascontrast, the expression of genes encoding enzyme APX in substrate for other secondary metabolites in plants.Halliwell–Asada cycle showed signiﬁcant increases in the VTE2, The condensation of PDP and HGA is catalyzed by the enzymeVTE1, and VTE4 over-expression transgenic lines (P < 0.01), and the VTE2. Because HGA and PDP are also used in the synthesis of otherexpression level of enzymes DHAR and MDAR was also observed secondary metabolites in plants, such as plastoquinones, phyllo-increasing, but the up-regulated extent was less than APX quinones, and chlorophylls, VTE2 has important effects onexpression. The VTE3 over-expressing lines presented signiﬁcant tocopherol pathway at the branch-point from these plastidincrease in the expression of APX compared with NC (P < 0.05), phytyllipid pathways [10,32]. In previous study , VTE2 over-while the expression of DHAR and MDAR showed no signiﬁcant expression resulted in a 4.4-fold increase in total tocopherolchange. In HPPD over-expression lines, there was no difference in content relative to wild type in the best transgenic line. In thisexpression level of the Halliwell–Asada cycle enzymes (P > 0.05) study, over-expression of VTE2 would increase total vitamin E to(Fig. 8A–C). (29.71 Æ 4.31) pmol mgÀ1 (FW) in average. Among ﬁfty independent transgenic lines, the best line produced 44.71 pmol mgÀ1 (FW)4. Discussion vitamin E in leaves. This result seemed lower compared with Collakova and DellaPenna’s report , because all of our results were4.1. Evaluating functions of tocopherol biosynthetic pathway compared with non-transgenic control, which were plants trans-enzymes in the view of metabolic engineering formed with the empty vector used in this study. In this way, the different tocopherol biosynthetic pathway genes could be compared In order to value the contribution of each enzyme in tocopherol under the same genetic background. a-Tocopherol composition in thebiosynthetic pathway, genes were constitutively over-expressed in best transgenic line was about 85.7%, showed slight decreaseArabidopsis. Elucidating changes of vitamin E contents or a- compared with NC. In previous report , the transgenic lines withtocopherol composition in transgenic lines were discussed below, highest total tocopherol content had an increase in g-tocopherolbased on the detailed dissection of the biosynthetic pathway. composition. Both results showed the downstream enzyme VTE4 had HGA is produced from p-hydroxyphenylpyruvate (HPP) by the limitation, which suggested co-expression of VTE4 with VTE2 mightcytosolic enzyme HPPD. It was reported that heterologous over- more obvious effect on a-tocopherol composition.expression of the barley HPPD gene under control of the 35S The enzymes VTE3, VTE1, and VTE4 have functions on thepromoter only elevated tocopherol content in seeds but not in proportion of tocopherol forms. The methylation reaction cata-
Y. Li et al. / Plant Science 178 (2010) 312–320 319lyzed by VTE3 determines the number and position of methyl Over-expressing two or more genes of one or several metabolicgroups in the ﬁnal products – without the catalysis of VTE3, the pathways will have multiple and accumulative functions intocopherol forms will be d- or b-; otherwise, it will be g- or a- increasing ﬂux of target product, which might be more efﬁcienttocopherol. Sole over-expression of VTE3 would not remarkably than transgenic plants harboring only one transgene. This resultaffect total vitamin E contents in this study, because the a- and g- provides a feasible strategy for metabolic engineering to cultivatetocopherol composition in Arabidopsis leaves is much higher than plants with higher nutrition values.b- and d-tocopherol composition. VTE4 plays important role onproducing of a-tocopherol. Over-expression of VTE4 gene solely 4.2. Vitamin E accumulation affected vitamin C and glutathione poolscould elevate a-tocopherol composition in Arabidopsis leaves, and in plant cellsresult in 20–30% increase in total vitamin E content. In previousstudy , over-expression of VTE4 gene in Arabidopsis seeds The opening question in metabolic engineering is whether thealtered the tocopherol composition but not the total tocopherol accumulation of target product might have effects on plantcontent. However, in this study total vitamin E content showed physiology or other metabolic products. AsA is important for plantmodest increase. Transcription data showed that genes encoding physiology and nutrition. Glutathione is water-soluble antioxidantupstream enzymes of tocopherol biosynthetic pathway, especially in plant, which was related with AsA by Halliwell–Asada cycle. InVTE2, had slightly elevated expression level in VTE4 transgenic line. this study, content and reductive ratio of AsA and GSH wasIt seems that the pulling force of the last enzyme in biosynthetic analyzed in transgenic plants, respectively. Our study proved thatpathway helped to induce the potential ability of the upstream in single-gene transgenic lines, the accumulation of vitamin Eenzymes. However, the endogenous expression of VTE2 was content or change in a-tocopherol composition, resulted in alimited , which gave hints that the combination of VTE2 and decrease in total ascorbate pool and a modest increase in theVTE4 might have both vitamin E content and a-tocopherol reductive ratio of AsA. Glutathione had similar changing trendscomposition elevated. The signiﬁcant change in VTE1 over- with AsA. In multiple-gene transgenic lines, whose vitamin Eexpression lines was the increase of g-tocopherol and its isoforms content and composition was both changed as described detailed(such as plastochromanol-8, the natural homologue of g-toco- above, the decreasing in AsA and GSH was also observed whichtrienol), which increased by 5.5-fold in average compared with NC, seemed to have lower contents than single-gene transgenic lines.while a-tocopherol did not show signiﬁcant increase. The All those observation indicated that vitamin E accumulation inaccumulation of g-tocopherol and its isoforms might be resulted plant cells might have effect on vitamin C and glutathione. Itfrom the limitation of endogenous VTE4, and this result was seemed that the lines with higher vitamin E content would havesimilar with report from Kanwischer et al. , who increased total lower contents AsA and GSH. In summary, increasing in vitamin Etocopherol levels of at least 7-fold in Arabidopsis leaves. The levels triggered compensatory changes in ascorbate and glutathi-different increase extents might be the different plant expression one levels, which was also demonstrated in former studies .vectors and promoters used in different studies. Because tocopherols, ascorbate and glutathione could be To further study the effect of co-operation of key enzymes on related with Halliwell–Asada cycle, further study suggested thevitamin E content and composition, dual-gene co-expression was changing of antioxidants might be affected by the activity ofconducted. Based on the results of single-gene over-expression Halliwell–Asada cycle in transgenic lines.lines, two different combinations (VTE2 + VTE4 and VTE3 + VTE4) Tocopherol and Halliwell–Asada cycle are closely associated inwere studied. In VTE3 + VTE4 lines, although vitamin E content chloroplasts. Tocopherol can protect PUFA and hydrosulﬁdecould not be obviously elevated, nearly all other tocopherol groups of proteins from oxidization with itself oxidized toforms switched to a-tocopherol. It was found that the tocopherol radical . Plants can convert the tocopherol radicalsVTE2 + VTE4 lines could increase vitamin E content and enhance back in time with the help of Halliwell–Asada cycle (Fig. 8D). Thea-tocopherol proportion at the same time. Vitamin E content in process can be generalized as follows. Tocopherol radical isVTE2 + VTE4 lines was higher than in VTE2 or VTE4 single reduced to tocopherol by AsA with the help of APX, and AsA istransgenic lines. It seems that the pulling force of the oxidized into monodehydroascorbic acid (MDA). MDAR reducesdownstream enzyme (the faucet enzyme) VTE4 plays a more MDA into AsA. On the other hand, dehydroascorbic acid (DHA) isimportant role in stimulating vitamin E accumulation whereas spontaneously produced from MDA and can be reduced to AsA bythe functioning of the upstream enzyme VTE2 is increased DHAR, with GSH oxidized to GSSG .proportionally. Although VTE2 is the enzyme which catalyzes the AsA plays important fuction in Halliwell–Asada cycle and iscondensation of HGA and PDP, its function seems ﬂux-limiting directly related with tocopherol. GSH participates Halliwell–Asadaand hence represents the ﬁrst committed step in tocopherol cycle and helps AsA regenerate. From the view of the overallbiosynthesis. VTE4 catalyzes the ﬁnal step of a-tocopherol, network, redox homeostasis is existed in plants which can keep thewhich is not inhibited by end product, it is suggested that it is balance between reactive oxygen species (ROS) and antioxidantscompletely desirable to fundamentally enhance the a-tocopher- . In this study, genetic engineering increased target geneol production. In previous report , seeds of VTE2 + VTE4 dual expression level, activated tocopherol biosynthesis, and theover-expression lines had a 12-fold increase in vitamin E activity increased tocopherols might active Halliwell–Asada cycle andrelative to wild type [Vitamin E activity was calculated as one affect the antioxidants associated with it, such as AsA and GSH.milligram of a-, b-, g-, and d-tocopherol corresponds to 1, 0.5, However, it was reported that the amount of endogenous DHAR0.1, and 0.03 mg of a-TE, respectively (Food and Nutrition Board, was limited , and APX showed higher expression level thanInstitute of Medicine, 2000)]. In other study we did in lettuce, the MDAR and DHAR in transgenic lines in this study, which couldVTE2 + VTE4 combination had a 19-fold increase in vitamin E accelerate the oxidization process of AsA to MDA and DHA. Theactivity relative to non-transgenic control (unpublished). All oxidized form of AsA would be quickly degenerated, resulting inthese data indicated that VTE2 + VTE4 dual transgenic lines are the total ascorbate pool decreasing. On the other side, the genes ofcrucial and useful for metabolic engineering. Also based on the AsA biosynthetic pathway were not up-regulated in transgenicsingle-expression and dual expression data, triple-expression of plants. As a result, the level of endogenous AsA might beVTE2, VTE1 and VTE4 is considered in leaves of lettuce and seeds decreasing. It indicated that elevating the activity of AsAof Brassica campestris L. in our lab, and it is looking forward to biosynthetic pathway might complement the decreasing in AsAmore expectant results. content.
320 Y. Li et al. / Plant Science 178 (2010) 312–3204.3. Outlook in tocopherol biosynthetic engineering  M. Rohmer, Mevalonate-independent methylerythritol phosphate pathway for isoprenoid biosynthesis. Elucidation and distribution, Pure Appl. Chem. 75 (2003) 375–387. It is concluded that, the ﬁve enzymes of tocopherol biosynthetic  D. DellaPenna, A decade of progress in understanding vitamin E synthesis inpathway have different functions, which can be divided into two plants, J. Plant Physiol. 162 (2005) 729–737.  D. DellaPenna, Progress in the dissection and manipulation of vitamin E synthesis,groups—the enzymes whose function are mainly on increasing Trends Plant Sci. 10 (2005) 574–579.vitamin E contents (HPPD, VTE2) and the enzymes effectively  S.R. Norris, T.R. Barette, D. DellaPenna, Genetic dissection of carotenoid synthesischange the composition of a-tocopherol (VTE3, VTE1, and VTE4). in Arabidopsis deﬁnes plastoquinone as an essential component of phytone desaturation, Plant Cell 7 (1995) 2139–2149.For the ﬁrst group, VTE2 has signiﬁcant effect in increasing vitamin  E. Collakova, D. DellaPenna, Isolation and functional analysis of homogentisateE content. For the second group, the enzymes have different roles phytyltransferase from Synechocystis sp. PCC 6803 and Arabidopsis, Plant Physiol.in changing the composition of one or two tocopherol form. In 127 (2001) 1113–1124.  D.K. Shintani, Z. Cheng, D. DellaPenna, The role of 2-methyl-6-phytylbenzoqui-practical application of metabolic engineering, it is recommended none methyltransferase in determining tocopherol composition in Synechocystisto choose two or more enzymes from the two groups to make sp. PCC 6803, FEBS Lett. 511 (2002) 1–5.multiple expression vectors for speciﬁc aims. Furthermore, ¨ ¨  S. Porﬁrova, E. Bergmuller, S. Tropf, R. Lemke, P. Dormann, Isolation of anproducts from other metabolic pathways might be inﬂuenced as Arbidopsis mutant lacking vitamin E and identiﬁcation of a cyclase essential for tocopherol biosynthesis, Proc. Natl. Acad. Sci. U.S.A. 99 (2002) 12495–12500.a result of increased target metabolic product. In this report,  D. Shintani, D. DellaPenna, Elevating the vitamin E content of plants throughaccumulation of vitamin E decreased AsA and GSH levels in metabolic engineering, Science 282 (1998) 2098–2100.transgenic lines. If the target of metabolic engineering is to  K.M. Oksman-Caldentey, D. Inze, Plant cell factories in the postgenomic era: new ways to produce designer secondary metabolites, Trends Plant Sci. 9 (2004) 433–increase both vitamin E and vitamin C, it is recommended to 440.regulate both tocopherol and AsA biosynthetic pathways at the  D. DellaPenna, B.J. Pogson, Vitamin synthesis in plants: tocopherols and carote-same time. If the target of metabolic engineering is to increase noids, Annu. Rev. Plant Biol. 57 (2006) 711–738. ¨  A. Trebst, B. Depka, H. Hollander-Czytko, A speciﬁc role for tocopherol and ofvitamin E levels only, such as genetic manipulation in the seeds of chemical singlet oxygen quenchers in the maintenance of photosystem II struc-oil crop, main attention should only be paid on tocopherol ture and function in Chlamydomonas reinhardtii, FEBS Lett. 516 (2002) 156–160.biosynthetic pathway. However, the relationship between vitamin ¨  M. Havaux, F. Eymery, S. Porﬁrova, P. Rey, P. Dormann, Vitamin E protects against photoinhibition and photooxidative stress in Arabidopsis thaliana, Plant Cell 17E and vitamin C is very complex, and the overall biochemical and (2005) 3451–3469.molecular regulation of the metabolic pathways in plants remains  H. Maeda, Y. Sakuragi, D.A. Bryant, D. DellaPenna, Tocopherols protect Synecho-limited. More research should be done to promise novel insights cystis sp. Strain PCC 6803 from lipid peroxidation, Plant Physiol. 138 (2005) 1422– 1435.into the regulation, integration, and interrelation of these path-  K. Asada, The water–water cycle in chloroplasts: scavenging of active oxygens andways in plants. dissipation of excess photons, Annu. Rev. Plant Physiol. Plant Mol. Biol. 50 (1999) This study set up an appraisable system in sole plant to value 601–639.and to compare different roles of enzymes in tocopherol  R. Mittler, Oxidative stress, antioxidants and stress tolerance, Trends Plant Sci. 7 (2002) 405–410.biosynthetic pathway played in changing vitamin E content and  S. Caretto, A. Paradiso, L. D’Amico, L.D. Gara, Ascorbate and glutathione metabo-a-tocopherol composition in plant leaves, which provides an lism in two sunﬂower cell lines of differing a-tocopherol biosynthetic capability,effective and referenced approach for tocopherol transgenic Plant Physiol. Biochem. 40 (2002) 509–513. ¨ ¨  M. Kanwischer, S. Porﬁrova, E. Bergmuller, P. Dormann, Alterations in tocopherolresearch. This work also sheds light on the relationship and cyclase activity in transgenic and mutant plants of Arabidopsis affect tocopherolregulation between lipid-soluble antioxidant vitamin E and water- content, tocopherol composition, and oxidative stress, Plant Physiol. 137 (2005)soluble antioxidants vitamin C and GSH, which will have very 713–723.  V.C. Collin, F. Eymery, B. Genty, P. Rey, M. Havaux, Vitamin E is essential for theimportant and pivotal function in plant physiology, especially ROS- tolerance of Arabidopsis thaliana to metal-induced oxidative stress, Plant Cellantioxidant network. Eviron. 31 (2008) 244–257.  J. Mao, Y.C. Zhang, Y. Sang, Q.H. Li, H.Q. Yang, A role for Arabidopsis cryptochromes and COP1 in the regulation of stomatal opening, Proc. Natl. Acad. Sci. U.S.A. 102Acknowledgments (2005) 12270–12275.  S.J. Clough, A.F. Bent, Floral dip: a simpliﬁed method for Agrobacterium-mediated We thank Professor Hongquan Yang (SIPPE, CAS, China) for transformation of Arabidopsis thaliana, Plant J. 16 (1998) 735–743.  F.M. Ausubel, R. Brent, R.E. Kingston, D.D. Moore, J.G. Seidman, J.A. Smith, K. Struhlsupplying pBSmyc and pHB plasmids for vector construction and (Eds.), Short Protocols in Molecular Biology, 3rd ed., John Wiley & Son, Inc., 1995 .helping with Arabidopsis transformation. Mr. Yuliang Wang  H.Q. Yang, Y.J. Wu, R.H. Tang, D. Liu, Y. Liu, A.R. Cashmore, The C termini of(Shanghai Jiao Tong University, China) helped in HPLC analysis. Arabidopsis cryptochromes mediate a constitutive light response, Cell 103 (2000)This research is supported by the National Basic Research Program 815–827.  N. Kato, M. Esaka, Changes in ascorbate oxidase gene expression and ascorbateof China (973 Program, 2007CB108805). levels in cell division and cell elongation in tobacco cells, Physiol. Plant 105 (1999) 321–329.  O.W. Grifﬁth, Determination of glutathione and glutathionedisulﬁde using glu- tathione reductase and 2-vinylpyridine, Anal. Biochem. 106 (1990) 207–212.Appendix A. Supplementary data  J. Falk, G. Andersen, B. Kernebeck, K. Krupinska, Constitutive overexpression of barley 4-hydroxyphenylpyruvate dioxygenase in tobacco results in elevation of Supplementary data associated with this article can be found, in the vitamin E content in seeds but not in leaves, FEBS Lett. 540 (2003) 35–40.  Y. Tsegaye, D.K. Shintani, D. DellaPenna, Overexpression of the enzyme p-hydro-the online version, at doi:10.1016/j.plantsci.2010.01.004. xyphenolpyruvate dioxygenase in Arabidopsis and its relation to tocopherol biosynthesis, Plant Physiol. Biochem. 40 (2002) 913–920.References  B. Savidge, J.D. Weiss, Y.H.H. Wong, M.W. Lassner, T.A. Mitsky, C.K. Shewmaker, D. Post-Beittenmiller, H.E. Valentin, Isolation and characterization of homogentisate H.M. Evans, K.S. Bishop, On the existence of a hitherto unrecognized dietary factor phytyltransferase genes from Synechocystis sp PCC6803 and Arabidopsis, Plant essential for reproduction, Science 56 (1922) 650–651. Physiol. 129 (2002) 321–332. P.M. Bramley, I. Elmadfa, A. Kafatos, F.J. Kelly, Y. Manios, H.E. Roxborough, W.  E. Collakova, D. DellaPenna, Homogentisate phytyltransferase activity is limiting Schuch, P.J.A. Sheehy, K.-H. Wagner, Vitamin E, J. Sci. Food Agric. 80 (2000) 913– for tocopherol biosynthesis in Arabidopsis, Plant Physiol. 131 (2003) 632–642. 938.  G.A. Pascoe, M.W. Fariss, K. Olafsdottir, D.J. Reed, A role of vitamin E in protection M.G. Traber, H. Sies, Vitamin E in humans: demand and delivery, Annu. Rev. Nutr. against cell injury, Eur. J. Biochem. 166 (1987) 241–247. 16 (1996) 321–347.  C.H. Foyer, H. Vanacker, L.D. Gomez, J. Harbinson, Regulation of photosynthesis R.R. Eitenmiller, Vitamin E content of fats and oils – nutritional implications, Food and antioxidant metabolism in maize leaves at optimal and chilling tempera- Technol. 51 (1997) 78–81. tures: review, Plant Physiol. Biochem. 40 (2002) 659–668. J. Soll, G. Schultz, Comparison of geranylgeranyl and phytyl substituted methyl-  Y. Li, Z. Wang, X. Sun, K. Tang, Current opinions on the functions of tocopherol quinols in the tocopherol synthesis of spinach chloroplasts, Biochem. Biophys. based on the genetic manipulation of tocopherol biosynthesis in plants, J. Integr. Res. Commun. 91 (1979) 715–720. Plant Biol. 50 (2008) 1057–1069.