Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
© 2001 Nature Publishing Group http://biotech.nature.com RESEARCH ARTICLE © 2001 Nature Publishing Group http://biotech.nature.com Overexpression of petunia chalcone isomerase in tomato results in fruit containing increased levels of flavonols Shelagh R. Muir1, Geoff J. Collins1, Susan Robinson1, Stephen Hughes2, Arnaud Bovy3, C.H. Ric De Vos3, Arjen J. van Tunen3, and Martine E. Verhoeyen1* Tomatoes are an excellent source of the carotenoid lycopene, a compound that is thought to be protective against prostate cancer. They also contain small amounts of flavonoids in their peel (∼5–10 mg/kg fresh weight), mainly naringenin chalcone and the flavonol rutin, a quercetin glycoside. Flavonols are very potent antioxidants, and an increasing body of epidemiological data suggests that high flavonoid intake is correlated with a decreased risk for cardiovascular disease. We have upregulated flavonol biosynthesis in the tomato in order to generate fruit with increased antioxidant capacity and a wider range of potential health benefit properties. This involved transformation of tomato with the Petunia chi-a gene encoding chalcone isomerase. Resulting transgenic tomato lines produced an increase of up to 78 fold in fruit peel flavonols, mainly due to an accumulation of rutin. No gross phenotypical differences were observed between high-flavonol transgenic and control lines. The phenotype segregated with the transgene and demonstrated a stable inheritance pattern over four subsequent generations tested thus far. Whole-fruit flavonol levels in the best of these lines are similar to those found in onions, a crop with naturally high levels of flavonol compounds. Processing of high-flavonol tomatoes demonstrated that 65% of flavonols present in the fresh fruit were retained in the processed paste, supporting their potential as raw materials for tomato-based functional food products. tive products10. In the approach detailed in this paper, heterologous overexpression in tomato of a gene from Petunia encoding chalcone isomerase, an enzyme involved in flavonol biosynthesis (Fig. 1), has been used to produce transgenic fruit with an increase in peel flavonols of up to 78 fold, mainly due to accumulation of quercetin glycosides. There is considerable interest in the development of food products from plants rich in protective vitamins or other compounds with potential health benefits. Several groups have already shown that genetic manipulation can be used to produce plants with improved nutritional traits; examples are Arabidopsis seeds with elevated vitamin E levels1 and provitamin A-enriched rice and tomato2,3. Flavonoids are a group of plant secondary metabolites thought to possess health-promoting properties. They occur naturally in fruits, vegetables, nuts, seeds, and flowers, and therefore form an integral part of the human diet. Flavonoids have been shown to exhibit a wide range of biological activities in animal cell culture and in vitro systems, including antioxidant and vasodilatory actions4. Results from several epidemiological studies suggest that increased consumption of flavonoids, as part of a balanced diet, may help to protect against cardiovascular disease5–7. However, the in vitro evidence for cardiovascular protection is particularly strong for one group of flavonoids, the flavonols (e.g., quercetin and kaempferol), making these compounds attractive targets for nutritional enhancement of food crops. The flavonoid biosynthetic pathway (Fig. 1) and its regulation have been the subject of many studies8. Commercial research in this area has thus far been directed mainly toward altering pigmentary flavonoids known as anthocyanins, which play an important role in flower and leaf color9. However, recent advances concerning pathway regulation and the biochemistry of specific enzymes and enzyme complexes have opened up several potential pathway modification strategies to enhance the plant’s ability to synthesize flavonoid bioac- 1Unilever 470 Results and discussion Flavonoid biosynthesis during tomato fruit development. Nonhydrolyzed extracts of red tomato fruit peel contained naringenin chalcone (narichalcone), in addition to the flavonol glycosides rutin (quercetin 3-rutinoside), kaempferol 3-rutinoside, and small amounts of a quercetin 3-trisaccharide (A.L. Davis, personal communication). Aglycones of quercetin, kaempferol, and naringenin were not detectable. Fruit flesh did not accumulate significant amounts of flavonoids; that is, rutin was present only at the limit of detection (0.5 mg/kg fresh weight tomato). Figure 2 shows the development-dependent accumulation of rutin and naringenin chalcone in fruit peel. Levels of rutin increased during ripening. Kaempferol rutinoside was present in small amounts (<10% of the amount of rutin) during all stages of development except green fruit (data not shown). Naringenin chalcone was absent in green peel but increased sharply during coloring of the fruit, reaching levels of ∼1% on a dry weight basis in turning peel before declining through the red and overripe stages. Naringenin chalcone represents an intermediate in the biosynthesis of flavonols and is converted into the next compound in the pathway, naringenin, by the enzyme chalcone isomerase (CHI). Research, Colworth House, Sharnbrook, Bedfordshire MK44 1LQ, UK. 2Department of Biology, University of Exeter, Exeter, EX4 4QG, UK. 3Plant Research International, PO BOX 16, 6700AA Wageningen, The Netherlands. *Corresponding author ([email protected]). nature biotechnology • VOLUME 19 • MAY 2001 • http://biotech.nature.com © 2001 Nature Publishing Group http://biotech.nature.com © 2001 Nature Publishing Group http://biotech.nature.com RESEARCH ARTICLE Initial screening of transgenic fruit involved hydrolyzing flavonoids to their respective aglycones before analysis by HPLC. The hydrolysis procedure also caused the nonenzymatic conversion of naringenin chalcone into naringenin11. Mean levels of up to 16.52 mg quercetin and 2.05 mg kaempferol per gram dry weight were observed in hydrolyzed peel extracts of fruit transformed with pBBC50, representing increases of 66 and 57 fold over control, respectively (Fig. 4A, B). There was a direct relationship between quercetin and kaempferol levels in hydrolyzed extracts of transformed fruit, with those plants displaying a highquercetin phenotype also possessing high levels of kaempferol (an approximate quercetin to kaempferol ratio of 10:1). Naringenin chalcone levels were severely depleted in high-flavonol transformants (Fig. 4C). Given that naringenin chalcone is a substrate for CHI, this observation suggests an increased level of CHI enzyme activity in pBBC50 transgenics. Total flavonoids did not accumulate above control levels in the fruit flesh or leaves of any of the pBBC50 transformants tested (data not shown). Nonhydrolyzed extracts were analyzed to determine the form in which flavonoids accumulated in the peel of pBBC50-transformed plants. Figure 5A and B show HPLC chromatograms obtained with nonhydrolyzed peel extracts from control and pBBC50transformed tomatoes, respectively. Rutin, with a retention time (RT) of 16.8 min, represented the major quercetin glycoside in peel of control and pBBC50-transformed tomatoes. In addition, significant amounts of isoquercitrin (quercetin 3-glucoside) (RT = 17.4 min) also accumulated in the peel of pBBC50transformed plants, which might suggest that a rhamnosyl transferase has become rate limiting in the production of rutin in highly expressing pBBC50 lines. Several peaks that were not detectable in controls were also found in pBBC50-transformed peel chromatograms. The peak at RT = 19.7 min appeared to contain a mixture of two compounds; the RT and absorbance spectra of the major component corresponded to that of kaempferol 3-rutinoside, whereas that of the minor component had an absorbance spectrum comparable to that of a quercetin glycoside. The small peak at 20.3 min had an absorbance spectrum comparable to kaempferol 3-rutinoside, but its higher RT value indicates a yet unknown kaempferol glycoside. Northern analysis revealed low levels of chi transcript in peel at all stages of development in the fruit of high-flavonol transformants (data not shown), in contrast to control tomatoes, for which no chi transcript could be detected. Further work is required to ascertain the exact mechanism by which upregulating CHI messenger RNA (mRNA) leads to increases in flavonoid end products in peel. CHI enzyme activity could represent the sole rate-limiting step. Alternatively, increases in CHI mRNA could bring about changes in activity of other enzymes in the pathway, Figure 1. Flavonoid biosynthetic pathway. Included are the structures of the major flavonol glycoside in tomato, rutin, and its precursor, isoquercitrin. Northern analysis of expression of the endogenous flavonoid biosynthetic genes chalcone synthase (chs), chi, and flavonol synthase (fls) with heterologous Petunia probes revealed that both chs and fls transcripts were detectable in peel in all developmental stages tested, with levels peaking during the breaker and turning stages before subsequently decreasing in the red stage (Fig. 3). The chi transcript level was below the limit of detection in peel of all developmental stages. This, together with the relatively high and specific accumulation of naringenin chalcone in turning peel, suggested that one of the rate-limiting steps in flavonol biosynthesis in peel could lie at the level of chi gene expression. Transcripts of the chs, chi. and fls genes were undetectable in fruit flesh, in agreement with high-performance liquid chromatography (HPLC) data, which only indicated the presence of flavonoids at the limits of detection in this tissue. Transcripts of the chs, chi, and fls genes were all present in young leaves. Transformation of tomato with the chi gene from Petunia and analysis of transgenic plants. Tomato was transformed with a binary vector (pBBC50) containing the Petunia hybrida chi gene under the control of the constitutive cauliflower mosaic virus (CaMV) double 35S promoter. Southern hybridization confirmed that all transgenic plants tested contained either one or two copies of the pBBC50 transgene (data not shown). Figure 2. Accumulation of rutin (dashed line) and naringenin chalcone (solid line) in developing tomato peel. Flavonoids were extracted under nonhydrolyzing conditions and analyzed by HPLC. The terms green, breaker, turning, red, and overripe refer to fruit color and represent stages at approximately 7, 28, 31, 39, 45, and 49 days post-anthesis, respectively. http://biotech.nature.com • MAY 2001 • VOLUME 19 • nature biotechnology 471 © 2001 Nature Publishing Group http://biotech.nature.com RESEARCH ARTICLE A © 2001 Nature Publishing Group http://biotech.nature.com A B B C C Figure 4. Flavonoid levels in the peel of transgenic tomato fruit. Flavonoids were extracted under hydrolyzing conditions and analyzed by HPLC. The hydrolysis procedure resulted in the chemical conversion of naringenin chalcone into naringenin. (A) Quercetin levels. (B) Kaempferol levels. Each bar represents the mean of two independent fruit readings from one transgenic plant. Black bars represent plant lines transformed with the control plasmid pSJ89. Unshaded bars represent plant lines transformed with pBBC50. (C) Relationship between quercetin and naringenin chalcone levels in peel. Each bar represents the mean of two independent fruit readings from one transgenic plant. Numbers prefixed with a G represent plants transformed with the control plasmid pSJ89. Numbers prefixed with a C represent plants transformed with pBBC50. The order in which plant line numbers appear does not correlate with parts A and B. Figure 3. Northern analysis of developing tomato fruit. Fruits were harvested at different developmental stages, denoted as follows: green (G), breaker (B), turning (T), and red (R), and separated into peel and flesh. Young leaves were harvested from mature tomato plants. RNA was separated on formaldehyde–agarose gels, blotted, and hybridized with Petunia (A) chs-a (Β) chi, and (C) fls probes. perhaps by feedback/forward regulation of mRNA or protein expression, or by stabilization of biosynthetic complexes containing CHI (ref. 12). Transcripts of chi were present at relatively high levels in leaf samples and in green, breaker, and turning flesh from pBBC50 transgenic plants (data not shown). Stability and phenotype of pBBC50 transgenic plants. A number of single-insert pBBC50 transgenic lines were selected for further evaluation over several subsequent generations. The high-flavonol phenotype was shown to segregate with the chi transgene in all generations tested so far (including fourth generation). There was no significant difference in the relative increase in flavonol levels over control in subsequent generations, or between hemizygous and homozygous populations (data not shown). Fruit flavonol levels remained elevated under open field trial and for the best lines approached those found in yellow onions, a crop that is naturally high in flavonols13 (Fig. 6). The vegetative phenotype of pBBC50 plants was indistinguishable from the parental variety. Fruit peel of pBBC50-transformed plants did not accumulate significant amounts of yellow-colored naringenin chalcone which may explain our observation of a slight dullness in the tone of red color in ripe fruit. Biochemical analysis indicated that the lycopene concentration of fruit had not been altered (data not shown). Properties of paste produced from high-flavonol tomato fruit. We have tested the effect of processing on the flavonol content of 472 nature biotechnology • VOLUME 19 pastes manufactured from pBBC50-transformed fruit using a benchtop scale processor, designed to mimic the factory scale operation. During the processing, peel and seeds are removed only after the initial heating step; as a consequence, the flavonols present in peel "leach" into the resulting paste. Concentrated paste manufactured from pBBC50-transformed fruit contained up to 1.9 mg flavonols per gram dry weight paste, a 21-fold increase over paste manufactured using the azygous control (data not shown). The paste retained 65% of the total flavonol content of the fresh fruit (data not shown), the remainder being destroyed during processing or discarded with the waste peel and seeds. We further established that these high-flavonol pastes were indistinguishable in their taste and flavor from azygous control pastes (data not shown). Conclusions. Constitutive overexpression of a Petunia gene encoding CHI in tomato resulted in elevated flavonol end products in the fruit peel. The tomato lines contained significantly increased levels of quercetin glycosides, and smaller but still substantial increases in kaempferol glycosides in fruit peel. These new varieties may offer opportunities for tomato-based products with an expanded range of potential health benefit properties. • MAY 2001 • http://biotech.nature.com © 2001 Nature Publishing Group http://biotech.nature.com RESEARCH ARTICLE © 2001 Nature Publishing Group http://biotech.nature.com A Figure 6. Flavonol levels of field trial transgenic fruit. Homozygous transgenic plants were grown in controlled open field trial conditions (1999 season, Stockton, California). Flavonols (quercetin plus kaempferol) were extracted under hydrolyzing conditions and analyzed by HPLC. Each bar represents 10 individual fruit (central columella and parenchyma included, seeds removed) readings from a transgenic plant line. Error bars representing the standard deviation of fruit in each line are shown. Black bars (–) represent fruit of azygous control plants. Unshaded bars (+) represent pBBC50-transformed plants. B diameter) from mature plants, separated fruit tissues, and whole fruit were harvested, frozen in liquid nitrogen, and stored at –80°C until use. Extraction and analysis of flavonoids. Flavonoids were determined either as aglycons or as flavonoid glycosides by preparing acid-hydrolyzed and nonhydrolyzed extracts, respectively. Frozen tissues were ground to a fine powder and lyophilized before extraction. Preparation of acid-hydrolyzed extracts was performed according to Hertog et al.17 with some modifications. Lyophilized material (50 mg) was hydrolyzed in 1.6 ml of 62.5% methanol and 0.4 ml 6 M HCl for 60 min at 90°C. Extracts were cooled on ice, diluted with 2 ml of 100% methanol, and sonicated at room temperature for 5 min. For nonhydrolyzed extracts, 50 mg of lyophilized material was extracted into 4 ml of 70% methanol by sonication at room temperature for 30 min. Extracts were filtered through 0.2 µm poly-tetrafluoroethylene (PTFE) filters before HPLC. HPLC was carried out on an HP1100 system (Hewlett Packard, Waldbronn, Germany) using a Nova-pak C18 (3.9 × 150 mm, particle size 4 µm) reverse-phase analytical column (Waters chromatography, Milford, MA). A photodiode array detector (Hewlett Packard) was used to record online spectra (from 200 nm to 550 nm) of compounds eluting from the column. HPLC separation of acid-hydrolyzed extracts was carried out under isocratic conditions of 25% acetonitrile in 0.1% trifluoroacetic acid (TFA) at a flow rate of 0.9 ml/min. Nonhydrolyzed extracts were separated using a gradient of acetonitrile in 0.1% TFA, at a flow rate of 1.0 ml/min: 5–25% linear in 30 min, then 25–30% in 5 min and 30–50% in 2 min, followed by a 3 min washing with 50% acetonitrile in 0.1% TFA. Peak purity, identification, and integration were carried out on Hewlett Packard Chemstations software version A.04.02 using commercially available flavonoid standards (Apin Chemicals, Ltd., Abingdon, Oxon, UK). The detection limit using this system was 0.05 µg flavonoid/ml extract, corresponding to ∼5 mg/kg dry weight and 0.5 mg/kg fresh weight tomato. RNA extraction and analysis. RNA was isolated according to the protocol of van Tunen et al16. For RNA gel blot analysis 10 µg of total RNA was loaded on formaldehyde agarose gels18 and electrophoresed overnight at 25 V. Separated RNA was blotted onto Hybond N+ membrane (Amersham Pharmacia Biotech, Bucks, UK). Petunia hybrida cDNA fragments encoding the flavonoid biosynthetic genes chs-a, chi, and fls were radiolabeled with 32P and purified using a RadPrime Labelling system (Gibco Life Technologies, Paisley, UK). Blots were hybridized overnight at 55°C and washed three times in 2 × sodium saline citrate, 1% sodium dodecyl sulfate, 55°C, 30 min before being subjected to autoradiography. Equal loading was checked by UV imaging of ribosomal RNA bands in all sample lanes before blotting. Figure 5. Chromatogram recorded at 370 nm, of nonhydrolyzed peel of control and pBBC50-transformed fruit. Tomatoes transformed with (A) control plasmid pSJ89 or (B) pBBC50 (plant no. C87). The major peaks corresponding to rutin (R), isoquercitrin (IQ), kaempferol rutinoside/quercetin glycoside (KR/QG)(co-eluting compounds), and a putative kaempferol glycoside (KG) are marked. Experimental protocol Plasmid construction. The Petunia chi gene was amplified from plasmid pMIP41, which contains the complete chi complementary DNA (cDNA) from P. hybrida inbred line V30 (ref.14). The chi coding sequence was cloned as a BamHI/SalI fragment into pFLAP10, a pUC derivative, containing a fusion of the double CaMV 35S promoter (Pd35S) and the Agrobacterium tumefaciens nos terminator (Tnos). A synthetic adapter fragment, consisting of a multiple cloning site (PacI/EcoRI/HindIII/AscI), was ligated into plasmid pGPTV-KAN (ref. 15) digested with EcoRI/HindIII, replacing the gusA-Tnos gene. The resulting plasmid was designated pBBC3. The Pd35S-chi-Tnos construct was transferred as a PacI/AscI fragment into pBBC3 and the resulting binary plasmid designated pBBC50. A GPTV-based binary plasmid (pSJ89) containing the β-glucuronidase gene under control of the CaMV 35S promoter and the nos terminator signal was used as a control plasmid. Generation of transgenic plants. Plasmids pBBC50 and pSJ89 were transferred to A. tumefaciens strain LBA4404, which was then used to transform tomato (var. FM6203, Unilever commercial variety) leaf disks according to standard protocols16. The transgenic status of kanamycin-positive explants was confirmed by PCR and Southern blot analysis. Plants were transferred to soil and grown in a glasshouse with a 16 h photoperiod and a 23°C day/18°C night temperature. Plants were allowed to self-pollinate for the production of subsequent generations. Homozygous lines were selected on the basis of PCR screening. Harvest of transgenic material. Fully red ripe glasshouse-grown fruits were harvested between 15 and 21 days post-breaker stage (breaker stage corresponds to the first flush of color on the green fruit). The outer layer of 2 mm thick (i.e. cuticula, epidermal layer plus some subepidermal tissue) was separated from the fruit and classified as peel. The columella, jellylike parenchyma, and seeds were removed and the remainder of the fruit was classified as flesh. Field trial fruits were harvested when visually ripe (i.e., fully red). The columella, parenchyma, and seeds were removed. The remainder of the fruit was classified as whole fruit. Young leaves (∼3–5 cm http://biotech.nature.com • MAY 2001 Acknowledgments The authors wish to thank Dr. Adel Elsheikh and Dr. Jinguo Hu (Lipton Innovation Centre, Stockton, CA) for their assistance with transgenic field trials. We also thank Bob Cowper, Carl Jarman, Tracey Macdonald, and Katherine Redwood (URC, Sharnbrook, UK) for their technical contributions. Received 16 October 2000; accepted 23 February 2001 • VOLUME 19 • nature biotechnology 473 © 2001 Nature Publishing Group http://biotech.nature.com © 2001 Nature Publishing Group http://biotech.nature.com RESEARCH ARTICLE 11. Krause, M. & Galensa, R. Determination of naringenin and naringenin chalcone in tomato skins by HPLC after solid phase extraction. Z. Lebensmittel 194, 29–32 (1992). 12. Burbulis, I.E. & Winkel-Shirley, B. Interactions among enzymes of the Arabidopsis flavonoid biosynthetic pathway. Proc. Natl. Acad. Sci. USA 96, 12929–12934 (1999). 13. Ewald, C., Fjelkner-Modig, S., Johansson, K., Sjoholm, I. & Akesson, B. Effect of processing on major flavonoids in processed onions, green beans, and peas. Food Chem. 64, 231–235 (1999). 14. van Tunen, A.J. et al. Cloning of the 2 chalcone flavanone isomerase genes from Petunia—co-ordinate light-regulated and differential expression of flavonoid genes. EMBO J. 7, 1257–1263 (1988). 15. Becker, D., Kemper, E., Schell, J. & Masterson, R. New plant binary vectors with selectable markers located proximal to the left T-DNA border. Plant Mol. Biol. 20, 1195–1197 (1992). 16. Fillatti, J.J., Kiser, J., Rose, R. & Comai, L. Efficient transfer of a glyphosate tolerance gene into tomato using a binary Agrobacterium tumefaciens vector. Bio Technology 5, 726–730 (1987). 17. Hertog, M.G.L., Hollman, P.C.H. & Venema, D.P. Optimization of a quantitative HPLC determination of potentially anticarcinogenic flavonoids in vegetables and fruits. J. Agric. Food Chem. 40, 1591–1598 (1992). 18. Sambrook, J. et al. Molecular cloning. A laboratory manual. (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY; 1989). 1. Shintani, D. & DellaPenna, D. Elevating the vitamin E content of plants through metabolic engineering. Science 282, 2098–2100 (1998). 2. Ye, X. et al. Engineering the provitamin A (beta-carotene) biosynthetic pathway into (carotenoid-free) rice endosperm. Science 287, 303–305 (2000). 3. Romer, S. et al. Elevation of the provitamin A content of transgenic tomato plants. Nat. Biotechnol. 18, 666–669 (2000). 4. Middleton, E. & Kandaswami, C. The impact of plant flavonoids on mammalian biology: implications for immunity, inflammation and cancer. In The flavonoids: advances in research since 1986. (ed. Harbourne, J.B.) 619–652 (Chapman and Hall, London; 1994). 5. Hertog, M.G.L. et al. Flavonoid intake and long term risk of coronary heart disease and cancer in the 7 Countries study. Arch. Intern. Med. 155, 381–386 (1995). 6. Hertog, M.G.L., Feskens E.J.M., Hollman P.C.H., Katan M.B. & Kromhout D. Dietary antioxidant flavonols and risk of coronary heart disease risk—the Zutphen elderly study. Lancet 342, 1007–1011 (1993). 7. Knekt, P., Jarvinen, R., Reunanen, A. & Maatela, J. Flavonoid intake and coronary mortality in Finland: a cohort study. Br. Med. J. 312, 478–481 (1996). 8. Weisshaar, B. & Jenkins, G.I. Phenylpropanoid biosynthesis and its regulation. Curr. Opin. Plant Biol. 1, 251–257 (1998). 9. Mol, J., Grotewold, E. & Koes, R. How genes paint flowers and seeds. Trends Plant Sci. 3, 212–217 (1998). 10. Dixon, R.A. & Steele, C.L. Flavonoids and isoflavonoids—a goldmine for metabolic engineering. Trends Plant Sci. 4, 394–400 (1999). 474 nature biotechnology • VOLUME 19 • MAY 2001 • http://biotech.nature.com