Genetically stable expression of functional ... - Wiley Online Library

Plant Biotechnology Journal (2007) 5, pp. 768–777

doi: 10.1111/j.1467-7652.2007.00283.x

Genetically stable expression of functional miraculin, a new type of alternative sweetener, in transgenic tomato plants Hyeon-Jin Original Stable expression Articles Sun et al. of miraculin in GM tomato Blackwell Oxford, Plant PBI © 1467-7644 XXX 2007 Biotechnology UK Blackwell Publishing Publishing Journal Ltd Ltd

Hyeon-Jin Sun, Hiroshi Kataoka, Megumu Yano and Hiroshi Ezura* Graduate School of Life and Environmental Sciences, Gene Research Center, University of Tsukuba, Tennodai 1-1-1, Tsukuba, Ibaraki 305-8572, Japan

Received 27 April 2007; revised 27 June 2007; accepted 29 June 2007. *Correspondence (fax +81 29 853 7263; e-mail: [email protected])

Summary Miraculin is a taste-modifying protein isolated from the red berries of Richadella dulcifica, a shrub native to West Africa. Miraculin by itself is not sweet, but it is able to turn a sour taste into a sweet taste. This unique property has led to increasing interest in this protein. In this article, we report the high-yield production of miraculin in transgenic tomato plants. High and genetically stable expression of miraculin was confirmed by Western blot analysis and enzyme-linked immunosorbent assay. Recombinant miraculin accumulated to high levels in leaves and fruits, up to 102.5 and 90.7 µg/g fresh weight, respectively. Purified recombinant

Keywords: low-calorie sweetener,

miraculin expressed in transgenic tomato plants showed strong sweetness-inducing activity,

miracle fruit, miraculin, sweetness-

similar to that of native miraculin. These results demonstrate that recombinant miraculin

inducing activity, taste-modifying

was correctly processed in transgenic tomato plants, and that this production system could

protein, transgenic tomato.

be a good alternative to production from the native plant.

Introduction Most taste-active substances are low-molecular-mass compounds, but proteins of higher molecular mass that are sweet or modify taste also exist (Witty, 1998). Nine sweet-tasting and taste-modifying proteins are known: thaumatin (van der Wel and Loeve, 1972), monellin (Morris and Cagan, 1972), mabinlin (Liu et al., 1993), brazzein (Ming and Hellekant, 1994), pentadin (van der Wel et al., 1989), egg white lysozyme (Masuda et al., 2001), curculin (Yamashita et al., 1990), neoculin (Shirasuka et al., 2004; Suzuki et al., 2004) and miraculin (Kurihara and Beidler, 1968; Theerasilp and Kurihara, 1988). With the exception of lysozyme, all of these proteins are derived from the fruits of tropical plants, and have long been in traditional use by indigenous peoples for sweetening foodstuffs. As these proteins can elicit the same sweetening effect as sucrose at very low concentrations, and are natural, they can be used in diabetic and dietetic foods, and are more acceptable to consumers than synthetic compounds (Witty, 1998; Faus, 2000). Although thaumatin is currently available to the public (Witty and Higginbotham, 1994), the commercial feasibility of these proteins is limited, because their natural

768

sources are tropical plants that are difficult to grow outside their normal environments. Therefore, many attempts have been made to produce these proteins in foreign hosts by biotechnological approaches (reviewed in Faus, 2000; Masuda and Kitabatake, 2006). Miraculin is a taste-modifying protein isolated from miracle fruit, the red berries of Richadella dulcifica, a shrub native to West Africa. Indigenous peoples often use these berries to improve the palatability of their acidic maize dishes and to sweeten sour beverages. The active principle of the berry is the protein miraculin, which by itself is not sweet. Unlike sweet proteins, it has the unusual property of being able to turn a sour taste into a sweet taste. Sour foods, such as lemons, limes and grapefruit, taste sweet when tasted together with this protein. The sweetness induced by citric acid after exposure to miraculin has been estimated to be around 3000 times that of sucrose on a weight basis (Kurihara and Beidler, 1969; Theerasilp and Kurihara, 1988; Gibbs et al., 1996; Kurihara and Nirasawa, 1997). As a result of this amazing property, there is increasing interest in this protein. In the 1970s, an attempt by Miralin Co. to commercialize miraculin as a dietetic aid in the USA was denied by the Food and Drug © 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd

Stable expression of miraculin in GM tomato 769

Administration (Gibbs et al., 1996). BioResources International, Inc. (Somerset, NJ, USA) is currently undertaking the commercial development of miraculin for use as a tastemasking agent, low-calorie sweetener and flavour enhancer. The company holds a patent for a method of purifying miraculin from miracle fruit (US Patent 5886155). In Japan, miracle fruit is marketed as a fruit and also in powder and tablet forms. The Japanese horticultural researcher Mitsuharu Shimamura has prepared tablets from miracle fruit to increase the availability of miraculin. The tablets are commercially available in Japan under the trademark ‘Miracle Fruit Tablet’ (Onishikatsumi-Shokai Corp., Osaka, Japan). These products are used and particularly popular amongst diabetics and dieters in Japan. The material currently on the market is a native form of miraculin that is derived from miracle fruit. Although miraculin has great potential as an alternative low-calorie sweetener for diabetic and dietetic purposes, there are limited natural sources of this protein. Therefore, attempts have been made to produce miraculin in foreign hosts, such as Escherichia coli (Kurihara, 1992), yeast (Kurihara and Nirasawa, 1997) and transgenic tobacco (Kurihara and Nirasawa, 1997). Although miraculin has been expressed in these organisms, the resulting recombinant miraculins do not have taste-modifying activity. Recently, we have expressed recombinant miraculin in transgenic lettuce plants (Sun et al., 2006a). The recombinant miraculin was correctly folded and had taste-modifying activity as powerful as that of native miraculin. Although transgene silencing occurred in later generations of the lettuce plants, the results suggest that transgenic plants can be good alternative vehicles for producing a biologically active form of miraculin. In this study, we expressed the taste-modifying protein miraculin as a potential alternative sweetener in an important agricultural crop, tomato. Miraculin was successfully produced in transgenic tomato leaves and fruit, and showed the same biological activity as native miraculin in both tissues. In addition, miraculin expression in transgenic tomato was genetically stable.

Results Transformation of tomato with the miraculin gene Tomato cotyledons were transformed by infection with the Agrobacterium tumefaciens strain GV2260 (Deblaere et al., 1985) containing the binary vector 35S-MIR (Figure 1a), which has been prepared previously (Sun et al., 2006a). In this construct, miraculin gene expression is driven by the constitutive cauliflower mosaic virus (CaMV) 35S promoter.

Figure 1 Map of the T-DNA region of the binary vector 35S-MIR adopted from Sun et al. (2006a) (a) and Southern blot analysis of tomato transformants (b). LB, T-DNA left border; miraculin, miraculin gene; NPTII, neomycin phosphotransferase gene; P35S, cauliflower mosaic virus (CaMV) 35S promoter; Pnos, nopaline synthase gene promoter; RB, T-DNA right border; Tnos, nopaline synthase gene terminator. Genomic DNAs from 14 transformants (lanes 2A–15A) and an untransformed tomato plant (Wt) were used for Southern blot analysis.

Fourteen kanamycin-resistant tomato lines were regenerated on selective medium, and the integration of the miraculin gene in these plants was confirmed by Southern blot analysis (Figure 1b). The DNA from each plant, including an untransformed tomato plant, contained two bands that migrated near the top of the membrane. This result indicates that tomato contains a miraculin homologue, possibly the LeMir gene, a tomato root-knot nematode-induced gene (Brenner et al., 1998). Therefore, the transgene copy numbers in the transgenic tomato genomes were estimated, excluding these two bands. The genomes of nine of the plants carried one copy of the miraculin gene, and the other plants contained at least two copies of the gene. Fourteen independent transgenic plant lines were obtained (Figure 1b). Measurement of the ploidy level of the transgenic tomato plants by flow cytometry revealed ten diploid lines and four tetraploid lines (data not shown). The tetraploid lines arose during the transformation procedure.

© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Plant Biotechnology Journal, 5, 768–777

770 Hyeon-Jin Sun et al.

Expression and stability of miraculin in transgenic plants To confirm the expression of miraculin in transgenic plants, total soluble protein was extracted from the leaves of T0 transgenic plants and analysed by Western blot. Figure 2a shows that miraculin accumulated to similar levels in ten of the 14 transgenic lines. The molecular mass of miraculin in these lines was very similar to that of native miraculin under both non-reducing (45 kDa; Figure 2a, top) and reducing (28 kDa; Figure 2a, bottom) conditions. However, the transgenic lines 14C and 20A showed lower expression levels, and the lines 3A and 11A failed to express miraculin to detectable levels (Figure 2a), suggesting that transgene silencing may have occurred in these lines. These four transgenic lines were tetraploid and expressed the miraculin gene at the transcriptional level (Figure 2b). These results suggest that the change in ploidy influenced transgene expression. Extracts from an untransformed tomato plant did not cross-react with an anti-miraculin antibody (Figure 2a). There was no correlation between the number of transgene copies in the genome and the level of miraculin accumulation. The T0 transgenic tomato lines expressing miraculin were transferred to soil and grown in a glasshouse. Transgenic plant seeds generated through self-pollination were collected from T0 transgenic lines containing a single copy of the miraculin gene and germinated in soil to analyse the stability of transgene expression and inheritance. The presence and expression of the transgene in 16 T1 seedlings of the transgenic line 56B were examined by genomic polymerase chain reaction (PCR) and Western blot analysis (Figure 3a). The miraculin expression level in the leaves of the 56B T1 progeny plants was similar to that of the T0 generation plant (Figure 3a, top). Genomic PCR analysis of the T1 progeny plants revealed the presence of the miraculin gene in a 3 : 1 Mendelian ratio (Figure 3a, bottom), indicating a single insertion of the transgene. There was a good correlation between miraculin expression, as monitored by Western blotting, and the presence of the miraculin gene in genomic PCR (Figure 3a). All of the T1 seedlings from different T0 plants with single-copy insertions of the transgene showed almost the same results as the transgenic line 56B (data not shown). The miraculin levels were also examined in fruits of T0 and T1 plants. Total soluble proteins extracted from untransformed tomato plants and T0 and T1 plants of the transgenic line 56B were subjected to Western blot analysis. Figure 3b shows that miraculin accumulated to similar levels in the T0 and T1 tomato plants. Moreover, miraculin was

Figure 2 Miraculin accumulation in the T0 generation of transgenic tomato plants. (a) Western blot analysis of transgenic tomato leaves under non-reducing (top) and reducing (bottom) conditions. DTT, dithiothreitol; MIR, purified native miraculin as a positive control; Wt, crude extracts from untransformed tomato plants; 2A–15A, crude extracts from transgenic tomato lines. Proteins were extracted from leaf tissues. The sizes of protein standards are shown in kilodaltons to the left. (b) Reverse transcriptase-polymerase chain reaction (RT-PCR) analysis of tetraploid plants. Miraculin, miraculin gene; Ribosomal, tomato 25S rRNA; Wt, untransformed tomato plants (diploid plant); 56B, transgenic tomato line (diploid plant); 3A, 11A, 20A, 14A, transgenic tomato lines (tetraploid plants).



Figure 3 Stable expression of miraculin in transgenic tomato plants. (a) Western (top) and genomic polymerase chain reaction (PCR) (bottom) analyses of T1 generation transgenic plants derived from the 56B line. Wt, crude extract from an untransformed tomato leaf; lane 1, crude extract from a T0 generation transgenic tomato leaf (56B line); lanes 2–17, crude extracts from T1 progeny leaves of the transgenic tomato line 56B. (b) Comparison of miraculin expression in T0 and T1 generation tomato plants. Crude extracts from tomato leaves (lanes 1, 3 and 5) and fruit (lanes 2, 4 and 6) were used. MIR, purified native miraculin (500 ng); lanes 1 and 2, untransformed plants; lanes 3 and 4, T0 generation transgenic plant line 56B; lanes 5 and 6, T1 generation plants of transgenic line 56B. (c) Analysis of miraculin expression during the development of transgenic tomato fruits. Wt, crude extract from untransformed tomato fruit; L, crude extract from a T1 generation transgenic tomato leaf of line 56B; lanes 6, 30, 45 and 60, crude extracts from the corresponding stages of T1 transgenic tomato fruit of line 56B. The sizes of protein standards are shown in kilodaltons to the left. (d) Comparison of miraculin expression in T1 and T2 generation tomato fruit. Crude extracts from T2 progeny tomato fruits of the 56B and 7C lines were analysed. Wt, crude extract of untransformed tomato plants; T1F56B, crude extract from T1 generation tomato fruit of line 56B; T1F7C, crude extract from T1 generation tomato fruit of line 7C; lanes 1–9, crude extracts from T2 progeny tomato fruits; MIR, purified native miraculin (500 ng).

present in both leaves and fruits. In addition, the molecular masses of the recombinant miraculins were similar to that of native miraculin, although that in fruit was slightly smaller than that in leaf. This difference in size may be the result of the presence of a glycoform, as miraculin is a glycoprotein (Theerasilp and Kurihara, 1988). Figure 3c shows the accumulation of miraculin during the development of transgenic tomato fruits. Tomato fruits (T1 progeny of the 56B line) were harvested at four different developmental

stages: 6 (cell division), 30 (cell expansion), 45 (mature green) and 60 (ripe red) days after flowering (DAF), and analysed by Western blotting. Miraculin accumulated to almost the same level throughout fruit development. As expected from the use of the constitutive CaMV 35S promoter, recombinant miraculin was present in all tissues (leaf, stem, root and fruit) (data not shown). Figure 3d shows the expression of miraculin in T2 generation tomato fruit from lines 56B and 7C. The miraculin expression levels in T2 fruit were similar to those in



Table 1 Concentration of miraculin in transgenic tomato plants Transgenic tomato lines 56B line T0 Leaf (Fruit) µg miraculin/g fresh weight* mg protein/g fresh weight µg miraculin/mg protein

7C line T1 Leaf (Fruit)

T2 Leaf (Fruit)

T0 Leaf (Fruit)

T1 Leaf (Fruit)

T2 Leaf (Fruit)

62.7 ± 7.1

102.5 ± 3.2

82.4 ± 8.8

85.2 ± 5.0

78.2 ± 5.7

90.1 ± 14.2

(55.4 ± 4.1)

(83.8 ± 9.8)

(90.7 ± 14.6)

(59.7 ± 18.3)

(45.3 ± 4.5)

(63.3 ± 8.6)

11.4

14.8

13.2

12.1

13.5

13.8

(12.3)

(15.2)

(12.6)

(13.6)

(14.5)

(14.7)

5.5

6.9

6.2

7.0

5.8

6.5

(4.5)

(5.5)

(7.2)

(4.4)

(3.1)

(4.3)

*The miraculin concentrations in protein extracts obtained from transgenic tomato plants (T0, T1 and T2 generations) were determined using the enzyme-linked immunosorbent assay (ELISA) of Sun et al. (2006a). Data represent means ± standard deviation (n = 3). Total soluble protein concentrations were determined using a BCA Protein Assay Kit (Pierce, Rockford, IL, USA).

fruit of the parent plant, indicating that miraculin expression is stable in succeeding generations. The miraculin contents in leaves and fruits of transgenic tomato plants (T0, T1 and T2 generations) were estimated using an enzyme-linked immunosorbent assay (ELISA). The results from two transgenic plants (lines 56B and 7C) are presented in Table 1. The miraculin concentrations in these lines were in the ranges 62.7–102.5 and 45.3–90.7 µg/g fresh weight in leaves and fruits, respectively. The level of miraculin in T2 progeny plants was similar to that in T0 and T1 generation plants. These results further confirm that the miraculin gene is stably expressed and inherited in transgenic tomato plants as the generations advance.

Characterization of post-translational modifications of miraculin expressed in transgenic tomato plants Miraculin has been reported to be a disulphide-linked dimer with two N-linked oligosaccharide chains attached to asparagine-42 (Asn42) and Asn186 (Theerasilp et al., 1989; Kurihara, 1992). In order to characterize the post-translational modifications of recombinant miraculin, miraculin was purified from the transgenic tomato plant lines. About 300 and 400 µg of purified miraculins were obtained from 100 g of fresh tomato fruit and leaves, respectively. The purification procedure was repeated multiple times to obtain sufficient amounts of purified miraculin for use in subsequent experiments. In addition, native miraculin was purified from the pulp of miracle fruit, as described previously (Theerasilp and Kurihara, 1988), resulting in about 5500 µg of purified miraculin from 20 g of fresh miracle fruit. The purified miraculins migrated as single bands in sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE)

(Figure 4a, lanes 1–3). To confirm whether miraculin formed a disulphide-linked dimer during post-translational processing, the three purified miraculins were subjected to non-reducing and reducing SDS-PAGE. As shown in Figure 4a, under nonreducing (lanes 1–3) and reducing (lanes 4–6) conditions, miraculin migrated at about 45 and 28 kDa, respectively, suggesting that miraculin expressed in transgenic tomato plants formed a disulphide-linked homodimer. The tomato fruit miraculin (lanes 3 and 6) was slightly smaller than the leaf miraculin (lanes 2 and 5). These different sizes may reflect glycoforms of the protein, as glycoproteins generally exist as populations of glycosylated variants. To ascertain whether the tomato miraculin was glycosylated, the purified miraculin was treated with peptide N-glycosidase A (PNGase A, Diagnostics GmbH, Mannheim, Germany). As shown in Figure 4b, after this treatment, both the tomato and native miraculins migrated as smaller bands, with molecular masses of approximately 24 kDa. This result indicates that the recombinant miraculin produced in transgenic tomato plants was N-glycosylated.

Assay of taste-modifying activity Miraculin is able to convert a sour taste into a sweet taste. To confirm whether the tomato-produced miraculin possessed taste-modifying activity, purified miraculin from tomato plants was subjected to sensory evaluation, as described previously (Kurihara and Beidler, 1969). Purified native miraculin was included in the assay as a reference. The results are summarized in Table 2. The sweetness intensity induced by 0.02 M citric acid after 0.4 µM purified miraculin solution was held in the mouth was equivalent to that of about 0.3 M sucrose. This value is equal to the maximum sweetness



content of these samples was higher than the 0.4 µM solution of purified miraculin, the induced sweetness intensity was lower than that of the purified protein. Tomato leaves showed relatively uniform taste-modifying activity, but the sweetness induced by tomato fruits was variable.

Discussion In this article, we have described the genetically stable expression of the taste-modifying protein miraculin in transgenic tomato plants. Transgenic plants have been developed as recombinant protein production systems (Howard and Hood, 2005), and the utility of these systems as bioreactors has been reviewed (Hood and Jilka, 1999; Yoshida and Shinmyo, 2000; Daniell et al., 2001; Hood et al., 2002; Horn

Figure 4 Purification and post-translational modification of recombinant miraculin produced in transgenic tomato plants. (a) Analysis of the dimerization of recombinant miraculin produced in tomato plants. Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed under non-reducing (lanes 1–3) and reducing (lanes 4–6) conditions. Lanes 1 and 4, purified native miraculin from miracle fruit; lanes 2 and 4, purified recombinant miraculin from transgenic tomato leaves; lanes 3 and 6, purified recombinant miraculin from transgenic tomato fruit. The sizes of protein standards are shown in kilodaltons to the right. (b) Analysis of the N-glycosylation of recombinant miraculin produced in transgenic tomato. SDS-PAGE was performed under reducing conditions. MIR, purified miraculin from miracle fruit; TL, purified recombinant miraculin from transgenic tomato leaves; TF, purified recombinant miraculin from transgenic tomato fruit. The sizes of protein standards are shown in kilodaltons to the left.

induced by miraculin (Kurihara and Nirasawa, 1997). In our assay, the sweetness-inducing activity of purified tomato miraculin was similar to that of purified native miraculin. These results clearly demonstrate that recombinant miraculin was accurately processed in tomato plants. The tastemodifying activity was also assayed using miraculin-expressing tomato leaves and fruits. The sweetness levels induced by 2 g of fruit and leaf tissues were equal to about 0.14 and 0.21 M sucrose, respectively (Table 2). Although the miraculin

et al., 2004; Streatfield, 2007). In brief, plant expression systems have several advantages, including low production costs, lack of animal pathogens and ease of increasing production to an agricultural scale. Moreover, plant cells are capable of complex post-translational modifications, such as glycosylation, that cannot be performed in bacteria. In this article, we have shown that a plant expression system can be used for the production of the taste-modifying protein miraculin. Because of this unusual property, the protein has potential use as a sweetener for diabetic and dietetic purposes. However, the use of this protein is restricted by the limited availability of its natural source (Witty, 1998). Therefore, we focused on transgenic plants as an alternative production system to increase the availability of the protein. We had previously expressed miraculin in the leaves of transgenic lettuce plants (Sun et al., 2006a). The miraculin in lettuce plants formed a disulphide-linked dimer and was glycosylated similarly to native miraculin. Moreover, the taste-modifying capability of this miraculin was as strong as that of native miraculin. However, the level of miraculin accumulation in T1 and T2 progeny was much lower than that in T0 plants. Many recent studies have shown that gene silencing occurs in transgenic plants (Finnegan and McElroy, 1994; Stam et al., 1997), and is thus a potential problem, one that needs to be overcome in transgenic plant systems. A high level of expression and the stability of the recombinant protein in transgenic plant systems are necessary for commercial applications. In plant-based antigen production, the expression levels of certain proteins vary widely with the expression system and plant species (Streatfield and Howard, 2003). Tomato is one of the most important vegetable crops and a genetic model for improving other crop plants (McCormick et al., 1986). Tomato is now grown worldwide for its edible fruits and has an efficient transformation protocol



Table 2 Sweetness intensities induced by native and recombinant miraculin Source material

Concentration of miraculin in taste sample (µg)*

Induced sweetness (SEV)† (M)

Miracle fruit (about 1 g of one fresh fruit)

178.8 ± 36.1

0.32 ± 0.01

Tomato leaf tissue (2 g of fresh weight)

161.0 ± 22.1

0.21 ± 0.02

Tomato fruit tissue (2 g of fresh weight)

133.2 ± 26.8

0.14 ± 0.03

Purified native miraculin

90

0.29 ± 0.02

Purified recombinant miraculin (leaf)

90

0.31 ± 0.02

Purified recombinant miraculin (fruit)

90

0.29 ± 0.01

*The miraculin concentration in miracle fruit and tomato tissues was determined using enzyme-linked immunosorbent assay (ELISA). Values are the means ± standard deviation (n = 10). The protein content in purified miraculin solutions was determined using a BCA Protein Assay Kit (Pierce, Rockford, IL, USA). †The sucrose equivalence value (SEV), corresponding to the sweetness intensity induced by 0.02 M citric acid, was evaluated by comparing the sweetness after exposure to the above samples with that of a series of standard sucrose solutions (0.1–0.5 M). Data represent means ± standard error (n = 10).

(Sun et al., 2006b). Moreover, tomato has well-established industrial culture and downstream processing. For these reasons, we decided to use the important edible plant, tomato, to produce miraculin. Recombinant miraculin was expressed in transgenic tomato plants, and was properly processed. Miraculin accumulated in transgenic tomato plants to levels higher than those in transgenic lettuce plants (Sun et al., 2006a). The sweetness induced by 0.02 M citric acid after exposure to purified miraculin from transgenic tomato plant tissues was equivalent to the sweetness of 0.3 M sucrose, the same as that induced by purified native miraculin. Most importantly, recombinant miraculin was stably expressed in the T1 and T2 generations of transgenic tomato plants, unlike the situation in transgenic lettuce plants (Sun et al., 2006a). Although the transgenes in the two expression systems were driven by the same promoter, some differences in expression were apparent. In transgenic lettuce plants, the level of miraculin accumulation was higher in single-copy transformants than in multicopy transformants, whereas the transgene copy number did not affect the amount of miraculin that accumulated in transgenic tomato plants. However, diploid tomato plants showed higher expression levels than tetraploid plants. These results suggest that the choice of plant species is an important factor in the stable production of specific proteins, and tomato (cultivar Moneymaker) is a more suitable host than lettuce (cultivar Kaiser) for miraculin production. The variability in the taste-modifying activity from fruit to fruit may have been caused by different miraculin concentrations in individual fruits. Even though miraculin appeared to accumulate to similar levels in succeeding generations of transgenic tomato lines, ELISA and activity assays showed that the expression levels varied. The reason for this effect is unknown, but there is at least one other published instance of variable expression of a protein in tomato fruits and

leaves of the same plant (Walmsley et al., 2003). This effect is inherent in transgenic plant systems and is an important factor to address. Interestingly, increases in fruit Brix (total soluble solids content, percentage sugar) were observed that roughly correlated with increases in miraculin activity. In tomato fruit, sugars and organic acids are the major metabolites that contribute to Brix (Grierson and Kader, 1986), and there is significant variability in the metabolite content (Schauer et al., 2005). This result suggests that a mechanism that causes high Brix affects miraculin activity. However, because the analysis was conducted on a limited number of fruits, a more detailed study needs to be performed, and we are therefore cultivating transgenic tomato plants subjected to stresses to produce high-Brix fruit. In addition, we are currently investigating new strategies for increasing miraculin production in transgenic plants using specific promoters and other plant species. As miracle fruit contains about 180 µg miraculin/g fresh weight (Table 2), and the fruit yield per 0.1 ha is about 500 kg, the possible miraculin production is 90 g per 0.1 ha. By contrast, miraculin-expressing transgenic tomato fruits contain about 90 µg miraculin/g fresh weight (Table 1), and the fruit yield per 0.1 ha in glasshouse production in Japan is about 10 000 kg (Nakano et al., 2006); therefore, the probable miraculin production in transgenic tomato is 900 g per 0.1 ha, giving a miraculin productivity ten times that of miracle fruits. For the fresh use of transgenic tomato, 1 g of fresh fruit contains 90 µg of miraculin, which is sufficient to induce maximum sweetness. However, the sweetnessinducing activity varies depending on the fruits, and therefore techniques to reduce the variability will be required for the fresh use of transgenic tomato. In conclusion, the production of recombinant miraculin in transgenic tomato plants has provided a new method to enhance the availability of the protein. These results open up



the possibility of the mass production of miraculin as a new type of low-calorie sweetener for diabetic and obese individuals.

Experimental procedures Transformation of tomato plants Tomato (Solanum lycopersicum cultivar Moneymaker) plants were transformed as described by Sun et al. (2006b) by infection with Agrobacterium tumefaciens strain GV2260 (Deblaere et al., 1985), harbouring the binary vector 35S-MIR, which contains the miraculin coding region (GENBANK accession number D38598), and was generated in a previous study (Sun et al., 2006a). After incubation in conditioned medium, putative transformants were selected by kanamycin resistance. The plants were maintained at 25 °C under a 16-h light/8-h dark photoperiod of fluorescent light at an intensity of 60 µmol/m2/s. Transgenic tomato plants were transferred to soil and grown in a glasshouse. Seeds were collected from two successive generations (T1 and T2) and were grown in a glasshouse for further experiments.

Genomic PCR and Southern blot analysis To confirm the presence of the miraculin gene in transgenic plants, genomic DNA was isolated from 0.5 g of fresh young tomato leaves, as described by Rogers and Bendich (1985). The presence of the miraculin gene was confirmed by PCR using specific primers (forward primer, 5′-T T T TCTAGAATGAAGGAAT TAACAATGCT-3′; reverse primer, 5′-TT TGAGCTCTTAGAAGTATACGGT T TTGT-3′). A sample of 100 ng of genomic DNA was used as the template, and the PCR cycling conditions were as follows: 35 cycles of 94 °C for 30 s, 55 °C for 30 s and 72 °C for 60 s. The PCR products were analysed on 1% agarose gels. For Southern analysis, 20 µg of genomic DNA was digested with XbaI, which cleaves only once outside the miraculin gene, and the fragments were resolved in a 1% agarose gel at 50 V for 4 h, and transferred to a Hybond-N+ nylon membrane (GE Healthcare UK Ltd., Amersham, Buckinghamshire, UK). A thermostable alkaline phosphatase (AP)-labelled miraculin gene-specific probe was generated using a CDP-Star AlkPhos Direct Labelling Kit, following the manufacturer’s instructions (GE Healthcare UK Ltd.). The membrane was hybridized overnight at 55 °C with the probe. Hybridization signals were detected by chemiluminescence using CDPstar (Roche Diagnostics GmbH, Mannheim, Germany), followed by exposure to X-ray film (Hyperfilm ECL, GE Healthcare UK Ltd.).

Analysis of ploidy level Leaves of approximately 14 days old from transgenic tomato plants, as well as an untransformed tomato plant, were used for the isolation of nuclei. Leaf pieces (0.2 g) were chopped individually on a plastic plate with a sharp razor blade in 400 µL of nucleus isolation buffer (Partec, Munster, Germany). After chopping, the samples were passed through a 30-µm nylon filter, and 4,6-diamino-2-phenylindole (DAPI, Partec) was added to stain the DNA. The DNA contents of the isolated nuclei were measured using a Partec PAS-II flow

cytometer. For calibration, the 2C peak from the nuclei of young leaves of diploid tomato seedlings was used.

Total RNA preparation and reverse transcriptasepolymerase chain reaction (RT-PCR) analysis Total RNA was isolated from tomato leaves using an RNeasy Plant Mini Kit (Qiagen, Hilden, Germany), according to the manufacturer’s instructions. First-strand cDNA was synthesized from total RNA using the SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen, Karlsruhe, Germany), following the manufacturer’s instructions. The resulting first-strand cDNA was amplified by PCR using the primer set described in the ‘Genomic PCR and Southern blot analysis’ section, and the following PCR cycling conditions: 94 °C for 9 min; 40 cycles of 94 °C for 30 s, 55 °C for 30 s and 72 °C for 60 s; and 72 °C for 5 min. As a reference, a tomato 25S ribosomal RNA gene (GENBANK accession number AW217955) was also amplified using the following primers: forward, 5′-GCT TGACTCTAGTCCGACT T-3′; reverse, 5′-CAGTCCTCGAAGAGTGGTAT-3′.

Protein extraction, Western blot analysis and ELISA The miraculin accumulation levels in transgenic tomato plants were determined using immunological measurement. Transgenic tomato plant tissues (leaf and fruit) were ground to fine powder in liquid nitrogen, and homogenized in three volumes of extraction buffer consisting of 20 mM Tris-HCl (pH 8.0), 500 mM NaCl and 2% polyvinylpolypyrrolidone (PVPP). The extracts were centrifuged at 15 000 g for 20 min at 4 °C, and the resulting supernatants were subjected to Western blot analysis and ELISA. The protein concentrations in the extracts were determined using a BCA Protein Assay Kit (Pierce, Rockford, IL, USA). The extracts (25 µL per lane) were resolved by SDS-PAGE in 12% gels and electrophoretically transferred to Hybond-P membrane (GE Healthcare UK Ltd.). The blots were reacted with an affinity-purified anti-miraculin antibody (Sun et al., 2006a), followed by incubation with anti-rabbit immunoglobulin G (IgG) coupled to horseradish peroxidase. Immunoreactive signals were detected using an Immun-Blot Assay Kit (Bio-Rad, Hercules, CA, USA), according to the manufacturer’s instructions. The amounts of miraculin in the transgenic tomato plants were determined by quantitative ELISA, as described by Sun et al. (2006a).

Purification of recombinant miraculin Native miraculin was purified from berries of R. dulcifica, as described previously (Theerasilp and Kurihara, 1988). Recombinant miraculin was purified from transgenic tomato tissues (leaf and fruit). Tissues (100 g fresh weight) were collected and ground to a fine powder in liquid nitrogen, and the proteins were extracted as described in the previous section. The proteins in the solutions were precipitated by the addition of solid ammonium sulphate to about 55% saturation. The precipitates were collected by centrifugation at 15 000 g for 20 min, resuspended in 20 mL of 0.02 M sodium phosphate (pH 6.8) and dialysed against the same buffer. The dialysed solutions were applied to a column (3.6 × 23 cm; bed volume, 234 mL) of CM-Sepharose Fast Flow (GE Healthcare UK Ltd.), which had been equilibrated with 0.02 M sodium phosphate (pH 6.8), at a flow



rate of 30 mL/h. The adsorbed substances were eluted with a linear gradient, starting with 200 mL of 0.02 M sodium phosphate (pH 6.8) in a mixing flask and 200 mL of the same buffer solution containing 1.0 M NaCl. The fractions containing miraculin were collected and dialysed against 0.02 M sodium phosphate (pH 6.8) containing 0.5 M NaCl. The dialysed solutions were applied to a ConA Sepharose 4B (GE Healthcare UK Ltd.) column (1.8 × 9 cm; bed volume, 23 mL), equilibrated with 0.02 M sodium phosphate (pH 6.8) containing 0.5 M NaCl, at a flow rate of 27 mL/h. The column was washed with the starting buffer, and the proteins were then eluted with a linear gradient, starting with 50 mL of the starting buffer in a mixing flask and 50 mL of the same buffer solution containing 0.15 M methyl-α-D-glucoside (Tokyo Kasei Co., Tokyo, Japan). The fractions containing miraculin were collected and concentrated. The concentrated solutions were applied to a Sephacryl S-200 HR column (2.6 × 98 cm; GE Healthcare UK Ltd.), equilibrated with 0.05 M sodium phosphate (pH 6.4) containing 0.15 M NaCl, at a flow rate of 60 mL/h. The active fractions were collected and used as purified recombinant miraculins. The purity of the miraculins was analysed by SDS-PAGE. The fractions of each purification step were monitored by measuring the absorbance at 280 nm, and by dot-blot analysis using the anti-miraculin antibody.

Analysis of the dimerization and N-glycosylation of recombinant miraculin To confirm whether the recombinant miraculin formed a disulphidelinked dimer, as does native miraculin, miraculins (500 ng per lane) purified from transgenic tomato plants and miracle fruit were subjected to SDS-PAGE under non-reducing and reducing conditions. The proteins were stained with a Silver Stain II Kit (Wako Pure Chemical Industries Ltd., Osaka, Japan). The N-glycosylation of the recombinant miraculin was analysed by Western blotting after treatment with PNGase A (Roche). Purified recombinant and native miraculins (500 ng per lane) were prepared in 0.01 M sodium acetate (pH 5.1) and boiled for 5 min, followed by the addition of 2 mU PNGase A (Roche). Enzyme digests were conducted at 37 °C for 24 h, and aliquots of the proteins were subjected to SDS-PAGE in the presence of dithiothreitol (DTT) and Western blotting with the anti-miraculin antibody.

Measurement of taste-modifying activity The taste-modifying activity of miraculin was assayed using six subjects, as described previously (Kurihara and Beidler, 1969; Sun et al., 2006a). Tomato leaves and fruit were washed with water. The seeds from tomato fruit were removed, and pericarp tissue was used. The subjects chewed 2 g of tomato plant tissues for 3 min, or held 5 mL of 0.4 µM purified miraculin solution in the mouth for 3 min. The subjects then spat out the material, rinsed their mouths with water, and tasted 0.02 M citric acid. The sweetness induced by 0.02 M citric acid was evaluated by comparison with that of a series of standard sucrose solutions (0.1– 0.5 M).

Acknowledgements We thank the members of the Ezura Laboratory for helpful discussions. This work was supported by the 21st Century

Center of Excellence Program of the Ministry of Education, Science, Sports, and Technology of Japan and a grant from The Asahi Glass Foundation to H.E.

References Brenner, E.D., Lambert, K.N., Kaloshian, I. and Williamson, V.M. (1998) Characterization of LeMir, a root-knot nematode-induced gene in tomato with an encoded product secreted from the root. Plant Physiol. 118, 237–247. Daniell, H., Streatfield, S.J. and Wycoff, K. (2001) Medical molecular farming: production of antibodies, biopharmaceuticals and edible vaccines in plants. Trends Plant Sci. 6, 219 –226. Deblaere, R., Bytebier, B., De Greve, H., Deboeck, F., Schell, J., van Montagu, M. and Leemans, J. (1985) Efficient octopine Ti plasmid derived vectors for Agrobacterium-mediated gene transfer in plants. Nucleic Acids Res. 13, 4777– 4785. Faus, I. (2000) Recent developments in the characterization and biotechnological production of sweet-tasting proteins. Appl. Microbiol. Biotechnol. 53, 145 –151. Finnegan, J. and McElroy, D. (1994) Transgene inactivation: plants fight back!. Biotechnology, 12, 883 –888. Gibbs, B.F., Alli, I. and Mulligan, C. (1996) Sweet and taste-modifying proteins: a review. Nutr. Res. 16, 1619 –1630. Grierson, D. and Kader, A.A. (1986) Fruit ripening and quality. In: The Tomato Crop. A Scientific Basis for Improvement (Atherton, J.G. and Rudich, J., eds), pp. 241–280. Dordrecht: Kluwer Academic Publishers. Hood, E.E. and Jilka, J.M. (1999) Plant-based production of xenogenic proteins. Curr. Opin. Biotechnol. 10, 382–386. Hood, E.E., Woodard, S.L. and Horn, M.E. (2002) Monoclonal antibody manufacturing on transgenic plants – myths and realities. Curr. Opin. Biotechnol. 13, 630 – 635. Horn, M.E., Woodard, S.L. and Howard, J.A. (2004) Plant molecular farming: systems and products. Plant Cell Rep. 22, 711–720. Howard, J.A. and Hood, E. (2005) Bioindustrial and biopharmaceutical products produced in plants. Adv. Agron. 85, 91–124. Kurihara, K. and Beidler, L.M. (1968) Taste-modifying protein from miracle fruit. Science, 161, 1241–1243. Kurihara, K. and Beidler, L.M. (1969) Mechanism of the action of taste-modifying protein. Nature, 222, 1176 –1179. Kurihara, Y. (1992) Characteristics of antisweet substances, sweet proteins, and sweetness-inducing proteins. Crit. Rev. Food Sci. Nutr. 32, 231–252. Kurihara, Y. and Nirasawa, S. (1997) Structures and activities of sweetness-inducing substances (miraculin, curculin, strogin) and the heat-stable sweet protein, mabinlin. FFI. J. Jpn. 174, 67–74. Liu, X., Maeda, S., Hu, Z., Aiuchi, T., Nakaya, K. and Kurihara, Y. (1993) Purification, complete amino acid sequence and structure characterization of the heat stable protein, mabinlin II. Eur. J. Biochem. 211, 281–287. Masuda, T. and Kitabatake, N. (2006) Developments in biotechnological production of sweet proteins. J. Biosci. Bioeng. 102, 375– 389. Masuda, T., Ueno, Y. and Kitabatake, N. (2001) Sweetness and enzymatic activity of lysozyme. J. Agric. Food Chem. 49, 4937– 4941. McCormick, S., Niedermeyer, J., Fry, J., Barnason, A., Horsch, R. and Fraley, R. (1986) Leaf disc transformation of cultivated tomato



(L. esculentum) using Agrobacterium tumefaciens. Plant Cell Rep. 5, 81–84. Ming, D. and Hellekant, G. (1994) Brazzein, a new high-potency thermostable sweet protein from Pentadiplandra brazzeana B. FEBS Lett. 355, 106 –108. Morris, J.A. and Cagan, R.H. (1972) Purification of monellin, the sweet principle of Dioscoreophyllum cumminsii. Biochim. Biophys. Acta, 261, 114–122. Nakano, Y., Watanabe, S., Kawashima, H. and Takaichi, M. (2006) The effect of daily applications on yield, fruit quality, and nutrient uptake of hydroponically cultivated tomato. J. Jpn. Soc. Hort. Sci. Jpn. 75, 421–429. Rogers, S.O. and Bendich, A.J. (1985) Extraction of DNA from milligram amounts of fresh, herbarium and mummified plant tissues. Plant Mol. Biol. 5, 69 –76. Schauer, N., Zamir, D. and Fernie, A.R. (2005) Metabolic profiling of leaves and fruit of wild species tomato: a survey of the Solanum lycopersicon complex. J. Exp. Bot. 56, 297–307. Shirasuka, Y., Nakajima, K., Asakura, T., Yamashita, H., Yamamoto, A., Hata, S., Nagata, S., Abo, M., Sorimachi, H. and Abe, K. (2004) Neoculin as a new taste-modifying protein occurring in the fruit of Curculigo latifolia. Biosci. Biotechnol. Biochem. 68, 1403 –1407. Stam, M., Mol, J.N.M. and Kooter, J.M. (1997) The silence of genes in transgenic plants. Ann. Bot. 79, 3 –12. Streatfield, S.J. (2007) Approaches to achieve high-level heterologous protein production in plants. Plant Biotechnol. J. 5, 2–15. Streatfield, S.J. and Howard, J.A. (2003) Plant-based vaccines. Int. J. Parasitol. 33, 479 – 493. Sun, H.-J., Cui, M.-L., Ma, B. and Ezura, H. (2006a) Functional expression of the taste-modifying protein, miraculin, in transgenic lettuce. FEBS Lett. 580, 620 – 626. Sun, H.-J., Uchii, S., Watanabe, S. and Ezura, H. (2006b) A highly efficient transformation protocol for Micro-Tom, a model cultivar for tomato functional genomics. Plant Cell Physiol. 47, 426 – 431.

Suzuki, M., Kurimoto, E., Nirasawa, S., Masuda, Y., Hori, K., Kurihara, Y., Shimba, N., Kawai, M., Suzuki, E. and Kato, K. (2004) Recombinant curculin heterodimer exhibits taste-modifying and sweet-tasting activities. FEBS Lett. 573, 135 –138. Theerasilp, S. and Kurihara, Y. (1988) Complete purification and characterization of the taste-modifying protein, miraculin, from miracle fruit. J. Biol. Chem. 263, 11 536 –11 539. Theerasilp, S., Hitotsuya, H., Nakajo, S., Nakaya, K., Nakamura, Y. and Kurihara, Y. (1989) Complete amino acid sequence and structure characterization of the taste-modifying protein, miraculin. J. Biol. Chem. 264, 6655 – 6659. Walmsley, A.M., Alvarez, M.L., Jin, Y., Kirk, D.D., Lee, S.M., Pinkhasov, J., Rigano, M.M., Arntzen, C.J. and Mason, H.S. (2003) Expression of the B subunit of Escherichia coli heat-labile enterotoxin as a fusion protein in transgenic tomato. Plant Cell Rep. 21, 1020 –1026. van der Wel, H. and Loeve, K. (1972) Isolation and characterization of thaumatin I and II, the sweet-tasting proteins from Thaumatococcus daniellii Bench. Eur. J. Biochem. 31, 221–225. van der Wel, H., Larson, G., Hladik, A., Hladik, C.M., Hellekant, G. and Glaser, D. (1989) Isolation and characterization of pentadin, the sweet principle of Pentadiplandra brazzeana Baillon. Chem. Senses, 14, 73 –79. Witty, M. (1998) New technologies for taste modifying proteins. Trends Food Sci. Technol. 9, 275 –280. Witty, M. and Higginbotham, J.D. (1994) Thaumatin. Boca Raton, FL: CRC Press. Yamashita, H., Theerasilp, A., Nakaya, T., Nakamura, Y. and Kurihara, Y. (1990) Purification and complete amino acid sequence of a new type of sweet protein with taste-modifying activity, curculin. J. Biol. Chem. 265, 15 770 –15 775. Yoshida, K. and Shinmyo, A. (2000) Transgenic expression systems in plants, a natural bioreactor. J. Biosci. Bioeng. 90, 353– 362.
