Journal of Plant Physiology 168 (2011) 11361141

Contents lists available at ScienceDirect

Journal of Plant Physiology

journal homepage: www.elsevier.de/jplph

Short communication

UDP-dependent glycosyltransferases involved in the biosynthesis of steviol glycosides

Amal A.A. Mohamed a,b , Stijn Ceunen b , Jan M.C. Geuns b, , Wim Van den Ende c , Marc De Ley d
a

Botany Department, Faculty of Science, 81528, South Valley University, Aswan, Egypt Laboratory of Functional Biology, K.U.Leuven, Kasteelpark Arenberg 31, 3001 Heverlee, Belgium c Laboratory of Molecular Plant Physiology, K.U.Leuven, Kasteelpark Arenberg 31, 3001 Heverlee, Belgium d Advanced Biochemistry, Molecular and Structural Biology, K.U.Leuven, Celestijnenlaan 200G, 3001 Heverlee, Belgium
b

a r t i c l e

i n f o

a b s t r a c t
A short-term experiment was designed to measure the transcript levels of downstream genes contributing to the biosynthesis of steviol glycosides. Stevia rebaudiana plants were subjected to longand short-day conditions for different time intervals. Samples from both lower and upper leaves were collected. Using quantitative real-time polymerase chain reaction, the transcript levels of three UDP-dependent glycosyltransferases, UGT85C2, UGT74G1 and UGT76G1, were studied. The results were compared with the steviol glycoside contents measured in the leaves, which were quantied by reversed phase HPLC. In the same daylength condition, steviol glycoside concentration and the transcript levels of the three UGT genes were higher in upper leaves than in lower leaves. Steviol glycosides accumulated more in plants under short-day conditions. Under these conditions, a highly signicant correlation was found between UGT85C2 transcription and total steviol glycoside accumulation in the upper leaves. This suggests that the glycosylation of steviol to form steviolmonoside is the rate-limiting step in the glycosylation pathway of steviol glycosides. In these upper leaves, a relatively high accumulation of rebaudioside A compared to stevioside was also observed, however, without correlation with the transcription of UGT76G1. 2011 Elsevier GmbH. All rights reserved.

Article history: Received 19 November 2010 Received in revised form 7 January 2011 Accepted 7 January 2011 Keywords: Glycosyltransferases Photoperiodic response Real-time quantitative PCR Steviol glycosides Stevia rebaudiana

Introduction The leaves of Stevia rebaudiana (Bertoni) accumulate numerous steviol glycosides (SVglys), the concentrations of which can vary widely depending on the genotype, growth conditions and developmental stage (Brandle and Rosa, 1992; Brandle et al., 1998; Starratt et al., 2002). Like all diterpenes, steviol, the aglycone of the SVglys, is synthesized from geranylgeranyl pyrophosphate, rst by protonation-initiated cyclization to ent-copalyl pyrophosphate (ent-CPP) by ent-CPS. Next, ent-kaurene is produced from ent-

Abbreviations: bps, base pairs; Dul A, dulcoside A; ent-CPP, ent-copalyl pyrophosphate; ent-CPS, ent-copalyl pyrophosphate synthase; ent-KAH, ent-kaurenoic acid 13-hydroxylase; ent-KAO, ent-kaurenoic acid oxidase; ent-KO, ent-kaurene oxidase; ent-KS, ent-kaurene synthase; LD, long day; Reb A, rebaudioside A; Reb C, rebaudioside C; ROS, reactive oxygen species; RP-HPLC, reversed-phase high performance liquid chromatography; RT-q PCR, real-time quantitative polymerase chain reaction; SD, short day; ST, stevioside; SVgly, steviol glycoside; UGT, uridine diphosphate-dependent glycosyltransferase. Corresponding author at: Kasteelpark Arenberg 31, Box 2436, 3001 Heverlee, Belgium. Tel.: +32 16 321510; fax: +32 16 321509. E-mail address: Jan.Geuns@bio.kuleuven.be (J.M.C. Geuns). 0176-1617/$ see front matter 2011 Elsevier GmbH. All rights reserved. doi:10.1016/j.jplph.2011.01.030

CPP by an ionization-dependent cyclization catalyzed by ent-KS. Ent-kaurene is subsequently oxidized in a three-step reaction to ent-kaurenoic acid, by ent-KO. From ent-kaurenoic acid, the biosynthesis of steviol deviates from gibberellin biosynthesis by direct hydroxylation at C-13 by ent-KAH. Steviol is further converted into the different glycosides by uridine diphosphate-dependent glycosyltransferases (UGTs) (Richman et al., 1999, 2005; Humphrey et al., 2006; Fig. 1). UGTs are found in all living organisms, catalyzing the transfer of a glycosyl moiety from an activated donor to an acceptor molecule, forming a glycosidic bond (Campbell et al., 1997). In S. rebaudiana, they trigger the downstream steps in the synthesis of SVglys. The aglycone steviol has two hydroxyl groups, one belonging to the C-19 carboxylic acid moiety and the other attached to the C-13, both of which can be glycosylated. Three candidate genes, UGT74G1, UGT76G1 and UGT85C2, of which the translation products were found to have an activity towards steviol or SVglys, were cloned by Richman et al. (2005). The addition of the C-13-glucose to steviol is catalyzed by UGT85C2, the C-19-glucose by UGT74G1 and nally glucosylation of the C-3 of the glucose at the C-13 position is catalyzed by UGT76G1. The UGT responsible for the glucosylation of the C-2 of the C-13-glucose has not yet been identied (Brandle and Telmer, 2007; Fig. 1).

A.A.A. Mohamed et al. / Journal of Plant Physiology 168 (2011) 11361141

1137

Fig. 1. The biosynthetic pathway of steviol glycosides, showing the rst steps which are shared with gibberellins

UGT74G1, UGT76G1 and UGT85C2 belong to family GT1 of the glycosyltransferase group according to the CAZy classication (http://www.cazy.org/fam/GT1.html), which describes more than 91 distinct families. The plant CAZy family 1 substrates include terpenoids, alkaloids, cyanogenic glucosides and glucosinolates as well as avonoids, isoavonoids and other phenylpropanoids. The present work is complementary to previous research conducted to prole the transcription of upstream genes in the SVgly pathway in relation to the accumulation patterns of SVglys (Mohamed et al., 2009, 2010).

Materials and methods Plant material and growth conditions Homogeneous groups of plants of the same age, grown in a greenhouse under long day (LD) conditions were selected and each week, two groups were transferred to a growth chamber and grown under LD or short day (SD) conditions, respectively. The LD groups received 16 h light and 8 h dark, while SD groups received 12 h light and 12 h dark. Both were placed under constant light of 15 W/m2 . This process was repeated for four weeks. This way, at the end of the experiment, the 8 groups of plants had been subjected to the

1138

A.A.A. Mohamed et al. / Journal of Plant Physiology 168 (2011) 11361141

applied conditions in the following manner: groups 1 and 2 for one week; 3 and 4 for 2 weeks; 5 and 6 for 3 weeks; 7 and 8 for 4 weeks. Quantication of steviol glycosides For analysis of steviol glycosides (SVglys), 50 mg of freeze-dried samples of both upper and lower leaves were extracted three times in 600 L of boiling water. A fourth extraction was not required as no signicant amount of SVglys could be recovered and to avoid possible losses, no purication step was used (Geuns, 2008, 2010). After pooling the extracts, 10 L was injected in HPLC on 2 Alltima C18 columns in series (250 mm 4.6 mm ID, 5 m particle size, UV detection at 200 nm, 10 mm path length). The SVglys were eluted with 27% AcCN:H2 O at 1 mL/min and quantied using stevioside as an external standard (Geuns, 2008, 2010). Extraction of RNA and cDNA synthesis Total RNA from both upper and lower leaves was extracted using the Plant RNA Mini Kit (Invitrogen), according to the manufacturers instructions. For cDNA synthesis, only RNA samples of high purity (A260/280 > 2.00) were used. An amount of 2.3 g of total RNA was reverse transcribed with 50 U AMV-reverse transcriptase in a total volume of 25 L of master mixture containing 2.5 mol/L oligo (dT)16 primer, 20 U RNase inhibitor, 1 PCR, 5 mmol/L MgCl2 , and 1 mmol/L of each dNTP (GeneAmp RNA PCR kit, Roche Molecular Systems, Branchburg, NJ, USA). The endogenous control (house keeping gene)

Primer design Primers for both target and house-keeping genes were designed using the OligoCalc software to achieve specic characters required for real-time quantitative polymerase chain reaction (RT-q PCR) (Table 1). Quantication using RT-q PCR SYBR green binds to double-stranded DNA, and upon excitation emits light. Thus, as PCR product accumulates, uorescence increases. The PCR reaction was adjusted to 25 L: 6 L milli-Qwater, 0.75 L of each primer (10 mol/L), 12.5 L SYBR green mix and 5 L of cDNA. Relative standard curves for target genes and house keeping genes were generated with serial (5) dilutions of previously prepared cDNAs. The reactions were run as triplicates in an ABI prism 7000 sequence detection system (ABI Prism 7000 SDS, Applied Biosystems) using the following thermal cycles: 50 C for 2 min, 95 C for 10 min, followed by 40 cycles of 95 C for 15 s and 60 C for 30 s. Statistical analysis Two-way analysis of variance (ANOVA) at p < 0.05 was applied to calculate for signicant differences in SVgly accumulation or gene expression between upper and lower leaves of plants subjected to the same daylength conditions or between the leaves of the same position (upper or lower) at different daylength conditions. In both cases, the time interval was chosen as the second variable. Results

The expression of the target gene should be normalized to the expression of another endogenous gene that is considered as a control. The endogenous control should be expressed at a constant level across all samples. For more reliability, in this experiment, -actin in addition to 18S rRNA was used to normalize the data individually.
Table 1 List of primers used in RT-q PCR for target and house keeping genes. Gene UGT85C2 UGT74G1 UGT76G1 18S rRNAa -Actin
a

Data from HPLC analysis of four major SVglys (stevioside (ST), rebaudioside A (Reb A), rebaudioside C (Reb C) and dulcoside A (Dul A)) are shown in Table 2 as percentages of dry weight. Under both LD and SD conditions,

Primer sequence (forward/reverse) 5 -TCGATGAGTTGGAGCCTAGTATT-3 /5 -CTAAACTGTATCCATGGAGACTC-3 5 -TGCATGAACTGGTTAGACGATAAG-3 /5 -GCATCCTACTGATTCGTGTGCTA-3 5 -GCAGCTTACTAGACCACGATC-3 /5 -CTCATCCACTTCACTAGTACTAC-3 5 -CCGGCGACGCATCATT-3 /5 -AGGCCACTATCCTACCATCGAA-3 5 -AGCAACTGGGATGACATGGAA-3 /5 -GGAGCGACACGAAGTTCATTG-3

Amplicon length 153 bp 274 bp 107 bp 59 bp 65 bp

Accession number AY345978 AY345982 AY345974 Cloned at the laboratory AF548026

Using specic primers designed for 18S rRNA, a segment of 400 bp of Stevia rebaudiana was obtained and used as a template to design primers convenient for RT-q PCR.

Table 2 Percentages of four steviol glycosides in upper and lower leaves of plants subjected to LD and SD conditions (mean standard deviation; n = 3). SVgly % Groups at LD conditions 1 Upper leaves ST Reb A Reb C Dul A Reb A/ST Total Lower leaves ST Reb A Reb C Dul A Reb A/ST Total 6.62 3.15 0.64 0.25 0.48 0.65 0.28 0.07 0.03 0.06 3 8.53 4.42 0.81 0.32 0.52 0.70 0.79 0.07 0.03 0.10 5 8.03 3.88 1.01 0.34 0.48 0.15 0.22 0.12 0.07 0.03 7 8.64 5.68 1.18 0.24 0.66 0.35 0.29 0.06 0.01 0.04 Groups at SD conditions 2 8.13 2.94 0.67 0.36 0.36 2.73 0.77 0.23 0.22 0.15 4 8.17 6.87 1.04 0.27 0.84 0.78 0.79 0.08 0.03 0.13 6 8.30 7.83 1.78 0.26 0.94 0.06 0.18 0.05 0.12 0.02 8 8.86 7.25 1.15 0.21 0.82 0.09 0.18 0.02 0.03 0.02

10.66 1.00 5.14 2.79 0.69 0.28 0.54 0.39 0.11 0.02 0.04 0.05

14.1 1.58 6.53 3.41 0.77 0.18 0.48 0.85 0.44 0.13 0.02 0.09

13.3 0.19 6.53 3.66 0.87 0.22 0.56 0.36 0.60 0.12 0.07 0.09

15.7 0.86 4.77 2.49 0.67 0.21 0.52 0.57 0.33 0.16 0.11 0.09

11.49 4.90 5.70 2.27 0.64 0.22 0.39 1.55 0.28 0.23 0.08 0.12

16.36 1.67 5.67 2.63 0.58 0.13 0.46 0.29 0.08 0.02 0.01 0.03

18.17 0.17 5.81 3.07 0.76 0.17 0.53 0.21 0.12 0.10 0.02 0.03

17.48 0.30 6.34 2.65 0.65 0.23 0.42 1.05 0.43 0.12 0.04 0.09

8.90 0.50

10.88 1.40

11.28 0.60

8.14 0.98

8.84 2.14

9.02 0.38

9.82 0.27

9.86 1.49

A.A.A. Mohamed et al. / Journal of Plant Physiology 168 (2011) 11361141

1139

Fig. 2. The expression of the three UGTs relative to actin in both upper and lower leaves which were subjected to LD and SD conditions for different time intervals.*Expression in lower leaves of group 3 could not be measured (n = 3). Normalization to 18S rRNA is not shown.

1140

A.A.A. Mohamed et al. / Journal of Plant Physiology 168 (2011) 11361141

the accumulation of SVglys in upper leaves was signicantly higher than in lower leaves (at p < 0.05) of the same plants. The ratio of the two major SVglys, Reb A and ST, is also shown in Table 2. In most conditions, Reb A accumulated in a more or less constant ratio to ST (about two times more ST than Reb A). Only in the upper leaves under SD conditions did Reb A accumulate to a higher ratio, closely approaching ST accumulation after more than 2 weeks (groups 4, 6 and 8) as the ratio nearly equalled 1. In RT-q PCR experiments, analysis of the dissociation curves from both experimental and -actin or 18S rRNA control samples revealed a single melting peak, indicating a specic signal, corresponding to the uridine diphosphate-dependent glycosyltransferase (UGT) target sequence and the endogenous control, respectively. House keeping genes, -actin and 18S rRNA, showed a similar expression pattern in both upper and lower leaves independent of the daylength or time interval. In all negative control samples, no amplication of the signal was detected, proving the purity of the samples. Very similar expression patterns were obtained from normalization of UGT expression by 18S rRNA and -actin. Fig. 2 shows the expression patterns of the three target UGTs in both upper and lower leaves in plants subjected to LD or SD conditions, respectively. In general, the transcript level of the three genes was higher in upper leaves than in lower leaves at the same daylength condition. The upper leaves under SD conditions showed a much higher transcription of UGT74G1 and UGT76G1 (p < 0.01) with the former having the greatest values. At the same time, these leaves had a greater accumulation of Reb A and Reb C than their counterparts (Table 2). In the upper leaves, transcription of UGT74G1 was signicantly greater in SD than in LD conditions (p < 0.01) while the transcription of the other two genes was not. In LD conditions, only UGT76G1 showed a signicant difference in its expression between upper and lower leaves. When comparing different groups in the same daylength condition, a highly signicant correlation (p < 0.01) was found only between UGT85C2 transcription and total steviol glycoside accumulation in the upper leaves of plants under short-day conditions. In these upper leaves, a relatively higher accumulation of rebaudioside A compared to stevioside was also observed, however without correlation with the transcription of UGT76G1.

Zaidan et al., 1980). Using the natural growth season in South Korea, a peak in ST content was observed in September, a transition month from LD to SD (Kang and Lee, 1981). In the same context, Yoshida (1986) measured Reb A in addition to ST and found similar accumulation patterns. Furthermore, the accumulation of Reb A never exceeded ST accumulation, except in September. In the present study, the transition of plants from LD to SD conditions accelerated the accumulation of Reb A in upper leaves to reach nearly that of ST. At the molecular level, upregulation of UGT74G1 and UGT76G1 was observed, which might be induced by SD conditions (Bassett et al., 2006). Overexpression of some glycosyltransferase genes led to a signicant increase in their respective glucosides (Jackson et al., 2001; Martin et al., 2001). In general, it is difcult to correlate a specic UGT and a specic derivative where plant UGTs are thought to be not highly specic and show broad substrate specicity based on a regio-selectivity (Hansen et al., 2003; Lim et al., 2003; Richman et al., 2005). However, there is a correlation between the higher SVgly accumulation in upper leaves and the transcription of certain UGT genes. This correlation can be found both in LD (p < 0.05 for UGT76G1) and SD groups (p < 0.01 for UGT85C2). The signicant correlation between UGT85C2 transcription and total SVgly accumulation suggests that the UGT85C2 enzyme, which adds a C-13-glucose to steviol to form steviolmonoside, is the rate-limiting step of the glycosylation pathway during SVgly biosynthesis. The reason why this correlation was only found in the upper leaves of SD plants might be because gene transcription and SVgly accumulation was much higher in these leaves when compared to lower leaves and LD groups. The overall increase in SVgly accumulation and UGT transcript levels under SD conditions probably correlates with the transition of the plants from LD to SD conditions as also observed earlier (Kang and Lee, 1981). This transition might be stressful for the plants (Gibon et al., 2009) possibly resulting in a higher production of more reactive oxygen species (ROS). The increased production of SVgly might then be a reaction of the plants to enhance their ROS scavenging capacity. Recently, it was shown that SVglys are very potent ROS scavengers (Stoyanova et al., 2010). Acknowledgements We would like to thank Mr. Tom Struyf and Mrs. Hilde Verlinden for technical assistance. Thanks extend to Mrs. Soe Van Soest for her assistance with RT-q PCR. References
Bassett CL, Wisniewski ME, Artlip TS, Norelli JL. J Am Soc Hortic Sci 2006;131:55163. Brandle JE, Rosa N. Can J Plant Sci 1992;72:12636. Brandle JE, Telmer PG. Phytochemistry 2007;68:185563. Brandle JE, Starratt AN, Gijzen M, Can J. Plant Sci 1998;78:52736. Campbell JA, Davies GJ, Bulone V, Henrissat B. Biochem J 1997;326:92942. Geuns JMC.Geuns JMC, editor. Proceedings of the 2nd EUSTAS stevia symposium Steviol glycosides: technical and pharmacological aspects. Heverlee: Euprint; 2008. p. 5978. Geuns JMC. Stevia and steviol glycosides. Heverlee: Euprint; 2010. Gibon Y, Pyl ET, Sulpice R, Lunn JE, Hhne M, Gnther M, et al. Plant Cell Environ 2009;32:85974. Hansen KS, Kristensen C, Tattersall DB, Jones PR, Olsen CE, Bak S, et al. Phytochemistry 2003;64:14351. Hsel W.Stumpf PK, Conn EE, editors. The biochemistry of plants. New York: Academic Press; 1981. p. 72553. Humphrey TV, Richman AS, Menassa R, Brandle JE. Plant Mol Biol 2006;61:4762. Jackson RG, Lim EK, Li Y, Kowalczyk M, Sandberg G, Hoggett J, et al. J Biol Chem 2001;276:43506. Kang KH, Lee KW. Hanjakji 1981;26:6989. Lim EK, Higgins GS, Li Y, Bowles DJ. Biochem J 2003;373:98792. Martin RC, Mok MC, Hebben JE, Mok DWS. Plant Physiol 2001;120:5537. Metivier J, Viana AM. J Exp Bot 1979;30:121122. Mohamed AAA, Geuns JMC, Van den Ende W, De Ley M.Jan MC, Geuns, editors. Proceedings of the 3rd EUSTAS stevia symposium Stevia in Europe. Heverlee: Euprint; 2009. p. 6383.

Discussion There are numerous genetic approaches available to investigate the action of the glycosyltransferases in the plant and how their catalytic activities may be related to physiological functions. Glycosylation and deglycosylation can be a regulatory mechanism altering levels of metabolites (Hsel, 1981). Previous studies based on RNA gel blot analysis, showed the expression of 12 UGTs, including the three UGTs (UGT85C2, UGT74G1 and UGT76G1), however not in a time-course study (Richman et al., 2005). It was concluded that these three UGTs were not more expressed than any of the others. Their expression was shown in the following order: UGT85C2 > UGT76G1 > UGT74G1. In the present study, we used RT-q PCR, a highly sensitive and specic method to accurately measure transcript levels of the three UGTs in a timecourse study and at different daylength conditions to search for a correlation between their expression and their respective biosynthetic products. Previous studies, focusing on the accumulation of SVglys in longterm experiments, obtained contradictory results. In two studies, the ST content was measured in plants grown under photoperiods of 16 h (LD) and 8 h light (SD). Its highest accumulation was found in plants grown under LD conditions (Metivier and Viana, 1979;

A.A.A. Mohamed et al. / Journal of Plant Physiology 168 (2011) 11361141 Mohamed AAA, Geuns JMC, Van den Ende W, De Ley M. Egypt J Exp Biol (Bot) 2010;6:718. Richman AS, Gijzen M, Starrat AN, Yang Z, Brandle JE. Plant J 1999;19:41121. Richman A, Swanson A, Humphrey T, Chapman R, McGarvey B, Pocs R, et al. Plant J 2005;41:5667. Starratt AN, Kirby CW, Pocs R, Brandle JE. Phytochemistry 2002;59:36770.

1141

Stoyanova S, Geuns J, Hideg E, Van den Ende W. Int J Food Sci Nutr 2010., doi:10.3109/09637486.2010.523416. Yoshida S. Jpn J Crop Sci 1986;55:18995. Zaidan LBP, Dietrich SMC, Felippe GM. Jpn J Crop Sci 1980;49:56974.