Cloning, expression and characterization of a functional cDNA clone encoding geranylgeranyl diphosphate synthase of Hevea brasiliensis

Biochimica et Biophysica Acta 1625 (2003) 214 – 220 www.bba-direct.com

Short sequence paper

Cloning, expression and characterization of a functional cDNA clone $ encoding geranylgeranyl diphosphate synthase of Hevea brasiliensis Akiyuki Takaya a, Yuan-Wei Zhang a,1, Kasem Asawatreratanakul b, Dhirayos Wititsuwannakul c, Rapepun Wititsuwannakul d, Seiji Takahashi a, Tanetoshi Koyama a,* a

Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, Katahira 2-1-1, Aoba-ku, Sendai 980-8577, Japan b Department of Chemistry, Faculty of Science, Thaksin University, Songkla 90000, Thailand c Department of Biochemistry, Faculty of Science, Mahidol University, Bangkok 10400, Thailand d Department of Biochemistry, Faculty of Science, Prince of Songkla University, Hat-Yai, Songkla 90112, Thailand Received 26 August 2002; received in revised form 19 November 2002; accepted 26 November 2002

Abstract Geranylgeranyl diphosphate (GGPP) synthase catalyzes the condensation of isopentenyl diphosphate (IPP) with allylic diphosphates to give (all-E)-GGPP. GGPP is one of the key precursors in the biosynthesis of biologically significant isoprenoid compounds. In order to examine possible participation of the GGPP synthase in the enzymatic prenyl chain elongation in natural rubber biosynthesis, we cloned, overexpressed and characterized the cDNA clone encoding GGPP synthase from cDNA libraries of leaf and latex of Hevea brasiliensis. The amino acid sequence of the clone contains all conserved regions of trans-prenyl chain elongating enzymes. This cDNA was expressed in Escherichia coli cells as Trx-His-tagged fusion protein, which showed a distinct GGPP synthase activity. The apparent Km values for isopentenyl-, farnesyl-, geranyl- and dimethylallyl diphosphates of the GGPP synthase purified with Ni2 +-affinity column were 24.1, 6.8, 2.3, and 11.5 AM, respectively. The enzyme shows optimum activity at approximately 40 jC and pH 8.5. The mRNA expression of the GGPP synthase was detected in all tissues examined, showing higher in flower and leaf than petiole and latex, where a large quantity of natural rubber is produced. On the other hand, expression levels of the Hevea farnesyl diphosphate synthase were significant in latex as well as in flower. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Hevea brasiliensis; Geranylgeranyl diphosphate; Rubber biosynthesis; Isoprenoid; Prenyl chain elongation; Prenyltransferase

Natural rubber is a mixture of high-molecular weight polymers consisting of isoprene units in cis-configuration derived from consecutive condensation of the so-called biologically active isoprene units, isopentenyl diphosphate (IPP). In the light of the common enzymatic prenyl chain elongation processes by prenyltransferases, an allylic shortchain diphosphate is required as co-substrate to initiate the prenyl chain elongation process in rubber formation [1– 3].

$ The complete cDNA sequence reported in this paper, designated ggps, is available from the DDBJ/GenBank/EBI database under accession number AB055496. * Corresponding author. Tel.: +81-22-217-5621; fax: +81-22-2175620. E-mail address: [email protected] (T. Koyama). 1 Present address: Department of Pharmacology, Yale University School of Medicine, New Haven, CT 06520-8066, USA.

The synthesis of the priming allylic short chain prenyl diphosphates is catalyzed by a trans-type prenyltransferase, whose enzymatic activities have been found in the bottom fraction as well as in the centrifuged fresh C-serum of Hevea latex [4]. The initiation step in rubber biosynthesis has been elucidated according to the configurations of the N-end of natural rubber as well as the substrate specificities of the rubber transferases from several rubber-producing plants, but detailed mechanism on the initiation process in rubber biosynthesis has not yet been clarified. Susceptibilities of several allylic prenyl diphosphates as initiating primer substrates in rubber synthesis in vitro with washed rubber particles (WRPs) from Hevea latex have shown that allylic prenyl diphosphates of chain lengths between C5 to C20 are effective in stimulating the rubber formation with efficiency of the increasing chain length in the order of C5 < C10 < C15 < C20 [1,5]. On the other hand, the structural elucidation of the natural rubber molecules from goldenrod

0167-4781/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. doi:10.1016/S0167-4781(02)00602-4

A. Takaya et al. / Biochimica et Biophysica Acta 1625 (2003) 214–220

or sunflower by 13C-NMR spectroscopy [6– 8], implied that the initiation step of rubber prenyl chain elongation could start from (E,E)-farnesyl diphosphate (FPP, C15) or (all-E)-geranylgeranyl diphosphate (GGPP, C20) as the allylic primer substrate, respectively. However, in the case

215

of Hevea rubber, the structural information about the oend has not yet been obtained distinctly. The Hevea FPP synthase was purified from the C-serum of Hevea latex [9,10], and the gene encoding Hevea FPP synthase has been cloned and characterized [11]. This gene

Fig. 1. Nucleotide and deduced amino acid sequences of hbgg. Underlined region is the proposal signal peptide predicted by PSORT program and ChloroP program.

216

A. Takaya et al. / Biochimica et Biophysica Acta 1625 (2003) 214–220

is highly expressed in latex-producing cells (laticifers) and in the epidermal region of the rubber plant. On the other hand, detailed information on the Hevea GGPP synthase have not yet been reported, although a predominant synthesis of GGPP as well as some polyprenyl products have been reported by incubation of IPP with freeze-dried bottom fraction, which is prepared from centrifuged fresh Hevea latex [4]. GGPP synthase catalyzes the condensation of IPP with an allylic prenyl diphosphate FPP to give GGPP, which is an essential precursor in the biosynthesis of several isoprenoids necessary for plant growth and development such as carotenoids, gibberellins, prenyl quinones, chlorophylls and formation of geranylgeranylated proteins [12]. Many plant GGPP synthase genes have been isolated and characterized from various sources including Sinapis alba [13], Capsicum annuum [14,15], Arabidopsis thaliana [16 – 18], Lupinus albus [19], Catharanthus roseus [20], Taxus Canadensis [21], Helianthus annuus [22], Scoparia dulcis and Croton sublyratus [23]. A detailed understanding of rubber biosynthesis in genetics and enzymology is important for bioengineering

of rubber production through genetic manipulation of the isoprenoid biosynthetic pathway. We are interested in the study of Hevea GGPP synthase with respect to the possibility that the enzyme participates in the formation of a possible initiator molecule in Hevea rubber biosynthesis. In this paper we describe functional cloning, expression and characterization of the cDNA encoding the GGPP synthase from Hevea brasiliensis. Comparison of the deduced amino acid sequences of GGPP synthases from many plants shows the presence of several conserved regions typical of trans-type prenyltransferases, while some divergence in N-terminal regions are assumed to be signal sequences responsible for differential subcellular localization [2]. Based on the conserved regions in the amino acid sequences of many plant GGPP synthases, four degenerate primers were designed. At first, two primers, GGPS-1S (according to the conserved sequence for EMIHTM, region II): 5V-GARATGATHCAYACNATG-3V and GGPS-1A (for DDILDV, region VI) 5V-ACRTCIARDATRTCRTC-3V, where R is A or G; H is A, C, or T; Y is C or T; N is A, C, G or T; S is G or C; D is G, A, or T; I is inosine, were employed to clone the gene from the leaf or

Fig. 2. Phylogeny of GGPP synthases from plants, bacteria, yeast, insect and mammals. A phylogenetic tree was drawn using the CLUSTAL W program. Accession numbers: AraGG1 – 4 and 6, A. thaliana GGPS1 – 4 and 6, L25813, U44876, AB023038, AC006135 and AB000835, respectively; C.annuumGG, C. annuum, X80267; C.roseusGG, C. roseus, X92893; Sulf, S. acidocaldarius, D28748; Ttherm, T. thermophilus, D87817; Drosop, D. melanogaster, AF049659; Yeast, S. cerevisiae, U31632; human, Homo sapiens, AB016043; HBGG, H. brasiliensis, AB055496 (this study).

A. Takaya et al. / Biochimica et Biophysica Acta 1625 (2003) 214–220

the latex cDNA of H. brasiliensis. However, the specific PCR products corresponding to those of a possible GGPP synthase, whose sizes were estimated to be approximately 450 bp, was not detectable by ethidium bromide staining. To obtain the specific PCR products by nested PCR, the region in the agarose gel corresponding to the DNA size of 400 – 600 bp were taken and subjected to DNA extraction, then the recovered DNA was used as the template for nested PCR with primers GGPS-1S, GGPS-1A, GGPS-2S (for MSLIHDD, region II); 5V-ATGWSIYTIATHCAYGAYGA3Vand GGPS-2A (for FQVVDDI, region VI); 5V-ATRTCRTCIACIACYTGRAA-3V to obtain PCR products of approximately 450 bp in length. The nucleotide sequences of the PCR products derived from the leaf mRNA were found to be identical to that from the latex mRNA. High degree of homologies was observed in the conserved regions between the PCR product and other plant GGPP synthases. The PCR product labeled with digoxigenin (DIG) was used as a probe for screening of Hevea cDNA libraries from the leaf or the latex, which were constructed by ZAP-cDNA Synthesis Kit (STRATAGENE). In total, 2  103 and 1  106 plaques were screened from leaf and latex cDNA library, resulting that one plaque in each cDNA library was hybridized to the probe. Each of them was subjected to in vivo excision and sequenced. Both clones were found to possess exactly the same DNA sequence in the open reading frame but showed some difference in either 3V-UTR or 5VUTR regions. The deduced amino acid sequence of the gene is shown in Fig. 1. Comparison of the amino acid sequences of the Hevea gene product with other plant GGPP synthases shows high homologies in all of the conserved regions (66.1%, 62.7%, 64.0% and 79.6% identity with the GGPP synthase-1 and -2 from A. thaliana, from C. annuum and from C. roseus, respectively). On the other hand, the protein showed low homology to the GGPP synthases from other species (25.5%, 22.1%, 21.9%, 23.7% and 16.2% identity with human, Drosophila melanogaster, Saccharomyces cerevisiae, Thermus thermophilus, and Sulfolobus acidocaldarius, respectively). According to the phylogenetic tree of GGPP synthases (Fig. 2), which is deduced by the CLUSTAL W program [24], the putative GGPP synthase from H. brasiliensis can be classified in the typical group of plant enzymes. We designated the putative GGPP synthase gene from H. brasiliensis as hbgg. To confirm whether the obtained cDNA actually encodes GGPP synthase, we constructed the expression vector systems by employing several fusion protein producing vectors, pET19b(+), pET22b(+) or pET32b(+) (Novagen, WI, USA). The plasmid termed pET19bGG was constructed by insertion of Nde I– BamH I fragment containing the cDNA of GGPP synthase into pET19b(+) to express the 43kDa protein fused with His-Tag at the N-terminal region. The plasmids termed pET22bGG or pET32bGG were constructed by insertion of Nco I –BamH I fragment containing the cDNA into pET22b(+) or pET32b(+), respectively. The plasmid pET22bGG was designed to express the 42-kDa

217

protein fused with pelB signal sequence for potential periplasmic localization at N-terminal region. The plasmid pET32bGG was designed to express the 57-kDa protein fused with the 109-amino-acid thioredoxin protein (TrxTag) for increasing the solubility of target protein and HisTag at N-terminal region. SDS-PAGE analyses of the expressed proteins (Fig. 3A) showed that the Escherichia coli transformants with pET19bGG and pET32bGG showed high expression of proteins with the expected molecular mass, but that with pET22bGG did not show detectable expression. Although most of the fusion-proteins were detected in the pellet fractions, small amounts of the

Fig. 3. SDS-PAGE analysis of overproduction of HBGG from E. coli cells harboring pET19bGG, pET22bGG or pET32bGG. (A) Total protein of E. coli BL21 (DE3) transformants. Lane M, molecular weight markers; lane 1, host cells, BL21 (DE3); lane 2, pET19bGG (IPTG ); lane 3, pET19bGG (IPTG+), lane 4, pET22bGG (IPTG ); lane 5, pET22bGG (IPTG+); lane 6, pET32bGG (IPTG ); lane 7, pET32bGG (IPTG+). (B) Soluble and pellet fractions. Lane M, molecular weight markers; lane 1, pET19bGG (soluble); lane 2, pET22bGG (soluble); lane 3, pET32bGG (soluble); lane 4, pET19bGG (pellet); lane 5, pET22bGG (pellet); lane 6, pET32bGG (pellet). (C) Purification of the expressed protein (pET32bGG) with a Ni2 +column. Lane M, molecular weight markers; lane 1, soluble fraction; lane 2, denatured pellet fraction.

218

A. Takaya et al. / Biochimica et Biophysica Acta 1625 (2003) 214–220

expressed proteins were detected in the soluble fractions (Fig. 3B). All of our efforts to obtain enzymatically active proteins by solubilization of the pellet fractions by use of detergents (Tween 20, Tween 80, Triton X-100, Triton X114, Nonidet P-40, octly-h-glucoside and CHAPS) or by osmotic shock, resulted in failure. The fusion protein in the soluble fraction derived from pET32bGG, however, was sufficient for further purification with Ni2 +-column as shown in Fig. 3C. The purified Trx-tagged fusion protein from the soluble fraction showed high GGPP synthase activity when assayed with 14C-labelled IPP and DMAPP, GPP or FPP as substrates (Fig. 4). The specific activities of the Trx-His-tagged GGPP synthases before and after purification were 5.4 and 125.3 nmol/min/mg, respectively. Even after digestion of the Trx-His-tag by thrombin, the GGPP synthase showed comparable specific activity of 123 nmol/min/mg. The pH profile of the recombinant GGPP synthase showed a broad ‘‘bell-shaped’’ curve with wide optimum pH range around 8– 9 (data not shown). The enzyme was inactive when assayed in metal-free buffer, and the activity

Fig. 4. Product analysis of the products from. (A) Product analysis by normal phase TLC on a Silica Gel G-60 plate (Merck) with a solvent system of 2-propanol/29% aqueous NH3/H2O (6:3:1). (B) Product analysis of alcohols derived by enzymatic hydrolysis of the reaction products by reversed phase TLC on an RP-18 TLC plate (Merck) with a solvent system of acetone/H2O (9:1). Lane 1, IPP+ DMAPP; lane 2, IPP + GPP, lane 3, IPP + FPP. The 0.2-ml incubation mixture for the GGPP synthase assay containing 25 mM Tris – HCl buffer (pH 8.0), 5 mM MgCl2, 25 mM NH4Cl, 25 mM h-mercaptoethanol, 1 mg/ml bovine serum albumin, 25 AM FPP, 25 AM [1-14C]IPP (185 GBq/mol) and 100 Ag enzyme protein was incubated at 37 jC for 30 min. The reaction products were immediately extracted with 1-butanol saturated with water, and applied for normal phase TLC. For the reversed phase TLC analysis, the radioactive products were hydrolyzed to the corresponding alcohols with potato acid phosphatase according to the method reported previously [25]. The TLC plates were exposed on a Fuji imaging plate for 12 h, and then the imaging plate was analyzed with Fuji BAS 1000 bioimage analyzer. The positions of authentic standards were visualized with iodine vapor.

Table 1 Kinetic parameters of Hevea GGPP synthase Km (DMAPP) (AM)

Km (GPP) (AM)

11.5 F 1.35

2.34 F 0.47 6.78 F 0.99 24.08 F 3.00 123

a

Km (FPP) (AM)

Km (IPP) (AM)

Vmax (nmol/min/mg)a

For the reaction with FPP.

was restored upon addition of Mg2 + or Mn2 + to the reaction mixture, with optimal levels of 5 mM for Mg2 + or 0.1 mM for Mn2 +, respectively. High concentration of Mn2 + caused inhibition of the GGPP synthase activity. With other metals, Co2 + showed a little activity (data not shown). The optimal temperature for the enzymatic reaction is around 40 jC (data not shown). Apparent Km values for DMAPP, GPP, FPP and IPP were calculated from the Lineweaver– Burk plots to be 11.5, 2.3, 6.8 and 24.1 AM, respectively (Table 1). These values are comparable to those of other plant GGPP synthases [21]. To examine the expression levels of the mRNAs for the GGPP synthase and for the FPP synthase from H. brasiliensis [11], RT-PCRs using total RNAs extracted from leaf, young leaf, latex and petiole were carried out. The primers designed for RT-PCR are 5V-CGTTAATCTCATCATGAGTTCAGTG3V and 5V-AACAACAACCCAATATCCCTAGC-3V for GGPP synthase gene or 5V-CCGTTTGAATCCATGGCGGATCTGA-3Vand 5V-TCCAGTCTTTGTCGACGTATCTGGAT-3Vfor FPP synthase gene. The cycle numbers used for RT-PCR were chosen to be within the linear range of the

Fig. 5. Analyses of mRNA expression patterns by RT-PCR. (A) Control (18S rRNA); (B) mRNA expression of hbgg (17 cycle and 20 cycle); (C) mRNA expression of Hevea FPP synthase (17 cycle and 20 cycle).

A. Takaya et al. / Biochimica et Biophysica Acta 1625 (2003) 214–220

levels of PCR products of the standard mRNA for 18S ribosomal RNA. As shown in Fig. 5, the predicted PCR products for the GGPP synthase were detected in all Hevea tissues examined, though the expression levels were different from each other. Expression of the GGPP synthase gene in leaf, flower and young leaf were higher than those in latex and petiole. On the other hand, expression levels of the FPP synthase gene were high in latex as well as in flower. These results indicate that the GGPP synthase and FPP synthase genes were differentially expressed in various tissues of H. brasiliensis, which is likely to reflect their roles and specific functions in each tissue. The deduced amino acid sequence of the Hevea GGPP synthase signifies a putative presence of a transit peptide sequence to chloroplast, which is predicted from both of PSORT (K. Nakai, Tokyo University, http://psort.ims.utokyo.ac.jp/) and ChloroP programs [26] (Fig. 1). Moreover, the mRNA of Hevea GGPP synthase was found to be expressed in all tissues investigated in this study. It is likely that the GGPP synthase has similar functions as that of A. thaliana GGPS-1, which locates in the chloroplast and engages in the biosynthesis of biologically important isoprenoids such as carotenoids, chlorophylls and gibberellins [12]. In H. brasiliensis, the expression of GGPP synthase was found to be higher in leaf and flower, both of which contain much amount of chloroplast or plastid, than in latex and petiole (Fig. 5). On the other hand, it has been reported that there is no chloroplast in mature latex of H. brasiliensis [27], where a large quantity of natural rubber is produced. Instead, lutoids and Frey – Wyssling particles have been known as the major specialized particles characteristic of Hevea latex [28]. It may be possible that the GGPP synthase in Hevea latex is engaged in the biosynthesis of h-carotene, which is responsible for the characteristic color of Frey – Wyssling particles. But little is known about the functions of Frey –Wyssling particles. On the other hand, as shown in Fig. 5, the mRNA expression level of the FPP synthase in Hevea latex is remarkable. Furthermore, as the GGPP synthase, the FPP synthase has also been detected predominantly in the Cserum, where most of the rubber particles are present [11]. It is likely that the FPP synthase in Hevea latex has a major role to produce the starter substrate FPP for the prenyl chain elongation in natural rubber biosynthesis.

Acknowledgements This work was supported in part by Grants-in-Aid for Scientific Research (12480169 to T.K. and 13680667 to Y.W.Z.) from Ministry of Education, Science and Culture of Japan, by the ‘‘Research for the Future’’ Program (JSPSRFTF 97I002302 to T.K.) from the Japan Society for the Promotion of Science, and by the Asahi Glass, Heiwa – Nakajima, Sumitomo, and Takeda Science Foundations.

219

References [1] B.L. Archer, B.G. Audley, New aspects of rubber biosynthesis, Bot. J. Linn. Soc. 94 (1987) 181 – 196. [2] S. Madhaven, G.A. Greenblatt, M.A. Foster, C.R. Benedict, Stimulation of isopentenyl pyrophosphate incorporation into polyisoprene in extracts from guayule plants (Parthenium argentaum Gray) by low temperature and 2-(3,4-dicholoro-phenoxy)triethylamine, Plant Physiol. 89 (1989) 506 – 511. [3] K. Cornish, R.A. Backhaus, Rubber transferase activity in rubber particle of guayule, Phytochemistry 29 (1990) 3809 – 3813. [4] J. Tangpakdee, Y. Tanaka, K. Ogura, T. Koyama, R. Wititsuwannakul, D. Wititsuwannakul, K. Asawatreratanakul, Isopentenyl diphosphate isomerase and prenyl transferase activities in bottom fraction and Cserum from Hevea latex, Phytochemistry 45 (1997) 261 – 267. [5] B.G. Audley, B.L. Archer, Biosynthesis of rubber, in: A.D. Roberts (Ed.), Natural Rubber Science and Technology, Oxford Univ. Press, London, 1988, p. 35. [6] Y. Tanaka, H. Sato, A. Kageyu, Structure and biosynthesis mechanism of natural cis-polyisoprene from Goldenrod, Rubber Chem. Technol. 56 (1983) 299 – 303. [7] Y. Tanaka, Structural characterization of naturally occurring cis- and trans-polyisoprenes, Proc. Int. Rubb. Conf. Kyoto Japan (1985). [8] Y. Tanaka, Structural characterization of naturally occurring cis- and trans-polyisoprenes by 13C-NMR spectroscopy, J. Appl. Polym. Sci., Appl. Polym. Symp. 44 (1989) 1 – 9. [9] B.L. Archer, E.G. Cockbain, Rubber transferase from Hevea brasiliensis latex, Methods Enzymol. 15 (1969) 476 – 480. [10] D.R. Light, R.A. Lazarus, M.S. Dennis, Rubber elongation by farnesyl pyrophosphate synthase involves a novel switch in enzyme stereospecificity, J. Biol. Chem. 264 (1989) 18598 – 18607. [11] K. Adiwilaga, A. Kush, Cloning and characterization of cDNA encoding farnesyl diphosphate synthase from rubber tree (Hevea brasiliensis), Plant Mol. Biol. 30 (1996) 935 – 946. [12] K. Okada, T. Sato, T. Nakagawa, M. Kawamukai, Y. Kamiya, Five geranylgeranyl diphosphate synthase expressed in different organs are localized into three subcellular compartments in Arabidopsis, Plant Physiol. 122 (2000) 1045 – 1056. [13] A. Laferriere, P. Beyer, Purification of Geranylgeranyl diphosphate synthase from Sinapis alba etioplasts, Biochim. Biophys. Acta 1077 (1991) 167 – 172. [14] R. Kunz, S. Romer, P. Hugueney, J.H. Weil, R. Schantz, B. Camara, Identification of a cDNA for the plastid-located geranylgeranyl pyrophosphate synthase from Capasicum annuum: correlative increase in enzyme activity and transcript level during fruit ripening, Plant J. 2 (1992) 25 – 34. [15] A. Badillo, J. Steppuhn, J. Deruere, B. Camara, M. Kuntz, Structure of a functional geranylgeranyl pyrophosphate synthase gene from Capsicum annuum, Plant Mol. Biol. 27 (1995) 425 – 428. [16] P.A. Scolnik, G.E. Bartly, Nucleotide sequence of an Arabidopsis cDNA for geranylgeranyl pyrophosphate synthase, Plant Physiol. 104 (1994) 1469 – 1470. [17] X.F. Zhu, K. Suzuki, K. Okada, K. Tanaka, T. Nakagawa, M. Kawamukai, H. Matsuda, Cloning and functional expression of a novel geranylgeranyl pyrophosphate synthase gene from Arabidopsis thaliana in Escherichia coli, Plant Cell Physiol. 38 (1997) 357 – 361. [18] X.F. Zhu, K. Suzuki, T. Sato, K. Okada, K. Tanaka, T. Nakagawa, H. Matsuda, M. Kawamukai, Geranylgeranyl pyrophosphate synthase encoded by newly isolated gene GGPS6 from Arabidopsis thaliana is localized in mitochondria, Plant Mol. Biol. 35 (1997) 331 – 341. [19] S.M. Aitken, S. Attucci, R.K. Ibrahim, P.J. Gulick, A cDNA encoding geranylgeranyl pyrophosphate synthase from white lupin, Plant Physiol. 108 (1995) 837 – 838. [20] B. Bantignies, T. Liboz, C. Ambid, Nucleotide sequence of a Catharanthus roseus geranylgeranyl pyrophosphate synthase gene, Plant Physiol. 110 (1995) 336.

220

A. Takaya et al. / Biochimica et Biophysica Acta 1625 (2003) 214–220

[21] J. Henfer, R.E. Ketchum, R. Croteau, Cloning and functional expression of a cDNA encoding geranylgeranyl diphosphate synthase from Taxus canadensis and assessment of the role of this prenyltransferase in cells induced for taxol production, Arch. Biochem. Biophys. 360 (1998) 62 – 74. [22] S.-K. Oh, I.J. Kim, D.H. Shin, J. Yang, H. Kang, K.-H. Han, Cloning, characterization, and heterologous expression of a functional geranylgeranyl pyrophosphate synthase from sunflower (Helianthus annuus L.), J. Plant Physiol. 157 (2000) 535 – 542. [23] W. Sitthithaworn, N. Kojima, E. Viroonchatapan, D.Y. Suh, N. Iwakami, T. Hayashi, M. Noji, K. Saito, Y. Niwa, U. Sankawa, Geranylgeranyl diphosphate synthase from Scoparia dulcis and Croton sublyratus. Plastid localization and conversion to a farnesyl diphosphate synthase by mutagenesis, Chem. Pharm. Bull. 49 (2001) 197 – 202. [24] T. Koyama, H. Fujii, K. Ogura, Enzymatic hydrolysis of polyprenyl pyrophosphates, Methods Enzymol. 110 (1985) 153 – 155.

[25] J.D. Thompson, D.G. Higgins, T.J. Gibson, CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice, Nucleic Acids Res. 22 (1994) 4673 – 4680. [26] O. Emanuelsson, H. Nielsen, G. Von Heijne, Chloro P, a neural network-based method for predicting chloroplast transit peptides and their cleavage sites, Protein Sci. 8 (1999) 978 – 984. [27] J.B. Gomez, Comparative ultracytology of young and mature latex vessels in Hevea brasiliensis, Proc. Int, Rubber Conf., vol. 2, RRIM, Kuala Lumpur, 1976, pp. 143 – 164. [28] D. Wititsuwannakul, R. Wititsuwannakul, Chapter 6, biochemistry of natural rubber and structure of latex, in: T. Koyama, A. Steinbu¨chel (Eds.), Biopolymers, Polyisoprenoids, vol. 2, Wiley-VCH, Weinheim, 2001, pp. 151 – 201.