Background

Forskolin, a labdane diterpene, is a major active compound isolated from tuberous roots of Coleus forskohlii Briq. (Lamiaceae) [1]. C. forskohlii has been used as an important folk medicine in India. Futher, forskolin has been found to be a potent activator of adenylate cyclase [2], leading to an increase in levels of c-AMP, which affects heart action, blood and intraocular pressure. Recently, forskolin has become commercially available as a drug for treating heart disease in Japan. Forskolin is not available by chemical synthesis due to its complicated structure. However, two groups have reported successful total synthesis of forskolin [3, 4].

Isoprenoids are essential for the normal growth and development processes in all living organisms. Isopentenyl diphosphate (IPP; C5) is a common metabolic precursor of all isoprenoids. Recently, several groups have demonstrated that two distinct pathways synthesise IPP in plants. The mevalonate (MVA) pathway occurs in the cytoplasm, and an alternative mevalonate-independent (2C-methyl-D-erythritol 4-phosphate; MEP) pathway occurs in plastids [5–7].

Geranylgeranyl diphosphate (GGPP) synthase catalyses the consecutive condensation of an allylic diphosphate with three molecules of IPP to produce GGPP, an essential linear precursor for biosynthesis of diterpenes, carotenoid, retinoids and side chain of chlorophyll [8]. GGPP synthase is an important branch point prenyltransferase enzyme in terpenoid biosynthesis.

GGPP synthase genes have been cloned in a number of organisms including; Arabidopsis thaliana [9, 10], Taxus canadensis [11], Helianthus annuus [12], Scoparia dulcis and Croton sublyratus [13], Sulfolobus acidocaldarius [14], Neurospora crassa [15], and mouse and human [16]. Amino acid sequence comparison has shown that GGPP synthases contain several domains of conserved amino acid residues including the first aspartate-rich motifs (FARM) and the second aspartate-rich motif (SARM) [17]. Futhermore, recent studies suggested that two amino acids at the four and five positions before FARM in the sequence, as well as an insertion in FARM of plant GGPP synthases play important roles in product length determination [13, 18].

Carotenoids arise from the coupling of two molecules of GGPP. The carotenoid biosynthetic gene cluster (crt genes) of Erwinia uredovora was elucidated [19], and is currently used to investigate the function of carotenoid related genes in a heterologous system. This crt gene cluster is composed of six genes; crtB (phytoene synthase), crtE (GGPP synthase), crtI (phytoene desaturase), crtX (zeaxanthin β-glucosidase), crtY (lycopene cyclase) and crtZ (β-carotene hydroxylase). Consequently, the production of carotenoids using E. coli harbouring the crt gene cluster can be used for the determination of GGPP synthase activity.

GGPP synthase is suggested to be a key enzyme in the biosynthesis of forskolin. Herein, we report the cDNA encoding C. forskohlii GGPP synthase and its heterologous expression in E. coli.

Results and discussion

cDNA cloning and sequencing of C. forskohlii GGPP synthasegene

The open reading frame (ORF) for full-length GGPP synthase gene encodes a protein of 359 amino acids, 1,077 nucleotides long, with a calculated molecular mass of 39.3 kDa. The amino acid sequence of C. forkohlii GGPP synthase revealed high homology throughout the entire coding region of Catharanthus roseus (75%), Arabidopsis thaliana (73%), Sinapis alba (72%), Croton sublyratus (69%), Scoparia dulcis (67%) and Mentha piperita (64%) (Fig. 1). However, comparison of the amino acid sequence with that of prokaryotic GGPP synthases showed a low level of homology (30–53%). Highly conserved residues were designated as domains I-VII. Two conserved aspartate-rich motifs, DDXX(X)D, were identifed. FARM and SARM have been shown to be important in substrate binding and catalysis [20–22].

Figure 1
figure 1

Amino acid sequence alignment of C. forskohlii GGPP synthase and other plant GGPP synthases. The residues boxed in black indicate the positions of at least five of the six compared sequences while the gray shading indicates similar amino acids. Dashes indicate gaps introduced for optimization of the alignment. Numbers of amino acids are indicated in the left and right margins. Asterisks show the seven domains that are highly conserved among prenyltransferases with two aspartate-rich motifs (domain II and VI). Species and accession numbers are C. forskohlii, AY515700; S. dulcis, AB034250_1; C. sublyratus, AB034249_1; C. roseus, T09966; S. alba, T10452; A. thaliana, F85434.

Full size image

Transient expression of putative localization signal of C. forskohliiGGPP synthase in tobacco cells

Sequence alignment of plant GGPP synthases showed that the N-terminal region has a low level of homology. It is reasonable to assume that these GGPP synthases have localization signals in their N-terminal regions to target them into specific subcellular compartments. The N-terminal region of C. forskohlii GGPP synthase was predicted to be localized in chloroplasts by the ChloroP 1.1 Prediction Server. In an effort to determine the localization of C. forskohlii GGPP synthase, the sequence coding for the 80 amino acid sequence at the N-terminus of C. forskohlii GGPP synthase was fused to the N-terminus of the GFP reporter gene and transformed into BY-2 tobacco cells. The pattern of putative localization signal of C. forskohlii GGPP synthase was identical to the positive chloroplast targeting signal [35SΩ-pt-sGFP(S65T)] (Fig. 2). The N-terminal region of C. forskohlii GGPP synthase was determined to contain a chloroplast localization signal. Recently, plant GGPP synthases have been determined to be translocated into plastids, mitochondria and cytosol [9, 23].

Figure 2
figure 2

Transient expression of GFP and GFP fusion proteins in BY-2 tobacco cells. (A) GFP protein [35SΩ-sGFP (S65T)]; (B) Arabidopsis chloroplast targeting signal (pt)-GFP fusion protein [35SΩ-pt-sGFP (S65T)]; (C) putative localization signal of GGPP synthase-GFP fusion protein [35SΩ-GGPP synthase-sGFP (S65T)]

Full size image

Heterologous expression and activity of C. forskohliiGGPP synthase

In order to express C. forskohlii GGPP synthase, the gene was constructed and cloned into the plasmid pBluescript II KS-. The fusion protein of GGPP synthase with lacZ had a calculated molecular mass of 41.6 kDa, was observed in the soluble fraction of E. coli carrying pGGPPS after IPTG induction (Fig. 3).

Figure 3
figure 3

Expression of GGPP synthase in pBluescript II KS-vector analysed by SDS-PAGE. M = molecular mass standards are indicated in kDa; lane 1 = control (E. coli transformed with pBluescriptII KS-); lane 2 = total protein extract after 6 hours of IPTG induction; lane 3 = soluble cytoplasmic fraction of cells treated with IPTG; and lane 4 = insoluble fraction of cells treated with IPTG

Full size image

Functional activity of expressed GGPP synthase was investigated by genetic complementation with the carotenogenic crt gene cluster. Carotenoids are produced in E. coli harbouring a crt cluster gene from E. uredovora. Replacements of a crt gene with an unknown gene with the same activity, can be used to determine the function of the gene [15]. Herein, the C. forskohlii GGPP synthase gene was cloned into pBluescript II KS- vector (pGGPPS) in order to produce a lacZ fusion protein. pGGPPS was then transformed into E. coli DH10B carrying the plasmid pACCAR25ΔcrtE in which the crtE encoding GGPP synthase had been deleted. The yellow color of carotenoid was observed in the transformant, indicating that pGGPPS carried the gene substituting the function of the crtE gene (Fig. 4). Carotenoid production of the transformants was compared with that of E. coli transformant carrying plasmid pACCAR25ΔcrtE and pBAA encoding mouse GGPP synthase (positive control) [16], and with transformant carrying plasmid pACCAR25ΔcrtE and a pBluescript II KS- (pBS) vector (negative control). This result suggested that the coding region of a cDNA of C. forskohlii GGPP synthase encodes a functional GGPP synthase.

Figure 4
figure 4

Carotenoid production of E. coli harboring plasmid pACCAR25ΔcrtE and plasmid expressing. (1) mouse GGPP synthase, pBAA; (2) plasmid expressing C. forskohlii GGPP synthase, pGGPPS; and (3) pBluescript II KS- vector, pBS

Full size image

Expression of GGPP synthase gene in organs of C. forskohlii

The expression of GGPP synthase gene was investigated by RT-PCR in different organs of C. forskohlii. Total RNA extracted from the roots, stems and leaves of an eight-month-old plant were analysed. The C. forskohlii GGPP synthase gene was strongly expressed in the leaves, whereas expression was decreased in stems and barely expressed in roots (Fig. 5). Therefore, the leaves are thought to be the primary location for forskolin synthesis. We previously reported the forskolin concentration in clonally propagated plant organs of C. forskohlii [24]. Tuberous roots and the stem base were determined to contain a higher concentration of forskolin than the organs. Moreover, the stem base, parts of the epidermis and cortex, the vascular bundle, and the pith were analysed separately. The highest concentration of forskolin was identified in the vascular bundle tissue. From these data, we proposed that GGPP synthase involved in biosynthesis of forskolin, is mainly synthesised in leaves, subsequently distributed to stems and finally accumulated in stem bases and roots.

Figure 5
figure 5

Expression of GGPP synthase in roots, stems and leaves of C. forskohlii culture. Ten microliters of PCR product was loaded in each lane. The lower panel shows 18S rRNA fragment as an internal control.

Full size image

Forskolin production via non-mevalonate pathway

In an effort to investigate the forskolin biosynthesis pathway by a non-mevalonate pathway, various concentrations of fosmidomycin, the specific inhibitor of 1-deoxy-D-xylulose-5-phosphate reductoisomerase (DXR) enzyme in the non-mevalonate pathway were applied to the C. forskohlii culture and the forskolin content of roots was determined (Fig. 6). Treatment led to a decrease in forskolin, whereas 10 μM fosmidomycin had no effect on forskolin production. At higher concentrations a dose-dependent inhibitory effect was observed. At 1000 μM fosmidomycin, the forskolin content was decreased by up to fifty percent in comparison to the control tissue without inhibitor treatment. Thus, forskolin was thought to be synthesised via a non-mevalonate pathway.

Figure 6
figure 6

Effect of fosmidomycin on forskolin production.

Full size image

A recent 13C-glucose feeding experiment using 13C-NMR analytical methodology suggested the biosynthetic pathway of forskolin via a non-mevalonate pathway [25]. In addition, the DXR gene regarding the specific enzyme in the first step of the non-mevalonate pathway was cloned from C. forskohlii [26].

Conclusions

C. forskohlii GGPP synthase was cloned and its subcellular localization was determined. The N-terminal region contained a signal which was localized in chloroplasts. Functional expression of GGPP synthase was investigated by genetic complementation with the carotenogenic crt gene cluster. Carotenoids were produced when the crtE gene was replaced with C. forskohlii GGPP synthase. GGPP synthase is thought to be involved in biosynthesis of forskolin, which is primary synthesised in the leaves, subsequently distributed to stems and finally accumulated in stem bases and roots.

Methods

Plant materials and reagents

C. forskohlii plantlets were cultured in hormone-free MS (Murashige and Skoog) medium at 25°C under a 16 hours light cycle. The light intensity was 3000 lux and the relative humidity was 60%. Shoot cuttings (10 mm in length) propagated by shoot tip culture were successively cultivated in vermiculite. BY-2 tobacco single cell suspension [27] was cultured in liquid modified LS (Leinsmaier and Skoog) medium supplemented with 0.2 mg l-1 of 2,4-D (2,4-dichlorophonoxy acetic acid) under dark conditions at 25°C on an orbital incubator. Restriction enzymes, ligase, and PCR-polymerase were purchased from Takara Shuzo Co., Ltd. (Tokyo, Japan) and Toyobo Co., Ltd. (Tokyo, Japan). Fosmidomycin (FR-3154) was purchased from Molecular Probes (Oregon, USA). Chemical reagents were purchased from Sigma Chemical Company (St. Louis, USA) and Nacalai Tesque Inc. (Tokyo, Japan)

Bacterial strains and plasmids

E. coli TOP10F' and E. coli DH10B carrying the plasmid pACCAR25ΔcrtE were used in the present investigation. The pUC119 vector was used for cDNA cloning and sequencing. The pBluescript II KS- vector was used as a GGPP synthase expression plasmid. The 35SΩ-sGFP(S65T) plasmid was used as a green fluorescent protein (GFP) reporter plasmid. The pBI121 plant vector and Agrobacterium tumefaciens LBA4404 were used for transformation of GFP and GFP-fusion genes to plant cells.

cDNA cloning and sequencing of C. forskohlii GGPP synthasegene

Total RNA was prepared from roots of the C. forskohlii culture using the acid guanidium-phenol-chloroform extraction procedure [28]. Single strand cDNA was synthesised using an oligo-dT adapter primer, M-MLV reverse transcriptase and total RNA as template. Degenerate primers were designed based on highly conserved amino acid sequences of previously cloned genes encoding plant GGPP synthases [13]. A 470 bp cDNA fragment was amplified using a nested PCR with Taq DNA polymerase and degenerate primers A, B, C and D (Table 1). The 3' end of cDNA was amplified using 3' rapid amplification of cDNA ends (RACE) with gene specific primers I and J, and adapter primer F. A 522 bp product was obtained by nested PCR. For 5' RACE, the first strand cDNA was polyadenylated at its 5' end by terminal deoxynucleotidyl transferase. The first and second PCR were performed with specific primers G and H and adapter primers E and F. A 285 bp product was obtained. The entire coding region of 1,077 bp was amplified by nested PCR using specific primers K, L, M and N designed from 5' and 3' RACE products.

Table 1 Primers for cDNA cloning of GGPP synthase.
Full size table

All amplified cDNA fragments were purified and digested with restriction enzymes at sites introduced via the PCR primers, and cloned into the vector pUC119. After transformation to E. coli TOP10F', clones harboring inserts were sequenced using a Model 310 Genetic Analyzer (PE Biosystems) using a BigDye Terminator Cycle Sequencing Kit.

The amino acid sequence deduced from the nucleotide sequence was compared with sequence databases in the Genome Net WWW server using the FASTA program. Multiple amino acid sequence alignment was performed using the CLUSTALW Multiple Sequence Alignment in the GenomeNet CLUSTALW Server.

Construction and expression of putative localization signal of C. forskohliiGGPP synthase

A 240 bp fragment of the N-terminal region of C. forskohlii GGPP synthase was PCR-amplified using primers P and Q and the PCR product was digested and cloned into the SalI-NcoI site of the 35SΩ-sGFP(S65T) plasmid. 35SΩ-pt-sGFP(S65T) was used as the positive control for chloroplast targeting [29, 30]. GFP, GGPP synthase-GFP fusion and pt-GFP fusion with CaMV35SΩ promoter and NOS3' terminator [35SΩ-sGFP (S65T), 35SΩ-GGPP synthase-sGFP (S65T) and 35SΩ-pt-sGFP (S65T), respectively] were subcloned into the HindIII-EcoRI site of the pBI121 vector and then transformed into A. tumefaciens LBA4404. The transformants were cultured at 28°C for two days in YEB liquid medium containing 25 μg/ml of kanamycin and 25 μg/ml of rifampicin. The transformants were washed twice and re-suspended in YEB medium. Agrobacterium transformants (108 cells) were applied to four ml of five-day-old BY-2 suspension culture. The culture was incubated at 28°C for two days under dark conditions. GFP and GFP fusion protein were analysed by fluorescence microscopy using Nikon Eclipse TE2000-U model. Cells were observed at a 400 × magnification.

Construction of plasmid for C. forskohliiGGPP synthase expression

The coding region of a cDNA of C. forskohlii GGPP synthase was amplified by PCR using specific primers M and O. A PCR product was digested; purified and cloned into the KpnI-SalI site of pBluescript II KS- vector, namely pGGPPS. This plasmid was transformed into E. coli XL1-Blue MRF' for over-expression. The transformants were cultured in LB liquid medium containing 50 μg/ml of ampicillin and 25 μg/ml of chloramphenicol. The culture was induced with 1 mM isopropyl-1-thio-β-D-galactoside (IPTG) and incubated for six hours at 37°C. The cells were harvested and washed with 50 mM Tris-HCl pH 8.0 by centrifugation. The pellet was re-suspended, lysozyme was added and the mixture was incubated for 30 minutes. The mixture was then sonicated for four cycles of 15 seconds at one minute intervals. The soluble fraction was obtained after centrifugation at 10,000 × g for 10 minutes. SDS-PAGE was conducted in order to detect the proteins [31].

Genetic complementation expression

The pACCAR25ΔcrtE plasmid contains the gene cluster crtB, crtI, crtX, crtY and crtZ encoding carotenoid biosynthetic enzymes with the exception of crtE (encoding GGPP synthase). The plasmid pBAA containing mouse GGPP synthase (positive control plasmid) and E. coli DH10B carrying the plasmid pACCAR25ΔcrtE was provided by Dr. M. Kawamukai, Shimane University, Japan [16]. pBluescript II KS- vector, pBS, was used as negative control. pGGPPS, pBAA and pBS were transformed into E. coli DH10B carrying the plasmid pACCAR25ΔcrtE. All transformants were plated on LB agar medium containing 50 μg/ml of ampicillin and 25 μg/ml of chloramphenicol and then incubated for two to three days at 25°C.

Reverse transcriptase-PCR (RT-PCR)

An eight-month-old C. forskohlii was analysed in twelve separate parts; leaf (L1–L4), stem (S1–S5) and root (R1–R3). The numbering is based on the maturation of organs. Total RNA was extracted from each part of plant. One microgram of total RNA was used as the template for the synthesis of the first strand cDNA (using SuperScript First-Strand Synthesis System for RT-PCR, Invitrogen). Primers M and O, the first strand cDNA and KOD-polymerase were used for the amplification of C. forskohlii GGPP synthase with the condition of denaturation, 98°C, 15 seconds; annealing, 60°C, 2 seconds and extension, 74°C, 5 seconds. The 18S rRNA fragment used as an internal control was amplified using primers R and S under the same conditions of C. forskohlii GGPP synthase amplification. The amplified PCR products were analysed by 1.0% agarose gel electrophoresis.

Analysis of forskolin production

C. forskohlii plantlets were treated with various concentrations of fosmidomycin and then investigated for forskolin content using the HPLC method as previously described [26]. Forskolin was detected by comparison with the retention time of a forskolin standard (Sigma) detected by UV absorption at 202 nm.