Phytochemistry98.pdf

Phytochemistry Vol. 49, No. 7, pp. 1929±1934, 1998 # 1998 Elsevier Science Ltd. All rights reserved GINSENOSIDE PRODUCTION BY GINSENG HAIRY ROOT VICTOR P. BULGAKOV*, MARIA V. KHODAKOVSKAYA, NATALI V. LABETSKAYA, GALINA K. CHERNODED and YURI N. ZHURAVLEV Laboratory of Biotechnology, Institute of Biology and Soil Science, Far East Branch of Russian Academy of Sciences, Vladivostok, 690022 Russia Key Word IndexÐPanax ginseng; Araliaceae; transgenic culture; rol genes; ginsenosides; AbstractÐPlasmid constructions containing rolA, rolB and rolC genes, isolated earlier from the TL-DNA of Agrobacterium rhizogenes were used to transform a cell culture (strain 1c) of Panax ginseng. The levels of ginsenosides were measured in the resulting transgenic tissues to evaluate the possible role of rol genes in ginsenoside formation. The ginsenoside content of the hairy root culture of P. ginseng, transformed by wild type A4 plasmid DNA and containing all rol loci, was higher than that of the control 1c culture (5.12±8.92 mg gÀ1 dry wt), being in the range of 13.23±21.27 mg gÀ1 dry wt. Ginseng tissue, transgenic for the rolA gene appeared to lose the ability to synthesize ginsenosides since only a trace amount of Re ginse- noside was found in 1c-rolA tissue. 1c-rolB cultures contained at least ®ve times lower ginsenoside levels compared to the initial 1c culture. The ginsenoside content of rolC transgenic roots was about three times higher than that of the respective control. Taking into account the di€erences in cell di€erentiation levels in tissues transformed by rol genes, we compared the ginsenoside levels in rolC roots and tumours. It was found that ginsenoside production in tissues with di€erent levels of di€erentiation is nearly the same. We have concluded that the plant oncogene rolC is responsible for increased ginsenoside formation in ginseng hairy rootcultures. # 1998 Elsevier Science Ltd. All rights reserved Yoshikawa and Furuya [8] performed ginsenoside determinations on several ginseng lines transformed RolA, rolB and rolC oncogenes isolated from T- by A4 A. rhizogenes strain and found the amount DNA of the Agrobacterium rhizogenes A4-Ri plas- of ginsenosides was two times higher than that of mids are known to be involved in the induction of untransformed cultured roots. Further, Coreopsis hairy roots in transformed plants. The expression of tinctoria hairy roots synthesized up to three times individual rol genes in transgenic plants not only as much 1'-acetoxy-eugenol-isobutyrate as ordinary induces rhizogenesis [1], but also a€ects the devel- cultured roots [9], and transformed culture roots of opment and physiology of the entire plant [2].
Chaenactis douglasii contained two times higher It was shown that hairy root cultures of a num- thiarubrine levels compared to ordinary cultured ber of plants had higher levels of secondary pro- roots [10]. However, the physiological relevance of ducts (e.g. nicotine, anabasine, cytisine, anagyrine, the level of cell di€erentiation with respect to sec- hyoscyamine, scopolamine, ginsenosides, thiarubrin ondary metabolite formation has never been and polyacetylenes) than wild type plants [for demonstrated, nor have the genes responsible for review see Refs [3] and [4]]. Ithas been proposed the increases in biosynthesis been identi®ed.
that increased biosynthesis of secondary metabolites To assess the possible role of rol genes in the in hairy root cultures is correlated with rhizogenesis phenomenon of secondary metabolite overproduc- caused by T-DNA integration [5±7]. In contrast to tion in hairy root cultures, we investigated the this conception, some data reveals that the pro- e€ects of these genes on ginsenoside levels by inocu- duction of secondary metabolites in hairy roots lating cultivated ginseng cells with bacteria harbour- exceeds that in ordinary cultured roots. Thus, ing rol genes. We chose to use a ginseng cell culture for these studies, since ginseng cultures have been *Author to whom correspondence should be addressed.
well studied [11±13]. The results of this work Table 1. Changes in properties of ginseng cells after transformation * Growth on the medium containing 100 mg lÀ1 kanamycin sulphate.
$ Five primary tumour cultures (1c-rolC-I. 1c-rolC-V).
% Five rootcultures (1c-rolC-I. 1c-rolC-V).
demonstrate that increased ginsenoside formation they displayed a friable, almost watery phenotype.
caused by A. rhizogenes, strain A4, can be explained Within the ®rst or subsequent passages, all estab- lished primary tumours transformed with pPCV002- CaMVC had formed adventitious roots. Transgenic calli produced roots on the medium containing 4- CPA as well as on hormone-free medium. Although Transformation of ginseng cells by A. rhizogenes and the eciency of root formation was di€erent in A. tumefaciens GV 3101 and regeneration of trans- di€erent tumour lines, none of the cultures of pri- mary tumours showed decreased formation of roots from one subculture to the next. However, it was Ginseng cell clusters were transformed by co-cul- easy to select a non-root-forming tumour line by tivation with A. tumefaciens GV 3101 strains har- selection for non-root-forming calli.
bouring rol genes, as well as with A. rhizogenes Adventitious roots emerging on tumours were strain A4. Of the initial several hundred primary excised and transferred into liquid media. In the tumours established, one rolA gene line, two rolB absence of phytohormones, as well as in the presence gene lines and ®ve rolC gene lines were resistant to of indole-3-acetic acid, root cultures were character- high kanamycin concentrations on subculturing and ized by slow growth and reduced lateral branching.
were thus con®rmed to be transformed (Table 1).
In the presence of 4-CPA, roots grew vigorously but Several hairy root cultures were established by wild- had a tendency to swell and form callus-like struc- tures over several subcultures. Fast growing rolC The rolA calli grew very slowly (Table 2) as com- root cultures with abundant lateral branching were pactyellow globular aggregates on hormone-free established using indole-3-butyric acid and these cul- medium as well as on W4CPA medium. These calli tures were used for further experiments.
Ginseng cell lines transformed with A. rhizogenes The rolB calli grew well as friable white-yellow strain A4 were also established which grew as pri- tissue on hormone-free medium and on W4CPA mary tumours and roots. One root clone showing medium containing 0.1 mg lÀ1 of 4-chlorophenoxya- rapid growth in WIBA medium for several subcul- cetic acid. No roots were observed on these calli tures was selected and used for analysis of ginseno- and our attempts to trigger rhizogenesis in this cul- The primary rolC-tumors grew rapidly on W4CPA Ginsenoside content of hairy roots, rolA, rolB and medium supplemented with kanamycin on which Table 2. Biomass accumulation by initial and transgenic Ginsenosides were extracted from transformed cultures growing in W4CPA medium for 28 days tissues and separated by HPLC. Transformed roots produced a setof ginsenosides which did notdi€er signi®cantly from those produced by the parent cul- culture, ginsenoside levels were nearly doubled in 1c-A4 roots, in accordance with the observations of Yoshikawa and Furuya [8]. Surprisingly, we found that 1c-rolA calli derived from transformed rolA gene cells of P. ginseng lost the ability to synthesize ginsenosides whereas the ginsenoside content of rolB cultures was low, in contrast to 1c cells. Over 1.5 years of analysis, the average amount of ginse- * Mean values2s.e. based on ®ve separate replicate samples.
1.2320.14 mg gÀ1 dry wt; that is 5.7 times lower Table 3. Production of ginsenosides by the initial and transformed cultures grown on liquid W4CPA medium for 28 days 0.8020.17 0.1320.02 0.1320.03 0.1720.02 7.1020.74 3.0021.02 1.1920.42 1.3320.26 1.0320.12 17.2524.02 0.6120.12 1.0520.14 2.6820.45 0.0320.005 1.2920.27 0.1220.04 0.5520.12 0.3520.04 6.0720.82 4.7821.10 0.5820.11 0.9220.19 0.7520.20 20.6325.19 0.5220.14 1c-rolC-III 3.0920.55 7.1422.12 1.6520.70 4.1320.67 0.7720.22 2.0620.22 0.8220.10 19.6624.06 0.6520.12 2.6520.21 0.6220.16 1.7020.20 0.5520.09 14.7822.25 0.5520.08 2.4520.55 0.5920.08 1.4920.26 0.5920.13 12.8222.12 0.6120.06 * Mean values2s.e. based on at least ®ve separate replicate samples.
than in the 1c culture. During this time, the 1c-rolA Rc, Rb2, Rd) in ginseng tissue. As a rule, above- ground ginseng parts have an index of less than 1, As indicated in Table 3, all rolC-rootlines except and underground parts more than 1 [11]. In the 1c for 1c-rolC-I contained ginsenosides concentrations callus culture, the Rb/Rg index is always less than 1 that exceeded 1.8±3 fold those of the untransformed (0.16±0.26); this is not surprising since the culture control culture. To determine whether this pattern originated from a stem. In contrast, the level of gin- of ginsenoside accumulation is stable between sub- senosides of Rb group increased in the all trans- cultures, we monitored ginsenoside levels in all root genic tissues (Table 3) which indicates that rol genes lines for 6 months. The results showed that di€er- strongly in¯uence ginsenoside biosynthesis.
ences in glycoside accumulation reported in Table 3 within the cultures remained constant during this Dependence of ginsenoside formation on the levels of time (data not shown). The 1c-rolC-II rootculture possessing the highest growth rate (Table 2) was We explored whether increased ginsenoside pro- duction in rolC cultures is a phenomenon associated Ginsenoside production by the 1c-rolC-II culture with rhizogenesis caused by T-DNA integration. A over long-term cultivation (two years) was found to single root tip was isolated from the 1c-rolC-II cul- vary from one subculture to another. The total gin- ture and cultivated in liquid WIBA medium for 3.5 senosides levels in this line ranged from 6.76 to months with 21 day subculture intervals. The result- 65.83 mg gÀ1 dry wt whereas those of the original ing tissue, which consisted of young lateral roots culture 1c varied between 5.12 and 8.92 mg gÀ1 dry (about 1 month old), mature lateral roots (2±2.5 wt. On average, 1c-rolC-II roots contained three month old), the main root (3±3.5 month old), and times more ginsenosides than control 1c cells. The secondary tumour tissue originating from the base ginsenoside content of wild-growing and plantation of the main root (3.5 month old) was analysed for plants were reported previously [14]. It was found ginsenoside content (Table 4). All tissues contained that roots of 19 P. ginseng plants collected in the 12 14±16 mg gÀ1 dry wtof ginsenosides exceptfor regions of the Russian Far East accumulated ginse- young roots which accumulated slightly decreased nosides ranging from 6.49 to 42.36 mg gÀ1 dry wt (an average value was 16.7722.24 mg gÀ1 dry wt).
The rolC-II tumour was grown separately and a While maximum amounts of ginsenosides in the culture of secondary tumours was established. As 1c-rolC-II roots exceeded those reported for natural shown in Fig. 1, the ginsenoside content of second- ginseng roots, the average content of ginsenosides ary tumour and rolC-II rootcultures varied over found corresponds to those occurring in natural several subcultures but on average they contained roots. In the transgenic 1c-rolC-II root culture, the accumulation of all ginsenosides correlated with the increase in biomass during the log phase of growth and maximum values were attained during the stationary phase of growth (data not shown).
The results presented in this report provide the ®rstevidence thatrol genes may have a strong e€ect Rol genes change the Rb group/Rg group ratio on secondary metabolism in plants. To determine The Rb/Rg index, which is often used to clarify the e€ects of rol genes on ginsenoside production, peculiarities in ginsenoside biosynthesis, indicates we transformed cell culture of P. ginseng using plas- the ratio between protopanaxadiol glycosides (Rg1, mid DNA containing the individual rol genes from Re, Rg2, Rf) and protopanaxatriol glycosides (Rb1, the TL-DNA of A. rhizogenes, A4 strain. The use Table 4. Ginsenoside content at di€erent stages of development and dedi€erentiation of the rolC-II roots Young lateral roots 1.3520.20 3.3720.22 0.5120.08 1.6820.20 3.3720.34 0.3720.06 0.5120.08 0.6720.10 11.821.15 1.7620.12 4.4820.67 0.9620.11 0.3220.04 4.4820.50 0.6420.09 0.8020.12 0.4820.07 13.9221.69 Main root2.1820.18 6.2520.55 1.0020.16 0.2020.03 4.6620.25 0.6920.09 1.0920.21 0.5920.16 16.6522.03 1.9420.35 8.2721.11 0.6420.06 0.5120.08 2.0720.19 Ð$ * Mean values2s.e. based on three independent determinations.
of callus culture was considered necessary in order action of rolC gene or is a phenomenon associated to ensure the homogeneity of the material for trans- with rhizogenesis caused by rolC gene integration, formation. While only a few tissues transgenic for we investigated the ginsenoside content of roots and rol genes were obtained and analyzed, signi®cant tumours of the same origin. As equal amounts of di€erences were found with respect to their growth, ginsenosides in tumour and root cultures were observed (Table 4, Fig. 1), itis dicultto accept Ginsenoside content in the hairy root culture estab- the suggestion that increased ginsenoside pro- lished after transformation of the 1c cell lines by duction in hairy roots could be attributed to root A4 A. rhizogenes was 2-fold higher then in the con- formation. This result is in agreement with the ob- trol culture (Table 3). We have shown that there is servation that pRiA4-derived calli of P. ginseng much variation between rolA, rolB and rolC cul- produce larger amounts of ginsenosides than non- tures in terms of their ability to produce ginseno- transformed calli obtained from the same plant [15].
sides. Cultures of rolC roots accumulated generally 1.8±3-fold more ginsenosides than control culture.
In contrast, minute amounts of ginsenosides were detected in rolA and rolB tissues (Table 3). Our results suggest that the rolC gene alone may play an important role in stimulating the biosynthetic ac- tivity of ginseng hairy root cultures.
CaMVBT and pPCV002-CaMVC) were kindly pro- In addition to the determination of glycoside pro- vided by Angelo Spena (Max-Planck-Institute fuÈr duction by the various rol cultures, preliminary ex- ZuÈchtungsforschung, Germany). These plasmids periments were undertaken in an attempt to contain the plant cassette vector pPCV002 [16] con- understand the mechanisms by which transformed taining the gene of interest: rolC and rolB under plant cells increase secondary metabolite pro- cauli¯ower mosaic virus (CaMV) 35S promoter duction. To assess whether stimulation of ginseno- control and rolA under the control of its own 5' side production in rolC roots is due to the direct promoter [1]. Constructions also carry a gene for Fig. 1. Ginsenoside production in Panax ginseng 1c culture, 1c-rolC-II rootculture and 1c-rolC-II sec- ondary tumour culture grown in W4CPA liquid medium for 4 weeks at258. Values are means2s.e. for kanamycin resistance (NPT-II) under eukaryotic fate. After 4±5 weeks of cultivation in the same control sequences. Standard techniques were used liquid medium, small white aggregates were for the construction of transformation systems, iso- observed. These 1.5±2 mm aggregates were trans- lation and analysis of DNA [17]. Esherchia coli ferred to W4CPA agarized medium with kanamycin TG2 strain was transformed by pPCV002 vectors.
to produce lines of primary kanamycin-resistant Further constructions were transferred from E. coli tumours designated as 1c-rolA (for the pPCV002-A strains TG2 to A. tumefaciens GV 3101 [16] as construction), 1c-rolB-I, 1c-rolB-II (for pPCV002- described [16]. The presence of rol genes in CaMVBT construction) and 1c-rolC-I, 1c-rolC-II,., Agrobacterium strains was con®rmed by restriction 1c-rolC-V (for pPCV002-CaMVC construction).
The 1c strain as well as the primary tumours were cultivated with 30 day subculture intervals in the dark at24±258 in 100 ml Erlenmeyer ¯asks.
E. coli TG2/pPCV002 strains were grown at 378 Primary 1c-rolC tumors were observed to spon- in LB medium with the addition of tetracycline taneously form adventitious roots. A number of (15 mg lÀ1) and ampicillin (50 mg lÀ1), GV 3101- transgenic ginseng root cultures were established by derived strains were grown in the LB medium con- placing root tips, isolated from adventitious roots, taining 50 mg lÀ1 kanamycin sulphate and 100 mg into liquid WIBA medium. These cultures, desig- lÀ1 carbenicillin at288. A4 A. rhizogenes strain was nated as 1c-rolC-I roots, 1c-rolC-II roots, etc., were grown under the same conditions without anti- further subcultured at 28 day intervals. Root cul- tures were cultivated in the dark at 258 in 500 ml Murashige and Skoog [18] medium was modi®ed Erlenmeyer ¯asks in an orbital shaker (100 r.p.m.; ium was supplemented with the following com- Transformation of 1c calli by A. rhizogenes A4 ponents (mg lÀ1): thiamine HCl (0.2), nicotinic acid was carried out using the standard feeder layer (0.5), pyridoxine HCl (0.5), meso-inositol (100), technique [19]. Cell suspension culture 1c, contain- peptone (100), sucrose (25000) and agar (6000) ing 2±3 g of cells in 30 ml W4CPA medium, was used as a feeder culture. Cells and bacteria were co-culti- with 0.4 mg lÀ1 4-chloro-phenoxyacetic acid (4- vated in W4CPA liquid medium for 7 days at188.
medium supplemented with 1.0 mg lÀ1 indole-3- with the addition of 250 mg lÀ1 cefotaxim. Growth of primary tumours was observed after 5±6 weeks Reagents were purchased from Sigma Chemical and root formation after 14 weeks from the begin- Co (MO, USA) and Serva Feinbiochemica GmbH ning of the experiment. Establishment of the root culture (designated as 1c-A4) was made as described The callus culture 1c was established in 1988 from the stem of a two-month old plant of Panax ginseng var. Mimaki C.A. Meyer [11]. Culture 1c Growth measurements were made on cultures was deposited at the Russian Collection of Plant which had been cultivated without kanamycin at ginsenosides [12]. Cultivation conditions, growth and hormonal requirements were as described previously [12±13]. The culture possessed cytokinin autonomy, and during the period of observation (more than 7 years), did not show any rhizogenic DNA was isolated from callus and root tissues Bendich [20]. Ten micrograms of DNA digested with EcoRI/HindIII was separated by agarose gel Calli of 1c culture (0.5 g) were transferred to the electrophoresis. DNA was blotted to a Zeta-Probe liquid W4CPA medium (10 ml) in Petri dishes and (Bio-Rad) membrane and hybridized with probe cultured at 248 in the dark on a rotary shaker. A suspension of A. tumefaciens GV3101 cells diluted pPCV002-CaMVC) which had been labeled with P-dATP using the Prime-a-Gene Labeling System day-old ginseng cell culture. After 2 days, cefotax- (Promega) according to Southern [17]. EcoRI ime was added to a ®nal concentration of 500 mg 2225 bp fragmentof pPCV002-CaMVBT was used lÀ1. After a 5 day interval, the cells were transferred as probe DNA for 1c-rolB culture. EcoRI 4480 bp to a fresh W4CPA medium supplemented with fragmentof pPCV002-ABC [1] was used as a probe 250 mg lÀ1 cefotaxim and 100 mg lÀ1 kanamycin sul- Neomycin phosphotransferase II (NPT) assay 6. Ko, K. S., Ebizuka, Y., Noguchi, H. and Enzyme activity was assayed following the proto- Sankawa, U., Chem. Pharm. Bull., 1988, 36, col developed by Reiss et al. [21]. The assay was 7. Mano, Y., Ohkawa, H. and Yamada, Y., Plant Eppendorf centrifuge tube with 40 ml 20 mM Tris- 8. Yoshikawa, T. and Furuya, T., Plant Cell NaCl, 15 mM dithiotreitol, and 2 mM phenylmethyl 9. Thron, U., Maresch, L., Beiderbeck, R. and sulphonil ¯uoride, pH 7. The homogenate was Reichling, J., Z. Naturforsch., 1989, 44c, 573.
cleared by centrifugation. Samples were fractionated 10. Constabel, C. P. and Towers, G. H. N., J.
on 10% polyacrylamide gel in non denaturating conditions. The position of enzymatically active NPT II proteins in the gel was determined by in Kozyrenko, M. M., Babkina, E. N., Uvarova, situ phosphorylation of kanamycin using [g32-P] N. I. and Makhankov, V. V., Rastit. Resurs.
(Plant Resources, Rus.), 1991, 27, 94.
12. Bulgakov, V. P., Zhuravlev, Y. N., Kozyrenko, Agropine and mannopine synthesized in the hairy roots were extracted and analyzed by high voltage Russia, A 01 H 4/00, C 12 N 5/00, 1993.
paper electrophoresis as described by Yoshikawa 13. Khodakovskaya, M. V., Bulgakov, V. P. and (Biotechnology, Rus.), 1995, 11, 40.
Isolation and determination of ginsenosides were Uvarova, N. I. and Eliakov, G. B., Khim. Prir.
Soedin (Chem. Nat. Prod., Rus.), 1993, 60, 237.
described [22, 23]. The lyophilized tissues were 15. Zhuravlev, Yu. N., Bulgakov, V. P., Moroz, L.
extracted with H2O±MeOH. The extract was evap- orated to dryness and sequentially extracted with Uvarova, N. I. and Eliakov, G. B., Doklady pentane and with butanol with H2O. The butanol residue was dissolved in MeOH (10 mg mlÀ1) and 16. Koncz, C. and Schell, J., Mol. Gen. Genet., 64 Â 2 mm; eluant: MeCN±H2O (gradientfrom 1:4 to 3:2 at a ¯ow rate of 100 ml minÀ1; detection: 17. Maniatis, T., Frisch, E. F. and Sambrook, J., Molecular cloning. A Laboratory Manual. Cold AcknowledgementsÐThis work was supported by the State Scienti®c and Technical Program of 18. Murashige, T. and Skoog, F., Physiol. Plant., Russia `Advanced Methods of Bioengineering' (`Gene and Cell Engineering' branch). We would also like to thank Dr. Angelo Spena for providing 19. Draper, J., Scott, R., Armitage, P. and Walden, R., Plant Genetic Transformation and Gene Expression, A Laboratory Manual. Blackwell 20. Rogers, S. O. and Bendich, A. J., Plant. Mol.
1. Spena, A., SchmuÈlling, T., Koncz, C. and 2. Schell, J., Koncz, C., Spena, A., Palme, K. and 21. Reiss, B., Sprengel, R., Will, H. and Schaller, 3. Bulgakov, V. P. and Zhuravlev, Y. N., Uspechi Sovr. Biol. (Adv. Mod. Biol., Rus), 1992, 112, Denisenko, V. A. and Isakov, V. V., Khim.
4. Ahn, J. C., Hwang, B., Tada, H., Ishimaru, K., Prir. Soedin. (Chem. Nat. Prod., Rus.), 1990, 23. Elkin, Y. N., Makhankov, V. V., Uvarova, N.
I., Bondarenko, P. V., Zubarev, R. A. and Knysh, A. N., Acta Pharmacol. Sinica, 1993,

Source: http://www.halt.ru/~ibss/gening/articles/1998-phytochemistry.pdf