Plant Cell Reports
Plant Cell Reports (1990) 9:181-184
Triterpenoid biosynthesis in tissue cultures of Glycyrrhiza glabra var. glandulifera
9 Springer-Verlag1990
S. Ayabe 1, 2, H. Takano 2, To Fujita 2, 4, T. Furuya 2, H. Hirota 3, and T. Takahashi 1 1 Department of Applied Biological Science, College of Agriculture and Veterinary Medicine, Nihon University, Fujisawa-shi, Kanagawa 252, Japan 2 School of Pharmaceutical Sciences, Kitasato University, Minato-ku, Tokyo 108, Japan 3 Department of Chemistry, Faculty of Science, University of Tokyo, Bunkyo-ku, Tokyo 113, Japan 4 Present address: Technical Research Institute, Snow Brand Milk Products, Kawagoe, Saitama 350, Japan Received March 22, 1990/Revised version received June 11, 1990 - Communicated by M. Tabata
ABSTRACT The incorporation of [l-14C]acetate and [214C]mevalonate into free and esterified triterpen-3ols was examined in original plant organs and tissue cultures of Glycyrrhiza ~labra var. glandulifera. Both substrates labeled B-amyrin, an oIeanane-type triterpene, and cycloartenol and 24-methylenecycloartanol, both of which are intermediates of phytosterol biosynthesis. The label in esterified triterpenes was distributed mainly in phytosterol intermediates, but not in ~-amyrin. The ratio of Bamyrin formation among the three triterpenes from [2-14C]mevalonate was relatively high in stolon segments and in root cultures, but negligible in callus cultures. Administration of a specific inhibitor of squalene-2,3-epoxide:cycloartenol (lanosterol) cyclase caused a marked increase of B-amyrin synthesis in root suspension cultures, and of 24methylenecycloartanol synthesis in cell suspension cultures, from [2-14C]mevalonate.
ABBREVIATIONS: GL, glycyrrhizin; GLA, glycyrrhetic acid; MVA, mevalonic acid; CCI, squalene-2,3-epoxide:cycloartenol cyclase inhibitor. INTRODUCTION Glycyrrhizin (GL) is a diglucuronide of a triterpenoid acid, glycyrrhetic acid (GLA), and is the main sweet principle of a crude drug, licorice. Several pharmacological/physiological activities of GL and GLA are known. The underground part (roots and stolons) of several Glycyrrhiza species is the major site of GL accumulation. Although several reports on tissue cultures of Glycyrrhiza have been published, no clear evidence of GL production by tissue culture systems has been available until recently (see Hayashi et al. 1988). GLA should be derived from a simple oleanane, Bamyrin, whose synthesis is catalyzed by an enzyme squalene-2,3-epoxide:B-amyrin cyclase. Squalene2,3-epoxide also affords other classes of triterpenols; e.g. cycloartenol which is an intermediate of phytosterol biosynthesis (Fig. i). Therefore, the transformation of squalene-2,3-epoxide by specific cyclases can be one of the regulatory points in the This paper is Part 72 in the series "Studies on Plant Tissue Cultures" from the laboratory at Kitasato University. For Part 71, see Furuya T, Koge K, Orihara Y (submitted for publication). Offprint requests to." S. Ayabe at Nihon University
biosynthesis of GL/GLA, phytosterols and other triterpenoids. Herein we describe the studies on triterpenoid biosynthesis in tissue cultures of G_:_.glabra var. glandulifera, whose stolons actually accumulate GL. We have examined the biosynthetic ability of oleanane- and lanostane-type triterpenes and their esters from radio-labeled precursors in morphologically different tissue cultures. The use of an inhibitor of specific cyclases with tissue cultures was also explored.
MATERIALS A~DMETflODS Tissue cultures: G. glabra var. glandulifera plants have been cultivated in the Pharmaceutical Plant Garden, Kitasato University. The tissue cultures were derived from sterilized leaf and petiole explants on Murashige and Skoog's (MS) agar media supplemented with 5 ppm IAA (I5 medium) or 1 ppm 2,4-D and 0.I ppm kinetin (D] KO.I medium). The cultures were maintained in the dark or under the 12 h illumination (4,000 lux) at 25~ From the callus tissue emerged from the explants on D1 KO.I medium, fragile yellow callus cultures under dark (the same medium) and hard green callus cultures under light (D0.1 K1 medium) have been established. On the medium with 10 ppm NAA and 0.2 ppm kinetin (NI0 KO.2 medium), the callus occasionally formed thick (ca. 1 mm) root tissue, and repeated selection yielded stable root + callus (ca. i0 %) cultures. Separately, adventitious roots appeared on the surface of the explants on 15 medium were transferred into a liquid medium of the same composition, and cultured in the dark. After ca. 1 month, fine roots (ca. 0.2 nun thick) in the medium were collected and placed onto the agar medium to establish fine root cultures. The established tissue cultures have been subcultured onto freshly prepared media every 4 weeks. Extraction an__~dseparation procedures: Plantmaterials (ca. 1 gFw each) were extracted with MeOH and successively with MeOH/CHCI 3 (I:I) under reflux for ca. 2 h. Combined extracts were evaporated in vacuo and participated between CHCI 3 and H20. The residues from MeOH/CHCI 3 extractions were further extracted with 50 % EtOH. The aqueous layers of CHCI3/H20 participation and 50 % EtOH extracts were combined, evaporated to a small volume, and extracted with n-BuOH saturated with HoO_ The aqueous layers were submitted to a Diaion H~-20 column and
182 eluted firstly with H20 and then with MeOH. With this separation method, GLA and GL were fractionated into CHCI$ layer and MeOH eluent of HP-20 column respectively. Feeding experiment s with [2-14C]MVA and sodium [lz~C]acetate: Two- to three-week-old tissue cultures or plant segments (ca. 1 g each) were transferred into a 30 ml vial containing ca. 2 ~Ci of [2-Z4C]MVA (55 mCi/mmol) or [l-14C]acetate (55 mCi/mmol) in 0.2 ml H20. After incubation in the dark or under 12 h light at 25~ for 48 h, the tissues were extracted and fractionated as above. The CHCI 3 layers were applied to silica gel TLC (0.25 mm thick, Merck precoated plates) and developed with toluene-EtOAc (4:1), and radioactive areas of non-polar lipids and triterpen-3-ols were scraped off. The recovered non-polar lipids were hydrolyzed with 4 % KOH/EtOH (w/v) and resultant triterpen-3-ols were collected by TLC ~s above. The triterpen-3-ols (either free or recovered from non-polar lipids) were acetylated with Ac20/pyridine (2:1), and acetates were further developed on silica gel/AgNO 3 plates (Baisted 1971) with toluene/n-hexane (2:3). The plates were examined by autoradiography (Hyperfilm B-max (Amersham); exposure time, 2 days), and radioactivity was measured with a liquid scintillation counter after separation of the spots. Administration of squalene-2~3-epoxide:cycloartenol cyclase inhibitor (CCI) t__oosuspension cultures: (• 5~,8aB~Dimethyl-2-(l,5,9-trimethyldecyl)-4a~H-decahydroisoquinolin-6~-ol, which was synthesized according to Taton et al. (1986), was used as CCI. An ethanolic solution of CCI (3 mg/6.25 ml) was added to I0- to 14-day-old cell- and root-suspension cultures (125 ml), and incubated for 2 h. [2-14C]MVA (5 ~Ci) in 5 ml medium was added and further incubated for the period indicated in Results. The tissues collected on Nylon cloth were washed thoroughly with distilled water, then extracted and analyzed for radioactivity in triterpenols and their esters as above.
Acetyl-CoA Mevalonic acid---~ ~
Squalene-2,3-epoxide
HO
H Cycloartenol
1
c00H
R=H: Glycyrrhetic a c i d
24-Methylenecycloartanol
R=(GlucA)2: Glycyrrhizin Phytosterols Fig. l-sterols.
Biosynthesis of glycyrrhizin and phyto-
RESULTS Absence of ~lycyrrhizin in tissue cultures No tissue cultures of G!ycyrrhiza glabra var. glandulifera exhibited the characteristic sweet taste of GL. Indeed, the MeOH eluents from HP-20 column (see Materials and Methods) prepared from the cultures failed to show the peak of GL in HPLC analysis. Incorporation o f 14C-labeled acetate and MVA into triterpen-3-ols and tri~erpene esters The TLC of the CHCI 3 extracts from the plant segments and tissue cultures which had been incubated with radio-labeled isoprenoid precursors showed several radioactive materials, among which were triterpen-3-ois, sterols and non-polar lipidso The non-polar lipids afforded on alkaline hydrolysis mainly radioactive triterpen-3-ols in addition to a small amount of sterols. No clear radioactive spot was seen around GLA area. Fig. 2 shows the autoradiograms of the radioactive triterpen-3-ols (either free or recovered from triterpene esters) separated on AgNO3-impreghated TLC after acetylation. The radioactivity in free triterpen-3-ols biosynthesized from [2-14C]MVA distributed in ~-amyrin, eycloartenol and 24-methylene-cycloartanol in addition to minor unidentified materials. In contrast, among the triterpen-3-ols
Fig. 2. TLC-Autoradiograms of free triterpenols and trite~ppene esters biosynthesized from [2-Z~C]MVA in G__n. ~labra var. ~landulifera plant segments and tissue cultures. The acetates of labeled triterpenol samples (either free or recovered from esters) were subjected to SiO2/AgNO 3 TLC. A, $-amyrin; C, cycloartenol; M, 24-methylenecycloartanol.
183 Radioactivity in CHCIa layer
Organ examined
d~m)
(xlO s
A
Stolon
2.3
Shoot
10.8
Root culture (15)
B
Distribution of radioactivity in CHCI 3 layer (%) 2 4 6 0 2 4 6
0
triterpenol
Triterpene ester
6.4
Root+callus (NIO KO.2)
15.1
Green callus (DO.I KI)
10.7
Callus (DI KO.I)
22.1
Stolon
1.2
Shoot
2.4
Root culture (I5)
3.1
Green callus (DO.I KI)
2.5
m |
8
[]~ Free triterpenol
[], B-Amyrin; ~ ,
IN []
Cycloartenol; ~ ,
Triterpene ester
24-Methylenecycloartanol
Fig. 3_t. Distribution of radioactivity in 8-amyrin, cycloartenol and 24-methylenecycloartanol and their esters biosynthesized from (A) [2-14C]mevalonate (2 ~Ci) and (B) [if14C]acetate (2 LLCi) in Ge_. glabra var. glandulifera plant segments and tissue cultures. recovered from triterpene esters, cycloartenol and 24-methylenecycloartanol were highly radioactive, whereas 8-amyrin showed only a negligible radioactivity. These radioactive spots were separated from TLC and dilution analysis with standard samples of triterpene acetates was carried out. The result (Table I) indicates constant specific radioactivity on repeated recrystallization. Thus, the radioactivity in the spots was predominantly, if not totally, due to the labeling of these triterpen-3-ols. The distribution of radioactivity among three triterpen-3-ols and their esters in the CHCI 3 extracts of the tissues examined is summarized in Fig. 3. Callus and root + callus cultures effectively converted MVA into CHCl3-soluble materials, while in stolon segments and root cultures the conversion ratio was low. Acetate was less effective precursor for triterpenoid synthesis than MVA. As expected from the autoradiography examination, a majority of B-amyrin newly synthesized from MVA exists as free form but not as esterified form. Noteworthily, the distribution of radioactivity in $-amyrin originated from MVA is higher in stolon and root cultures than in callus or root + callus cultures. In contrast, in shoot and callus cultures, more effective formation of the intermediates of sterol biosynthesis is observed. ZuC-Acetate, even in spite of its low Table i. Dilution analysis of [1~C]-8-amyrin, ~y~loartenol and [14C]-24-methylenecycloartanol isolated from tissue cultures of G. glabra var. glandulifera incubated with [2-14C]me--valonic acid. ~ B-Amyrin acetate Number of reSpec. act. crystal. (nCi/mmol) 1 2 3 4 5
8.26 8.33 8.01 8.09
Cycloartenol acetate
24-Methylenecycloartanol acetate
Spec. act. (>Ci/mmol)
Spec. act. (~Ci/mmol)
0.107 0.103 0.103 0.104
0.185 0.169 0.161 0.162 0.163
* Each triterpenol was acetylated, diluted with corresponding acetate of a standard sample, and repeatedly recrystallized.
incorporation, prominently labeled the oleanane in stolon and shoot of the original plant compared to root and callus cultures, Effect of cycloartenol cyclase the biosynthesis of oleananetriterpenols
inhibitor (CCI) o__nn and lanostane-type
5~,8aa-Dimethyl-2-(l,5,9-trimethyldecyl)-4a~Hdecahydroisoquinolin-6B-ol, which mimics the high energy intermediate in the enzymatic cyclization of squalene-2,3-epoxide into lanostane skeleton, has been demonstrated to inhibit the formation of cycloartenol, but not of ~-amyrin, from l~C-labeled acetate in bramble cells (Taton et al. 1986). This cycloartenol (and lanosterol) cyclase inhibitor (CCI) was expected to alter the triterpenoid biosynthesis in G__~.glabra var. glandulifer a tissue cultures. The results shown in Fig. 4 indicate a characteristic difference in the effect of CCI on free triterpenol formation among the suspension cultures examined. In root suspension cultures (I5 medium), 0.06 mM CCI increased the ratio of ~-amyrin formation from [2-14C]MVA and decreased the ratio of cycloartenol and 24-methylenecycloartanol formation (Fig. 4A). The effect was most clear at 24 h incubation, but declined somewhat in 1 week. In the cultures fed with CCI the incorporation of radioactivity into CHCI$ layer was higher (up to 1.5 fold at 1 week culture) than in control cultures; actual formation of B-amyrin is ca. 2 fold in 24 h incubation. The autoradiograms of CHCI 3 layers of the root cultures with or without CCI exhibited almost the same pattern, except for appearance of a spot near triterpen-3-ols in CCI-treated cultures (data not shown). In cell suspension cultures, in contrast, no increase in the ratio of ~-amyrin synthesis after the addition of CCI was observed (Fig. 4B). However, characteristically, the ratio of 24-methylenecycloartanol formation increased. The ratio of triterpene ester formation from MVA in the presence of CCI was lower than that in the absence of CCI throughout the experiments conducted. However, because the incorporation of MVA into the CHCI 3 extracts was always higher in CCItreated cultures, actual formation of triterpene esters was not largely affected by CCI.
184 Time
A
4h
24 h
Additive
Radioactivity in CHCI 3 layer (xlO s dpm)
None
1.8
CCI
1.9
None
0
Distribution of radioactivity in CHCI 3 layer (%) 4 8 12 16 20
Free triterpenol
16.6
CCI
19.1
None
ii.0
CCI
15.6
1 week
B
None
1.3
CCI
4.7
Free triterpenol
4h None
8.2
24 h CCI
22.8 N,
~-Amyrin; [], Cycloartenol; [], 24-Methylenecycloartanol
F i ~ ~__u. Effect of cycloartenol cyclase inhibitor (CCI) on biosynthesis of ~-amyrin, eycloartenol and 24methylenecycloartanol and their esters from [2-14C]MVA (5 pCi) in (A) root suspension and (B) cell suspension cultures of Glycyrrhiza glabra var. glandulifera.
DISCUSSION Recently Hayashi et al. (1988) have reported that cell cultures of G_~.~labra fail to produce GL, although they contain several triterpenes including B-amyrin. Similarly, Wu et al. (1974) and Henry et al. (1984) have reported the absence of GL in G__~. glabra, var. typica cell cultures. In contrast, two Japanese patents (Tamaki et al. 1975; Fujita et al. 1978) have briefly described the production of GL by tissue cultures of Glycyrrhiza spp. However, those patents lacked the details of the culture conditions and analytical procedures, and thus unambiguous evidence of GL production has not been available. Ko et al. (1989) have reported the production of GL in transformed hairy root cultures of G. uralensis, but no production by transformed roots of the same plant species has also been described (8aito et al. 1990). In our callus and root cultures, the production of GL could not be observed. Nevertheless, the incorporation experiments with Z4C-labeled acetate and MVA indicate the metabolic alteration in different organs and tissue cultures of G_:_.~labra var. glandulifera. The callus cultures convert MVA mainly into the sterol pathway, while stolons of the original plant and root cultures do have the pathway to oleanane-type triterpenes. This implies that cycloartenol cyclase is expressed ubiquitously, while ~-amyrin cyclase is expressed in organ-specific manner. However, in our G__~.glabra var. glandulifera root cultures, the biosynthetic pathway from B-amyrin to GL/GLA seems to be blocked. In GLproducing hairy root cultures of G. uralefisis, this pathway expresses (Ko et al. 1989). The discrepancy may be explained by the difference of plant species, additional factors required for GL synthesis other than root organ differentiation, or the special features of transformed cells. In undifferentiated Callus cells of G_m _. glabra vat. glandulifera, however, $-amyrin accumulates (unpublished results). Our experiments were conducted with 2- to 3-week-old cultures (later logarithmic or stationary phase), and it is possible that at some stage of growth cycle, B-amyrin cyclase activity may he expressed even in undifferentiated cells. In peas, B-amyrin synthesis is implicated to be developmentally controlled (Baisted 1971; Fang and Baisted 1975). The effects of CCI were characteristically
different within the tissues examined (Fig. 4). The results with root suspension cultures were in agreement with the anticipated inhibitory property of the decahydroisoqninolinol derivative as CCI, and indicate that the syntheses of lanostane- and oleananetype triterpenols are competitive in the root tissues. In cell suspension cultures, in contrast, a marked increase in 24-methylenecycloartanol formation was the major effect of CCI. While this could be explained by the inhibition of enzymes involved in the later steps of sterol biosynthesis (e.g. cycloeucalenol-obtusifoliol isomerase; Taton et al. 1987), the reason of differential effects on the morphologically different recipient tissues is unclear. Another interesting aspect found here is the effective formation of esters of cycloartenol and 24-methylenecycloartanol in both tissue cultures and original plants, while the synthesis of B-amyrin ester was negligible. These results are in agreement with the differential pathway of free and esterified triterpenes found in Euphorbia latex (Nemethy et al. 1983) and in Ilex leaves (Niemann
1985). REFERENCES Baisted DJ (1971) Biochem J 124: 375-383. Fang T-Y, Baisted DJ (1975) Biochem J 150: 323-328. Fujita Y, Teranishi K, Furukawa T (1978) Japanese Patent (52-91188). Hayashi H, Fukui H, Tabata M (1988) Plant Cell Rep 7: 508-511. Henry M, Edy A-M, Marty B (1984) C R Acad Sc Paris Serie III 299: 899-903. Ko KS, Noguchi H, Ebizuka Y, Sankawa U (1989) Chem Pharm Bull 37: 245-248. Nemethy EK, Skrukrud C, Piazza GJ, Calvin M (1983) Biochim Biophys Acta 760: 343-349. Niemann GJ (1985) Plants 166: 51-56. Saito K, Kaneko, H, Yamazaki M, Yoshida M, Murakoshi I (1990) Plant Cell Rep 8: 718-721. Tamaki E, Morishita I, Nishida K, Kato K, Matsumoto T (1975) Japanese Patent (50-16440). Taton M, Benveniste P, Rahier A (1986) Bioehem Biophys Res Commn 138: 764-770. Taton M, Benveniste P, Rahier A (1987) Phytochemistry 26: 385-392. Wu C-H, Zabawa EM, Townsley PM (1974) J Inst Can Sci Technol Aliment 7: 105-107.
Plant Cell Reports (1990) 9:181-184
Triterpenoid biosynthesis in tissue cultures of Glycyrrhiza glabra var. glandulifera
9 Springer-Verlag1990
S. Ayabe 1, 2, H. Takano 2, To Fujita 2, 4, T. Furuya 2, H. Hirota 3, and T. Takahashi 1 1 Department of Applied Biological Science, College of Agriculture and Veterinary Medicine, Nihon University, Fujisawa-shi, Kanagawa 252, Japan 2 School of Pharmaceutical Sciences, Kitasato University, Minato-ku, Tokyo 108, Japan 3 Department of Chemistry, Faculty of Science, University of Tokyo, Bunkyo-ku, Tokyo 113, Japan 4 Present address: Technical Research Institute, Snow Brand Milk Products, Kawagoe, Saitama 350, Japan Received March 22, 1990/Revised version received June 11, 1990 - Communicated by M. Tabata
ABSTRACT The incorporation of [l-14C]acetate and [214C]mevalonate into free and esterified triterpen-3ols was examined in original plant organs and tissue cultures of Glycyrrhiza ~labra var. glandulifera. Both substrates labeled B-amyrin, an oIeanane-type triterpene, and cycloartenol and 24-methylenecycloartanol, both of which are intermediates of phytosterol biosynthesis. The label in esterified triterpenes was distributed mainly in phytosterol intermediates, but not in ~-amyrin. The ratio of Bamyrin formation among the three triterpenes from [2-14C]mevalonate was relatively high in stolon segments and in root cultures, but negligible in callus cultures. Administration of a specific inhibitor of squalene-2,3-epoxide:cycloartenol (lanosterol) cyclase caused a marked increase of B-amyrin synthesis in root suspension cultures, and of 24methylenecycloartanol synthesis in cell suspension cultures, from [2-14C]mevalonate.
ABBREVIATIONS: GL, glycyrrhizin; GLA, glycyrrhetic acid; MVA, mevalonic acid; CCI, squalene-2,3-epoxide:cycloartenol cyclase inhibitor. INTRODUCTION Glycyrrhizin (GL) is a diglucuronide of a triterpenoid acid, glycyrrhetic acid (GLA), and is the main sweet principle of a crude drug, licorice. Several pharmacological/physiological activities of GL and GLA are known. The underground part (roots and stolons) of several Glycyrrhiza species is the major site of GL accumulation. Although several reports on tissue cultures of Glycyrrhiza have been published, no clear evidence of GL production by tissue culture systems has been available until recently (see Hayashi et al. 1988). GLA should be derived from a simple oleanane, Bamyrin, whose synthesis is catalyzed by an enzyme squalene-2,3-epoxide:B-amyrin cyclase. Squalene2,3-epoxide also affords other classes of triterpenols; e.g. cycloartenol which is an intermediate of phytosterol biosynthesis (Fig. i). Therefore, the transformation of squalene-2,3-epoxide by specific cyclases can be one of the regulatory points in the This paper is Part 72 in the series "Studies on Plant Tissue Cultures" from the laboratory at Kitasato University. For Part 71, see Furuya T, Koge K, Orihara Y (submitted for publication). Offprint requests to." S. Ayabe at Nihon University
biosynthesis of GL/GLA, phytosterols and other triterpenoids. Herein we describe the studies on triterpenoid biosynthesis in tissue cultures of G_:_.glabra var. glandulifera, whose stolons actually accumulate GL. We have examined the biosynthetic ability of oleanane- and lanostane-type triterpenes and their esters from radio-labeled precursors in morphologically different tissue cultures. The use of an inhibitor of specific cyclases with tissue cultures was also explored.
MATERIALS A~DMETflODS Tissue cultures: G. glabra var. glandulifera plants have been cultivated in the Pharmaceutical Plant Garden, Kitasato University. The tissue cultures were derived from sterilized leaf and petiole explants on Murashige and Skoog's (MS) agar media supplemented with 5 ppm IAA (I5 medium) or 1 ppm 2,4-D and 0.I ppm kinetin (D] KO.I medium). The cultures were maintained in the dark or under the 12 h illumination (4,000 lux) at 25~ From the callus tissue emerged from the explants on D1 KO.I medium, fragile yellow callus cultures under dark (the same medium) and hard green callus cultures under light (D0.1 K1 medium) have been established. On the medium with 10 ppm NAA and 0.2 ppm kinetin (NI0 KO.2 medium), the callus occasionally formed thick (ca. 1 mm) root tissue, and repeated selection yielded stable root + callus (ca. i0 %) cultures. Separately, adventitious roots appeared on the surface of the explants on 15 medium were transferred into a liquid medium of the same composition, and cultured in the dark. After ca. 1 month, fine roots (ca. 0.2 nun thick) in the medium were collected and placed onto the agar medium to establish fine root cultures. The established tissue cultures have been subcultured onto freshly prepared media every 4 weeks. Extraction an__~dseparation procedures: Plantmaterials (ca. 1 gFw each) were extracted with MeOH and successively with MeOH/CHCI 3 (I:I) under reflux for ca. 2 h. Combined extracts were evaporated in vacuo and participated between CHCI 3 and H20. The residues from MeOH/CHCI 3 extractions were further extracted with 50 % EtOH. The aqueous layers of CHCI3/H20 participation and 50 % EtOH extracts were combined, evaporated to a small volume, and extracted with n-BuOH saturated with HoO_ The aqueous layers were submitted to a Diaion H~-20 column and
182 eluted firstly with H20 and then with MeOH. With this separation method, GLA and GL were fractionated into CHCI$ layer and MeOH eluent of HP-20 column respectively. Feeding experiment s with [2-14C]MVA and sodium [lz~C]acetate: Two- to three-week-old tissue cultures or plant segments (ca. 1 g each) were transferred into a 30 ml vial containing ca. 2 ~Ci of [2-Z4C]MVA (55 mCi/mmol) or [l-14C]acetate (55 mCi/mmol) in 0.2 ml H20. After incubation in the dark or under 12 h light at 25~ for 48 h, the tissues were extracted and fractionated as above. The CHCI 3 layers were applied to silica gel TLC (0.25 mm thick, Merck precoated plates) and developed with toluene-EtOAc (4:1), and radioactive areas of non-polar lipids and triterpen-3-ols were scraped off. The recovered non-polar lipids were hydrolyzed with 4 % KOH/EtOH (w/v) and resultant triterpen-3-ols were collected by TLC ~s above. The triterpen-3-ols (either free or recovered from non-polar lipids) were acetylated with Ac20/pyridine (2:1), and acetates were further developed on silica gel/AgNO 3 plates (Baisted 1971) with toluene/n-hexane (2:3). The plates were examined by autoradiography (Hyperfilm B-max (Amersham); exposure time, 2 days), and radioactivity was measured with a liquid scintillation counter after separation of the spots. Administration of squalene-2~3-epoxide:cycloartenol cyclase inhibitor (CCI) t__oosuspension cultures: (• 5~,8aB~Dimethyl-2-(l,5,9-trimethyldecyl)-4a~H-decahydroisoquinolin-6~-ol, which was synthesized according to Taton et al. (1986), was used as CCI. An ethanolic solution of CCI (3 mg/6.25 ml) was added to I0- to 14-day-old cell- and root-suspension cultures (125 ml), and incubated for 2 h. [2-14C]MVA (5 ~Ci) in 5 ml medium was added and further incubated for the period indicated in Results. The tissues collected on Nylon cloth were washed thoroughly with distilled water, then extracted and analyzed for radioactivity in triterpenols and their esters as above.
Acetyl-CoA Mevalonic acid---~ ~
Squalene-2,3-epoxide
HO
H Cycloartenol
1
c00H
R=H: Glycyrrhetic a c i d
24-Methylenecycloartanol
R=(GlucA)2: Glycyrrhizin Phytosterols Fig. l-sterols.
Biosynthesis of glycyrrhizin and phyto-
RESULTS Absence of ~lycyrrhizin in tissue cultures No tissue cultures of G!ycyrrhiza glabra var. glandulifera exhibited the characteristic sweet taste of GL. Indeed, the MeOH eluents from HP-20 column (see Materials and Methods) prepared from the cultures failed to show the peak of GL in HPLC analysis. Incorporation o f 14C-labeled acetate and MVA into triterpen-3-ols and tri~erpene esters The TLC of the CHCI 3 extracts from the plant segments and tissue cultures which had been incubated with radio-labeled isoprenoid precursors showed several radioactive materials, among which were triterpen-3-ois, sterols and non-polar lipidso The non-polar lipids afforded on alkaline hydrolysis mainly radioactive triterpen-3-ols in addition to a small amount of sterols. No clear radioactive spot was seen around GLA area. Fig. 2 shows the autoradiograms of the radioactive triterpen-3-ols (either free or recovered from triterpene esters) separated on AgNO3-impreghated TLC after acetylation. The radioactivity in free triterpen-3-ols biosynthesized from [2-14C]MVA distributed in ~-amyrin, eycloartenol and 24-methylene-cycloartanol in addition to minor unidentified materials. In contrast, among the triterpen-3-ols
Fig. 2. TLC-Autoradiograms of free triterpenols and trite~ppene esters biosynthesized from [2-Z~C]MVA in G__n. ~labra var. ~landulifera plant segments and tissue cultures. The acetates of labeled triterpenol samples (either free or recovered from esters) were subjected to SiO2/AgNO 3 TLC. A, $-amyrin; C, cycloartenol; M, 24-methylenecycloartanol.
183 Radioactivity in CHCIa layer
Organ examined
d~m)
(xlO s
A
Stolon
2.3
Shoot
10.8
Root culture (15)
B
Distribution of radioactivity in CHCI 3 layer (%) 2 4 6 0 2 4 6
0
triterpenol
Triterpene ester
6.4
Root+callus (NIO KO.2)
15.1
Green callus (DO.I KI)
10.7
Callus (DI KO.I)
22.1
Stolon
1.2
Shoot
2.4
Root culture (I5)
3.1
Green callus (DO.I KI)
2.5
m |
8
[]~ Free triterpenol
[], B-Amyrin; ~ ,
IN []
Cycloartenol; ~ ,
Triterpene ester
24-Methylenecycloartanol
Fig. 3_t. Distribution of radioactivity in 8-amyrin, cycloartenol and 24-methylenecycloartanol and their esters biosynthesized from (A) [2-14C]mevalonate (2 ~Ci) and (B) [if14C]acetate (2 LLCi) in Ge_. glabra var. glandulifera plant segments and tissue cultures. recovered from triterpene esters, cycloartenol and 24-methylenecycloartanol were highly radioactive, whereas 8-amyrin showed only a negligible radioactivity. These radioactive spots were separated from TLC and dilution analysis with standard samples of triterpene acetates was carried out. The result (Table I) indicates constant specific radioactivity on repeated recrystallization. Thus, the radioactivity in the spots was predominantly, if not totally, due to the labeling of these triterpen-3-ols. The distribution of radioactivity among three triterpen-3-ols and their esters in the CHCI 3 extracts of the tissues examined is summarized in Fig. 3. Callus and root + callus cultures effectively converted MVA into CHCl3-soluble materials, while in stolon segments and root cultures the conversion ratio was low. Acetate was less effective precursor for triterpenoid synthesis than MVA. As expected from the autoradiography examination, a majority of B-amyrin newly synthesized from MVA exists as free form but not as esterified form. Noteworthily, the distribution of radioactivity in $-amyrin originated from MVA is higher in stolon and root cultures than in callus or root + callus cultures. In contrast, in shoot and callus cultures, more effective formation of the intermediates of sterol biosynthesis is observed. ZuC-Acetate, even in spite of its low Table i. Dilution analysis of [1~C]-8-amyrin, ~y~loartenol and [14C]-24-methylenecycloartanol isolated from tissue cultures of G. glabra var. glandulifera incubated with [2-14C]me--valonic acid. ~ B-Amyrin acetate Number of reSpec. act. crystal. (nCi/mmol) 1 2 3 4 5
8.26 8.33 8.01 8.09
Cycloartenol acetate
24-Methylenecycloartanol acetate
Spec. act. (>Ci/mmol)
Spec. act. (~Ci/mmol)
0.107 0.103 0.103 0.104
0.185 0.169 0.161 0.162 0.163
* Each triterpenol was acetylated, diluted with corresponding acetate of a standard sample, and repeatedly recrystallized.
incorporation, prominently labeled the oleanane in stolon and shoot of the original plant compared to root and callus cultures, Effect of cycloartenol cyclase the biosynthesis of oleananetriterpenols
inhibitor (CCI) o__nn and lanostane-type
5~,8aa-Dimethyl-2-(l,5,9-trimethyldecyl)-4a~Hdecahydroisoquinolin-6B-ol, which mimics the high energy intermediate in the enzymatic cyclization of squalene-2,3-epoxide into lanostane skeleton, has been demonstrated to inhibit the formation of cycloartenol, but not of ~-amyrin, from l~C-labeled acetate in bramble cells (Taton et al. 1986). This cycloartenol (and lanosterol) cyclase inhibitor (CCI) was expected to alter the triterpenoid biosynthesis in G__~.glabra var. glandulifer a tissue cultures. The results shown in Fig. 4 indicate a characteristic difference in the effect of CCI on free triterpenol formation among the suspension cultures examined. In root suspension cultures (I5 medium), 0.06 mM CCI increased the ratio of ~-amyrin formation from [2-14C]MVA and decreased the ratio of cycloartenol and 24-methylenecycloartanol formation (Fig. 4A). The effect was most clear at 24 h incubation, but declined somewhat in 1 week. In the cultures fed with CCI the incorporation of radioactivity into CHCI$ layer was higher (up to 1.5 fold at 1 week culture) than in control cultures; actual formation of B-amyrin is ca. 2 fold in 24 h incubation. The autoradiograms of CHCI 3 layers of the root cultures with or without CCI exhibited almost the same pattern, except for appearance of a spot near triterpen-3-ols in CCI-treated cultures (data not shown). In cell suspension cultures, in contrast, no increase in the ratio of ~-amyrin synthesis after the addition of CCI was observed (Fig. 4B). However, characteristically, the ratio of 24-methylenecycloartanol formation increased. The ratio of triterpene ester formation from MVA in the presence of CCI was lower than that in the absence of CCI throughout the experiments conducted. However, because the incorporation of MVA into the CHCI 3 extracts was always higher in CCItreated cultures, actual formation of triterpene esters was not largely affected by CCI.
184 Time
A
4h
24 h
Additive
Radioactivity in CHCI 3 layer (xlO s dpm)
None
1.8
CCI
1.9
None
0
Distribution of radioactivity in CHCI 3 layer (%) 4 8 12 16 20
Free triterpenol
16.6
CCI
19.1
None
ii.0
CCI
15.6
1 week
B
None
1.3
CCI
4.7
Free triterpenol
4h None
8.2
24 h CCI
22.8 N,
~-Amyrin; [], Cycloartenol; [], 24-Methylenecycloartanol
F i ~ ~__u. Effect of cycloartenol cyclase inhibitor (CCI) on biosynthesis of ~-amyrin, eycloartenol and 24methylenecycloartanol and their esters from [2-14C]MVA (5 pCi) in (A) root suspension and (B) cell suspension cultures of Glycyrrhiza glabra var. glandulifera.
DISCUSSION Recently Hayashi et al. (1988) have reported that cell cultures of G_~.~labra fail to produce GL, although they contain several triterpenes including B-amyrin. Similarly, Wu et al. (1974) and Henry et al. (1984) have reported the absence of GL in G__~. glabra, var. typica cell cultures. In contrast, two Japanese patents (Tamaki et al. 1975; Fujita et al. 1978) have briefly described the production of GL by tissue cultures of Glycyrrhiza spp. However, those patents lacked the details of the culture conditions and analytical procedures, and thus unambiguous evidence of GL production has not been available. Ko et al. (1989) have reported the production of GL in transformed hairy root cultures of G. uralensis, but no production by transformed roots of the same plant species has also been described (8aito et al. 1990). In our callus and root cultures, the production of GL could not be observed. Nevertheless, the incorporation experiments with Z4C-labeled acetate and MVA indicate the metabolic alteration in different organs and tissue cultures of G_:_.~labra var. glandulifera. The callus cultures convert MVA mainly into the sterol pathway, while stolons of the original plant and root cultures do have the pathway to oleanane-type triterpenes. This implies that cycloartenol cyclase is expressed ubiquitously, while ~-amyrin cyclase is expressed in organ-specific manner. However, in our G__~.glabra var. glandulifera root cultures, the biosynthetic pathway from B-amyrin to GL/GLA seems to be blocked. In GLproducing hairy root cultures of G. uralefisis, this pathway expresses (Ko et al. 1989). The discrepancy may be explained by the difference of plant species, additional factors required for GL synthesis other than root organ differentiation, or the special features of transformed cells. In undifferentiated Callus cells of G_m _. glabra vat. glandulifera, however, $-amyrin accumulates (unpublished results). Our experiments were conducted with 2- to 3-week-old cultures (later logarithmic or stationary phase), and it is possible that at some stage of growth cycle, B-amyrin cyclase activity may he expressed even in undifferentiated cells. In peas, B-amyrin synthesis is implicated to be developmentally controlled (Baisted 1971; Fang and Baisted 1975). The effects of CCI were characteristically
different within the tissues examined (Fig. 4). The results with root suspension cultures were in agreement with the anticipated inhibitory property of the decahydroisoqninolinol derivative as CCI, and indicate that the syntheses of lanostane- and oleananetype triterpenols are competitive in the root tissues. In cell suspension cultures, in contrast, a marked increase in 24-methylenecycloartanol formation was the major effect of CCI. While this could be explained by the inhibition of enzymes involved in the later steps of sterol biosynthesis (e.g. cycloeucalenol-obtusifoliol isomerase; Taton et al. 1987), the reason of differential effects on the morphologically different recipient tissues is unclear. Another interesting aspect found here is the effective formation of esters of cycloartenol and 24-methylenecycloartanol in both tissue cultures and original plants, while the synthesis of B-amyrin ester was negligible. These results are in agreement with the differential pathway of free and esterified triterpenes found in Euphorbia latex (Nemethy et al. 1983) and in Ilex leaves (Niemann
1985). REFERENCES Baisted DJ (1971) Biochem J 124: 375-383. Fang T-Y, Baisted DJ (1975) Biochem J 150: 323-328. Fujita Y, Teranishi K, Furukawa T (1978) Japanese Patent (52-91188). Hayashi H, Fukui H, Tabata M (1988) Plant Cell Rep 7: 508-511. Henry M, Edy A-M, Marty B (1984) C R Acad Sc Paris Serie III 299: 899-903. Ko KS, Noguchi H, Ebizuka Y, Sankawa U (1989) Chem Pharm Bull 37: 245-248. Nemethy EK, Skrukrud C, Piazza GJ, Calvin M (1983) Biochim Biophys Acta 760: 343-349. Niemann GJ (1985) Plants 166: 51-56. Saito K, Kaneko, H, Yamazaki M, Yoshida M, Murakoshi I (1990) Plant Cell Rep 8: 718-721. Tamaki E, Morishita I, Nishida K, Kato K, Matsumoto T (1975) Japanese Patent (50-16440). Taton M, Benveniste P, Rahier A (1986) Bioehem Biophys Res Commn 138: 764-770. Taton M, Benveniste P, Rahier A (1987) Phytochemistry 26: 385-392. Wu C-H, Zabawa EM, Townsley PM (1974) J Inst Can Sci Technol Aliment 7: 105-107.
