Plant Cell Reports

Plant Cell Reports (1991) 10:256-259

9 Springer-Verlag1991

Detection of ginkgolide A in Ginkgo biloba cell cultures D a n i e l l e - J u l i e C a r r i e r 1, N a t h a l i e C h a u r e t 2, M i c h a e l M a n c i n i 2, P i e r r e C o u l o m b e 3, R o n a l d N e u f e l d ~ M a r t i n W e b e r 1, a n d J e a n A r c h a m b a u l t 2 x McGill University, Chemical Engineering Department, 3480 University St., Montreal, Canada H3A 2A7 2 Biotechnology Research Institute, National Research Council of Canada, 6100 Royalmount Ave., Montreal, Canada H4P 2R2 3 Semoptics, 148 Colonnade Road Unit 7, Nepean, Canada K2E 7R7 Received September 21, 1990/Revised version received April 30, 1991 - Communicated by J. M. Widholm

Abstract Ginkgo biloba cells were cultured in two 500 mL shake flasks and in 2 L and 6 L immobilization bioreactors using MS medium supplemented with 1 mg.L"1NAA, 0.1 mg.L-~K and 30 g.L 1 sucrose. Specific growth rates were 0.06 d-1, 0.11 d"1and 0.07 d-1for the 2 L and 6 L bioreactors and shake flask cultures, respectively. Extracellular phosphate, nitrate, ammonium and carbohydrate uptake rates of the bioreactor cultures were approximately 17 to 39 % slower than those of shake flask cultures. The specific oxygen uptake and carbon dioxide transfer rates of immobilized Ginkgo biloha cells ranged from 0.027 to 0.041 mmol O2.g-l.d.w.hra (maximum uptake at 14 days) and 0.020 to 0.057 mmol CCh.g.a .d.w.hr4 (maximum production at 14 days). Extracts from the biomass of the two immobilized and shake flask suspension cultures were analysed for ginkgolide A by GC-MS. Yields of 7, 17, 19 and 7 ng.g.-Id.w, of ginkgolide A were determined for shake flask 1, shake flask 2 and the 2 L and 6 L immobilized cultures, respectively. Traces of ginkgolide B were detected with the signal to noise ratio, however, being too low for positive confirmation og this last product. Abbreviations CTR: Carbon dioxide transfer rate; DO: Dissolved oxygen; g.d.w.: Gram dry weight; GA: Ginkgolide A; GB: Ginkgolide B; GC: Gas chromatography; GC-MS: Gas chromatography-mass spectrometry; HPLC: High performance liquid chromatography; K: Kinetin; MS: Murashige and Skoog salt medium; NtKtMS: Complete Murashige and Skoog medium supplemented with 1 mg.L4 NAA, 0.1 mg.L-~ K and 30g.U 1 sucrose; NAA: Naphthaleneacetic acid; OTR: Oxygen transfer rate; PAF: Platelet Aggregating Factor; qc~: Specific carbon dioxide productionrate; qo~: Specific oxygenuptake rate; u: Specific growth raw; Introduction G i n k g o l i d e s are unique C20 lactone cage molecules (Fig. 1) extracted f r o m the leaves o f the slow g r o w i n g G i n k g o b i l o b a L. tree (Drieu 1986). These diterpenes are formed o f six five-membered rings including a spiro-nonane system, a tetrahydrofuran cycle and three lactonic groups (Braquet et al. 1985). The synthesis o f these c o m p l e x products requires multi-step procedures which are difficult to scale up economically (Core), et al. 1988). These phytochemicals show significant pharmacological activity against Platelet A g g r e g a t i n g F a c t o r ( P A F ) (Braquet et al. 1985). G i n k g o l i d e B (GB) is the most efficient P A F antagonist. Offprint requests to." R. Neufeld

0

0

I I i I

ej(~/

_

ok Go oc

I ,1 1 1 io.io.i.I io.ioHio.

o-

el)u~

v

I

Fig. 1. Ginkgolides A, B and C

Nakanishi and Habaguchi (1971) were the first to grow G_._ biloba cultures with the objective o f elucidating the biosynthetic pathway o f ginkgolides. H o w e v e r , they were unable to detect ginkgolide A (GA) or GB from their cultures using H P L C analysis. Since their pioneering work, no report has been published on ginkgolide production b y G. b i l o b a cell culture. In previous work, G. b i l o b a cell suspension cultures carried out in shake flasks were characterized for growth and main extracellular nutrient uptake patterns (Carrier et al. 1990). Early attempts to detect ginkgolides in these cultures using the protocol p r o p o s e d b y Lobstein-Guth et al. (1983) for ginkgolide purification from G. biloba leaf extracts failed. In o r d e r to lower the limit o f detection for determination o f ginkgolide presence in G. biloba cells, G C - M S was used. This analytical instrument offered the possibility o f analyzing samples in the Selected Ion M o n i t o r i n g mode thus alleviating the need for purification steps. Oinkgolide C was not monitored because the mass o f the silylated molecule exceeded the settings o f the GCMS. L a r g e r quantities o f cells, to be extracted, were produeed in 2 L and 6 L immobilization bioreactors (Archambault et al. 1990). Plant cell culture is an

257 interesting alternative to produce ginkgolides. However, the biosynthetic ability of in vitro grown G. biloba cells to produce ginkgolides must be assessed and information on the growing conditions of the cells must be gathered before proceeding to yield improvement. This represents the first published report confirming the presence of GA in G....~. biloba cell culture. Materials and methods Materials Leaves, picked from field grown G. biloba trees (Botanical Gardens, Montreal, Canada) in late August 1988, were stored at -80' C for subsequent extraction. Prof. W. Yates ( U n i v e r s i t y o f Indiana,Indianapolis, USA) donated seeds for initiation o f callus cultures. Ginkgolide standards GA and GB were kindly provided by Dr. P. Braquet 0nstitut Henri Beaufour, Les Ulis, France).

Suspension cultures G. biloba cell line 27-4 was maintained in MS medium (Murashige and Skoog 1962) supplemented with 1 m g . L a NAA, 0.1 m g . L "t K and 30 g . L a sucrose as previously reported (Carder et al. 1990). The pH of this medium was adjusted to 6.0 prior to sterilization (121" C, 1 bar, 25 minutes). Delong flasks (500 mL) containing 100 mL of medium were inoculated with 30 m L o f 10 day old suspension. The suspensions were agitated at 120 RPM and maintained at 25~ C under constant incandescent illumination (continuous illumination with two General Electric F20T12/CW lights placed 60 cm from the cultures).

Immobilized bioreactor cultures G. biloba cells were grown in 2 L and 6 L immobilization bioreaetors (Archambault et al. 1990) in N~K~MS. The 6 L immobilization bioreactor was equipped with a Ingold dissolved oxygen (DO) polarographic probe connected to a Pegasus Multimeter calibrated prior to inoculation using nitrogen and air at a flow rate o f 0.1 VVM. lnocula for the 2 L and 6 L bioreactors consisted of 500 and 1200 m L of 10 day old G. bUoba suspension prepared in NaKIMS. More than 95% of the cells were immobilized on the support material (immobilizing area of 1728 and 6500 cm2 for the 2 L and 6 L respectively) within 24 hours. The immobilized cultures, maintained at 26* C under constant illumination, were aerated at 0.1 VVM with air supplemented with 2% (V/V) CO2. The 6 L bioreactor inlet and outlet gas composition and oxygen and carbon dioxide transfer rates were measured and calculated as outlined by Archambault (1991). Samples were taken every second day. At the end o f the culture period, the bioreactors were dismantled. The immobilization structures loaded with wet biomass were drained of excess medium and weighed. The cells were scraped, collected, weighed and lyophylized.

Analytical Biomass (only for shake flask cultures) and extraeellular carbohydrate, phosphate and nitrate concentrations were analysed as previously described (Carder et al. 1990). Ammonium ion concentrations were determined according to (Weatherburn 1967). Biomass samples from the 2 L and 6 L bioreactors (4.37 and 9.73 g.d.w, respectively) and shake flask grown ceils (1.94 and 1.83 g.d.w.) as well as G. biloba leaves (6.82 g.d.w.) were lyophylized and extracted in 500 mL acetone for 16 hr at room temperature. The solvent was evaporated to dryness. One m L of silylating agent (Trisil BSA DMF, Pierce Chemicals, Rockford, IL, USA) was added to the residue. The mixture was vortexed vigorously and heated for 1 hr at 73* C. GC-MS analysis o f the extracts were performed with a Varian 6000 GC equipped with a 30 m x 0.25 m m I.D. DBI W.C.O.T. (J&W Scientific) capillary column and an on-column injector (J&W Scientific). The GO was directly interfaced to a VG ZAB HF mass spectrometer. Helium was used as carrier gas at a pressure of 1 bar. The temperature program used increased the temperature from 200 ~ C to 285 ~ C at a rate of 10" C per rain. The mass spectrometer conditions were: 70 -el/electron-impact ionization; trap current 100 hA; source temperature 275* C; all re-entrants and transfer lines were maintained at 250' C.

Resolution was 5000 (10% valley definition). Spectra were acquired in the Selected Ion Monitoring (SIM) mode using accelerating voltage switching using a dwell time of 120 ms each and a 10 ms delay switching between ions trdz 537 and 552. Quantitation was based on the peak area from the m/z 537 mass chromatograms in the SIM mode. The concentration of silylated GA was determined by comparing its response to that of coinjected deuterated trimethylsilylated GA (50 ul deutero regisil (Regis Chemical, Morton Grove, IL, USA) added to 95 ng of GA and heated at 75* C for 1 hour).

Results and Discussion Shake flask cultures G. biloba cells (line 27-4) were grown in two shake flasks. Biomass growth and extracellular nutrient concentrations in the cultures were monitored. Average nutrient uptake and biomass production rates were calculated as previously described (Carrier et al. 1990). Growth characteristics are summarized in Table 1 and compared to the previously cultivated cell line 32-27 which displayed 26 to 58 %higher growth and nutrient consumption rates. Neither cell line entirely consumed the 40 mM nitrate originally contained in MS, with 10raM residual extracellular nitrate remaining unconsumed by the end of the culture (27 days). The two suspension cultures were harvested and biomass lyophylized when extracellular nitrate consumption had halted and the extracellular ammonium and phosphate were depleted (27 days). Bioreactor cultures G. biloba cells (line 27-4) were simultaneously grown in 2 L and 6 L immobilization bioreactors for 27 and 31 days respectively. This cultivation technique was used because it was found to be simpler than suspension cultures (Archambault et al. 1990). Complete attachment of the inoculated cells was observed 24 hours after inoculation yielding initial biomass concentrations of 3.75 and 1.52 g.d.w.L "1for the 2 L and 6 L bioreactors. Nutrient uptake rates are presented in Table 1. Nutrient consumption curves of 6 L immobilized cultures are shown in Figure 2. Extracellular phosphate and ammonium were consumed by day 26 at average rates of 0.002 mM.d 1 and 0.63 mM.d 1 respectively. As of day 6, the pH dropped to a value of 4.3 and increased to 4.8 by day 26 remaining flmre until the end of the culture. The extracellular nitrate consumption rate was estimated as 1.0 mM.d ~ and began when sucrose hydrolysis was completed. As for shake flask cultures, only 77 % of the extracellular nitrate was taken up by the immobilized cells. After 31 days of culture, 5 g.L"1 of extracellular glucose remained unconsumed. Most extracellular nutrient consumption rates were lower (carbohydrate:39 %, nitrate:26%,ammonium:17%,) than those of shake flask cultures. The immobilization technique prevented biomass sampling and rendered impossible the characterization of the biomass growth kinetics of the cultures. The nutrient consumption patterns of shake flask grown cells of lines 27-4 and 32-27 indicated that the lag phase ended upon complete extraceUular sucrose hydrolysis. As seen from Figure 2, complete extracellular sucrose hydrolysis of the

258 immobilized culture occurred on the 8~ day. From the inoculated and final harvested bioreactor biomass quantities and this estimate of the lag phase, the average specific growth rate was estimated as 0.11 day-l. As seen from Table 1, this value is 36 % greater than the specific growth rate of shake flask grown cells. The specific growth rate of the 2 L culture was 0.06 day'L Better oxygen transfer may account for the higher specific growth rate obtained for the 6 L culture. Overall yields (final-initial biomass content/initial-finalcarbohydrate content) of 0.40, 0.47 and 0.38 were calculated for the shake flask cultures and the 2 and 6 L immobilization bioreactors, respectively. The patterns of the rates of exchange of oxygen and carbon dioxide of the 6 L immobilized culture (oxygen (OTR) and carbon dioxide transfer rates (CTR)) are presented in Figure 3. The oxygen concentration of the

exerted in comparing maximum biomass concentrations obtained from the bioreactor and shake flask grown cells. Sampling of immobilized cultures results in removing cell free medium as compared to suspension culture sampling. 0.9

~

'

~

&--

~0.7

"W~7 ~VN

E.5 ~ 0.4

~ 700 w + ~u.u=

/

A~'It

~

0.2 00.1

1900

ACPR

00

200 10. 0

D

O.O

~

i 8

4

l 12

i t 15 20 Tbme (doy)

i

24

~

28

32

0.0

Fig. 3 0 T R , ~ and media oxygen concentration from 6 L immobilized G. biloba culture.

This may contribute to increasing the fmal biomass concentration obtained in immobilized cultures. Ginkgolide analysis

::

LI\ 20

t"

..... I:

v:

?

\

~

_

_:~ . . . .

I,o

el

,

,

~

~_*-~o

ooo

Pure GA was silylated and analysed by GC-MS. As seen in Figure 4, silylated GA displays a weak molecular ion [m] + at m/z 552 in its 70 -eV electron impact (El) mass spectrum. The major fragmentation ion at m/z 537 results from the loss of a methyl group. Figure 5 shows the mass chromatograms of pure silylated GA with a retention time of about 11 minutes which varied from injection to injection by less than 0.1 minute.

Fig. 2. Extracellular nutrient consumption by G. biloba cells cultivated in 6 L immobilization bioreactor. medium decreased over time to a value of 60% of

air

saturation. Maximum OTR and CTR of 0.54 mmol O2.L" 1.hr-1 (day 31) and 0.58 mmol CO2.L'Lhr"t (day 26) were observed. Maximum qco2 and qo2 were estimated at 0.041 mmol O2.g.-]d.w.dayl (day 14) and 0.057 mmol CO2.g." ld.w.day-~ (day 14). The specific rates calculated for the immobilized G. biloba cultures were 10-fold lower than those reported for Catharanthus roseus cultures immobilized in a similar 20 L bioreactor; 0.62 mmol CO2.g.'ld.w.hr4 and 0.43 mmol O2.g.ld.w.hr "1 (Archambault 1991). Results shown in Figure 3 suggest high viability of the culture and define oxygen transfer requirements of immobilized G. biloba cultures. As shown by the increase in OTR and CTR and the decrease in O2 concentration, the cells were physiologically active throughout the culture duration. The time at which the bioreactor cultures were terminated was arbitrarily chosen and may not have coincided with maximum ginkgolide production. The 2 L and 6 L cultures were stopped when extracellular nitrate consumption was halted and extracellular ammonium and phosphate were exhausted. The immobilizing support was covered by emerald green cells superimposed to form a biofilm having a thickness, on average, of 0.5 cm. Cultures were completely dedifferentiated showing no partial differentiation such as shoot or root primordia. Biomass concentrations of 14.24 and 14.82 g.d.w.L"1 were measured for the 2 L and 6 L immobilized cultures, respectively. Caution must be

W

i=

,,

Fig. 4. Partial mass spectrum of pure silylated GA measured at 70 -eV electron impact (El). IlL 9'~' ~1

m/=.

9:: II1.~

537 lld - C H a ] * '

..... i;I I ; I

1~ + .

l ; l b:l ~ , i ~=i I:i b,I i : l

Fig. 5. Mass chromatograms ofpure si~lated GA. Each trace normalized

to the largest peak within the time window shown.

Bioreactor and shake flask biomass samples were lyophylized, extracted with acetone, silylated and analysed by GC-MS. Figure 6 represents the mass spectral chromatogram of extracted cells harvested from the 2 L immobilization bioreactor. Yields of 7, 17, 19 and 7 ng.g." ld.w. of GA were determined for shake flask 1, shake flask 2 and the 2 L and 6 L immobilized cultures respectively. A small peak was observed at 12.3 minutes corresponding to the retention time for GB standard (unpublished results). However, the signal to noise ratio was too low to allow positive confirmation of GB from the cell extracts. Steric hindrance may have prevented complete silylation or limited the efficiency of silylation of GB while the higher temperature required to elute silylated GB off the GC column may have induced thermal

259 Table 1: Biomass production and nutrient consumption rates of 500 mL shake flasks and 2 L and 6 L immobilized bioreactor (line 27-4).

u (day1)

32-27" 0.13 27-4 Shake F.0.07 6L 0.11 2L 0.06

lag phase (days)

duration of culture (days)

max. biomass concentration g-t.d.w.L-~

average total carbohydrate consumption rate g.La.d "a

average phosphate consumption rate mM.d "~

average ammonium consumption rate mM.d a

average nitrate con. rate mM .d-~

2

18

13.42

1.95

0.004

1.75

1.84

2 8 6

27 31 27

10.26 14.82 14.24

1.42 0.87 1.04

0.002 0.002 0.007

0.76 0.63 0.62

1.36 1.00 0.61

* Cell line 32-27. Data obtained from Carrier et al. (1990)

decomposition rendering the detection of this compound more difficult. Control experiments were conducted which indicated that contamination from the glassware or chemicals used was not responsible for the positives results. Extraction and analysis of G. biloba leaves gave yields of 2.5 mg.g.ld.w. GA and 0.8 mg.g.'ld.w. GB in agreement with results reported by Okabe et al. (1967). Nakanishi and Habaguchi (1971) did not detect ginkgolides in G. biloba cell cultures utilizing HPLC or TLC analytical techniques. The low concentrations in the cultures coupled with the difficulties associated with their detection may have rendered positive confirmation impossible. ~W

9-

Ill

....

.

];il ],1

.

.

tl

.

.

il;I ~;i i4;i[8;I i1:1 il,I

Fig. 6. Mass spectral chromatograms of G. biloba cells grown in 2 L immobilized cultures," a. monitoring fragment ion at m/z 537 (silylated GA with loss of a methyl group); b. monitoring molecular ion at m/z 552 (intact silylated GA); Conclusions This study confirms that G. biloba cells can be cultured in the bioreactors developed by Archambault (1990). Extracellular nutrient uptake rates of 2 L and 6 L immobilized cultures were lower than shake flask cultured cells. The specific oxygen uptake and carbon dixoide transfer rates calculated for the 6 L culture were a factor of 10 lower than those of C___~roseus . cells cultivated in a similar system. This study shows that in vitro grown G.._~. biloba cells produce ginkgolides, although at much lower yields than observed in the plant. Yields of 7, 17, 19 and 7 ng.g.'~d.w, of ginkgolide A were determined for shake flask I, shake flask 2 and the 2 L and 6 L immobilization

bioreators, respectively. From these first results, no difference was observed between both modes of culture. In order to determine the potential of the present system, productive cell lines must be selected and induction schemes (elicitation, ginkgolide production medium, transformed cultures etc.) have to be devised and evaluated. Future work comprises aspects of purification of the acetone extracts as well as the determination of production kinetics of shake flask and immobilized bioreactor cultures. Acknowledgments The authors wish to thank Dr. Ian Reid for reviewing the manuscript. D.J.C. was a recipient of a F.C.A.R. scholarship. References Arehambault, J., Volesky, B. and Kurz, W. (1990) Biotechnol. Bioeng. 35, 702-711. Arehambault, J. (1991) Enz. Microbiol. TechnoL (in press) Braquet, P., 5pinnewyn, B., Braquet, M., Bourgaln, R., Taylor, J., Etienne, A. and Drieu, K. (1985) Blood. Vess. 16, 558-572. Carder, D., Cosentino, G., Neufeld, R., Rho, D., Weber, M. and Arehambault, J. (1990) Plant Cell. Rep. 8, 635-638. Corey, E., Kang, M., Desai, M., Ghosh, A. and Houpis, I. (1988) J.Am. Chem.Soc. 110, 649-651. Drieu, K. (1986) La Pres,~e M~dicale. 15, 1455-1457. Lobstain-Guth, A., Brian~on-Scheid F. and Anton R. (1983) J. Chromatog., 267, 431-438. Murashige, T.and 5koog, F.(1962) Physiol.Plant. 15, 473-493. Hakanishi, K.and Habagtlchi, K. (1971) J. Am. Chem. Soc. 93, 35463547. Okabe, K., Yamada, K., Yamamura, S. and Takada, 5.(1967) J. Chem. Soc. (c), 2201-2206. Weatherburn, M. (1967).I. Anal. Chem. 39, 971-974.

Detection of ginkgolide A in Ginkgo biloba cell cultures.

Ginkgo biloba cells were cultured in two 500 mL shake flasks and in 2 L and 6 L immobilization bioreactors using MS medium supplemented with 1 mg.L(-1...
413KB Sizes 0 Downloads 0 Views