Journal of Biotechnology, 21 (199'1) 1-20 © 1991 Elsevier Science Publishers B.V. All rights reserved 0168-1656/91/$03.50 ADONIS 016816569100141A BIOTEC 00677

Effect of culture process on alkaloid production by C a t h a r a n t h u s roseus cells. I. Suspension cultures Rosanne Tom

1, Barbara

Jardin t, Claude Chavarie and Jean Archambault

Biotechnology Research Institute, National Research Council Canada, Montreal, Canada and t Ecole Polytechnique of Montreal, Chemical Engineering Department, Montrea~ Canada (Received 1 March 1990; revision accepted 15 May 1991)

Summary The processes for production of indole alkaloids in shake flask suspension cultures of Catharanthus roseus cells using Zenk's alkaloid production medium (APM) were evaluated. The 1-stage process consisted of inoculating APM and incubating for 15 days. The 2-stage process involved 6 d of cultivation in growth medium followed by 15 d of incubation in APM. Growth, main nutrient consumption and alkaloid production were monitored. Both culture processes produced ~ 20 g dw per I of biomass. However, 2-stage cultures yielded an inorganic nutrient richer and more active plant cell biomass, richer in inorganic nutrients, as indicated by higher (> 70%) nutrient availability and consumption. Total and individual indole alkaloid production were 10 times higher (740 mg 1-~ and 25 to 4000 /xg per g dw, respectively) for 2-stage than for 1-stage cultures. For both processes, highest alkaloid productivity coincided with complete extracellular consumption of major inorganic nutrients, especially nitrate, by the cells. Complete carbohydrate consumption in 2-stage cultures resulted in a 40% decline in produc-

Correspondence to: J. Archambault, Biotechnology Research Institute, National Research Council Canada, 6100 Royalmount Ave., Montreal, Canada H4P 2R2. Abbreviations: 2,4 D: 2,4 dichlorophenoxyacetic acid; APM: alkaloid production medium; AB5:B5 growth medium with IAA; B5: plant cell basal growth medium; KLa: mass transfer coefficient; TIA: total indole alkaloid content; SM: secondary metabolites.

2 tion. Small but significant ( ~ 10%) product release was observed for both culture regimes, which seemed not to be related to cell lysis. Catharanthus roseus; Suspension; Cultures; Indole alkaloids

Introduction

The difficulties and associated high cost of producing certain highly valuable phyto-pharmaceuticals through agriculture and lengthy polluting extraction and purification processes have been reviewed (Payne and Shuler, 1988). The use of large scale plant cell culture as an alternate source for these compounds has been suggested since the early 1960's but with no real commercial success yet on the horizon (Curtin, 1983). Means to induce the biosynthesis of desired products by cultured cells have still not been devised (for example: vincristine and vinblastine) or product yields are considered too low for economical commercialization and/or cannot be scaled up satisfactorily. However, plant cell culture may be the only feasible production route for some of these high value chemicals. The well publicized case of shikonin production is particularly revealing. High product yields [up to 23% of biomass dry weight (d.w.)] have been obtained in a large scale (750 1) batch process. The production rate of this process is said to be 5 kg of shikonin per 3 weeks of culture for $20000 total revenue per batch (Curtin, 1983). Delays in approval of this product, limited market size (150 kg per year) and the doubtful economics of this process, especially when considering the high operating cost of such a culture system (> $500-1000 per day), may all have contributed to Mitsui's decision to stop this production (Rajnchapel-Messai, 1988). Obviously, other parameters and different approaches must be considered to render this technology commercially viable. Market size (volume × price) must be sufficient to justify either expensive low volume expensive production runs or the economies of scale or larger culture systems and a multi (plant) product business strategy. Many factors influence the production of secondary metabolites by cultured plant cells. Cell line selection, product induction schemes such as production medium formulation (Zenk et al., 1977; Kutney et al., 1983; Stafford et al., 1985; Kurz, 1986; Morris, 1986), precursor feeding, elicitation (Eilert et al., 1985), transformed and organ cultures (Rhodes et al., 1986) and on-line product harvesting (Payne et al., 1987; Payne and Shuler, 1988) have been studied for various plant cell systems. Fewer studies have been reported on the effect of the physical environment (dissolved oxygen and carbon dioxide, Ducos and Pareilleux, 1986), pH etc., technique (suspension and immobilization, Brodelius, 1985) and regime (batch, fedbatch, perfusion, chemostat, 1- or 2-stage, Sahai and Shuler, 1984) on the performance of plant cell cultures. This may partly explain the variability of the results obtained. These physical factors have a profound effect on the productivity of large scale (> 0.5 1) plant cell processes (Scragg et al., 1987).

Bioengineering research must be directed at integrating these many parameters into economical process schemes and systems involving, for example, the reuse of the (immobilized) plant cell biomass in numerous optimized production cycles with product release and on-line harvesting in mind. Such a process would allow spreading the high cost of the inoculum preparation for the main production bioreactor (> 50% of the shikonin process culture costs) over a larger quantity of product, increase productivity and reduce downstream processing costs. This process scheme, which has already been proposed (Brodelius, 1985), has now come of age. Studies have been published on the feasibility of its various steps. The objectives of this work were to evaluate the production of indole alkaloids by C. roseus cell suspension cultures in shake flasks, performed according to 1-stage [production in Zenk's traditional alkaloid production medium (APM) (Zenk et al., 1977) only] and 2-stage (growth followed by production) regimes. Results from this study will be compared to those of surface immobilized cultures performed in 2-1 bioreactors (Archambault et al., 1990) using similar regimes in a subsequent publication (Tom et al., 1991).

Materials and Methods Plant cell culture

The Catharanthus roseus cell line MCR17 was generated from the leaf of a Periwinkle plant in 1984 (Archambault et al., 1989). Suspension culture of this cell line has been maintained since that time in Gamborg's basal growth B5 medium supplemented with 4.5/zM of 2,4-dichlorophenoxyacetic acid (2,4 D) and 20 g lsucrose (1B5 medium). The pH value of this medium was adjusted to 5.5 before sterilization (121°C, 1 bar, 1 h). The 200 ml suspension cultures contained in 500-ml Delong flasks were maintained at 28 °C under constant illumination on a rotary shaker operating at 150 rpm. The plant cell line was subcultured weekly at a 10% (v/v) inoculation. Production cultures in APM were maintained under identical conditions. Alkaloid production regimes

Two culture regimes were evaluated for the production of alkaloids by C. roseus cells. The 1-stage process consisted of inoculating 200 ml APM in a 500-ml Delong flask with 20% (v/v) of a 6-d old 1B5 grown C. roseus cell suspension culture (late exponential phase ~ 12 g dw per 1). The 2-stage process involved inoculating 200 ml of growth medium with 10% (v/v) of a 6-d-old late exponential phase C. roseus cell suspension culture grown in 1B5 medium. On the 6th day of this growth stage, most of the medium (90%) was decanted from the cell suspension and replaced by APM to a final volume of 200 ml to initiate the production stage. Originally, 1B5 was used for the growth phase of the 2-stage regime. Due to the inhibitory effect of 2,4 D on alkaloid production

(Roustan et al., 1982), this hormone was replaced with 1 /zM of indole acetic acid (IAA) in B5 medium with 20 g 1-1 sucrose to make the AB5 growth medium of the 2-stage process. This IAA concentration is the same as found in APM. The duration of the growth phase (6 d) was selected from immobilized cultures production studies to be published subsequently (Tom et al., 1991) for comparison purposes. In both culture regimes, the production phase lasted 15 d. All media were steam sterilized (121 ° C, 15 psi, 1 h). Results presented are averaged values (+ 10%-15%) of four replicate shake flask cultures.

Analytical: biomass production and nutrient consumption The suspension cultures were sampled (30 ml) every 2 or 3 d. The biomass was separated from the medium by filtration (5 /~m). The cells were rinsed with distilled water. Part of this sample (10 ml) was used for dry weight measurement (60 ° C, 24 h) while the residual sample was used for alkaloid determination. The collected medium sample was assayed for pH, conductivity as well as extracellular carbohydrates, nitrate, ammonium, inorganic phosphate and alkaloid content. The medium conductance was measured with a Yellow Springs Instrument (YSI) glass dip cell connected to a YSI model 35 conductance meter. Nitrate ion concentration was measured with an Orion nitrate electrode (model 930700) connected to a pH meter (Fisher model 805MP). Ammonium ion concentration was measured with an Orion ammonium electrode (model 951205) connected to a pH meter. The inorganic phosphate content of the medium was determined according to the ascorbic acid-phosphate analysis method described in the Standard Methods for the Examination of Water and Waste Water (1985). Carbohydrate concentration was measured using a Waters High Performance Liquid Chromatography (HPLC) system (model U6K injector, model 590 pump, model 410 refractometer) equipped with a Biorad Aminex Carbohydrate HPX-87C column. The column was maintained at 80 ° C. The mobile phase was HPLC grade water. The flow rate and back pressure range were 1.0 ml min-~ and 900-1500 psi, respectively.

Analytical." indole alkaloid production Indole alkaloids were extracted from the biomass and the medium according to the liquid-liquid partitioning protocol of Kutney et al. (1983). They were assayed by three methods. Products were identified by thin layer chromatography according to Farnsworth et al. (1964) and by high HPLC by comparison to available standards. Total indole alkaloids in biomass and medium extracts were evaluated at 280 nm according to Lee et al. (1981). A mixture of known alkaloids was made (Table 1) and diluted to various ratios (Table 2). The UV absorbance (AU), measured at 280 nm using a Varian model DMS100 spectrophotometer, of these solutions was

TABLE 1 Composition of the standard indole alkaloid solution used for TIA measurement Indole alkaloid

Concentration (mg I- 1) ,

Tryptamine Strictosidine lactam Yohimbine 19-Epivindolinine Vindolinine Vindoline Ajmalicine Lochnerinine Catharanthine Vincristine Serpentine Vinblastine Tabersonine Vincadifformine

51.62 54.69 59.38 34.43 37.32 35.06 52.25 19.25 63.63 33.06 3.35 34.31 72.19 53.25

* Solution was a mixture of the alkaloid standards dissolved in a 1:1 methanol-ethyl acetate solvent. linearly c o r r e l a t e d ( c o r r e l a t i o n coefficient o f 0.99) to t h e i r total i n d o l e alkaloid c o n t e n t ( T I A : mg k g - 1 o f solution) by the following e q u a t i o n : T I A = 0.644 * A U + 0.0288

(1)

This r e l a t i o n s h i p was u s e d to assess T I A o f C. roseus cell cultures. T h e c o n c e n t r a t i o n of individual a l k a l o i d s was m e a s u r e d using a W a t e r s H P L C system ( m o d e l U 6 K injector, m o d e l 590 p u m p , m o d e l 680 g r a d i e n t g e n e r a t o r a n d m o d e l U V 4 9 0 E m u l t i p l e w a v e l e n g t h d e t e c t o r ) e q u i p p e d with a S p e c t r a Physics m o d e l SP4270 i n t e g r a t o r . A l k a l o i d s e p a r a t i o n was a c h i e v e d using a B r o w n l e e L a b s Spheri-5 R P 8 c o l u m n m a i n t a i n e d at 48 ° C. T h e flow r a t e a n d b a c k p r e s s u r e r a n g e w e r e 2 ml m i n - l a n d 3 0 0 0 - 5 0 0 0 psi. T h e m o b i l e p h a s e was a mixture o f H P L C g r a d e w a t e r c o n t a i n i n g 2.5 m M o f W a t e r s P I C A r e a g e n t a n d m e t h a n o l . A l i n e a r g r a d i e n t m e t h o d , lasting 35 min, starting at 55% m e t h a n o l in w a t e r , was a d o p t e d f r o m V e r z e l e et al. (1981) for b e t t e r s e p a r a t i o n . T h e d e t e c t i o n w a v e l e n g t h was

TABLE 2 Indole alkaloids solutions and UV absorbance measured at 280 nm used for TIA determination Solution TIA concentration (mg I- t )

Absorbance (relative units)

1.20 0.60 0.37 0.30 0.15

1.790 0.920 0.558 0.448 0.128

7.0

T o.

6-01~

• "--~'---'-~0--

•

-0

•

•

•

•

5.0

0 0

500.

A-

1•.

101 o_

8'.

~

6.

°7

2-

30. mo,..~ 20.

"V

10.

U3 ~

201

/

/ , - -

'_ 151 =

I / I I

e

5 1 / OT

0

.

.

.

3

.

.

.

6 .

.

.

9 .

.

.

.

12 .

.

15

TiME (d) Fig. 1. Growth curves of 1-stage C. roseus cultures, o: nitrate concentration; ~: ammonium concentration; v : phosphate concentration.

primarily 254 nm. A chart recorder was also connected to the UV detector to measure absorbance at 280 nm and 330 nm to facilitate tabersonine and lochnerinine identification.

Results

Growth of I-stage cultures Growth and major nutrient consumption curves of suspension cultures performed according to the 1-stage regime are presented in Fig. 1. Characteristic rates are summarized in Table 3. Ammonium and phosphate were taken up by the cells within 3 d from inoculation. Nitrate was depleted from the medium by day 6. The appearance of 0.3 mM of ammonium in the medium at the end of the culture (day 15) indicated a decline in cell viability. Medium conductivity closely paralleled the

TABLE 3 Growth characteristics of 1 and 2-stage suspension cultures Growth parameter Maximum biomass, dw (g 1- l) Specific growth rate ( d - l ) Average growth rate, dw (g i - l d-1) Average carbohydrate consumption rate (g 1- t d -1) Average biomass yield, dw per sugar (g g - t ) Nitrate consumption rate (mM d - t ) Ammonium consumption rate (mM d - l )

1-Stage APM

2-Stage AB5

APM

22.0 0.25 1.32

13.0 0.43 2.00

21.0 0.18 1.15

1.84 0.72 1.80 2.71

4.17 0.47 4.15 0.90

4.44 0.35 5.15 1.94

consumption of the inorganic nutrients, dropping from 2300 to 650/zmho by day 3 (NH 4 and PO 4 depletion) and to 300/zmho by day 6 (NO 3 depletion) until the end of the culture. The medium pH increased from 4.3 at inoculation to 6.0 by day 3 and remained at this value until the end of the culture. Extracellular carbohydrates were consumed at an average rate of 1.84 g 1-1 d - i from day 3 onward, after complete sucrose hydrolysis in the medium. They were not depleted from the medium by day 15. Dried biomass concentration increased faster during the first nine days (20 g 1-1) and attained 22 g 1-1 by day 15. Final dried biomass yield relative to carbohydrates was 72%.

Product spectra of 1-stage cultures The identifiable indole alkaloids found in these cultures were tryptamine, strictosidine lactam, ajmalicine, serpentine and epivindolinine. Their respective production patterns are illustrated in Fig. 2. The precursor tryptamine and the first alkaloid of the pathway, strictosidine lactam, were detected in the cells during the whole culture duration. Traces of both compounds were detected in inoculated cells. Tryptamine yield per dw seemed to decline with time, from 150/zg g-1 at day 3 to 65/zg g-1 at day 15. Strictosidine lactam remained at ~ 20-25/zg g-X up to day 12 and increased to 80/zg g-I by day 15. Ajmalicine appeared only on day 9 at a low level of 3-5/zg g - 1 . Culture content of serpentine increased between day 9 and day 15, from 8 to 23 /zg g-X. Surprisingly, the late occurring product epivindolinine was found early from day 3 to 6, ~ 35/zg g-l, and disappeared by day 9. The pattern of products released into the medium is illustrated in Fig. 3. Of the two precursors detected, only tryptamine was found in the medium in low concentrations between 40 and 90 /zg l -x, which represented only 2-8% of total production. Ajmalicine (10/zg 1-1) and serpentine (5/zg 1-1) appeared extracellularly only at the end of the culture when declining cell viability was suspected. The low quantity of ajmalicine detected in the culture at that time (~ 1.6 /zg) was mostly in the medium while little serpentine (1% of total production) was found

PRODUCT YIELD

(~g g-ldw) 150100. 50. O.

30.r~lajmalicinE I serpenfinc 20.

10. o

40. I ~ epivindolinine

20 10 0

0

3

6

9 TIME (d)

12

15

Fig. 2. Yields of identified alkaloids observed in 1-stage cultures. extracellularly. It appeared that the onset of cell lysis did not yield high release of all intracellular alkaloids into the medium, unless specific products were degraded faster than others. Epivindolinine (240/~g 1-1) was released early (day 3 and 6) to significant levels (90-95% of its total production) into the medium.

Productivity of 1-stage cultures C. roseus cell cultures, including the one discussed in this study, produce a wide range of indole alkaloids, many of which are unknown. In order to better assess their biosynthetic performance, the total indole alkaloid (TIA) method described previously was used to evaluate the total productivity of the cultures under study. Identifiable and quantifiable alkaloids of 1-stage cultures discussed in the previous section represented 2 to 8% of their TIA illustrated in Fig. 4. The TIA of 1-stage cultures reached a maximum value of ~ 55 mg 1-1 at day 3 and remained more or less constant ( + 1 5 mg 1-1) afterwards. Total product release into the medium was low ( < 10 mg 1- t or < 20% of total) before day 9 and increased up to 30% of TIA produced by day 15 as a result of possible partial cell lysis at that time (Fig. 1). Total product yield per dw increased to 6 mg g - t at day 3 and then declined to a constant value of 3 + 1 mg g-1. Apparently, no increase in overall indole alkaloid biosynthesis occurred after day 3 - 6 of the culture while

IrXTRACELLULAR CONCENTRATION (/~g I - 1culture) 100" ~

tryptamlne

80. 60. 40. 20: 0

r'-'-lalm~llclne

10. ~serpentlne

5.

0" 250. ~-~ eplvlndollnlne 200. 150. 100. 50. 0

0

3

6

9 TIME ( d )

12

15

Fig. 3. Extracellular alkaloid concentration detected in i-stage cultures. volumetric and specific productivity (Table 4) declined after this time period. This coincided with complete uptake of phosphate, ammonium and nitrate by the cells (Fig. 1). Product turnover, and onset of culture degeneration indicated by ammonium release, occurred after day 6.

Growth of 2-stage cultures Two-stage culture regimes have long been suggested for the production of secondary metabolites (SM) by plant cell culture. The first stage would yield rapid growth of the productive plant cell biomass to high concentration and the second stage would induce SM under nutrient stress, limitation or elicitation. The APM was designed for nutrient stress conditions resulting in slow growth (/z ~ 0.25 d-x), high biomass concentration (22 g 1-1) and alkaloid production (Figs. 1-3, Table 4). Two-stage cultures were performed in an attempt to increase their productivity (Table 4) by accelerating their growth phase which was carried out in B5 medium supplemented with 20 g 1- l sucrose and 1/zM IAA as mentioned in Materials and Methods. The growth and main nutrient consumption patterns of 2-stage cultures are presented in Fig. 5. Obviously, growth and nutrient consumption rates during the first stage were 50 to 130% faster than for 1-stage (APM only) cultures, except for

10 =~ 5

100-

~-~ 6o.

'%40:/--

fl~o, 2oi. R d

108-

/

,

80: 60: %4o: i2o:

i

t

i

i

i

~_..~_____~.._...._...._....~ •

R

0

3

6

9

12

15

TIME (d) Fig. 4. Total indole alkaloid estimations of 1-stage cultures. (a) Total produced alkaloids averaged at 55 + 14 mg 1- l culture between day 6 and 15; (b) extracellular TIA content; (c) total alkaloid content on a per g dry weight basis where the yield averaged 3 5:1 mg g - i.

ammonium uptake (Table 3). The second stage performed in APM showed faster carbohydrate (77%) and nitrate (186%) consumption rates but slower specific growth (72%) and ammonium uptake (72%) rates than for the 1-stage cultures.

TABLE 4 Comparison of volumetric and specific productivities Time (d)

1-Stage process

2-Stage process

Volumetric productivity (mg1-1 d -1)

3 6 9 12 15 18 21

16.4 8.3 4.1 6.7 3.9 -

18 29 22 57 49 25 19

Specific productivity, per dw (mgg -l d -l)

3 6 9 12 15 18 21

2.1 0.55 0.21 0.35 0.18

2.5 2.4 1.2 2.7 2.6 1.2 0.99

11 7.0

]I--GROWTH---~

.PRODUCTION. . . . . . .

I

!

: 5.5 ~

•

:

•

"

"

-

15

18

•/'-

E 20 t

0

3

6

9

12

21

T~Ur (d) Fig. 5.

Growth curves of 2-stage

C. roseus cultures. ©: nitrate concentration; tion; v : phosphate concentration.

A; ammonium concentra-

Overall, the net growth rate and the maximum biomass concentration of 2-stage and 1-stage culture processes were comparable. However, biomass yield of the former process (72%) was significantly higher than found for the latter (47% and 35%). Inorganic nutrients were consumed later (day 9 for 2-stage cultures as compared to day 6 for 1-stage cultures (Fig. 1)). No ammonium was released into the medium of 2-stage cultures. Carbohydrates were totally consumed by day 15 compared to a residual extracellular concentration of 10 g I-~ for 1-stage cultures (Fig. 1). In summary, 2-stage cultures did not show growth improvement over 1-stage cultures, except that the higher nutrient availability (100 to 140%) and consumption rates, and lower biomass yield observed may indicate a more active plant cell biomass.

Product spectra of 2-stage cultures In addition to the 5 known alkaloids found in 1-stage cultures, four other known products, yohimbine, tabersonine, lochnerinine and catharanthine, were detected

12

PRODUCT YIELD (/j.g g - 1dw) I---GROWTH-~

...........

PRODUCTION . . . . . . . .

I

4000

3000 2000 1000 0 60O 400

200 0 150 100 t 50 eplvlndollnlno

b

5OO

J

400 3OO

20O 100 0 40 ~s~ chthorcinthine 30 20 10 0

.3

6

9 12 15 TiME (d)

18

21

Fig. 6. Yields of identified alkaloids observed in 2-stage cultures.

in 2-stage cultures. As shown in Fig. 6, no significant quantity of known products could be detected during the growth phase. Relatively low tryptamine yields per dw ( < 200/~g g-l at day 18 and 21) were found during the production phase. Strictosidine lactam production was significant (1430 to 3960 /~g per g dw) as compared to the other alkaloids. Ajmalicine and serpentine were present at yields of 95-340/xg g-1 and 170-680/~g g-1, respectively, during the whole production phase. A trend of progressive partial biotransformation of the former into the latter and eventually into the hydroxylated corynanthe alkaloid yohimbine (350/zg per g dw) on day 21 of the production phase is suggested from Fig. 6. Aspidospermatan alkaloids lochnerinine and tabersonine were detected at day 9 and at day 21 at yields of 30 to 50 /zg g-I and 80 to 150 /zg g-l, respectively. Catharanthine appeared on day 21 at 35 /zg g-t. Unfortunately, these cultures were not pursued beyond day 21 to assess if commercially interesting products may have been biosynthesized. Epivindolinine was detected at day 18 and was produced

13 F'.XTRACELLULAR CONCENTRATION OJ,g I - 1culture) --GROWTH--I . . . . . . . . 1000- "---I tryptamlne HUmstrictosldlne lactam

PRODUCTION. . . . . . . . . .

500

o 800" r-~ almotlei~e U

i

,

-m

I

m

18

21

serpentine

I

600. 400. 200 0 E~yohimbine 4000 epivindolinlne 3000 2000 1000 0

3

6

9

12

15

TIME (d)

Fig. 7. Extracellularalkaloidconcentrationdetectedin 2-stagecultures. at a relatively high yield of 560 ~g g-1 on day 21. Catharanthine and the aspidospermatan alkaloids were not found in 1-stage cultures. The pattern of products detected in the medium of 2-stage cultures is shown in Fig. 7. This product spectrum which differs from but concurs with product yields (Fig. 6), seems to indicate significant release by C. roseus cells (at least for line MCR17) cultured under the specified conditions (Fig. 5). Tryptamine was detected only in the cells in relatively low yields (< 200/~g per g dw) at the end of the cultures (day 18 and 21) while it was found mainly in the medium at concentrations of 0.44 and 1.05 mg 1-~ early in the production phase (day 9 to 12). Strictosidine lactam was found in the medium only from day 15 to 21 at concentrations ranging from 70 to 400/zg 1-], which represented less than 1% of total produced. Little ajmalicine was detected extracellularly (~ 50 /~g l-t). Released serpentine occurred mainly on day 12 at a concentration of 0.7 mg lwhich represented 19% of total production. Extracellular yohimbine and epivindolinine were detected on day 21 at concentrations of 4.6 and 0.4 mg l-t or 43% and 2% of total production respectively. No lochnerinine, tabersonine nor catharanthine were detected extracellularly. This may be partly related to the lower pK a (4.2 and 5.5) of these aspidospermatan compounds as compared to other alkaloids detected (pKa > 6.0) in relation to the pH of the medium (6.0) and to the low quantity of catharanthine (pK~ 6.8) produced (35/~g per g dw) at the end of the

14

culture. Except for yohimbine (pK, 6.3) and epivindolinine, no clear relationship was observed between produced (Fig. 6) and released (Fig. 7) alkaloids detected in these cultures. The more metabolically active plant cell biomass of 2-stage cultures (Fig. 5) yielded a richer spectrum of known (and unknown) indole alkaloids (9 products, Fig. 6) as compared to 1-stage cultures (5 products, Fig. 2). Individual alkaloid yields from 25 to 4000 ~g per g dw, in comparison to 5 to 150/xg g- ~ (Fig. 2), and extracellular products ranging from 10 to 4600 tzg 1- i, as compared to 5 to 280/xg 1-1 (Fig. 3) of the former process were one order of magnitude higher than observed for the latter. Known product spectra resulting from both culture regimes differed significantly. Strictosidine lactam largely dominated over tryptamine in 2-stage cultures as compared to 1-stage cultures. Both ajmalicine, serpentine, and eventually yohimbine, were present during the whole production phase of the former whereas mainly, serpentine was detected from day 9 onward in the latter. Four additional known and more complex products appeared in 2-stage cultures. Finally, epivindolinine was found in both types of culture, but much sooner (day 3-6) in 1-stage than in 2-stage cultures (day 12-15 of the production phase). Consequently, 2-stage cultures resulted in subsequent transformation of precursor compounds into more complex and interesting alkaloids.

1000 ~ I ---GROwTH--]. . . . . . . . . . . .

PRODUCTION .........

lO

!,:°o:I ~

200

d

IO~:T

'

'

'

"

b

~}~ 804 ~ 6o~

t

'

'

'

'

'

15

18

21

c

~T,, 601 ~ 401 0

3

6

9

12 TIME (d)

Fig, 8. TIA estimates of 2-stage cultures. (a) Total produced alkaloids representing intra- and extracellular quantities; (b) extracellular TIA concentrations; (c) total alkaloid content expressed on a dry weight basis.

15

Productivity of 2-stage cultures The identified and quantified alkaloids of 2-stage cultures discussed in the previous section represented 1%, 58%, 10% and 37% of their TIA concentrations illustrated in Fig. 8 at day 6, 9, 12 to 15 and 18 to 21, respectively. This figure shows some production ( ~ 25% of maximum observed) occurring at the end of the growth phase in AB5 medium. Production increased significantly to a maximum of 700 mg l-: between day 12 and 15 of the culture and declined to 400 mg l-] between day 18 and 21. Total product release was low ( < 65 mg 1-t of culture) representing about 10% of total indole alkaloids produced during the whole culture. The total alkaloid yield per dw increased to 40 mg g-: by day 15 following the trend noted for total production. Obviously, no net biosynthesis occurred after day 15 and products were biotransformed into more complex alkaloids. Thereafter, as shown in Fig. 6, volumetric and specific productivity (Table 4) peaked between day 12 and 15. Highest increase in production (200 to 700 mg l -l or 22 mg l-: d -] to 57 mg 1-l d -~) which occurred from day 9 to 12, coincided with complete uptake by the plant cell biomass of main inorganic nutrients (NO3, NH 4 and PO 4) from the medium (Fig. 5). A similar trend was noted for 1-stage cultures (Figs. 1 and 4). The plateau in total production and yield (day 12-15) occurred at complete disappearance of carbohydrates from the medium (day 15). This was followed by a decrease in productivity and the biosynthesis of more complex alkaloids. As observed for individual alkaloids, total alkaloid production of 2-stage cultures was one order of magnitude higher than found for 1-stage cultures. This is summarized in Fig. 9. The 2-stage culture regime lengthened the process by 40% and did not increase in growth nor biomass concentration. However, it seemed to generate a significantly more productive plant cell biomass.

TOTAL INOOL[ ALKALOIDS ( mQ I - t oollu~)

1 O01 l-Stage Procoss 801 • 6°1

......

,0 t 201/

o.___o~o

O.~o--m~', 1000- 2 - S t a g e Process 800 i 6001 4O0 2001 . 0

3

6

,

,

9 12 15 18 21 TIME (d)

Fig. 9. Comparison of total alkaloid production in 1- and 2-stage cultures, e: Intracellular; O: Extracellular.

16 Discussion

Two processes were compared for the production of indole alkaloids by shake flask grown C a t h a r a n t h u s roseus cell suspension cultures. This biosystem has been extensively evaluated for secondary metabolite production (Zenk et al., 1977; Kutney et al., 1983; and others). Numerous indole alkaloids are produced, with few known a n d / o r of commercial interest, making its production kinetics difficult to assess. In this work, the kinetics of alkaloid production by cultured C. roseus cells (line MCR17) was evaluated by following the biosynthesis of individual known alkaloids and by estimating the totality of UV (280 nm) absorbing products (Lee et al., 1981) of these extracted cultures. Obviously, this last estimate is less accurate than chromatographic methods. The results presented in the previous section confirm the difficulty of evaluating the overall production, performance of this system when following only known indole alkaloids. The TIA approach offered a more comprehensive measure of the global productivity of the system. However, little relationship was found between TIA and individual (or total quantity of) known alkaloids, except for a similar increase in production upon inorganic nutrient depletion, when comparing 1 and 2-stage cultures. At that time (day 9), the correspondence between TIA measurements and total known alkaloids detected was the highest in both processes (8 and 58% respectively). The 2-stage cultures resulted in biomass production (growth rate and highest biomass concentration) comparable to that of 1-stage cultures. However, main inorganic nutrients were supplied and taken up in larger quantities (50% to 300%) and consumed, during the production phase of the former, significantly faster (> 70%) (especially nitrate) as compared to the latter. In addition, the lower biomass yield suggests higher respiratory activity of 2-stage cultures. This is indicative of a nutrient richer (higher nitrogen-to-carbohydrate ratio) and a metabolically more active plant cell biomass in 2-stage cultures whose productivity was 10 times higher than observed for 1-stage cultures. This increase in production could be attributed partly to the lower inoculum volume (10%) and lower 2,4 D carryover of 2-stage cultures, which were diluted twice (in AB5 and APM), as compared to the 20% inoculum used for 1-stage cultures. However, it is doubtful that a 0.1 ppm difference in final concentration of 2,4 D carried over between both culture regimes resulted in the significant increase in production observed for 2-stage cultures (Roustan et al., 1982). In addition, a lower inoculum volume would have yielded slower growing and consequently less productive 1-stage cultures. The requirement for an actively growing inoculum for alkaloid production by C. roseus cells has been reported (Stafford et al., 1985; Rokem and Goldberg, 1985). Two-stage processes have long been suggested for plant cell cultures, whereby rapid growth to high cell density would be followed by a resting and producing stage (Tabata and Fujita, 1985; Kurz, 1986; Morris, 1986; Sakuta and Komamine, 1987). The results presented in this study for 2-stage cultures show further that a nutritionally primed plant cell biomass yields significant increases in productivity

17 of the culture system. This aspect was not optimized for suspension cultures since results of this study will be compared to those obtained from immobilized cultures performed according to Archambault et al. (1990) and to similar schemes. For both culture regimes, highest total alkaloid concentration (Figs. 4 and 8) and production (Table 4), as measured by the TIA method, coincided with complete uptake of major inorganic nutrients (especially nitrate: day 3 to 6 of the production phase) and lowest conductivity of the medium (Figs. 1 and 5). This is also reflected in the appearance of ajmalicine and serpentine in 1-stage cultures (Fig. 2) and in the appearance of all known alkaloids in 2-stage cultures (Fig. 6). Nitrate assimilation has been shown to produce significant changes in cultured plant cell metabolism including increases in secondary metabolic enzymes such as phenylalanine ammonia lyase (Hahlbrock, 1974). This explains the increased productivity observed. The additional stress of carbohydrate disappearance from the medium of 2-stage cultures (Fig. 5: day 15) resulted in rapid ( < 3 days) and significant ( ~ 43% of highest TIA) products turnover as shown by Wink (1987) and formation of more complex alkaloids (Fig. 6: yohimbine, tabersonine, lochnerinine, epivindolinine and catharanthine). Carbon and/or nitrogen retrieval for survival may explain this alkaloid degradation, although starch was most probably still present intracellularly in high quantity. The quantities of indole alkaloids detected in the medium of these cultures, although small, were significant. Other alkaloids and/or larger quantities may have been released and degraded rapidly before they could be measured. Extracellularly detected TIA were 10 to 20% of total production, which represented 20 to 65 mg 1-1 for 1- and 2-stage cultures, respectively. Little tryptamine, strictosidine lactam or serpentine (1% of total produced) were released while most epivindolinine was found extracellularly on day 3 and 6 of 1-stage cultures. Declining cell viability toward the end of 1-stage cultures, as indicated by ammonium release (MacCarthy et al., 1980), did not result in high release of total ( ~ 10 mg 1-~ additional) and/or known products. Old, non-producing cells would be most probably lysing initially. The release pattern observed in 2-stage cultures differed somewhat. Most tryptamine (pK a 10.2) produced initially was released ( ~ 1 mg 1-1). Yohimbine (pK a 6.3) and serpentine (pKa 10) were the only products detected extracellularly in significant relative quantities (40% and 19% of total produced respectively). The closeness of the culture pH (6.0) to the PKa values of yohimbine and ajmalicine may justify their presence in the medium. It is likely that ajmalicine may have been partly oxidized to serpentine in the medium which may explain the predominance of the latter alkaloid extracellularly (Rosazza and Duff, 1986). A companion paper (Tom et al. 1991) compares the alkaloid production of 200 ml C. roseus cell suspension cultures to that of 2 1 surface-immobilized bioreactor cultures performed according to Archambault et al. (1990).

Acknowledgements The authors would like to thank the following people for their technical support: C. B6dard, N. Chauret, Dr. D. Rho, M. Pacheco-Oliver, M. Trani and B.

18

Chatson, Dr. I. Reid for editorial comments and F. Landry for typing the manuscripts.

References Archambault, J., Volesky, B. and Kurz, W.G.W. (1989) Surface immobilization of plant cells. Biotechnol. Bioeng. 33, 293-297. Archambault, J., Volesky, B. and Kurz, W.G.W. (1990) Development of bioreactors for surface immobilized plant cells. Biotechnol. Bioeng. 35, 702-711. Brodelius, P. (1985) The potential role of immobilization in plant cell biotechnology. Trends Biotechnol. 3, 280-285. Curtin, M.E. (1983) Harvesting profitable products from plant tissue culture. Bio/Technology 1, 649-659. Ducos, J.P. and Pareilleux, A. (1986) Effect of aeration rate and influence of pCO 2 in large-scale cultures of Catharanthus roseus cells. Appl. Microb. Biotechnol. 25, 101-105. Eilert, U., Kurz, W.G.W. and Constabel, F. (1985) Stimulation of sanguinarine accumulation in Papaver sornniferum by fungal elicitors. J. Plant Physiol. 119, 65-76. Farnsworth, N.R., Blomster, R.N., Damratoski, D., Meer, W.A. and Cammarato, L.V. (1964) Studies on C a t h a r a n t h u s alkaloids VI. Evaluation by means of Thin Layer Chromatography and CAS reagent. Lloydia 27(4), 302-314. Hahlbrock, K. (1974) Correlation between nitrate uptake, growth and changes in metabolic activities of cultured plant cells. In: Street, H.E. (Ed.), Tissue Culture and Plant Science, Academic Press, pp. 363-377. Kurz, W.G.W. (1986) Biological and environmental factors of product synthesis. Accumulation and biotransformation by plant cell culture. N.Z.J. Technol. 2, 77-81. Kutney, J.P., Aweryn, B., Choi, L.S.L., Honda, T., Kolodziejczyk, P., Lewis, N.G., Sato, T., Sleigh, S.K., Stuart, K.L., Worth, B.R., Kurz, W.G.W., Chatson, K.B. and Constabel, F. (1983) Studies in plant tissue culture. The synthesis and biosynthesis of indole alkaloids. Tetrahedron 39(22), 3781-3795. Lee, S.L., Cheng, K.D. and Scott, A.I. (1981) Effects of bioregulators on indole alkaloid biosynthesis in C. roseus cell culture. Phytochemistry 20, 1841-1843. MacCarthy, J.J., Ratcliffe, D. and Street, H.E. (1980) The effect of nutrient medium composition on the growth cycle of C. roseus G. Don cells grown in batch culture. J. Exp. Bot. 31, 1315-1325. Morris, P. (1986) Regulation of product synthesis in cell cultures of C. roseus - Effect of culture temperature. Plant Cell Rep. 5, 427-432. Payne, G.F., Shuler, M.L. and Brodelius, P. (1987) Large scale plant cell culture. In: Lydersen, B.K. (Ed.), Large Scale Cell Culture Technology, Hanser Pub., Munich, pp. 194-229. Payne, G.F. and Shuler, M.L. (1988) Selective adsorption of plant products. Biotechnol. Bioeng. 31, 922-928. Rajnchapel-Messai, J. (1988) Cellules v~g&ales: enqu~te de m~tabolites. Biofutur, juillet-aofit, pp. 23-34. Rhodes, M.J.C., Robins, R.J., Hamill, J. and Par, A.J. (1986) Potential for the production of biochemicals by plant cell cultures. N.Z.J. Technol. 2, 59-70. Rokem, J.S. and Goldberg, I. (1985) Secondary metabolites from plant cell suspension culture: methods for yield improvements. Adv. Biotechnol. Proc. 4, 241-274. Rosazza, J.P.N. and Duff, M.W. (1986) Metabolic transformations of alkaloids. In: Brossi, A. (Ed.), The Alkaloids. Vol. 27, Academic Press Inc., Orlando, p. 323-368. Roustan, J.P., Ambid, C. and Fallot, J. (1982) Influence de l'acide 2,4 dichloroph6noxyac6tique sur l'accumulation de certains alcaloides indoliques dans les cellules quiescentes de Catharanthus roseus cultiv6es in vitro. Physiologic v6g6tale 20, 523-532. Sahai, O.P. and Shuler, M.L. (1984) Multistage continuous cultures to examine secondary metabolite formation in plant cells: phenolics from Nicotiana tabacurn. Biotechnol. Bioeng. 26, 27-33.

19 Sakuta, M. and Komamine, A. (1987) Cell growth and accumulation of secondary metabolites. In: Constabel, F. and Vasil, I. (Eds.), Cell Culture and Somatic Cell Genetics of Plants. Vol. 4, Academic Press, NY., pp. 97-264. Scragg, A.H., Morris, P., Allan, E.J., Bond, P. and Fowler, M.W. (1987) Effect of scale-up on serpentine formation by Catharanthus roseus suspension cultures. Enzyme Microb. Technol. 9, 619-264. Stafford, A., Smith, L. and Fowler, M.W. (1985) Regulation of product synthesis in cell cultures of Catharanthus roseus (L.) G. Don. Plant Cell Tissue Organ Culture 4, 83-94. Standard Methods for the Examination of Water and Wastewater (1985) 16th edit., Greenberg, A.E., Trussel, R.R. and Clesceri, L.S. (Eds.), American Public Health Association, Washington, D.C., pp. 437-438. Tabata, M. and Fujita (1985) Production of shikonin by plant cell cultures. In: Zaitlin, M., Day, P. and Hollaender, A., (Eds.), Biotechnology in Plant Science. Relevance to Agriculture in the Eighties. Academic Press Inc., New York, pp. 207-218. Tom, R., Jardin, B., Rho, D., Chavarie, C. and Archambault, J. (1991) Effect of culture process on indole alkaloid production by C. roseus cells. II: Immobilized cultures. J. Biotechnol., 21, 21-42. Verzele, M., DeTaeye, L., Van Dyck, J. and DeDecker, G. (1981) HPLC of l/inca rosea alkaloids and the correlation plate height and molecular weight. J. Chromatogr. 214, 95-99. Wink, M. (1987) Physiology of the accumulation of secondary metabolites with special reference to alkaloids. In: Constabel F. and Vasil, 1. (Eds.), Cell Culture and Somatic Cell Genetics of Plants. Vol. 4, Academic Press, N.Y., pp. 17-42. Zenk, M.H., EI-Shagi, H., Arens, H., Stockigt, J., Weiler, W. and Deus, B. (1977) Formation of the indole alkaloids serpentine and ajmalicine in cell suspension cultures of Catharanthus roseus. In: Barz, W., Reinhard, E., Zenk, M.H. (Eds.), Plant Tissue Culture and its Biotechnological Application. Springer Verlag, pp. 27-43.