Journal of Biotechnology, 21 (1991) 43-62
43
© 1991 Elsevier Science Publishers B.V. All rights reserved 0168-1656/91/$03.50 ADONIS 0168165691001432 BIOTEC 00684
Stimulated indole alkaloid release from C a t h a r a n t h u s roseus immobilized cultures. Initial studies B. J a r d i n 1, R. T o m
l
C. C h a v a r i e 1, D. R h o 2 a n d J. A r c h a m b a u l t 2
1 Ecole Polytechnique of MontreaL Chemical Engineering Department, Montreal, Canada and 2 Biotechnology Research Institute, National Research Council Canada, Montreal, Canada
(Received 1 March 1990; revision accepted 26 May 1991)
Summary Vacuolar sequestration of valuable secondary metabolites remains the major limitation to the use of immobilization technology for large scale plant-cell-based bioprocesses, which otherwise may be a more efficient culture system than suspension for this biomass. In this initial study, the release of indole alkaloids produced by immobilized C a t h a r a n t h u s roseus cells cultured in Zenk's Alkaloid Production Medium was evaluated. Unstimulated alkaloid 'release in immobilized cultures reached levels of 10 to 50% of total production or 3 to 100% of known alkaloid content (30 to 4700 /zg l - l ) , which was higher than that found for suspension cultures of the cell line used (10 to 25% of total production) without apparent cell lysis. Modifications of the medium p H value of immobilized cultures were explored in order to improve this release. Periodical additions of acid (HC1 0.1 N) or base ( K O H 0.1 N) solutions (2% v / v ) to different cultures resulted in rapid ( < 3 h) and transient variations in extracellular p H value from 5.5 to 4.3, and 5.8 to 8.5, respectively. In both cases, these variations provoked significant increases in total alkaloid (from ~ 5 - 1 0 mg 1-1 to 15 mg l - l ) , ajmalicine (from 0 to ~ 0.29 mg 1-1)
Correspondence to: J. Archambault, Biotechnology Research Institute, National Research Council
Canada, 6100 Royalmount Ave., Montreal, Canada H4P 2R2. Abbreviations: B5: plant cell basal growth medium; APM: alkaloid production medium; AB5:B5
medium containing IAA; IAA: indole acetic acid; IA: immobilizedairlift bioreactor; KLa: mass transfer coefficient (h-l); TIA: total indole alkaloids.
44 and serpentine (from 0 to ~ 0.20 mg 1-1) release, without apparent cell lysis or decrease in the culture viability. This product release was estimated to represent 100% of alkaloids produced. Catharanthus roseus; Immobilization; Culture; Bioreactor; Indole alkaloids; Prod-
uct release
Introduction
The development of immobilization techniques for plant cell culture is motivated by the problematic nature of their large scale suspension cultivation and by economic considerations. Cultured plant cells are vulnerable to the mixing stresses created in conventional larger stirred tank bioreactors (Payne et al., 1987). High density (> 10-20 g 1-1) cultures are viscous (~200-500 + centipoises), shear thinning fluids (Scragg et al., 1986; Tanaka, 1987) whose homogeneous mixing and oxygen transfer can rarely be achieved efficiently in conventional equipment, including airlift bioreactors (Piehl et al., 1988). Immobilization may circumvent these problems by isolating and segregating the cells from the medium (Payne et al., 1987). The doubling times of plant cells range from 24 to 100 h. This lengthens the raising of the productive biomass to large scale. Immobilization culture systems can stabilize the productive biomass so that long cultivation within properly designed semi-continuous processes can be achieved, thus decreasing the requirement for continued renewal of the biocatalyst (Brodelius, 1985). The immobilization of plant cells also has limitations. The concentration of the biocatalyst is limited by the volume occupied by the immobilizing matrix, and by the high wet-to-dry weight ratio (10 to 35 + ) characteristic of this type of biomass (Archambault et al., 1990a), in relation to the overall mixing and oxygen transfer efficiency of the bioreactor. Aggregation may cause micro mass transfer problems and induce cell population heterogeneity due to nutrient concentration gradients within the biofilm (Shuler, 1985). This implies that nutrition of immobilized plant cells must be closely monitored and controlled to avoid starvation and/or overgrowth while maintaining the biomass at its best physiological state for high productivity of secondary metabolites. Another limitation of immobilized cultures is that the targetted products must be released extracellularly by the cells and be soluble in the aqueous medium for recovery. The intravacuolar sequestration of indole alkaloids in C. roseus cells (Renaudin and Guern, 1982) renders questionable the use of immobilization. However, there is evidence that the culture system used to grow plant cells may influence product release into the medium (Majerus and Pareilleux, 1986; Yeoman, 1987; Barnabas and David, 1988; Archambault et al., 1990b; Tom et al., 1991a, b). The development of mild techniques to stimulate alkaloid release has
45 been suggested to circumvent this problem (Tanaka et al., 1985; Parr et al., 1987; Berlin et al., 1988; Brodelius, 1988; Kobayashi et al., 1988). It was proposed that the compartmentation of indole alkaloids in the vacuole of C. roseus cells may be governed primarily by ions trapping (Renaudin and Guern, 1982; Guern et al., 1987) although this mechanism has been questioned (DeusNeumann and Zenk, 1986). This model is based on the lipophilic nature of the alkaloids which, upon biosynthesis in the cytoplasm, are believed to permeate the tonoplast freely by diffusion. Once inside the vacuole, the alkaloids are protonated due to the acidity of this organelle (Guern et al., 1987) a n d / o r conjugation of alkaloids to various chemical residues (phenolics, polyphosphates, magnesium, glycosyl etc.) forms derivatives which are trapped within the vacuole due to loss of their lipophilic properties. It was proposed that this mechanism of sequestration may be reversed by acting on the acidity of the vacuole in relation to the pK a value of the trapped alkaloids (Renaudin and Guern, 1982). This may be achieved by exploiting the homeostatic function of the vacuole, which regulates the cytoplasmic pH (~ 7.0-7.5) of the cell within the culture environment (Boiler and Wiemken, 1986). In fact, modification of the extracellular pH was suggested as a more viable alternative to induce alkaloid release from cultured C. roseus cells than solvent permeabilization (Brodelius, 1988). Previous work indicated that acidification or dilution of C. roseus suspension cultures induced alkaloid release (Renaudin and Guern, 1982). First studies on the production of alkaloids by surface immobilized C. roseus cells cultured in a modified 6 1 airlift bioreactor showed that a moderate increase in extracellular pH from 5.0 to 5.5, a n d / o r partial dilution (20%) of the medium provoked a significant (1000%) increase in alkaloids release (Archambault et al., 1990b). This work presents detailed results on the basal release of indole alkaloids into the medium of C. roseus cell suspension and surface immobilized cultures in 2 1 bioreactors, using Zenk's Alkaloid Production Medium (Zenk et al., 1977), reported previously in a comparative production study of these two culture systems (Tom et al., 1991a, b). Subsequently, transient modifications of the medium pH value are examined in order to increase the release of indole alkaloids produced by immobilized C. roseus cells. Results from this initial study are discussed in the context of immobilized plant cell-based bioprocesses. Materials and Methods Plant cell culture Catharantus roseus cell line MCR17 was generated and maintained as previously described (Tom et al., 1991a, b). Basal alkaloid release study C. roseus cells were cultured in suspension in 500 ml Delong flasks as above and in 2 1 surface immobilized bioreactors (Archambault et al., 1990a) according to two
46 alkaloid production regimes using Zenk's (1977) Alkaloid Production Medium (Tom et al., 1991a, b). Study of both production regimes in the suspension and immobilized systems was performed by simultaneously inoculating, growing and monitoring 4 shake flasks and a series of bioreactor cultures. Every 3 d, suspension cultures were sampled (,-, 30 ml) and two immobilized cultures were stopped and dismantled to monitor biomass content, main nutrient consumption and alkaloid production. Each series of bioreactor experiments was performed twice. Results presented are averages for suspension cultures ( + 10%) and for bioreactor cultures ( + 10%). Relative alkaloid release was defined as the ratio of the alkaloid(s) found extracellularly, divided by the total amount of product detected in the culture.
Stimulated alkaloid release study Stimulated alkaloid release experiments were carried out using the 6 1 surface immobilization Bioreactor IA (Archambauit et al., 1990a) because of the larger quantity of medium available for sampling as compared to the 2 i bioreactor. The inert immobilizing material A07 (Archambauit et al., 1989) was cut into a strip of 24 × 130 cm (immobilizing area ~ 6270 + 150 cm 2) which was wound around a stainless steel cage in a square spiral configuration. The matrix-covered cage (15 x 15 × 24 cm) was placed vertically into the 6 I glass vessel of the modified airlift Bioreactor IA. The bioreactor was assembled and fitted with associated accessories (sterile air filter, sampling port and calibrated pH probe) and sterilized (121°C, 1 bar, 1 h). The media used for 6 1 cultures were filter sterilized using 0.22 ~ m Millipak 60 disposable filter units (by Millipore). The suspension inocula were prepared in 500 ml Delong flasks containing 200 ml 1B5 medium. These suspensions were inoculated and cultured as described previously, for 5-6 d (late exponential phase ~ 12 g 1- ~) prior to their transfer (5 flasks giving 1 i of suspension) into a 2 1 flask. At inoculation of the immobilized culture, the medium container, the 2 ! inoculum flask and a sterile Friedrich condenser were connected to the bioreactor which was subsequently installed on its holding structure. The bioreactor was inoculated and filled to the operational level (5.2 !). The aeration rate was set at 0.3-0.5 vvm (kL a ~ 8-10 h-~). The temperature of the culture was maintained at 28 + 0.5 °C using a control system which included a resistance temperature detector and two heating bands attached to the glass tube of the bioreactor, all connected to a PID controller (model TC300 by Viconics Inc.). The 2-stage alkaloid production regime (6-d growth period in AB5 medium followed by production in APM) was used for the 6 ! immobilized cultures. Product release experiments involving pH modifications of the medium were performed by adding to the cultures 100 ml ( ~ 2% (v/v)) of sterile 0.1 N HCI (acid experiments) or 0.1 N KOH (base experiments). During the following 3 h, the medium was sampled ( ~ 100 ml) frequently. These samples were frozen and analyzed subsequently for nutrient and alkaloid content. This sampling schedule was necessary to pinpoint maximal product release which was reported to occur shortly after
47 stimulation (Renaudin, 1981; Archambault et al., 1990b). Two days following each pH modification experiment, 1 1 of fresh APM was added to the culture ( ~ 20% (v/v)) to replenish for subsequent repeated release experiments which were performed at 3 d intervals, or 1 d after the 1 1 APM addition. This replenishment and experimental schedule and the use of APM were arbitrarily chosen. Analytical." biomass production and nutrient consumption Samples (~ 30 ml) from shake flask suspension cultures were analyzed for pH, conductivity, extracellular content of residual major nutrient carbohydrates and nitrate, dried biomass and intra- and extracellular alkaloid concentration. The medium of each immobilized culture was sampled (30-100 ml) regularly (10 min to 2 d) and assayed for pH, conductivity, extracellular content of residual major nutrients and alkaloids. At the end of an immobilized culture, the bioreactor was dismantled. The immobilizing structure loaded with the wet biomass was drained of excess medium, weighed (Wv) and extracted for alkaloids. The unattached biomass was collected, weighed and dried (60 o C) to constant weight. The residual medium was collected, weighed and analyzed as described above. The net amount of immobilized biomass (W l) contained in the bioreactor was calculated as described previously (Tom et al., 1991b). Medium composition, and intracellular and extracellular indole alkaloids were analyzed as described previously (Tom et al., 1991a).
Results
Basal alkaloid release Shake flask (0.2 1 liquid volume) suspension and 2 1 surface immobilized C. roseus cultures were performed according to the 1-stage (in APM only) and 2-stage (6-d growth phase in AB5 medium followed by production in APM) regimes. Results from this comparative production study have been reported previously (Tom et al., 1991a, b). Two stage cultures were found 5- to 10-fold more productive than 1-stage cultures. In the present study, we report on the release without stimulation (defined as basal release) of total alkaloids, as measured by the TIA method, of these cultures and on the release performance of 2-stage cultures for the typical alkaloids ajmalicine and serpentine. These results will be compared to that obtained from immobilized cultures stimulated for product release. As shown in Table 1 and Fig. 1, total product release in 2-stage suspension cultures reached 30 to 40% of total production during the growth phase and on day 6 when the shock from medium exchange (AB5 to APM) occurred. Low product release (10 to 65 mg 1-1 or 10% of total production) was observed thereafter, although intracellular alkaloid accumulation increased significantly
48 TABLE 1 Basal release of total alkaloids in C, roseus cell cultures Culture Time (d)
Suspension cultures Alkaloid concentration Medium (mg I - l )
Cells (mg I - i )
Immobilized cultures Alkaloid content Product release
Medium (mg 14)
Cells (mg I - t )
(%)
Product release
(%)
1-Stage 0 3 6 9 12 c 15 c 2-Stage 0 3 6 6d 9 12 15 18 21
0 6.2 3.0 7.6 12.0 19.0
0 46.0 46.0 29.0 68.0 45.0
0 13.0 6.0 21.6 15,0 32.0
0 1.6 16.0 12.0 3.1 6,2
3.9 9.7 13.0 21.0 40.0 46,0
0 14.0 56.0 36.0 7.2 12.0
0 17,0 4,5 26.0 9.0 N/A 65.0 25.0 39.0
0 37.0 170.0 40.0 190.0 680.0 670.0 430.0 380.0
0 31.0 3,0 40.0 5.0 N/A 9,0 6.0 17,0
0.5 52.0 19.0 18.0 23.0 22.0 21,0 15.0
1.9 8.4 12.0 18.0 24.0 30.0 160.0 32.0
21.0 76.0 60.0 50.0 48.0 42,0 12.0 40.0
a 1-stage: one stage cultivation in APM; b 2-stage: two stage cultivation first for 6 d in AB5 then in APM; c suspected partial cell lysis as indicated by the release of 0.3 m M NH~- in to the medium; d after medium exchange from AB5 to APM.
(Table 1) during the production phase. Product release did not seem to parallel production. These results concur with the limited release observed for ajmalicine (day 6 only after medium exchange: 60 ~g l-~ or 20% of production). Serpentine was diluted on day 6 (from 60/.Lg I-1 or
43
© 1991 Elsevier Science Publishers B.V. All rights reserved 0168-1656/91/$03.50 ADONIS 0168165691001432 BIOTEC 00684
Stimulated indole alkaloid release from C a t h a r a n t h u s roseus immobilized cultures. Initial studies B. J a r d i n 1, R. T o m
l
C. C h a v a r i e 1, D. R h o 2 a n d J. A r c h a m b a u l t 2
1 Ecole Polytechnique of MontreaL Chemical Engineering Department, Montreal, Canada and 2 Biotechnology Research Institute, National Research Council Canada, Montreal, Canada
(Received 1 March 1990; revision accepted 26 May 1991)
Summary Vacuolar sequestration of valuable secondary metabolites remains the major limitation to the use of immobilization technology for large scale plant-cell-based bioprocesses, which otherwise may be a more efficient culture system than suspension for this biomass. In this initial study, the release of indole alkaloids produced by immobilized C a t h a r a n t h u s roseus cells cultured in Zenk's Alkaloid Production Medium was evaluated. Unstimulated alkaloid 'release in immobilized cultures reached levels of 10 to 50% of total production or 3 to 100% of known alkaloid content (30 to 4700 /zg l - l ) , which was higher than that found for suspension cultures of the cell line used (10 to 25% of total production) without apparent cell lysis. Modifications of the medium p H value of immobilized cultures were explored in order to improve this release. Periodical additions of acid (HC1 0.1 N) or base ( K O H 0.1 N) solutions (2% v / v ) to different cultures resulted in rapid ( < 3 h) and transient variations in extracellular p H value from 5.5 to 4.3, and 5.8 to 8.5, respectively. In both cases, these variations provoked significant increases in total alkaloid (from ~ 5 - 1 0 mg 1-1 to 15 mg l - l ) , ajmalicine (from 0 to ~ 0.29 mg 1-1)
Correspondence to: J. Archambault, Biotechnology Research Institute, National Research Council
Canada, 6100 Royalmount Ave., Montreal, Canada H4P 2R2. Abbreviations: B5: plant cell basal growth medium; APM: alkaloid production medium; AB5:B5
medium containing IAA; IAA: indole acetic acid; IA: immobilizedairlift bioreactor; KLa: mass transfer coefficient (h-l); TIA: total indole alkaloids.
44 and serpentine (from 0 to ~ 0.20 mg 1-1) release, without apparent cell lysis or decrease in the culture viability. This product release was estimated to represent 100% of alkaloids produced. Catharanthus roseus; Immobilization; Culture; Bioreactor; Indole alkaloids; Prod-
uct release
Introduction
The development of immobilization techniques for plant cell culture is motivated by the problematic nature of their large scale suspension cultivation and by economic considerations. Cultured plant cells are vulnerable to the mixing stresses created in conventional larger stirred tank bioreactors (Payne et al., 1987). High density (> 10-20 g 1-1) cultures are viscous (~200-500 + centipoises), shear thinning fluids (Scragg et al., 1986; Tanaka, 1987) whose homogeneous mixing and oxygen transfer can rarely be achieved efficiently in conventional equipment, including airlift bioreactors (Piehl et al., 1988). Immobilization may circumvent these problems by isolating and segregating the cells from the medium (Payne et al., 1987). The doubling times of plant cells range from 24 to 100 h. This lengthens the raising of the productive biomass to large scale. Immobilization culture systems can stabilize the productive biomass so that long cultivation within properly designed semi-continuous processes can be achieved, thus decreasing the requirement for continued renewal of the biocatalyst (Brodelius, 1985). The immobilization of plant cells also has limitations. The concentration of the biocatalyst is limited by the volume occupied by the immobilizing matrix, and by the high wet-to-dry weight ratio (10 to 35 + ) characteristic of this type of biomass (Archambault et al., 1990a), in relation to the overall mixing and oxygen transfer efficiency of the bioreactor. Aggregation may cause micro mass transfer problems and induce cell population heterogeneity due to nutrient concentration gradients within the biofilm (Shuler, 1985). This implies that nutrition of immobilized plant cells must be closely monitored and controlled to avoid starvation and/or overgrowth while maintaining the biomass at its best physiological state for high productivity of secondary metabolites. Another limitation of immobilized cultures is that the targetted products must be released extracellularly by the cells and be soluble in the aqueous medium for recovery. The intravacuolar sequestration of indole alkaloids in C. roseus cells (Renaudin and Guern, 1982) renders questionable the use of immobilization. However, there is evidence that the culture system used to grow plant cells may influence product release into the medium (Majerus and Pareilleux, 1986; Yeoman, 1987; Barnabas and David, 1988; Archambault et al., 1990b; Tom et al., 1991a, b). The development of mild techniques to stimulate alkaloid release has
45 been suggested to circumvent this problem (Tanaka et al., 1985; Parr et al., 1987; Berlin et al., 1988; Brodelius, 1988; Kobayashi et al., 1988). It was proposed that the compartmentation of indole alkaloids in the vacuole of C. roseus cells may be governed primarily by ions trapping (Renaudin and Guern, 1982; Guern et al., 1987) although this mechanism has been questioned (DeusNeumann and Zenk, 1986). This model is based on the lipophilic nature of the alkaloids which, upon biosynthesis in the cytoplasm, are believed to permeate the tonoplast freely by diffusion. Once inside the vacuole, the alkaloids are protonated due to the acidity of this organelle (Guern et al., 1987) a n d / o r conjugation of alkaloids to various chemical residues (phenolics, polyphosphates, magnesium, glycosyl etc.) forms derivatives which are trapped within the vacuole due to loss of their lipophilic properties. It was proposed that this mechanism of sequestration may be reversed by acting on the acidity of the vacuole in relation to the pK a value of the trapped alkaloids (Renaudin and Guern, 1982). This may be achieved by exploiting the homeostatic function of the vacuole, which regulates the cytoplasmic pH (~ 7.0-7.5) of the cell within the culture environment (Boiler and Wiemken, 1986). In fact, modification of the extracellular pH was suggested as a more viable alternative to induce alkaloid release from cultured C. roseus cells than solvent permeabilization (Brodelius, 1988). Previous work indicated that acidification or dilution of C. roseus suspension cultures induced alkaloid release (Renaudin and Guern, 1982). First studies on the production of alkaloids by surface immobilized C. roseus cells cultured in a modified 6 1 airlift bioreactor showed that a moderate increase in extracellular pH from 5.0 to 5.5, a n d / o r partial dilution (20%) of the medium provoked a significant (1000%) increase in alkaloids release (Archambault et al., 1990b). This work presents detailed results on the basal release of indole alkaloids into the medium of C. roseus cell suspension and surface immobilized cultures in 2 1 bioreactors, using Zenk's Alkaloid Production Medium (Zenk et al., 1977), reported previously in a comparative production study of these two culture systems (Tom et al., 1991a, b). Subsequently, transient modifications of the medium pH value are examined in order to increase the release of indole alkaloids produced by immobilized C. roseus cells. Results from this initial study are discussed in the context of immobilized plant cell-based bioprocesses. Materials and Methods Plant cell culture Catharantus roseus cell line MCR17 was generated and maintained as previously described (Tom et al., 1991a, b). Basal alkaloid release study C. roseus cells were cultured in suspension in 500 ml Delong flasks as above and in 2 1 surface immobilized bioreactors (Archambault et al., 1990a) according to two
46 alkaloid production regimes using Zenk's (1977) Alkaloid Production Medium (Tom et al., 1991a, b). Study of both production regimes in the suspension and immobilized systems was performed by simultaneously inoculating, growing and monitoring 4 shake flasks and a series of bioreactor cultures. Every 3 d, suspension cultures were sampled (,-, 30 ml) and two immobilized cultures were stopped and dismantled to monitor biomass content, main nutrient consumption and alkaloid production. Each series of bioreactor experiments was performed twice. Results presented are averages for suspension cultures ( + 10%) and for bioreactor cultures ( + 10%). Relative alkaloid release was defined as the ratio of the alkaloid(s) found extracellularly, divided by the total amount of product detected in the culture.
Stimulated alkaloid release study Stimulated alkaloid release experiments were carried out using the 6 1 surface immobilization Bioreactor IA (Archambauit et al., 1990a) because of the larger quantity of medium available for sampling as compared to the 2 i bioreactor. The inert immobilizing material A07 (Archambauit et al., 1989) was cut into a strip of 24 × 130 cm (immobilizing area ~ 6270 + 150 cm 2) which was wound around a stainless steel cage in a square spiral configuration. The matrix-covered cage (15 x 15 × 24 cm) was placed vertically into the 6 I glass vessel of the modified airlift Bioreactor IA. The bioreactor was assembled and fitted with associated accessories (sterile air filter, sampling port and calibrated pH probe) and sterilized (121°C, 1 bar, 1 h). The media used for 6 1 cultures were filter sterilized using 0.22 ~ m Millipak 60 disposable filter units (by Millipore). The suspension inocula were prepared in 500 ml Delong flasks containing 200 ml 1B5 medium. These suspensions were inoculated and cultured as described previously, for 5-6 d (late exponential phase ~ 12 g 1- ~) prior to their transfer (5 flasks giving 1 i of suspension) into a 2 1 flask. At inoculation of the immobilized culture, the medium container, the 2 ! inoculum flask and a sterile Friedrich condenser were connected to the bioreactor which was subsequently installed on its holding structure. The bioreactor was inoculated and filled to the operational level (5.2 !). The aeration rate was set at 0.3-0.5 vvm (kL a ~ 8-10 h-~). The temperature of the culture was maintained at 28 + 0.5 °C using a control system which included a resistance temperature detector and two heating bands attached to the glass tube of the bioreactor, all connected to a PID controller (model TC300 by Viconics Inc.). The 2-stage alkaloid production regime (6-d growth period in AB5 medium followed by production in APM) was used for the 6 ! immobilized cultures. Product release experiments involving pH modifications of the medium were performed by adding to the cultures 100 ml ( ~ 2% (v/v)) of sterile 0.1 N HCI (acid experiments) or 0.1 N KOH (base experiments). During the following 3 h, the medium was sampled ( ~ 100 ml) frequently. These samples were frozen and analyzed subsequently for nutrient and alkaloid content. This sampling schedule was necessary to pinpoint maximal product release which was reported to occur shortly after
47 stimulation (Renaudin, 1981; Archambault et al., 1990b). Two days following each pH modification experiment, 1 1 of fresh APM was added to the culture ( ~ 20% (v/v)) to replenish for subsequent repeated release experiments which were performed at 3 d intervals, or 1 d after the 1 1 APM addition. This replenishment and experimental schedule and the use of APM were arbitrarily chosen. Analytical." biomass production and nutrient consumption Samples (~ 30 ml) from shake flask suspension cultures were analyzed for pH, conductivity, extracellular content of residual major nutrient carbohydrates and nitrate, dried biomass and intra- and extracellular alkaloid concentration. The medium of each immobilized culture was sampled (30-100 ml) regularly (10 min to 2 d) and assayed for pH, conductivity, extracellular content of residual major nutrients and alkaloids. At the end of an immobilized culture, the bioreactor was dismantled. The immobilizing structure loaded with the wet biomass was drained of excess medium, weighed (Wv) and extracted for alkaloids. The unattached biomass was collected, weighed and dried (60 o C) to constant weight. The residual medium was collected, weighed and analyzed as described above. The net amount of immobilized biomass (W l) contained in the bioreactor was calculated as described previously (Tom et al., 1991b). Medium composition, and intracellular and extracellular indole alkaloids were analyzed as described previously (Tom et al., 1991a).
Results
Basal alkaloid release Shake flask (0.2 1 liquid volume) suspension and 2 1 surface immobilized C. roseus cultures were performed according to the 1-stage (in APM only) and 2-stage (6-d growth phase in AB5 medium followed by production in APM) regimes. Results from this comparative production study have been reported previously (Tom et al., 1991a, b). Two stage cultures were found 5- to 10-fold more productive than 1-stage cultures. In the present study, we report on the release without stimulation (defined as basal release) of total alkaloids, as measured by the TIA method, of these cultures and on the release performance of 2-stage cultures for the typical alkaloids ajmalicine and serpentine. These results will be compared to that obtained from immobilized cultures stimulated for product release. As shown in Table 1 and Fig. 1, total product release in 2-stage suspension cultures reached 30 to 40% of total production during the growth phase and on day 6 when the shock from medium exchange (AB5 to APM) occurred. Low product release (10 to 65 mg 1-1 or 10% of total production) was observed thereafter, although intracellular alkaloid accumulation increased significantly
48 TABLE 1 Basal release of total alkaloids in C, roseus cell cultures Culture Time (d)
Suspension cultures Alkaloid concentration Medium (mg I - l )
Cells (mg I - i )
Immobilized cultures Alkaloid content Product release
Medium (mg 14)
Cells (mg I - t )
(%)
Product release
(%)
1-Stage 0 3 6 9 12 c 15 c 2-Stage 0 3 6 6d 9 12 15 18 21
0 6.2 3.0 7.6 12.0 19.0
0 46.0 46.0 29.0 68.0 45.0
0 13.0 6.0 21.6 15,0 32.0
0 1.6 16.0 12.0 3.1 6,2
3.9 9.7 13.0 21.0 40.0 46,0
0 14.0 56.0 36.0 7.2 12.0
0 17,0 4,5 26.0 9.0 N/A 65.0 25.0 39.0
0 37.0 170.0 40.0 190.0 680.0 670.0 430.0 380.0
0 31.0 3,0 40.0 5.0 N/A 9,0 6.0 17,0
0.5 52.0 19.0 18.0 23.0 22.0 21,0 15.0
1.9 8.4 12.0 18.0 24.0 30.0 160.0 32.0
21.0 76.0 60.0 50.0 48.0 42,0 12.0 40.0
a 1-stage: one stage cultivation in APM; b 2-stage: two stage cultivation first for 6 d in AB5 then in APM; c suspected partial cell lysis as indicated by the release of 0.3 m M NH~- in to the medium; d after medium exchange from AB5 to APM.
(Table 1) during the production phase. Product release did not seem to parallel production. These results concur with the limited release observed for ajmalicine (day 6 only after medium exchange: 60 ~g l-~ or 20% of production). Serpentine was diluted on day 6 (from 60/.Lg I-1 or
