Lanatoside 15′-O-Acetylesterase from Digitalis lanata Plants and Cell Cultures

J.PlantPhysiol. Vol. 136.pp. 289-294 (1990)

Lanatoside 15'-O-Acetylesterase from Digitalis lanata Plants and Cell Cultures RENATE SUTOR, HUBERT HOELz,

and

WOLFGANG KREIS

Pharmazeutisches Institut, Universitat Tiibingen, Auf der Morgenstelle 8, Tiibingen, FRG Received September 1S, 1989 . Accepted January 9,1990

Summary The enzymatic deacetylation of 15'-O-acetylated cardenolides was investigated. A partially purified enzyme extract prepared from suspension-cultured Digitalis lanata cells (line K 3 OHD) showed high levels of lanatoside-hydrolysing activity. The enzyme responsible is located in the cell walls, from which it could be solubilized with 0.1 M Na-citrate buffer, pH 6.0. A standard enzyme assay was established based on an HPLC method and the apparent Km values of several cardenolide substrates were determined. Lanatoside-specific esterase could be demonstrated in leaves and several callus and suspension cultures of Digitalis lanata but not in the tissues of two other Digitalis species, namely D. purpurea and D. heywoodii.

Key words: Digitalis lanata, cardiac glycosides, cell suspension culture, cell wall, esterase, specific esterase. Abbreviations: HPLC

=

High performance liquid chromatography.

Introduction To date more than 60 different cardiac glycosides have been isolated from Digitalis lanata tissues. The lanatosides, which are the predominant cardenolides in Digitalis lanata leaves (up to 0.6 % of the dry matter), represent more than 60 % of the total cardenolides in this tissue (Wagner, 1988). Some of the Digitalis glycosides are important medicines used in the treatment of cardiac insufficiency brought on by hypertension and atherosclerosis (Mutschler, 1986). Digitoxin and digoxin, two of the cardenolides most frequently used in medicine, can be prepared from Digitalis lanata leaves. The production process involves fermentation and saponification steps during which the primary glycosides, such as the lanatosides A and C, are deglucosylated and deacetylated to their respective secondary glycosides, namely digitoxin and digoxin. Until recently, attempts to solubilize these enzymes from the leaf material had failed. In the meantime, however, cardenolide-specific glucohydrolases have been solubilized from various Digitalis tissues and the enzymes have been partially characterized (Kreis et aI., 1987, 1990). Although cardenolide-specific acetylesterase activity was found to be associated with cell wall preparations from a variety of Digitalis species (Buhl, 1984), at© 1990 by Gustav Fischer Verlag, Stuttgart

tempts to solubilize and characterize this enzyme have not been reported. Suspension cultures of Digitalis lanata provide a unique material for studying cardenolide conversions on the cellular and enzymatic levels. Recently it was possible to isolate and localize two enzymes using such cultures, namely NADPH: digitoxin 12,a-hydroxylase (Petersen and Seitz, 1985) and UDPG: digitoxin 16'-O-glucosyltransferase (Kreis et aI., 1986). In our present study, a selected Digitalis lanata cell line that is known to deacetylate exogenous cardenolides in vivo was used to investigate the deacetylation of lanatosides on the cellular and enzymatic level.

Materials and Methods Plants and cell cultures Seeds of Digitalis lanata Ehrh., Digitalis purpurea L. and Digitalis heywoodii P. et M. Silva (Scrophulariaceae) were either obtained from Boehringer Mannheim GmbH, FRG (D. purpurea, D. lanata, D. heywoodii) or collected in the Black Forest (D. purpurea; Sepasgosarian, 1987). Leaves were taken from Digitalis plants (rosette stage) grown in the institute's greenhouse.

290

RENATE SUTOR, HUBERT HOELz, and WOLFGANG KREIS

The D. lanata cell lines K30HL and K40HD (Kreis, 1987) were used to isolate, charaCierize and localize the lanatoside 15'-0acetylesterase. The suspensions were subcultured in modified Murashige and Skoog (1962) medium and grown under conditions described elsewhere (Kreis and Reinhard, 1985, 1988). In 1978 callus was initiated from the filaments of a D. heywoodii plant obtained from Boehringer Mannheim GmbH, F.R.G. (Line F). In 1987, suspension cultures were established from stock callus cultures of line F and cultivated in modified Murashige and Skoog medium (Kreis and Reinhard, 1985) supplemented with the following phytohormones: 2.0 mg/L 6-furfurylaminopurine (kinetin), O.S mg/L 2,4-dichlorophenoxyacetic acid (2,4-D) and 1 mg/L indole-3-acetic acid (IES). Subcultivation was done every 10.5 days as described for D. lanata. The D. purpurea cell lines used were established in 1983 and cultivated as described by Sepasgosarian (1987).

Table 1: Deacetylation of lanatoside A in vivo in cell cultures of Digitalis lanataa _ Cell line

Lanatosides (LA + LC)b Deacetyllanatosides (PA + DC) (mg/kg fresh weight) (mg/kg fresh weight) 250 22 172 219 307 91 580 22

K10HD K20HD K30HD K40HD K10HL 138 5 K20HL 152 61 K30HL 128 225 K40HL 307 30 a Lanatoside A (40 mg/L) was fed on day 3 of the culture cycle and the cells analysed after 24 hours of incubation. b LA = lanatoside A; LC - lanatoside C; PA = purpureaglycoside A; DC = deacetyllanatoside C.

Deacetylation in vivo Cells (20g fresh weight per 300ml culture medium) were incubated together with lanatoside A (40 mg/L) in l-L Erlenmeyer flasks under standard culture conditions. After 24 hours of incubation, samples of the suspension were withdrawn. The cells were separated from the culture medium and analysed by HPLC for their cardenolide content as described previously (Kreis et al., 1986).

10 min_ The cell walls were washed several times with water and once with 0.5 % Triton X-l00 solution until the supernatant was protein-free. Cell wall preparations were routinely checked for malate dehydrogenase activity (Bergmeyer, 1974), which served as the marker for intracellular enzymes.

Chemicals Enzyme extraction Cultured cells were separated from the medium, frozen with liquid nitrogen and extracted with 0.1 M Na-citrate buffer (pH 6.0). The homogenate was filtered through Mira cloth and the cell residues were then pelleted by centrifugation at 20,000 g. Proteins were precipitated from the supernatant by adding solid ammonium sulfate up to 9S % saturation. The precipitate was centrifuged, dissolved in extraction buffer and desalted on Sephadex G-2S. Acetone dry powders were prepared from Digitalis leaves. The powders were extraCied with 0.1 M Na-citrate buffer and the extracts then processed as described above.

Lanatoside 15'·0-acetylesterase assay in vitro Standard assays for lanatoside lS'-O-acetylesterase were performed in 1.5 mL Eppendorf vials containing 300 ILL of the Sephadex G-2S eluate and 0.6 mg lanatoside C (2 mM). After 30 minutes at 37 °C the reaction was terminated by adding 300 ILL chloroform: isopropanol (3: 2). After vortexing the vials for 30 s the organic phase was removed and the buffer phase once again extracted with chloroform/isopropanol. The organic phases were combined and evaporated. The residue was dissolved in 300 ILL methanol and subjected to HPLC analysis. In experiments where a-acetyldigoxin was used as the substrate for testing lanatoside 15'-0-acetylesterase activity the reaction was terminated by adding 300 p.L methanol. The precipitate was centrifuged and the supernatant was analysed by HPLC.

Localization of lanatoside 15'-0-acetylesterase Protoplasts and vacuoles were prepared as described in a previous publication (Kreis and Reinhard, 1985)_ Cell walls were prepared and purified using the method of Masuda et al. (1988) with slight modifications_ Three grams of wet cells were homogenized in two volumes of 10 mM Na-phosphate buffer, pH 7_0_ The preparations were inspeCied microscopically for complete cell breakage_ The insoluble cell fragments were centrifuged at 1,000 g for 10 min_ The residue was suspended in distilled water and centrifuged at 270 g for

Lanatoside A and a-acetyldigoxin were obtained from Carl Roth GmbH & Co KG (Karlsruhe, FRG). The lanatosides Band C were gifts from Boehringer Mannheim, FRG.

Analytical procedures The conditions for the HPLC analysis of enzyme assays containing cardenolides have been described earlier (Kreis et aI., 1986). Protein was determined according to Bradford (1976). Bovine serum albumin served as the protein standard.

Results and Discussion Deacetylation in vivo Several Digitalis lanata cell lines were checked for their ability to deacetylate lanatoside A (Table 1). The cell lines K 3 OHL and K 2 OHD converted lanatoside A to purpureaglycoside A very rapidly. In either case more than 50 % of the substrate were deacetylated after 24 hours of incubation. The cell lines K 4 OHD and K 1 OHD, on the other hand, exhibited only weak deacetylation activity in vivo (Table 1). Therefore, protein extracts from line K3 OHL were used to establish an enzyme assay in vitro.

Deacetylation in vitro Although cell line K 3 OHL cells showed strong lanatoside deacetylation in vivo, only weak enzyme activity was detected in particle-free extracts from cells homogenized in 0.1 M phosphate buffer (pH 7.0). Most of the enzyme activity was still associated with the 20,000 g pellets. Several extraction buffers were tested and optimal solubilization of the enzyme was achieved when a 0.1 M Na-citrate buffer was used for cell extraction. It was reported that citrate stimulates the activity of the membrane-bound chickpea malonyl-

Lanatoside 15'-O-acetylesterase

reaction was linear for about 30 minutes when a protein concentration of 0.25 mg/mL was used in the assay. Reducing agents, such as ascorbate and dithiothreitol, or the addition of polyvinylpolypyrrolidone (PVPP) had no stimulatory effect on the acetylesterase activity. Ethylenediaminetetraacetic acid (Na-EDT A) did not affect enzyme activity itself but promoted enzyme solubilization (data not shown).

Table 2: Influence of different extraction buffers and buffer exchange on lanatoside 15' -O-acetylesterase activity. Extraction buffer (pH 6.0) Citrate-phosphate Citrate

Buffer for enzyme assay (Sephadex G-25 eluate, pH 6.0)

Relative esterase activity (%t 84 100

Sodium phosphate Citrate 12 Citrate Citrate 100 Citrate Phosphate 84 a The esterase activity found in citrate buffer was set to equal 100 %. It ranged from about 300 to 400 /'katlkg protein depending on the enzyme preparation used.

Substrate specificity Most of the carboxylesterases known are able to hydrolyse a large variety of natural or artificial acylesters (Krisch, 1971). Only few acylhydrolases have been reported to show high specificity towards one single group of structurally related substrates (e.g. Siitfeld and Towers, 1982; Polz et al., 1987; Hinderer et al., 1986). Unspecific esterase activity (determined with p-nitrophenylacetate as the substrate) could be detected in cell cultures from both Digitalis lanata and D. purpurea, but only the extracts from D. lanata hydrolysed lanatoside C to its deacetylated derivative. In addition, citrate buffer extracts from twice-washed walls from D. lanata K 3 OHD cells showed strong lanatoside hydrolysing activity together with almost no unspecific esterase activity. On the basis of this and other indirect evidence we concluded that the wall-bound esterase from Digitalis lanata cell cultures is a cardenolide-specific enzyme. Apparent Km-values were determined for several acetylated cardenolides, namely a-acetyldigoxin and the lanatosides A, Band C (Fig. 1). Lanatoside A had the lowest apparent Km-value, but the enzyme reaction was inhibited at

esterase by a factor of 2 (Hinderer et al., 1987). In the case of our Digitalis cells the lanatoside-hydrolysing esterase activity detected in citrate-buffer extracts was about 20 - 30 times higher than that found when phosphate buffer was used (T able 2). In order to find out whether citrate is necessary only for efficient enzyme extraction or also important for catalytic activity, cells were extracted with either phosphate or citrate buffer and the buffers then exchanged on Sephadex G-25. Phosphate-buffer-extracted protein was now in citrate buffer and vice versa. It was shown that citrate is needed for enzyme extraction, but not for the enzyme activity itself (Table 2) and that 4 volumes of buffer were necessary for optimal enzyme extraction from cultured K 3 OHL cells. Optimal esterase activity occurred at pH 6.0 with half maximal activities at pH 5.0 and pH 7.2, respectively. The temperature optimum was at 37°C and the deacetylation

Km= 0.50 mM

Km=0.13 mM

60

60

40

40

20

20

,

..

O(-Acetyldigoxin

..,

o

L E

10

20

Lanatoside A

o

30

Km=042 mM

10

60

40

40

20

20

30

Lanatoside C

Lanatoside B 10

20

Km=0.91 mM

60

o

291

20

30

1. [mW1]

0

10

20

30

S

Fig. 1: Lineweaver-Burk plots of kinetic data obtained with esterase preparations from the Digttalts lanata cell line K 3 OHL. Apparent Kmvalues for various cardenolide substrates were graphically determined. Each data point represents the mean of three replicates.

292

RENATE SUTOR, HUBERT HOELz, and WOLFGANG KREIS

lanatoside A concentrations higher than about 0.33 mM (twice the Km). Lanatoside B (Km = 0.42 mM) is not available in sufficient quantities and a-acetyldigoxin (Km = 0.5 mM) is not stable in solution. Therefore lanatoside C (Km = 0.91 mM) was selected to be the standard substrate when the cardenolide-specific acetylesterase, from now on referred to as lanatoside 15'-O-acetylesterase, was examined in cell-suspension cultures.

Localization 0/ lanatoside 15 '-Q·acetylesterase So far, our knowledge about the cellular organization of cardenolide biotransformation in suspension-cultured Digi· talis lanata cells was restricted to the subcellular localization of glucosylated and non-glucosylated cardiac glycosides and the related 16'-O-glucosyltransferase (see Kreis and Reinhard, 1987; Diettrich et al., 1987 for proposed models). In order to further contribute to our understanding of the cellular mechanisms involved in cardenolide modification and storage, we here investigated the localization and cellular distribution of lanatoside 15'-O-acetylesterase. Protoplasts, cell walls and vacuoles were prepared from cultured K30HD cells. Lanatoside-hydrolysing activity could be detected neither in protein preparations from isolated protoplasts nor in purified vacuoles. Citrate buffer extracts obtained from cell walls, on the other hand, showed a specific lanatoside esterase activity up to 20 times higher than extracts prepared from whole cells. This demonstrates that all of the cardenolide-specific esterase present in suspension-cultured Digitalis lanata cells is localized in the cell wall fraction. This is remarkable since other specific acylesterases have been characterized as either membrane-bound (Hinderer et al., 1987) or soluble (Polz et al., 1987) cellular enzymes. In the case of the K 3 OHD cells that were used in the present compartmentalization study, about 7 - 8 volumes of citrate buffer were necessary to completely solubilize the enzyme from the cell walls. Cell line K3 OHD was also used for cardenolide biotransformation experiments in vivo. Lanatoside C rapidly loses its acetyl moiety when administered to cultured cells. The product of this reaction, deacetyllanatoside C, is found in the medium and is only gradually taken up by the cells (Kreis, 1987). This observation was surprising, since in the same cell line deacetyllanatoside C never appeared in the culture medium when produced via the biotransformation of digitoxin or digoxin. It is now clear that lanatoside C is hydrolysed extracellularly by the wall-bound esterase, making the situation completely different from that during digoxin biotransformation, where the substrate diffuses into the cells (Kreis and Reinhard, 1987) and deacetyllanatoside C is then formed by a cytosolic glucosyltransferase (Kreis et al., 1986). Lanatoside esterase could be solubilized from living D. lanata cells. Digitalis cells were infiltrated with citrate buffer and after preset times cells and washing fluid were separated and both analyzed for their esterase activities. After 10 and 15 min, 55 % and 88 % of the overall enzyme activity was found in the washing buffer. In both set-ups cell growth was reduced compared to the non-treated control. The cells, however, recovered from the buffer treatment and after about 5 days the cells' acetyl esterase activity increased again,

1.4,.-----------------, 12

e

"I

~ 1.0

-- 0.8 n. ~

C\

e

0.6 ;S 0.4 E

C\

E

02

_

.-e_ .----:::::

./'-~_e./ /'

:7

/'/

..

/"

;.--O . O~--_=r=..",,--;=___._-.....-_r___.-_._-_._-..__--r--' o

2

4

6

8

10 12

14 16

18 20

Time [d]

Fig. 2: Solubilization in vivo of wall-bound lanatoside esterase from cultured Digitalis lanata cells. Cells were infiltrated for 10 min (.6.) or 15 min (A) with citrate buffer and then transferred back into culture medium. The time courses of esterase activity were monitored over a period of 21 days and compared with that of a non-treated control (e). DC, deacetyllanatoside C.

"I c

.Qj

-0 d:

C\

E

2.5

K3 OHL K40HD 0-0 0-0

2.0 1.5

l5 to g'

0.5

o

2

4

6

8

Time [d]

10

12

14

Fig. 3: Time course of lanatoside 15'-O-acetylesterase in vitro. The Digitalis lanata cell lines K30HL (e) and K40HD (0) were examined. DC, deacetyllanatoside C.

indicating that the enzyme was synthesized de novo and transported to its place of action, the cell wall (Fig. 2).

Lanatoside 15 '-O-acetylesterase in cell cultures and leaf tissues Once the enzyme assay was established, several cell cultures and plant tissues were assayed for lanatoside 15'-O-acetylesterase activity. Considerable differences were found among the D. lanata cell lines tested. K3 OHL, the cell line used to characterize the lanatoside esterase, showed the highest deacetylation rates of all the cell lines tested so far (Table 3}. The changes in lanatoside esterase activity over the course of one cell culture cycle were examined with two different cell lines. As expected after the experiments in vivo, line K 3 OHL showed stronger enzyme activities than line K 4 OHD (Fig. 3). In both cell lines the transfer of cells into fresh culture medium caused an increase in enzyme activity. After 3 days of culture a cell-line-specific activity level was reached.

Lanatoside 15'-O-acetylesterase Table 3: Lanatoside 15' -O-acetylesterase activity in cell cultures and leaves from Digitalis lanata, Digitalis purpurea and Digttalis hey·

woodii.

Species

Tissue

Lanatoside esterase activity (Jtkat/kg Protein)

D. lanata

Suspension K 3 OHL Suspension K 4 OHD Suspension K 1 OHD Suspension W. 1.4 Callus W. 1.4 Young leaves Old leaves Suspension 21 B Callus 21 B Callus 33 Leaves Suspension F Callus Leaves

401.2 90.9

D. purpurea

D. heywoodii

17.7

88.5 123.9 149.1 17.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Lanatoside 15'-O-acetylesterase was not detectable in the suspension cultures of D. heywoodii or D. purpurea examined in the present study (Table 3). The respective callus cultures also showed no lanatoside esterase activity. Digitalis lanata leaves contain large amounts of lanatosides or other cardenolides, that could interfere with the esterase assay. Therefore, acetone dry powders were used when measuring lanatoside 15'-O-acetylesterase in leaf tissue in order to ensure cardenolide-free enzyme preparations. Lanatoside C, which served as the substrate when assaying lanatoside esterase in suspension cultures, could not be used to determine the enzyme activity of leaves, because lanatosides are also good substrates for the cardenolide-specific glucohydrolase known to be present in leaves of various Digitalis species (Stoll et aI., 1935, Kreis et aI., 1987, 1990). Therefore a-acetyldigoxin was used as the substrate when enzyme preparations from leaves were analysed for esterase activity. As already mentioned, a-acetyldigoxin is eventually degraded under standard assay conditions. For that reason the assay conditions were changed in such a way that 1 volume of methanol was used to stop the enzyme reaction and that the samples were analysed immediately. Using the modified assay, lanatoside esterase was demonstrated in D. lanata leaves, young leaves showing higher levels of enzyme activity than older ones. The enzyme could not be detected in citrate-buffer extracts of D. heywoodii or Digitalis purpurea (Table 3). In addition, cell walls were prepared from callus cultures or leaves from either species and then suspended in citrate buffer. Again, no conversion of aacetyldigoxin to digoxin was observed. These last findings contrast sharply with results reported by Buhl (1984), who incubated cell walls prepared from the leaves of various Digi· talis species with a-acetyldigoxin. He reported equally low conversion rates for both D. lanata and D. purpurea and high digoxin formation with cell wall preparations of D. hey· woodii and two other Digttabs species. It could be that the plants used in Biihl's investigations and our own were not in a comparable physiological or developmental stage and surely investigations into the variations in esterase activity during plant development have to follow. Since, on the other

293

hand, D. purpurea plants do not contain lanatosides but, instead, produce large quantities of the purpureaglycosides A and B, it might well be that enzymes metabolizing lanatosides are absent from D. purpurea tissues. We were able to demonstrate, for instance, that lanatosides are not deglucosylated by the cardenolide-specific glucosidase solubilized from D. purpurea leaves but that they are by the glucosidases from both D. lanata and D. heywoodii (Kreis et aI., 1990). Acknowledgements This work was supported by a grant from the Deutsche Forschungsgemeinschaft (Re 131/11-2).

References BERGMEYER, H. U.: Methoden der enzymatischen Analyse, 3. Aufl. Verlag Chemie, Weinheim (1974). BRADFORD, M. M.: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. AnaL Biochem. 72, 248-254 (1976). BUHL, W.: Enzyme in BHittern von Digitalis-Arten (unter besonderer Beriicksichtigung von herzglykosidspaltender Glucosidase und Esterase). Doctoral thesis, University of Marburg (1984). DIETTRICH, B., U. ASTER, N. GREIDZIAK, W. Roos, and M. LUCKNER: Glucosylation of digitoxin and other cardenolides in cell cultures of Dzgttalzs lanata. Biochem. PhysioL Pflanzen 182, 245-255 (1987). HINDERER, W., J. KOSTER, and W. BARZ: Purification and properties of a specific isoflavone 7-0-glucoside-6/1-malonate malonylesterase from roots of chickpea (Ocer arietinum L.). Arch. Biochern. Biophys. 248,570-578 (1986). - - - Further characterization of chickpea isoflavone 7-0-glucoside-6/1-0-malonate: malonylesterase: evidence for a highly specific, membrane-bound enzyme in roots of Cicer arietmum L. Z. Naturforsch. 42c, 251-257 (1987). KREIS, W.: Untersuchungen zur Kompartimentierung der Cardenolid-Biotransformation in Digitalzs lanata Zellkulturen. Doctoral thesis, University of Tiibingen (1987). KREIS, W. and U. MAy: Cardenolide glucosyltransferases and glucohydrolases in leaves and cell cultures of three Dzgitalis (Scrophulariaceae) species. J. Plant PhysioL (in press). KREIS, W., U. MAY, and E. REINHARD: UDP-glucose: digitoxin 16'O-glucosyltransferase from suspension cultured Dzgttalis lanata cells. Plant Cell Rep. 5, 442-445 (1986). - - - Cardiac glycoside glucosyltransferases and glucohydrolases from three Digttalis species. Pharmaceutisch Weekblad Scientific Edition 9, 238 (1987). KREIS, W. and E. REINHARD: Rapid isolation of vacuoles from suspension-cultured Dzgltalis lanata cells. J. Plant PhysioL 121, 385 - 390 (1985). - - Selective uptake and vacuolar storage of primary cardiac glycosides by suspension-cultured Digitalzs lanata cells. J. Plant PhysioL 128, 311- 326 (1987). - - 12 i3-hydroxylation of digitoxin by suspension-cultured Digz· talzs lanata cells. Production of deacetyllanatoside C using a two-stage culture method. Pianta Medica 1988, 95 -100 (1988). KRISCH, K.: Carboxylic ester hydroiases. In: BOYER, P. D. (ed.): The Enzymes, Volume 5, 2nd ed., 43-69. Academic Press, New York (1971).

294

RENATE SUTOR, HUBERT HOELz, and WOLFGANG KREIS

MASUDA, H., S. KOMIYAMA, and S. SUGAWARA: Extraction of enzymes from cell walls of sugar beet cells grown in suspension culture. Plant Cell Physiol. 29, 623 -627 (1988). MURASHIGE, T. and F. SKOOG: A revised medium for rapid growth and bio-assays with tobacco tissue cultures. Physiol. Plant. 15, 473-497 (1962). MUTSCHLER, E.: Arzneimittelwirkungen, 5. Auflage. Wissenschaftliche VerlagsgesellschafI mbH, Stuttgart (1986). PETERSEN, M. and H. U. SEITZ: Cytochrome P-450-dependent digitoxin 12 fj-hydroxylase from cell cultures of Digitalis lanata. FEBS Lett. 188, 11-14 (1985).

POLZ, L., H. SCHUBEL, and J. STOCKlGT: Characterisation of 2 fj (R)17-0-acetylajmalan: acetylesterase - a specific enzyme involved

in the biosynthesis of the Rauwolfia alkaloid ajmaline. Z. Naturforsch. 42c, 333-342 (1987). SEPASGOSARIAN, J.: 16fj-Hydroxylierung durch Zellkulturen von Digitalis purpurea L. Doctoral thesis, University of Tiibingen (1987). SUTFELD, R. and G. H. N. TOWERS: 5-(4-acetoxy-1-butinyl)-2,2'bithiophene : acetate esterase from Tagetes patula. Phytochem. 21, 277 -279 (1982). STOLL, A., A. HOFFMANN, and W. KREIS: Dber glucosidspaltende Enzyme der DigitalisbHitter. 12. Mitteilung iiber Herzglucoside. Hoppe-Seyler's Z. Physiol. Chern. 235, 249-264 (1935). WAGNER, H.: Pharmazeutische Biologie, 2. Drogen und ihre Inhaltsstoffe, 4. Auflage. Gustav Fischer-Verlag, Stuttgart, New York (1988).