Enzyme Activity and Shikonin Production in Litbospermum er ytbrorbizon Cell Cultures

Venkatesh Srinivasan Department of Chemical Engineering, State University of New York, Buffalo, New York 14260

D.D.Y. Ryu" Department of Chemical Engineering, University of California, Davis, California 95616

Received March 5, 1991/Accepted January 7, 1992

The activities of the biosynthetic enzymes phenylalanine ammonia lyase (PAL) and 3-hydroxy-3-methylglutaryl-CoA- reductase (HMGR) were measured in cells transferred from growth to production medium in a two-stage batch culture. It was found that both these enzymes showed transient increases, PAL (three- to fourfold) and HMGR (two- to four- fold), at or near the point of exhaustion of nitrogen source (NO3). Production of shikonin derivatives also started at this time. The addition of excess nitrate to the medium shortly before nitrate exhaustion (days 6-8) markedly reduced the final product yield (by 70-80%) while addition of excess ni- trate in the later stationary growth phase (days 14-16) had no significant effect. When the production rate of shikonin derivatives was correlated with PAL activity, it was ob- served that production rate is very low (less than 1 mg/L day) at low levels of PAL activity (below 0.1 unit/mg pro- tein). Once a threshold level of PAL activity (about 0.15 unit/ mg protein) is reached, the biosynthetic rate of shikonin derivatives increases. Such a relationship could not be deduced for HMGR activity. It was concluded that the pro- duction of shikonin derivatives may be limited at the phenylalanine deaminating step at low levels of PAL activity. Key words: Lithospermum erythrorhizon nitrate PAL and HMGR activity shikonin derivative production

INTRODUCTION

A considerable amount of research has gone into the examination of commercially useful secondary metabo- lites from plants. Of these, relatively few have been exploited on a commercial scale. The production of shikonin derivatives from cell cultures of Lithosper- mum erythrorhizon is a good example of a successfully scaled-up bioprocess for the commercial production of a plant secondary me tab~ l i t e .~~ Shikonin derivatives are naphthoquinone pigments of the following form:

OH 0 R

* To whom all correspondence should be addressed.

Biotechnology and Bioengineering, Vol. 40, Pp. 6 9 4 (1992) 0 1992 John Wiley 84 Sons, Inc.

R = H (deoxyshikonin) R = OCOCH=C(CH3)2 R = OH (shikonin) R = OCOCH2CH(CH3)2 R = OCOCH3 R = OCOCH(CH3);!

R = OCOCH(CH3)CH2CH3 R = OCOCH2C(OH) (CH3)z

In this article the term shikonin derivatives implies the existence of one or more of these derivatives in the product. The quantitation of product has been done by measuring them collectively as an equivalent quantity of pure shikonin (R = OH).

Currently, a two-stage batch process involving sepa- rate growth and production stages is being employed. Medium optimized individually for growth (MG-5) and production (M-9) is used and a product yield of 1400 mg/ L of shikonin derivatives has been reported.23 The reac- tors are of the stirred airlift type with 200 L working volume in the growth stage and 750 L in the production stage with residence times of 9 and 14 days, respectively. Laboratory-scale immobilized cell culture systems em- ploying either a polyurethane foam matrix" or calcium alginate beads13 have also been used, and these methods have reportedly resulted in highly improved yields.

Environmental factors that influence the growth of L. erythrorhizon cells and the production of shikonin derivatives have been well studied.6 It was found that nitrate, phosphate, copper, sulfate, and sucrose had es- pecially marked effects. Also, the type of nitrogen source used had dramatic effects. The inclusion of as low as 0.09 mM ammonium resulted in complete inhibition of production of shikonin derivatives by cell line M-18 while nitrate was a suitable nitrogen source with 6.7 mM being optimum for production. Organic nitrogen sources such as peptone, casein hydrolysate, and yeast extract caused significant decreases in shikonin deriva- tive content, and urea delayed the initiation of produc- tion. Mizukami et a1.16 suggested that in the presence of high nitrogen, L-phenylalanine is preferentially incorpo- rated into protein, and after nitrogen was exhausted, phenylalanine was converted into shikonin derivatives. They also indicated that high sucrose concentration and low nitrate concentrations promoted production. Be- sides, by using streptomycin sulfate, a noted inhibitor of

CCC 0006-3592/92/01069-06$0400

protein synthesis, they found that production of shi- konin derivatives was promoted.

The biosynthetic pathway for the production of shi- konin derivatives as elucidated by Inouye et a1.12 is shown in Figure 1. Hahlbrock and c o - w o r k e r ~ ~ ~ * ~ ~ ~ ' ~ found in their work with parsley cultures that the increase in phenylalanine ammonia lyase (PAL) activity was a func- tion of the growth stage of the cells and that the activity peaked when nitrate was completely exhausted from the medium. They also showed that various enzymes of phenolic synthesis were coordinately regulated and that all responded similarly to nitrate exhaustion. PAL was therefore chosen as a representative enzyme of the phenylpropanoid pathway in this study. It has been shown that radiolabel from labeled phenylalanine is incorporated into shikonin derivatives in root systems of L. erythrorhizon. 21 However, the biosynthesis of shikonin derivatives proceeds from the combination of intermedi- ates p-hydroxybenzoic acid (PHB) from the phenyl- propanoid pathway and geranyl pyrophosphate (GPP) from the isoprenoid pathway, an important enzyme in the regulation of the latter pathway being 3-hydroxy-3- methylglutaryl coenzyme A reductase (HMGR).' We therefore decided to study (1) the activities of PAL and HMGR in the production stage of a two-stage batch culture system, (2) the correlation between enzyme ac- tivity and shikonin biosynthesis, and (3) the significance of external nitrate levels in the synthesis of shikonin derivatives.

MATERIALS AND METHODS

Cell Line

Agrobacterium tumefaciens transformed Lithospermum erythrorhizon cells. The cell line was obtained as a gift from the stock culture collection of the Genetic Research Institute, Taeduck, South Korea.

Growth Medium

Schenck and Hildebrandt basal salts mixture with 5% (w/v) sucrose, vitamins (thiamine HCI, 5 mg/L; pyri- doxine HCl, 0.5 mg/L; and nicotinic acid, 5 mg/L), and inositol (1 g/L) was the growth medium. This medium is referred to as SH-SS hereafter. The transformed cell line could be successfully cultivated in hormone-free medium.

Production Medium

The M-923 without the hormones indoleacetic acid and kinetin (M-9h-) and with 5% (w/v) sucrose (M-9h--5S) was used in all production-stage studies unless specifi- cally indicated. All chemicals were obtained from the Sigma Chemical Company, St. Louis, MO. Liquid paraf- fin (Malinkrodt, Paris, KY) 1:6 (v/v) was included as a

PHB 1 ' ) GPP ( 9 )

4 m-GERANYL PHB ( " )

PHB- GLCiis) 4 c

SHIKONINS( 1 2 )

Figure 1. Biosynthetic pathway of shikonin derivatives: (1) L-phenylalanine, (2) phenylalanine ammonia lyase, ( 3 ) trans- cinnamic acid, (4) p-hydroxybenzoic acid, (5) acetyl coenzyme A, (6) 3-hydroxy-3-methylglutaryl coenzyme A, (7) HMG CoA reduc- tase, (8) mevalonic acid, (9) geranyl pyrophosphate, (10) GPP-PHB geranyl transferase, (11) rn-geranyl-p-hydroxybenzoic acid, (12) shikonin derivatives, (13) PHB-o-glucosidase, (14) PHB-o- glucosylase, (15) PHB glucoside.

product extractant before sterilization in all production- stage experiments."

Cell Cultivation

Suspension cell cultures were obtained by transferring small portions of calli to liquid growth medium. Subse- quently, the cells which sloughed off the larger clumps and remained in suspension after the clumps settled down following the mixing of the flask's contents (fine cells) were subjected to repeated subcultures. The cells were subcultured every 9 days using 30-40% inoculum by volume to maintain them in an actively growing state.

Batch cultures were conducted in shake flasks and in 1.4-L working volume NBS-Bioflo reactors (New Bruns- wick Scientific Co., Edison, NJ) with temperature and dissolved oxygen control. Dissolved oxygen was mea- sured with a polarographic probe (Cole Parmer Instru- ment Co., Chicago, IL) and controlled between 6 and 8 ppm in the growth stage and between 4 and 6 ppm in the production stage (see below) by a dissolved oxygen controller (Cole Parmer Instrument Co., Chicago, IL). These values were determined as optimal for the cell line M-18 in earlier ~tudies .~ Filtered air and pure oxy- gen were used as the control gases. The aeration rate was 0.14-0.15 wm. The impellers were changed from f lat-bladed turbine impellers to propeller-type impellers (3 in. diameter) to minimize shear and provide better mixing at lower agitation rates. The agitation speed was 75-100 rpm.

Two-stage batch cultures were conducted by growing cells in SH-5S for about 15 days or until nitrate was

70 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 40, NO. 1, JUNE 5, 1992

exhausted. The cells were then transferred to the pro- duction medium (M-9h--5S) (30% inoculum by volume).

Cell growth was monitored by periodically taking samples. The samples were filtered in a buchner funnel using preweighed Whatman No. 1 filter paper. The fresh weight was calculated from the combined weight of the filter paper and cells. Nitrate was assayed in cell- free medium by a nitrate electrode (Orion Research, Cambridge, MA). Phosphate was measured by the Fiske- Subbarow method as described by Umbreit et al.24 Sucrose was measured using an HP 1090 liquid chro- matograph (Hewlett-Packard Co. , Waldbronn, Germany) with an on-line refractive index detector (model HP 1037A, Hewlett-Packard Co.). The column was of the cation exchange type (Aminex 87H, 300 X 7.8 mm, Bio-Rad Laboratories, Richmond, CA). The solvent used was 0.005M H2S04. Elution conditions were 0.5 mL/min and 29"C, a temperature lower than 35°C being used to prevent sucrose inversion. Shikonin derivatives in the medium and that associated with the cells were assayed by the method of Mizukami et a1.16 They are reported as equivalent quantities of pure shikonin (R = OH) (TCI Chemicals, Portland, OR).

All cultures were conducted at 25°C and in the dark. The pH of the media was adjusted to 5.9 before sterilization.

Enzyme Assays

Phen ylalanine Ammonia L yase

Cell extracts were prepared and assayed for enzyme as described by Rhodes et aL2'

HMG- CoA Reductase

Cell extracts were prepared and assayed for enzyme as described by Chappell and Nable.3 Enzyme activity is calculated as the total of the activities measured in the 10,OOOg and 100,OOOg fractions. Protein content in en- zyme extracts was measured by the Bradford method.' Enzyme activity is reported in units per milligram of protein where one unit is the amount of enzyme that is needed for the formation of one micromole of product per minute under assay conditions.

RESULTS AND DISCUSSION

Since PAL and HMGR are representative enzymes di- recting the flow of precursors into the biosynthetic path- way of shikonin derivatives, we studied the change in the activities of these enzymes in the production stage of a two-stage batch culture system. The technique and cul- tivation conditions are described in the materials and methods section.

Figure 2 shows the observed variation in PAL and HMGR activities in the production stage. We found that

0 .4 -

- .E 0 . 5 - - P

g 0 . 2 - . I

2 2 0 . 1 - < L

100 1 10.04

0 io 2 0 3 0

TIME (days)

Figure 2. Enzyme activity in crude cell extracts and product pro- files in the production stage: (A) PAL, (A) HMGR, (.) shikonin derivatives, (0) nitrate. Data represent the mean and variation as observed in six experiments.

PAL activity increased rapidly following nitrate deple- tion and a peak in the activity was observed between days 16 and 18. Also following the depletion of nitrogen source, the concentration of shikonin derivatives was found to increase and a maximum concentration of 281 mg/L was observed around day 22. We found that the maximum concentration of shikonin derivatives varied over a range of 160-300 mg/L in various experi- ments. Figure 2 also illustrates the variation in HMGR activity in the production stage. Although the enzymes of the phenylpropanoid pathway have been known to re- spond to nitrate depletion in a characteristic manner as shown by Hahlbrock and other^:,^,^,'^ it was surprising that HMGR, an enzyme in the terpenoid pathway, showed similar behavior. HMGR activity was found to rise two- to threefold over its activity prior to nitrate depletion. A peak was observed around day 15 after which HMGR activity dropped to roughly 50% of the peak activity over a 5-day period.

It can be observed (Fig. 2) that as PAL activity in- creases, presumably proportional to the amount of ac- tive enzyme present in the cells, the rate of shikonin synthesis also increases and the production rate de- creases as PAL activity decreases. On this basis, we at- tempted to correlate PAL activity with the rate (mg/L . day) of shikonin production. The results are as illus- trated in Figure 3. It can be seen that when PAL activity

3 PAL (Unilslmg protein)

Figure3. Variation in production rate with increasing PAL activity. Note that production rate is very low at low levels of PAL activity.

SRlNlVASAN AND RYU: ENZYME ACTIVITY AND SHlKONlN PRODUCTION 71

is below 0.1 unit/mg, the shikonin derivative production rate is very low and within experimental variability, es- sentially constant. However, as PAL activity increases beyond 0.15 unit/mg, shikonin synthesis rate appears to change drastically with the apparent rate changing al- most twofold for a 0.005-unit/mg increase in PAL activ- ity. This suggests that when PAL activity exceeds a certain threshold value (0.15-0.16 unit/mg in this case), the biosynthetic rate of shikonin derivatives increases in a manner not correlated with PAL activity, implying that PAL is probably not the rate limiting enzyme in shikonin biosynthesis. When a similar correlation is at- tempted for HMGR activity and shikonin production rate (Fig. 4), the variation of production rate is even more random and no relationship between enzyme ac- tivity and production rate could be discerned.

To test the specificity of the secondary metabolic re- sponse to nitrogen source depletion, excess amounts of nitrate (5 mL of 1000 mg/L nitrate added to -80-mL culture) (Fig. 5) or sucrose (data not shown) were added

50 1

0.02 0 . 0 3 H M G R (Unils/mg.protein)

Figure 4. Variation in production rate with increasing HMGR activity.

- 5 0

4 0

'30

- 20 - 1 0

- 21

Time(dayr)

Figure 5. Effect of adding excess nitrate on production of shikonin derivatives. Data represent the mean of duplicate experi- ments. Five milliliters of 1000 mg/L potassium nitrate solution was added at times indicated by arrows. Open symbols correspond to control conditions, closed symbols to nitrate addition on day 6 (dashed lines) and to nitrate addition on day 14 (dotted lines). Cell growth (solid line without data points) is shown diagrammatically to indicate phases of growth; (O,., X) nitrate, ( 0 , m ) sucrose, (A,A, +) shikonin.

to the culture medium at various times during the culti- vation. The results were highly dependent upon the time of addition. If excess nitrate was added on days 6-8, i.e., shortly before nitrate exhaustion, the final yield of shikonin derivatives was markedly reduced to 15-25% of control levels. However, if excess nitrate was added later in the stationary phase (days 14-16), the yield was not significantly affected (typically greater than 80% of control levels). It was observed (Fig. 5) that once nitro- gen is completely depleted in control cultures, sucrose uptake slows down considerably. The addition of excess nitrate on days 6-8 resulted in accelerated sucrose con- sumption (Fig. 5). In response to nitrate addition on days 6-8, sucrose was depleted between days 10 and 12 compared to between days 14 and 16 in control condi- tions. When excess nitrate was added on day 14, most of the initially supplied sucrose had already been con- sumed. We also found (data not shown) that the ad- dition of excess nitrogen on days 6-8 did not prevent the transient increase in PAL activity, although only a 2-2.5-fold increase was observed. The difference be- tween these levels and control levels (3-4-fold increase) is probably not significant. HMGR activity was not mea- sured in this experiment. In independent experiments (manuscript under preparation) we have found that the effect of adding nitrogen-rich medium to stationary- phase cultures is to restimulate growth and the effect of adding sucrose-rich medium is to further stimulate pro- duction. The effect of excess nitrate on the production of shikonin derivatives therefore appears to be an indi- rect effect mediated by sucrose consumption. The lack of a significant effect on PAL activity also suggests that at high external nitrate levels, production may be lim- ited by factors other than PAL activity. Mizukami et a1.I6 have suggested that the level of free intracellular phenyl- alanine is important for production and that in the pres- ence of high nitrogen, phenylalanine is preferentially incorporated into protein. It is possible that our findings corroborate this reasoning. Excess sucrose (without other nutrients) added either on day 8 or day 14 had no effect on product yield and was even found to negatively affect cell growth. It thus appears that nitrate depletion to exhaustion or to a very low level is an important re- quirement for the onset of production. Although in sev- eral it has been found that secondary metabolic activity starts soon after the exhaustion of ni- trogen source in the culture medium, to our knowledge, this is the first report of secondary metabolic response in plant cell cultures to the addition of excess nitrogen at or near the point of nitrate exhaustion. Following ni- trate depletion, secondary metabolic activity, indicated by rapid increases in PAL and HMGR activities, starts. Shikonin derivative production also begins at this stage. There appears to be a critical level of PAL activity that is necessary for the production of shikonin derivatives. Once PAL activity increases beyond this value, shikonin is synthesized at rates not proportional to PAL activity.

72 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 40, NO. 1, JUNE 5, 1992

Therefore, the rate of production may be limited at the phenylalanine deaminating step at low levels of PAL ac- tivity. Presumably, once sufficient intracellular enzyme activity is attained, this limitation is lifted and some other step in the biosynthetic pathway is then rate con- trolling. We have measured the total soluble phenolics in the culture medium (data not shown), and the lag time before their synthesis begins coincides fairly well with the time required to achieve this minimum PAL activity. Thus, the depletion of nitrate appears to signal the start of secondary metabolic activity in L. eryth- rorhizon cell cultures.

Very little is known about the regulation of PAL ac- tivity in cultured cells of L. erythrorhizon. To date, only the effect of light on PAL activity has been reported. Heide et a1.l’ showed that PAL activity in dark grown cells (M-9 production medium) was 1.3 times higher that in cells grown in light, indicating that light does not cause a significant increase in PAL activity. However, when cells were transferred from the growth medium (Linsmaier and Skoog)’’ to the production medium (M-9), PAL activity increased more than twofold.26 At this point we are not certain of the role of the intracel- lular nitrate pool in this scheme. In addition, the de- tailed mechanisms of increase in enzyme activity are not known. For instance, enzyme activity could increase by de novo synthesis or by activation of stored inactive enzyme. Even less is known about the regulation of HMGR activity in L. erythrorhizon cell cultures. Stud- ies with potato tuber by Oba et al.” have shown that induced HMGR activity in potato tuber is primarily due to an increase in its rate of synthesis. Even though the data presented here appear to indicate that neither PAL nor HMGR may be the rate limiting enzyme, their role in the biosynthesis of shikonin derivatives is still of in- terest. For instance, detailed kinetic studies with puri- fied enzyme preparations and appropriate studies with labeled precursors would greatly help in quantifying substrate fluxes through these pathways. Such studies remain to be carried out.

The foregoing results suggest that the initial biosyn- thetic steps in the biosynthesis of shikonin derivatives, viz., the phenylalanine deaminating step mediated by PAL and the MVA production step mediated by HMGR, may not be the rate limiting steps. The onset of produc- tion, the rapid increase in the activities of PAL and HMGR following nitrate depletion, and the reduction in product yield by the addition of excess nitrogen indicate that the concentration of nitrogen source greatly influ- ences the production process. Under normal cultivation conditions, PAL appears to limit the rate of shikonin biosynthesis at low levels of PAL activity, but under high nitrate conditions, it is possible that the availability of precursors for secondary metabolism may control the level of production. Based on the data presented, it is suggested that the level of the carbon source may have a role in this effect.

References

1. Bradford, M.M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein using the prin- ciples of protein-dye binding. Anal. Biochem. 72: 248.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

Chandler, S., Dodds, J. 1983. The effect of phosphate, nitrogen and sucrose on the production of phenolics and solasodine in callus cultures of Solanurn laciniaturn. Plant Cell Rep. 2: 205. Chappell, J., Nable, R. 1987. Induction of sesquiterpinoid bio- synthesis in tobacco cell suspension cultures by fungal elicitor. Plant Physiol. 85: 469. Ebel, J. , Schaller-Hekeller, B., Knobloch, K-H., Wellman, E., Grisebach, H., Hahlbrock, K. 1974. Coordinated changes in enzyme activities of phenylpropanoid metabolism during the growth of soybean cell suspension cultures. Biochim. Biophys. Acta 362: 417. Fujita, Y. 1988. Shikonin production by plant (Lithosperrnurn erythrorhizon) cell cultures. p. 225 In: Y. P. S. Bajaj (ed.), Bio- technology in Agriculture and Forestry, vol. 4, Medicinal and Aromatic Plants I. Springer-Verlag, Berlin, Heidelberg. Fujita, Y., Hara, Y., Suga, C., Morimoto, T. 1981. Production of shikonin derivatives by cell suspension cultures of Lithosper- mum erythrorhizon, 11. A new medium for the production of shikonin derivatives. Plant Cell Rep. 1: 61. Gershenzon, J., Croteau, R. 1990. Regulation of monoterpene biosynthesis in higher plants. p. 99. In: G. H. N. Towers and H. A. Stafford (eds.), Biochemistry of the Mevalonic Acid Path- way to Terpenoids. Plenum, New York. Hahlbrock, K. 1974. Correlation between nitrate uptake, growth and changes in metabolic activities of cultured plant cells. p. 363. In: H. E. Street (ed.), Tissue Culture in Plant Sci- ence. Academic, New York. Hahlbrock, K., Ebel, J., Ortmann, R. O., Sutter, A., Wellman, E., Grisebach, H. 1971. Regulation of enzyme activities related to the biosynthesis of flavone glycosides in cell cultures of pars- ley (Petroselinum hortense). Biochim. Biophys. Acta 244: 7. Heide, L., Nishioka, N., Fukui, H., Tabata, M. 1989. Enzymatic regulation of shikonin biosynthesis in Lithospermurn erythrorhi- zon cell cultures. Phytochemistry 28: 1873. Heide, L., Tabata, M. 1987. Enzyme activities in cell-free ex- tracts of shikonin producing Lithosperrnurn erythrorhizon cell suspension cultures. Phytochemistry 26: 1645. Inouye, H., Ueda, S., Inoue, K., Matsumara, H. 1979. Biosyn- thesis of shikonin in callus cultures of Lithospermurn eryth- rorhizon. Phytochemistry. 18: 1301. Kim, D-J., Chang, H-N. 1990. Enhanced shikonin production from Lithosperrnurn erythrorhizon by in situ extraction and cal- cium alginate immobilization. Biotechnol. Bioeng. 36: 460. Knobloch, K., Bast, G., Berlin, J. 1982. Medium and light in- duced formation of serpentine and anthocyanins in suspension cell cultures of Catharanthus roseus, Phytochemistry 21: 591. Linsmaier, E. M., Skoog, F. S. 1965. Organic growth factor re- quirements of tobacco tissue cultures. Physiol. Plant. 18: 100. Mizukami, H., Konoshima, M., Tabata, M. 1977. Effect of nu- tritional factors on shikonin derivative formation in Lithosper- mum callus cultures. Phytochemistry 16: 1183. Oba, K., Kondo, K., Doke, N., Uritani, I . 1985. Induction of 3-hydroxy-3-methylglutaryl CoA reductase in potato tubers after slicing, fungal infection or chemical treatment, and some properties of the enzyme. Plant Cell Physiol. 26: 873. Okazaki, M., Hino, F., Nagusawa, K., Miura, Y. 1982. Effects of nutritional factors on formation of scopoletin and scopolin in tobacco tissue cultures. Agric. Biol. Chem. 46: 601. Ragg, H., Kuhn, D. N., Hahlbrock, K. 1981. Coordinated regu- lation of 4-coumarate: CoA ligase and phenylalanine ammonia- lyase mRNAs in cultured plant cells of parsley. J. Biol. Chem. 256: 10061. Rhodes, M. J.C., Wooltorton, L.S.C. 1971. The effect of ethylene on the respiration and the activity of phenylalanine

SRlNlVASAN AND RYU: ENZYME ACTIVITY AND SHlKONlN PRODUCTION 73

ammonia-lyase in swede and parsnip root tissue. Phytochem- istry 10: 1989.

21. Schmid, H.V., Zenk, M. H. 1971. p-Hydroxybenzoic acid and mevalonic acid as precursors of the plant naphthoquinone alka- nin. Tetrahedr. Lett. 44: 4151.

22. Seo, W-T., Liu, J-R., Park, Y-H. 1989. Enhanced production of shikonin using polyurethane entrapped Lithosperrnurn eryth- rorhizon cells. Kor. J. Appl. Microbiol. Bioeng. 17: 343.

23. Tabata, M., Fujita, Y. 1985. Production of shikonin by plant cell cultures. pp. 207-218 In: M. Zaitlin, P. Day, and A. Hol- laender (eds.), Biotechnology in Plant Science. Academic, Or- lando, FL.

24. Umbreit, W.W., Burris, R.H., Stauffer, J.F., 1972. p. 267 In: Manometric and Biochemical Techniques. Burgess, Minneapolis.

25. Westcott, R., Henshaw, G. 1976. Phenolic synthesis and phenyl- alanine ammonia-lyase activity in cell suspension cultures of Acerpseudoplatanus L. Planta. 131: 67.

26. Yazaki, K., Fukui, H., Tabata, M. 1986. Accumulation of p-o- beta-D-glucosylbenzoic acid and its relation to shikonin biosyn- thesis in Lithosperrnurn cell cultures. Phytochemistry 25: 1629.

27. Zenk, M., El-Shagi, H., Shulte, U. 1975. Anthraquinone pro- duction by cell suspension cultures of Morinda citrifolia. Planta Med. Suppl., p. 79.

74 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 40, NO. 1, JUNE 5, 1992