185

Plant Cell, 1issue and Organ Culture 42:185-193, 1995. © 1995 Kluwer Academic Publishers. Printed in the Netherlands.

Media optimization for maximum biomass production in cell cultures of pacific yew R.E.B. Ketchum, D.M. G i b s o n & L. Greenspan GaUo Plant Protection Research Unit, USDA, ARS, NAA, U.S. Plant, Soil, and Nutrition Laboratory, Tower Road, CorneU University, Ithaca, NY 14853, USA Received 23 March 1995; accepted in revised form 8 May 1995

Key words: taxol, tissue culture, Taxus brevifolia

Abstract

Three cell lines of Taxus brevifolia Nutt. with differing growth rates were used to assess the effects of basal salt mixtures, carbohydrates, organic nitrogen additives, vitamin formulations, and plant growth regulators on callus growth. Gamborg's B5 major salts provided significantly better growth than all other salt formulations tested. The greatest biomass was obtained with 1% total carbohydrate. The best carbohydrate combination, 0.5% fructose + 0.5% sucrose, was significantly better than all other combinations of carbohydrates tested. A complex vitamin mixture was significantly better than any one previously published vitamin formulation. Greatest rates of callus growth were obtained with 4.14 #M (1 mg 1-1 picloram, 0.46 #M (0.1 mg 1-1 kinetin, and 0.38 #M (0.1 mg 1-1) abscisic acid or 0.29 #M (0.1 mg 1-1 gibberellic acid. Our final medium, TM5, is superior to published methods for the general callus culture of T. brevifolia. This medium has improved growth in three tested cell lines to provide doubling times of 3.5 to 5.6 days, an average 5.3-fold increase over our previously published medium. Abbreviations: 2,4-D- 2,4-dichlorophenoxyacetic acid, 2,4,5-T- 2,4,5-trichlorophenoxyacetic acid, 2ip - 6-(7,'~dimethylamino)-purine, ABA - abscisic acid, BA - 6-benzyladenine, GA3 - gibberellic acid, IAA - indole-3acetic acid, IBA-indole-3-butyric acid, kinetin - 6-furfurylarninopurine, NAA - napthaleneacetic acid, picloram 4-amino-3,5,6-trichloropicolinic acid. Introduction

The importance of taxol as a chemotherapeutic drug has resulted in tremendous interest in the compound (for reviews see Cragg &Snader 1991; Kingston 1991). Limits to the supply of taxol from natural sources such as bark and needles (Cragg & Snader 1991; U. S. Forest Service 1993) have stimulated work into developing alternative sources of the drug, including semi-synthesis (Denis et al. 1988; Holton et al. 1992), total synthesis (Holton et al. 1994; Nicolaou et al. 1994), and plant cell culture. Production of taxol from cell cultures of Taxus brevifolia was first demonstrated in this laboratory by Christen et al. (1989; 1991). Although other investigators have described taxol production in callus and suspension cultures of Taxus, production of rapid, reproducible,

and economically significant quantities of taxol via plant cell culture has not been demonstrated (Fett-Neto et al. 1992; Wickremesinhe & Arteca 1993a; Fett-Neto et al. 1993; Fett-Neto et al. 1994). One key to achieving this goal is the isolation of rapidly growing cell lines capable of producing taxaries. Our laboratory noted slow growth rates in T. brevifolia cell lines; approximately 14 days for a doubling in fresh weight (Gibson et al. 1993). Fett-Neto et al. (1993) reported slow growth rates of Taxus cuspidata Sieb. & Zucc. cell lines, with doubling times of 17 and 20 days for fresh and dry weight, respectively. Wickremesinhe & Arteca (1993a) described a doubling time of 13 days for their fastest growing Taxus x media Rehd. callus line. We believe that optimization of taxol production in cell culture will ultimately be a consequence of opti-

186 mizing both conditions for rapid growth of the cells as well as rapid production oftaxol. We have concentrated solely on optimizing growth in this set of experiments. This detailed examination of medium constituents was conducted to analyze which factors have the most profound effects on maximizing biomass production. An empirically-selected basal medium has been developed that has greatly improved growth rates for Taxus cell cultures.

Media preparation

Materials and methods

All media were prepared from individual chemicals with the exception of the vitamin mixtures which were purchased premixed from Sigma Chemical (St. Louis, MO). All components were added, and pH adjusted to 5.5, prior to autoclaving. Trace elements were from B5 medium (Gamborg et al. 1968). Iron was added as EDTA-ferric-sodium salt, 50 mg 1-1. Sigma purified agar was added at a concentration of 1%. References and formulations for all other media are given in the Results section.

Plant material

Statistical analyses

Cell lines OC1, CC2010, and 4aYAC were isolated from cambium of three separate, wild-grown, T. brevifolia Nutt. trees as previously described (Gibson et al. 1993). These cell lines had rapid, moderate, and slow growth rates, respectively, and have been previously shown to be non-taxane producing genotypes (Gibson et al. 1993).

Each treatment had 8-10 replicates for each cell line. Experimental results were analyzed using a two-way ANOVA for balanced data. Differences between treatments within single cell lines were determined with a one-way ANOVA followed by Tukey's multiple comparison test with a family error rate of 0.05. Statistics were calculated using Minitab software, release 8 (Addison-Wesley, Reading, MA). Data is presented in tables with the treatment means, number of sampies analyzed for each treatment, and 95% confidence intervals. Means with the same letter, and within the same experiment, are not significantly different. When graphed, the data is presented as treatment means with error bars representing the 95% confidence intervals. In the text, all experimental results and differences are statistically significant, unless noted otherwise.

Chemicals

All chemicals and reagents were purchased from Sigma Chemical1 (St. Louis, MO) except for picloram, which was purchased from Aldrich Chemical (Milwaukee, WI). Growth measurements

Prior to inoculating medium for an experiment, all of the callus from the cell line to be used was pooled. An exact weight of callus inoculum was determined following placement of approximately 200 mg callus on 10 ml of fresh medium in 30 ml capped QorpaekTM jars (Fisher Scientific, Pittsburgh). Cultures were grown in the dark at 25 °C for 28 days. At the end of the growth period, callus was weighed. To minimize differences in growth that may have been the result of variation in inoculum size, results are presented as a growth index: G r o w t h I n d e x = [WtF~,~at - Wtznitial]

(1)

Wtlnitial At the end of each experiment, we pooled the best growing callus to use for inoculation of the subsequent experiment. The basal medium for each experiment was determined from the results of the previous experiment.

Results and discussion

It is extremely important to note that in the present study our basal medium changed from one experiment to the subsequent experiment. If a treatment was found to give a significant improvement, that treatment was incorporated into the basal medium of subsequent experiments. Our rationale is based on results from preliminary experiments testing effects of various auxins. When a sub-optimal m o u n t of sucrose in the basal medium was used, no significant differences were seen between treatments; when an optimal amount of sucrose was used, significant differences were noted. Thus, in this study, it is crucial that comparisons be limited to treatments within an experiment.

187 3.5

Table 1. Effect of major salt formulations on growth of T. breviJblia callus cell lines OC1, 4aYAC, and CC2010.

32.5

Major salt formulations*

N

Mean growth index 95% C.I.

Anderson

20 2.81bc

(3.24, 2.37)

DBM2

20 1.28cd

(1.71, 0.83)

1

MS

20 1.99c

(2.42, 1.55)

0.5

NN

20 2.89 b

(3.32, 2.45)

KM

20 3.60 b

(4.03, 3.16)

B5

20 5.14 a

(5.57, 4.70)

LM

20 0.70 d

(1.13, 0.26)

2

Growth Index

1.5

0 Control

A.

Growth Index

O.2%CH 2%S Initial B5 Media

2D

il

* Abbreviations are given in text.

3-

ly better growth on 2 mg 1- l 2,4-D, growth of CC2010 was reduced, and growth of 4aYAC was not different from the control.

2"

1-

C Control

B.

0.2%CH

2%S

2D

Major salt formulations

Initial B5 Media

Fig. 1. Growth of T. brevifolia cell lines on previously tested media (Gibson et al. 1993). A) Treatment means with 95% confidence intervals (error bars) from all three cell lines following two-way ANOVA. B) Individual cell line growth index means. Bars are the standard error of the means. Control = B51CA medium (Table l) without casein hydrolysate; 0,2%CI-I = Control + 0.2% casein hydrolysate; 2%S = Control with 2% sucrose instead of 1% sucrose; 2D = Control medium with 9.05/~m 2,4-D instead of 4.52/zm 2,4-D.

Cell lines selected for study

Multiple cell lines were tested within each experiment to assess individual genotype responses as well as a generalized group response in order to develop an optimized medium. Cell line or genotype-specific responses to a particular treatment were noted previously (Gibson et al. 1993). In the present study we used three cell lines, initiated from different trees, and with differing growth characteristics (Fig. 1). Our control was B5 medium supplemented with 4.52 #M (1 mg 1-1 2,4-D and 1% sucrose (Gamborg, et al. 1968). The results of the two-way ANOVA for all cell lines combined indicated that this medium supported better callus growth than media with either 2% sucrose or 9.05 #M (2 mg 1-1) 2,4-D, but was not significantly different from control medium + 0.2% casein enzymatic hydrolysate (Fig. 1A). Results of individual cell lines (Fig. 1B) each showed reductions of growth on 2% sucrose and no appreciable effect with addition of casein hydrolysate. Only cell line OC 1 had significant-

A number of formulations were tested to see which major salts supported the best callus growth. The salt mixtures were chosen for their wide range of differences in ionic composition, and only the major salts were varied in each medium. The formulations tested were Anderson (1978), DBM2 (Gresshoff & Doy 1972), MS (Murashige & Skoog 1962), NN (Nitsch & Nitsch 1969), KM (Kao & Michayluk 1975), B5 (Gamborg et al. 1968), and LM (Litvay et al. 1985). There was a difference in callus growth for the best major salt formulation, B5, when compared to all other mixtures (Table 1). Our results are similar to those of Wickremesinhe & Arteca (1993a) who found that both B5 and MS were superior for T. × media callus. Flores & Sgrignoli (1991) found B5 medium to be less effective than MS in promoting rooting in T. × media cv. Hicksii and T. brevifolia, but their results were obtained on the complete basal media, not just the major salts. When comparing basal salt formulations, there was a strong correlation between decreasing NH4 + concentration and increasing growth index (r 2 = 0.85). There was also a strong correlation between decreasing NH4+:NO3 - molar ratio and increasing growth index (r 2 = 0.86). The correlation coefficient between growth and other ions in the major salt mixtures was never greater than 0.53 for ionic species that were present in all of the formulations tested. Flores et al. (1993) noted that White's medium (1934) promoted callus formation in isolated embryos of Taxus spp., when compared to

188 Table 2. Carbohydrate influences on growth of T. brevifolia callus cell lines OCI, 4aYAC, and CC2010. Carbohydrate type and concentration (%)*

N

Mean growth index

95% C.I.

0.5 Suc + 0.5 Gin

20

4.30b

(4.50, 4.08)

0.5 Suc + 0.5 Fru 0.33 Suc + 0.33 Fru + 0.33 Gin 1 Gin

20 20 20

4.95 a 4.49 b 3.50 e

(5.16, 4.74) (4.70, 4.28) (3.71, 3.29)

1 Fru 0.5 Gin + 0.5 Flu 1.5 Sue 1 Sue

20 20 20 20

4.69 ~ 3.73 c 3.48c 4.16b

(4.90, (3.93, (3.68, (4.36,

4.48) 3.51) 3.26) 3.95)

* S u c - sucrose; Glu - glucose; Fru - fructose.

DCR medium (Gupta & Durzan 1985) and suggested that callusing might be due to the reduced nutrient level and lower NH4+:NO3 - ratio of White's (1934) medium.

Carbohydrates Various combinations and concentrations of sucrose, glucose, and fructose were examined for their effect on callus growth (Table 2). The control medium contained 1% sucrose, which was previously shown to be preferable to 2% sucrose (Fig. 1). An intermediate sucrose level of 1.5% was also tested. All remaining carbohydrates were added in equal proportions to total 1% (w/v) of the medium (e.g., 3.33 g 1-1 sucrose + 3.33 g 1-1 fructose + 3.33 g 1-1 glucose). Media with 1% glucose, 1.5% sucrose, or 0.5% glucose + 0.5% fructose, all inhibited callus growth when compared to the 1% sucrose control. Media conraining 0.5% sucrose + 0.5% glucose or 0.33% sucrose + 0.33% glucose + 0.33% fructose were not different from the control. Callus growth on 1% fructose or 0.5% fructose + 0.5% sucrose was better than on the control medium. The most rapid callus growth was on medium with 0.5% fructose + 0.5% sucrose. Fructose improved growth of callus in almost every treatment examined except when added in equal combination with glucose. Perhaps even more remarkable is the observation that the cell cultures grew better when fructose was substituted for sucrose as the sole carbohydrate. Our results with fructose were quite different than those obtained by Westgate et al. (1991) who found significantly lower cell yields in cultures of a related genus in the Taxaceae, Cephalotaxus har-

ringtonia (Forbes) K. Koch, when fructose was used as the sole carbon source. Our optimal carbohydrate concentration of 1% is lower than that used by many other investigators when culturing cells of Taxus. Fett-Neto et al. (1992) used 3% sucrose for T. cuspidata and Taxus canadensis Marsh. tissue cultures. Wickremesinhe & Arteca (1993a) used total carbohydrate concentrations from 2 to 4%, with 2% sucrose having superior results, and additions of glucose and fructose at either 0.25% and 1% suppressing growth.

Organic nitrogen supplements In an effort to better define our medium, we tested amino acids as possible organic nitrogen replacements for the addition of casein hydrolysate. Six representative amino acids were tested together with a control and an additional treatment with 0.2% casein enzymatic hydrolysate. Amino acids were selected that had little or no degradation when sterilized at 121 °C for 20 minutes, and that were representative of acidic or basic amino acids, amino acids with polar and non-polar R groups, and amino acids involved in transamination reactions. Individual amino acids were added at a concentration of 2 mM which is close to the upper range of individual amino acid concentrations in medium containing 0.2% casein enzymatic hydrolysate (Technical Bulletin, Sheffield Products, Memphis, TN). From the two-way ANOVA of all cell lines combined, media supplemented with 0.2% casein hydrolysate, 2 mM proline, or 2 mM aspartic acid supported greater growth of callus than the unsupplemented control (Fig. 2A). Arginine, asparagine, or glycine supplements all resulted in greater treatment

189

Growth Index

Control CH

A.

Growth Index

Pro Gly Asp Asn Amino Acid (2 raM)

il lh1i I

Contro CH

B.

Pro Gly Asp Ash Amino Acid (2 raM)

Trio

Trp

Arg

Arg

Fig. 2. Growthof T. brevifolia cell lines on media with amino acids or casein hydrolysate. A) Treatment means with 95% confidence intervals (error bars) from all three cell lines following two-way ANOVA. B) Individual cell line growth index means. Bars are the standard error of the means. All amino acids were added at a concentration of 2 mM. Control = no supplemental organic nitrogen; CH = 0.2% casein enzymatic hydrolysate, Pro = proline, Gly = glycine; Asp = L-aspartic acid; Asn= L-asparagine; Trp = tryptophan; Arg = L-arginine.

Table 3. Amino acid concentrationeffects on growth of T. brevifolia callus cell lines OC1 and 4aYAC. Amino acid concentration*

N Meangrowth 95% C.I. index

1 mM Asp + Arg + Gly + Pro 20 3.48ab 2 mM Asp + Arg + Gly + Pro 20 3.39ab 5 mM Asp + Arg + Gly + Pro 20 3.21b

(3.74, 3.21) (3.65, 3.12) (3.47, 2.94)

*Concentration is for each individual amino acid.

means than the control, but differences were not significant. Tryptophan reduced the growth of callus when compared to the control. The effect o f supplemental amino acids on callus growth illustrates some of the differences that can be observed among T. brevifolia cell lines. For OC1, glycine, aspartic acid, tryptophan, or arginine all reduced callus growth when compared to the unsupplemented control (Fig. 2B), and no treatment improved

growth. These results were in contrast to CC2010, which showed an improvement with every amino acid or casein hydrolysate supplement, except tryptophan, when compared to the unsupplemented control (Fig. 2B). Both asparagine and tryptophan resulted in a reduction in callus growth in cell line 4aYAC. All other treatments were not different from the unsupplemented control medium. The minor improvements seen with supplementation when all cell line data were combined is the result of the major improvement of growth in CC2010; only one treatment, tryptophan, was detrimental to all cell lines. The second amino acid experiment examined the effects of concentration of individual amino acids in a mixture, and their effect on growth of callus lines OC 1 and 4aYAC. Aspartic acid, glycine, proline, and arginine were added to callus medium, each at a concentration of 1 mM, 2 mM, or 5 mM. Treatment means of growth indices from the two-way A N O V A for both cell lines decreased as the concentration of the amino acids increased, although the differences were not significant (Table 3). These results did not change when data for the individual cell lines were analyzed.

Vitamin formulations and mixtures The B5 (Gamborg et all 1968), NN (Nitsch & Nitsch 1969), KM (Kao & Michayluk 1975), and APS6 (Provasoli 1968) vitamin formulations, mixtures of those formulations, and various concentrations were examined for their influence on growth of callus cell lines. Results were not statistically significant (data is not shown). However, a mixture o f N N + K M vitamins provided improved growth of callus when compared to other vitamin formulations.

Auxins Several experiments were conducted to assess the role of various plant growth regulators on the growth of T. brevifolia cell lines. In the first experiment, auxins N A A and picloram were compared to the control of 2,4-D at a concentration of 4 . 5 2 / I M and the treatment of 4.52 # M 2,4-D with 0.44/~M BA. The treatment means for the two-way ANOVA from the growth indices of all three cell lines indicated that the addition of BA improved growth of the cell lines, as did the substitution of picloram for 2,4-D (Table 4, Experiment 1). There was no difference when N A A was substituted for 2,4-D in any o f the cell lines.

190

Table 4. Combination and concentration effects of plant growth regulators on growth of T. brevifolia callus cell lines OC1, 4aYAC, and CC2010. Treatment (mg !- 1)

N

Mean growth index

95% C.I.

Experiment 1. Auxins and cytokinin

1 2,4-D 1 2,4-D + 0.1 BA 1 NAA 1 Pie

24 24 24 24

1.28b 1.64a 1.28b 1.47ab

(1.37, (1.73, (1.37, (1.56,

Experiment 2. Auxins with BA

1 2,4,5-T + 0.1 BA

1 IAA + 0.1 BA 1 Pie + 0.1 BA 1 IBA + 0.1 BA

30 30 30 30

3.24a 2.32b 3.23a 2.50b

(3.44, 3.03) (2.51, 2.11) (3.42, 3.02) (2.69, 2.29)

Experiment 3. Cytokinin~

1 2,4-D 1 2,4-D + 0.1 Kinetin 1 2,4-D + 0.1 2ip 1 2,4-D + 0.1 BA

30 30 30 30

2.24b 3.04a 2.81a 2.24b

(2.52, 1.96) (3.31, 2.76) (3.08, 2.53) (2.52, 1.96)

Experiment 4. Plant growth regulator ratios

1 Pic + 0.1 KJnetin 1 Pie + 0.2 Kinetin 1 Pie + 0.4 Kinetin 1 Pie + 1 Kinetin 0.4 Pie + 1 Kinelin 0.2 Pie + 1 Kinetin 0.1 Pie+ 1 Kinetin 0.75 Pie + 0.075 Kinetin

24 24 24 24 24 24 24 24

3.02a 2.84~ 3.14a 1.68c 2.42b 1.77c 1.38e 3.03a

(3.26, 2.76) (3.09, 2.59) (3.39, 2.89) (1.92, 1.42) (2.67, 2.17) (2.02, 1.51) (1.63, 1.12) (3.27, 2.77)

Experiment 5. GA3 and ABA

1 Pic + 0.1 BA 1 Pic + 0.1 BA + 0.1 GA3 1 Pie + 0.1 BA + 1 GA3 1 Pie + 0.1 BA + 0.1 ABA 1 Pie + 0.1 BA + 1 ABA

27 27 27 27 27

3.79c 4.73a 3.98be 4.50ab 3.56e

(3.99, 3.58) (4.93, 4.52) (4.18, 3.78) (4.70, 4.30) (3.76, 3.36)

T h e second series o f plant growth regulator experiments e x a m i n e d the effect o f substituting additional auxins, in c o m b i n a t i o n with 0.44 # M BA, o n callus growth. I n this e x p e r i m e n t picloram and 2,4,5-T were better than I A A a n d I B A for callus growth, b u t n o t significantly different from o n e another (Table 4, Experim e n t 2). Observations o f other cell lines m a i n t a i n e d o n m e d i a c o n t a i n i n g 2,4,5-T indicated that most T. brevifolia cell lines, as well as cell lines o f other Taxus species, grow m u c h worse o n 2,4,5-T than o n picloram (data n o t shown). Cells typically turned b r o w n , had extensive necrotic regions, and grew very slowly if the cells survived at all. A l t h o u g h a n initial shortterm gain in growth was observed, cultures did n o t

1.18) 1.54) 1.18) 1.36)

survive long-term subculture o n m e d i u m c o n t a i n i n g 2,4,5-T. Picloram has b e e n an excellent a u x i n for the routine culture o f Taxus cells, and has b e e n used c o n t i n u o u s l y for m o r e than three years. I n all cases, 4.14 # M picloram provided m u c h better callus growth, in l o n g - t e r m cultures, than the other a u x i n s that we tested.

Cytokinins C y t o k i n i n effects were e x a m i n e d at a c o n c e n t r a t i o n o f 0.1 m g 1-1 in c o m b i n a t i o n with 4.52 # M 2,4-D and the e n h a n c e d v i t a m i n mixture ( N N + K M ) . K i n e t i n a n d 2ip increased callus growth w h e n c o m p a r e d to 2,4-D alone, or 2,4-D with B A (Table 4, E x p e r i m e n t 3). W i t h i n this

191 experiment, 2,4-D + BA did not improve growth when compared to 2,4-D alone. Overall, cytokinin supplementation in combination with auxins is beneficial for growth. Medium supplemented with kinetin supported the best callus growth for all three cell lines tested; however, 2ip also supported good callus growth and should be considered as an alternate cytokinin. The influence of the ratio and concentration of auxin and cytokinin on callus growth was examined in another experiment. Picloram and kinetin were used in varying ratios between concentrations of 0.1 mg 1-1 and 1 mg 1-1 as determined from the previous experiments. For the combined results of all three cell lines, ratios of 1.0:0.1, 1.0:0.2, 1.0:0.4, and 0.75:0.075 (mg 1-l picloram: mg 1-l kinetin) were not different from one another but were better than all other treatments (Table 4, Experiment 4). Our optimum auxin:cytokinin ratio for callus growth is 10:1, although the exact concentration of plant growth regulators warrants further study. Ratios of picloram:kinetin of 1:0.1 and 0.75:0.075 resulted in nearly identical treatment means. These plant growth regulator concentrations are considerably lower than those reported by Fett-Neto et al. (1993). Working with T. cuspidata tissue cultures, they found best growth at a 2,4-D concentration of 8 mg 1-1 and kinetin concentration of 1 mg 1-1.

Gibberellic acid and abscisic acid The final plant growth regulator experiment examined the effect of either gibberellic acid (GA3) or abscisic acid (ABA) on growth of T. brevifolia callus. Either GA3 or ABA was added to the medium at concentrations of 1.0 or 0.1 mg 1-1. Neither 1 mg 1-1 ABA (3.78 #M) or 1 mg 1-1 GA3 (2.89 #M) provided a significant difference in cell growth, compared to the control (Table 4, Experiment 5). However, both GA3 and ABA increased growth of callus when added to medium at a concentration of 0.1 mg 1-1. Fett-Neto et al. (1993) found a similar increase in growth present in T. cuspidata cell lines treated with 0.5 mg 1-1 GA3, although data was not presented indicating that their results were significantly different from other concentrations tested or their GA3-free control.

New media formulations New media formulations, created as a result of the previously describedexperiments, were compared to our original B51CA medium (Table 5). Medium TM5 was

Growth

3-

Index

B51CA

A,

TM5

TM6

TM7

Initial vs. Final Media

Growth Index

B51CA

R.

TM5

TM6

TM7

Initial vs. Final Media

Fig. 3. Comparisonof growth of T. brev!tblia cell lines on previously tested media and latest media formulations. A) Treatment means fromall three cell lines B) Individualcell line growth index means. Bars are the standard error of the means. Completemedium formulations are included in Table5.

better than all other media formulations, while TM6 was worse (Fig. 3A). B51CA and TM7 were not significantly different from one another. Medium B51 CA was the original medium that resulted from our earlier tissue culture studies of T. brevifolia (Gibson et al. 1993). TM5 is a medium that incorporated most of the positive results obtained in this series of experiments. TM6 medium included 0.2% casein hydrolysate, and was tested as a 'more nutritionally complete' version of TM5 medium. Two of our new media formulations, TM5 and TM7, contain a mixture of proline, glycine, arginine and aspartic acid, each at 1 mM concentration. TM7 medium was similar to TM5 except that the combination of KM + NN vitamins in TM5 was substituted with a vitamin mixture that resembles a modified Nitsch & Nitsch (1969) formulation. The vitamins for TM7 were selected from the vitamins that seemed to have the greatest influence on growth, as determined from comparative studies of all of our vitamin experiments. Casein hydrolysate did not make a difference in growth in our first experiment and seemed to enhance growth in a later experiment. However, both of our

192 Table 5. Composition of B51 CA compared to that of media developed as a result of optimization experiments.

B51CA

mgl - l

B5 Major & Trace Salts EDTA-Fe-Na Salt 2,4-D

Normal* B5 Major & Trace Salts Normal* 50 EDTA-Fe-Na Salt 50 1 Picloram 1 Kinetin 0.1 GA3 0.1 ABA 0.1 Normal* NN Vitamins Normal* KM Vitamins Normal*

B5 Vitamins

Casein Hydrolysate (2%) 2000

Sucrose(1%)

10000

Agar (1%) pH

10000 5.5

TM5

Aspartic Acid (1 raM) Ar~nine (1 raM) Glycine (1 mM)

Proline(1 raM) Sucrose(0.5%) Fructose(0.5%) Agar (1%) pn

mg1-1

133 175 75 115 5000 5000 10000 5.5

TM6

mg1-1

B5 Major & Trace Salts Normal* EDTA-Fe-Na Salt 50 Picloram 1 Kinetin 0.1 GA3 0.1 ABA 0.1 NN Vitamins Normal* B5 Vitamins Normal*

Casein Hydrolysate

2000

Sucrose(0.5%) Fructose(0.5%) Agar (1%) pH

5000 5000 10000 5.5

TM7

mg1-1

B5 Major & Trace Salts EDTA-Fe-Na Salt Picloram Kinetin GA3 ABA d-Biotin Folic Acid Myo-lnositol Nicotinic Acid Pyridoxine HCI Thiamine HC1 Aspartic Acid (1 raM) Arginine (1 raM) Glycine (1 raM) Proline (1 raM) Sucrose (0.5%) Fructose (0.5%) Agar (1%) pH

Normal* 50 1 0.1 0.1 0.1 0.05 0.5 150 5 1 0.05 133 175 75 115 5000 5000 10000 5.5

* 'Normal' refers to concentration of component as it appeared in the original reference (in text).

media formulations that contained casein hydrolysate, B51CA and TM6, resulted in worse callus growth than media with just individual amino acids added, TM5 and TM7. Hydrolyzed casein contains tryptophan which repressed growth in these callus cell lines. Casein hydrolysate, therefore, has been removed from our final medium and those amino acids giving enhanced growth have been substituted for a defined organic nitrogen source. Tests of our previously published medium (B51CA) revealed almost identical treatment means both at the beginning and end of this series of experiments. Some loss of vigor was noted in cell line OC1 during the later stages of this study. For cell lines CC2010 and 4aYAC, the final TM5 medium provided the best callus growth for both of these cell lines (Fig. 3B). This medium has improved growth in cell lines CC2010 and 4aYAC to give growth indices of 5.0 and 5.2 respectively, or cell doubling times of 5.6 and 5.3 days. The growth indices on this medium represent a 4.7-fold increase over our previous report for cell line 4aYAC, and a 7.1-fold increase for cell line CC2010(Gibsonetal. 1993). The best results for OC1, before a loss in vigor, was a growth index of 8.0 or a

cell doubling time of 3.5 days. This represents a 4.0fold increase over our previous report (Gibson et al. 1993). Wickremesinhe & Arteca (1993b) suggest that a fresh weight doubling time of 7 days in cultures of C. harringtonia is rapid for woody species. The most rapid growth that we observed in OC1 was twice that observed by Wickremesinhe & Arteca (1993b). FettNeto et al. (1992) reported a 2.8-fold increase in fresh weight after 25 days of growth in one T. cuspidata cell suspension, a doubling time of 8.9 days. With the cell lines in this study, as well as other cell lines currently in culture, TM5 is a significantly improved medium that may provide comparable increases in biomass accumulation for other Taxus cultures.

Conclusions

When evaluating plant cell culture as a means to produce taxol, both growth and taxane production must be optimized. Our goal was to develop a tissue culture medium that would increase the growth of Taxus cell lines. The results of these experiments have led to a medium optimized for growth of T. brevifolia callus.

193 We are presently using TM5 medium for the routine culture of over 300 genetically distinct cell lines representing 6 different Taxus species, including T. brevifolia, Z x media, Z cuspidata, Z brevicata, Z baccata L., and T. canadensis. This medium has been used to maintain culutures of these species for over three years for some cell lines. While the media presented here are by no means optimal for all Taxus cell lines, they should be, at the very least, a good starting point for the tissue culture of most species of Taxus. The same variables that have led to significant increases in growth of cell cultures of Taxus species are currently being studied for their effect on production of taxol and other taxanes.

Acknowledgements This work was partially supported by the National Cancer Institute, grant #CA55138-02. We wish to thank Judy Luong and Thanh 'Tom' Do for their excellent technical assistance, Dr. Tom Hirasuna for reviewing the manuscript and for thoughtful and insightful discussions, and Christopher David Ketchum for inspiration and motivation.

Notes 1 Mention of a trademark, proprietary product, or vendor does not constitute a guarantee or warranty of the product by the U.S. Department of Agriculture and does not imply its approval to the exclusion of other products or vendors that may also be suitable.

References Anderson WC (1978) Rooting of tissue cultured rhododendrons. Proc. Int. Plant Prop. Soc. 28:135-139 Christen AA, Bland J & Gibson DM (1989) Cell culture as a means to produce taxol. Proc. Am. Assoc. Cancer Res. 30:566 Christen AA, Gibson DM, & Bland J (1991) Production of taxol or taxol-like compounds in cell culture. US Patent 5,019,504 Cragg GM & Shader KM (1991) Taxol: the supply issue. Cancer Cells 3:233-235 Denis J-N, Greene AE, Guenard D, Gueritte-Voegelein F, Mangatal L & Potier P (1988) A higMy efficient, practical approach to natural taxol. J. Amer. Chem. Soe. 110:5917-5919 Fett-Neto AG, DiCosmo F, Reynolds WF & Sakata K (1992) Cell culture of Taxus as a source of the neoplasticdrug taxol and relatedtaxanes.Biotechnology 10:1572-1575 Fett-Neto AG, Melanson SJ, Sakata K & DiCosmo F (1993) Improved growth and tax'olyield in developing calliof Taxus cusp/data by medium composition.Biotechnology II: 731-734

Fett-Neto AG, Zhang W Y & DiCosmo F (1994) Kinetics of taxol production, growth, and nutrientuptake in cell suspensions of Taxus cuspidata. Biotechnol.and Bioengineer.44:205-2 I0 Hores H E & SgrignoliPJ 0991) In vitro culture and precocious germination of Taxus embryos. In Vitro Cell. Dev. Biol. 27P: 139-142 Flores T, Wagner LJ & Flores HE (1993) Embryo culture and taxane production in Taxus spp. In Vitro Cell. Dev. Biol. 29P: 160-165 Gamborg OL, Miller RA & Ojima K (1968) Nutrient requirements of suspension cultures of soybean root cells. Exp. Cell Res. 50: 148-151 Gibson DM, Ketchum REB, Vance NC & Christen AA (1993) Initiation and growth of cell lines of Taxus brevifolia (Pacific Yew). Plant Cell Rep. 12:479--482 GresshoffPM & Doy CH (1972) Development and differentiation of haploid Lycopersicon esculentum (tomato). Planta 107:161-170 Gupta P & Durzan D (1985) Shoot multiplication from mature trees of Douglas fir (Pseudotsuga menziesii) and Sugar pine (Pinus lambertiana). Plant Cell Rep. 2:177-179 Holton RA, Liu JH, Gentile LN & Beidiger RJ (1992) Semi-synthesis of taxol. Proceedings of the Second National Cancer Institute Workshop on Taxol and Taxus. Holton RA, Somoza C, Kim HB, Liang F, Biediger RJ, Boatman PD, Shindo M, Smith CC, Kim Set al. (1994) First total synthesis of taxol. 1. Functionalization of the B ring. JACS 116:1597-1598 Kao KN & Michayluk MR (1975) Nutritional requirements for growth of Vicia hajastana cells and protoplasts plated at a very low population density in liquid medium. Planta 126:105-110 Kingston, DGI (1991) The chemistry of taxol. Pharmac. Ther. 52: 1-34

Litvay JD, Verma DC & Johnson MA (1985) Influence ofa loblolly pine (Pinus taeda) culture medium and its components on growth and somatic embryogenesis of the wild carrot (Daucus carota). Plant Cell Rep. 4:325-328 Murashige T & Skoog F (1962) A revised medium for rapid growth and bioassays with tissue cultures. Physiol. Plant 15:473-497 Nicolaou KC, Yang Z, Liu JJ, Ueno H., Nantermet PG, Guy RK, Claiborne CF, Renaud J, Couladouros EA., Paulvannan K & Sorensen F.J (1994) Total synthesis of taxol. Nature 367: 630634 Nitsch JP & Nitsch C (1969) Haploid plants from pollen grains. Science 163:85-87 Provasoli L (1968) Media and prospects for the cultivation of marine algae In: Watanabe A & Hattori A (F_As)Cultures and Collections of Algae, Proc US-Japan Conf., Hakone, 1966, pp. 63-75, Tokyo. Jap. Soc. Plant Physiol US Forest Service (1993) Pacific Yew Draft Environmental Statement. US Department of Agriculture Westgate PJ, Emery AH, Hasegawa PM & Heinstein PF (1991) Growth of Cephalotaxus harringtonia plant cell cultures. Appl. Microbiol. Biotechnol. 34:798-803 White P (1934) Potentially unlimited growth of excised tomato root tips in a liquid medium. Plant Physiol. 9:585-600 Wickremesinhe ER & Arteca RN (1993a) Taxus callus cultures: initiation, growth optimization, characterization, & taxoi production. Plant Cell, Tissue, & Organ Culture 35:181-193 Wickremesinlae ER & Arteca RN (1993b) Establishment of fastgrowing callus & root cultures of Cephalotaxus harringtonia. Plant Cell Rep. 12:80-83