Effects of Wounding on Taxol and Baccatin III Levels in Taxus media cv Hicksii

TWO experiments were carried out to test the hypothesis that herbivory induces higher levels of taxanes in the bark of two year old Taxus media cv Hicksii cuttings. For the first experiment bark was cut 1.0 mm deep every 5.0 mm in two groups of cuttings; with 0.1% 2,4-D applied over the wounded bark in one group. Both groups and a control were harvested one week later. In the second experiment bark was wounded (as above) and treated cuttings with their respective control groups analyzed one and three weeks after treatment. Bark extracts were analyzed by high performance liquid chromatography. Cutting significantly decreased taxol levels in the second, but not the first, experiment. Cutting tended to increase baccatin III in both experiments, though not significantly. Application of 2,4-D decreased baccatin III, but did not significantly alter the taxol concentrations. These results suggest that taxol and baccatin III are not induced after wounding Taxus bark, or that the time course of induction is shorter than three weeks. OHIO J. SCI. 96 (3): 41-44, 1996 INTRODUCTION Many secondary compounds produced by plants are inducible, that is their concentration in all or part of the plant increases in response to an external stimulus such as herbivory (Lewison et al. 1991a, Lewison et al. 1991b). However, it is not known whether the medically important taxanes are inducible in Taxusspp. (yews). These taxanes include taxol, which is approximately 30% effective in ovarian cancer patients (Stone 1993) and baccatin III, which is used in the hemi-synthesis of taxol (Bombardelli et al. 1992). We hypothesized that taxanes are inducible defenses against herbivory or pathogens in yews, in part because other diterpenes have been shown to deter herbivory (Robinson 1991) and to be inducible in conifers (Rosenthal and Berenbaum 199DTaxus spp. are subject to little insect herbivory, perhaps due to deterrence by taxanes (Perusse et al. 1977). We further hypothesized that such induction can result from mechanical injury of the bark. This hypothesis is based on studies that showed terpene concentration increased after wounding in diverse taxa (Kahl 1978). Furthermore, the activity of monoterpene cyclase (an enzyme responsible for the formation of terpenes) increased seven days after mechanical wounding in three species of Abies and one species of Picea (Lewison et al. 1991a). This is important because Lewison et al.(1991b) found a direct correlation between the amount of monoterpene cyclase activity and monoterpene content. We also hypothesized that taxane production can be increased through mechanical wounding of the bark coupled with the application of 2,4-dichlorophenoxyacetic acid (2,4-D). Fahn and Zamski (1970) demonstrated that 2,4-D in lanolin was effective at inducing resin flow and Manuscript received 18 August 1995 and in revised form 5 June 1996 (#95-15). Present Address: Department of Environmental and Plant Biology, Ohio University, Athens, OH 45701 increasing the number of vertical resin ducts and tracheids in P. halepensis. Although Taxus spp. lack normal and traumatic resin canals and vertical tracheids where resins are stored (Brown and Panshin 1940), it still may be possible that 2,4-D may induce proliferation of taxane-producing cells. MATERIALS AND METHODS Plant Material Taxus media cv Hicksii (Nuttal) was used for both experiments because it was shown to have similar or higher amounts of taxol than T. brevifolia (Vidensek et al. 1990). Cuttings which had been rooted for two years were donated by Kern Nursery (Liberty Township, OH; Butler Co.). Cuttings for both experiments were about 35 cm tall and moved to a greenhouse 12 weeks prior to treatment. The cuttings for the 1993 experiment were moved to a greenhouse on March 22, treated June 8, and harvested June 15. The mean low and high temperatures were 15° and 32° C, respectively, for the 1993 experiment. The rooted cuttings for the 1994 experiment were moved to the greenhouse January 13, treated April 7, and harvested April 14 and 28. The mean low and high temperatures for 1994 were 13° and 24° C. Plants were fertilized with Peter's Solution® (Milpitas, CA) (N20 P10 K20) as needed (up to two times daily) with a natural photoperiod. Voucher specimens were deposited in the Willard Sherman Turrell Herbarium (MU) of Miami University. Treatments Experiment One (1993) One hundred seventeen rooted cuttings were assigned to three treatment groups using a random number table. Unfortunately, due to centrifuge tubes cracking, only 13 control, 16 cut, and 34 cut and 2,4-D treated samples were analyzed. For treatment one, a 5.0 mm horizontal cut through the bark was made every 5.0 mm from the soil level to where the bark was approximately half EFFECTS OF WOUNDING TAXUS VOL. 96 woody. Each cut was approximately 25% around the stem. Wounding was done with a new razor blade for each plant, which was sanitized with 99-9+% methanol. In treatment two, the bark was cut into as above and a 0.1% concentration of 2,4-D in lanolin applied to the trunk with a brush. Treatment three, the control group, was not cut with a razor blade, or treated with 2,4-D. The trees were grown for one week after treatment as this time interval was shown to be sufficient for the amount of monoterpene cyclase activity to increase 5-15 fold (Lewison et al. 1991b), then cut off at soil level and immediately stripped of their bark. Experiment Two (1994) In 1994, 120 rooted cuttings were randomly assigned to four groups. Trees in two groups were cut with a razor blade as in the 1993 experiment; one group was allowed to grow for one week after treatment; the second, three weeks. The other two groups were controls that were harvested synchronously with their respective treatment group. No 2,4-D was used in any of these treatments. Stems were cut and stripped of their bark as in 1993. Extraction Procedure Bark samples were dried at 23° C then ground in liquid nitrogen with a mortar and pestle. Taxanes were extracted in methanol by sonication for 30 minutes followed by centrifugation for 10 minutes. Each bark sample was extracted three times; the three extracts were pooled and concentrated by evaporating off methanol with nitrogen while heating to 40° C. The residue in each of the centrifuge tubes was dissolved in 1.5 ml methanol:water (30:70) and purified through PrepsepTM C-18 columns under the vacuum of a 60 cc syringe. This procedure followed Auriola et al. (1992). Analysis The High Performance Liquid Chromatography (HPLC) conditions followed Auriola et al. (1992). The HPLC system used to analyze the samples included a Scientific Systems Inc. model 300 LC pump and a LoPulse® model LP-21 pulse dampener used to propel the solvent (mobile phase) at a steady rate of 0.2 ml/min. A Rheodyne® injector equipped with a 20 uX sample loop delivered the sample to the column. A Phenomenex® Curosil-G-6Li model 03A-3121-G0 guard column (30 X 4.6 mm) and model 00G-3121-R0 analytical column (250 X 3-2 mm) were employed to separate the taxanes. The first experiment used a Waters® 490 E programmable multiwavelength detector to measure amounts of taxanes, while the second experiment employed a LDC® SpectroMonitor III model 1204 A detector. Both detectors were set at 228 nm. Both experiments employed a Linear® 1200 strip chart recorder, and the second experiment also employed a Shimadzu® Chromatopac C-R6A electronic integrator to automatically measure peak area. Statistical analysis was accomplished using SAS (1990). More specific methods are provided by Egan (1994). RESULTS Experiment One (1993) Taxol levels did not differ among the treatments (Table 1) . A one way ANOVA demonstrated a significant effect of treatment on baccatin III concentrations (Fig. 1, Table 1). Based on Bonferroni multiple comparison procedure, the only significant pairwise difference was between the cut treatment and the treatment in which the bark was cut then treated with 2, 4-D. There was a 45% increase in baccatin III for the cut treatment in relation to the control group, and a 56% decrease in the cut, 2,4-D treatment compared to the control group.