Regio- and stereo-selective biotransformation of 2α,5α,10β,14β-tetra-acetoxy-4(20), 11-taxadiene by Ginkgo cell suspension cultures

Article abstract of DOI:10.1016/S0040-4020(02)00529-X

Ginkgo biloba cell suspension cultures were used to bioconvert sinenxan A, 2α,5α,10β,14β-tetra-acetoxy-4(20), 11-taxadiene, a taxoid isolated from callus tissue cultures of Taxus spp. Besides two major products, 9α-hydroxy-2α,5α,10β,14β-tetra-acetoxy-4(20

Full text of DOI:10.1016/S0040-4020(02)00529-X

TETRAHEDRON
Pergamon
Tetrahedron 58 .2002) 5659±5668
Regio- and stereo-selective biotransformation of
2a,5a,10b,14b-tetra-acetoxy-4ꢀ20), 11-taxadiene by Ginkgo cell
suspension cultures
Jungui Dai,a Min Ye,a Hongzhu Guo,a Weihua Zhu,b Dayong Zhang,b Qiu Hu,b Junhua Zhenga
and Dean Guoa,p
aThe State Key Laboratory of Natural and Biomimetic Drugs, School of Pharmaceutical Sciences, Peking University, Xueyuan Road #38,
Beijing 100083, People's Republic of China
bInstitute of Materia Medica, Chinese Academy of Medical Sciences and Peking Union Medical College, Xiannong Tan Street #1,
Beijing 100050, People's Republic of China
Received 6 December 2001;revised 7 May 2002;accepted 23 May 2002
AbstractÐGinkgo biloba cell suspension cultures were used to bioconvert sinenxan A, 2a,5a,10b,14b-tetra-acetoxy-4.20), 11-taxadiene, a
taxoid isolated from callus tissue cultures of Taxus spp. Besides two major products, 9a-hydroxy-2a,5a,10b,14b-tetra-acetoxy-4.20),
11-taxadiene 1 and 9a,10b-dihydroxy-2a,5a,14b- triacetoxy-4.20), 11-taxadiene 2, additional six minor products were obtained and ®ve
of them identi®ed as new compounds. On the basis of chemical and spectral data, their structures were identi®ed as 9a,14b-dihydroxy-
2a,5a,10b-triacetoxy-4.20), 11-taxadiene 3, 6a,10b-dihydroxy-2a,5a,14b-triacetoxy-4.20), 11-taxadiene 4, 6a,9a,10b-trihydroxy-
2a,5a,14b-triacetoxy-4.20), 11-taxadiene 5, 9a,10b-O-.propane-2,2-diyl)-2a,5a,14b-triacetoxy-4.20), 11-taxadiene 6, 9a-hydroxy-
2a,5a,10b,14b-tetra-acetoxy-4.20), 11-taxadiene formate 7, 10b-hydroxy-2a,5a,9a,14b-tetra-acetoxy-4.20), 11-taxadiene formate 8,
respectively. Investigation of the properties of the enzymes responsible for the biocatalysis process of sinenxan A to 1 and 2 revealed
that the enzymes were extracellular and constitutive. Using sinenxan A and the two major products .1 and 2) as indicators, the stage and
concentration of sinenxan A added and the kinetics of the biotransformation reaction were investigated. The results showed that: .1) the
optimal stage for sinenxan A addition was the logarithmic phase of the cell growth period, in which sinenxan A was almost completely
bioconverted, and the biotransformation rates were up to 60 and 20% for 1 and 2, respectively;.2) the optimal concentration of sinenxan A
added was 60 mg/L;.3) the substrate was mainly converted into 1 and 2 in the ®rst 48 h after addition and then into the minor products.
q 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
brevifolia, which yielded low amounts of paclitaxel
.approximately 0.01% of the dry weight). These slow grow-
ing, woody gymnosperms were harvested in large amounts
to provide paclitaxel for clinical trials: approximately
10,000 kg of dry weight of bark are required to yield
1.0 kg of paclitaxel, in other words, more than 3600 trees
must be cut down for the treatment of only 500 patients.10
This situation illustrates a serious resource crisis that has to
be alleviated by using various approaches to produce alter-
nate sources of paclitaxel. Alternative approaches for
production of paclitaxel by cell suspension cultures and
by semi-synthetic process are being evaluated by various
groups.11 Bristol-Myers Squibb reported alternative
methods for the production of paclitaxel, a semisynthetic
approach and the application of biocatalysis in enabling
the semisynthesis of paclitaxel. Three novel enzymes were
discovered that converted a variety of taxanes to a single
molecule, 10-deacetyl baccatin III, a precursor for paclitaxel
semisynthesis. The concentration of 10-deacetyl baccatin III
was increased 5.5±24 fold in the extracts treated with the
enzymes, depending upon the type of Taxus cultivars used,
and the dif®cult separation of a group of closely related
The diterpenoid paclitaxel .Taxolw, Scheme 1), originally
isolated from the Paci®c yew .Taxus brevifolia Nutt.) in
1971,1 exhibited remarkably high cytotoxicity and strong
antitumor activity against different tumors resistant to exist-
ing anticancer drugs.2 It has been approved for the treatment
of advanced ovarian and breast cancer,3,4 and it is currently
in clinical trials for treatment of lung, skin, head and neck
cancers with encouraging results.5 Since the discovery of
paclitaxel in the late 1960s, its unique chemical structure,
signi®cant biological activity, as well as its novel mechan-
ism of action6,7 have led to research by scientists from
different ®elds.8,9
Initially the development of paclitaxel was a serious
problem since its only approved source was the bark of T.
Keywords: taxane;sinenxan A;biotransformation;enzymes;cell suspen-
sion cultures; Ginkgo biloba.
Corresponding author. Tel.: 186-10-62091516;fax: 186-10-62092700;
p
e-mail: gda@mail.bjmu.edu.cn
0040±4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020.02)00529-X

5660
J. Dai et al. / Tetrahedron 58 52002) 5659±5668
Scheme 1.
molecules can be avoided. Biocatalytic processes have also
been reported for the preparation of C-13 paclitaxel side
chain synthons.11
its ef®cacy against the resistant human breast cancer cells
when TRAs are co-administered at 1.0 mM.15
Sinenxan A, 2a,5a,10b,14b-tetra-acetoxy-4.20), 11-taxa-
diene, is a taxoid isolated from the callus cultures of
Taxus spp. in high yields .ca. 1±2% of dry weight).16,17
The rich resources and its taxane-skeleton give it valuable
potential for the semisynthesis of paclitaxel or other struc-
turally related bioactive compounds, such as `second-
generation' taxoid anticancer agents and TRAs mentioned
above. A number of studies on its structural modi®cation by
chemical and biocatalytic approaches were reported.18±21
Recently, we also reported its highly regio- and stereo-
selective hydroxylation at C-9 and deacetylation at C-10
by Ginkgo cell suspension cultures to form two new major
products.22 This paper describes the formation of six addi-
tional products. Detailed results concerning the timing and
amount of substrate addition, substrate concentration,
kinetics of the biotransformation reaction and the basic
properties of the enzymes responsible for the biotransforma-
tion of the two major products are also provided.
Recently, attention was also paid to the rapid emergence of
multidrug resistant .MDR) cancer cells in the chemotherapy
of various types of advanced solid tumor, which is mainly
caused by the overexpression of P-glycoprotein .pgp), a
cell-membrane transporter for a variety of hydrophobic
substrates during the course of the clinical use of vincristine,
paclitaxel and etoposide which are very effective hydro-
phobic antitumor drugs for cancer chemotherapy.12 The
overexpression results in decreased accumulation of the
drug within the cancer cells, since the cell can ef®ciently
pump out the drug molecules.13,14 Ojima and co-workers
have discovered and developed the `second-generation'
taxoid anticancer agents-paclitaxel congeners that would
not be recognized by pgp and would still display potent
anticancer activity. Also they discovered taxane-based
MDR reversal agents .TRAs, such as SB-RA-131012, SB-
RA-4001, etc. Scheme 1). Paclitaxel recovers 95±99.8% of
Scheme 2. The biotransformation of sinenxan A by Ginkgo cell suspension cultures.

J. Dai et al. / Tetrahedron 58 52002) 5659±5668
Table 1. H NMR spectral data of compounds 3, 5±8 .CD3Cl, 500 MHz; dH mult., J, Hz)
5661
1
Proton
3
5
6
7
8
1
2
3
2.00 .brs)
5.41 .dd, 6.0, 2.0)
2.88 .d, 6.0)
1.90 .d, 2.0)
5.37 .dd, 2.5, 6.0)
2.96 .d, 6.0)
1.91 .d, 1.5)
5.44 .dd, 6.0, 2.0)
2.82 .d, 5.5)
1.94 .d, 2.5)
5.42 .dd, 2.5, 6.5)
3.00 .d, 6.5)
1.95 .d, 2.5)
5.42 .dd, 2.5, 6.5)
2.97 .d, 6.5)
4
5
6
7
5.28 .brs)
1.81 .m)
1.54 .m), 1.77 .m)
5.08 .brs)
3.93 .m)
1.62 .m), 2.00 .m)
5.32 .brs)
1.83 .m), 1.75 .m)
1.73 .m), 1.61 .m)
5.31 .brs)
1.86 .m)
1.76 .m)
5.31 .brs)
1.86 .m)
1.71 .m)
8
9
4.22 .d, 10.0)
5.82 .d, 10.0)
4.01 .d, 9.5)
4.75 .d, 9.5)
4.19 .d, 9.0)
4.83 .d, 9.5)
5.88 .d, 10.5)
6.03 .d, 10.5)
5.85 .d, 10.5)
6.14 .d, 10.5)
10
11
12
13
2.67 .dd, 19.0, 9.0), 2.53
.dd, 5.0, 19.0)
4.09 .dd, 9.0, 5.0)
2.82 .dd, 18.6, 8.6), 2.44
.dd, 5.1, 18.6)
4.98 .dd, 9.3, 5.1)
2.82 .dd, 9.5, 19.0), 2.50
.dd, 4.5, 19.0)
4.96 .dd, 4.5, 9.0)
2.85 .dd, 9.0, 19.5),
2.45 .dd, 4.0, 19.5)
4.98 .dd, 4.5, 9.0)
2.86 .dd, 9.5, 19.0),
2.45 .dd, 4.0, 19.5)
4.98 .dd, 4.5, 9.0)
14
15
16
17
1.59 .s)
1.17 .s)
1.64 .s)
1.24 .s)
1.70 .s)
1.21 .s)
1.73 .s)
1.13 .s)
1.72 .s)
1.13 .s)
18
19
2.11 .s)
1.04 .s)
2.20 .s)
1.20 .s)
2.17 .s)
0.98 .s)
2.15 .s)
0.89 .s)
2.15 .s)
0.86 .s)
20
OCOCH3
OCHO
5.31 .brs), 4.95 .brs)
2.11 .s), 2.08 .s), 2.15 .s)
5.01 .s), 5.46 .s)
2.02 .s), 1.96 .s), 2.06 .s)
5.31 .brs), 4.89 .brs)
2.05 .s), 2.02 .s), 2.01 .s)
5.34 .brs), 4.90 .brs)
2.18, 2.05, 2.01, 2.01
8.12 .s)
5.34 .brs), 4.89 .brs)
2.18, 2.05, 2.01, 2.01
8.05 .s)
1.47 .s), 1.41 .s)
2. Results
Compound 4 was identi®ed as a known structure. Its NMR
spectral data, IR spectral data and optical rotation were in
good agreement with those of 6a,10b-dihydroxy-2a,5a,
14b-triacetoxy-4.20), 11-taxadiene reported by Hu et al.21
Sinenxan A was ef®ciently bioconverted by Ginkgo cell
suspension cultures. The substrate was administered to 15-
day-old cell cultures and eight products were isolated by
chromatographic methods after an additional 6 days of incu-
bation. On the basis of the spectral and chemical data, their
structures were identi®ed as 9a-hydroxy-2a,5a,10b,14b-
tetra-acetoxy-4.20), 11-taxadiene 1, 9a,10b-dihydroxy-
2a,5a,14b-triacetoxy-4.20), 11-taxadiene 2, 9a,14b-dihy-
The FAB mass spectrum of 5 showed a quasi molecular ion
peak [M1Na]1 at m/z 517, consistent with the molecular
formula of C26H38O9. The 1H NMR spectrum of 5 was simi-
lar to that of 4, except that signals corresponding to H-9a .d
1.67, dd, Jˆ5.5, 15.0 Hz) or H-9b .d 2.39 m) in 4 had
disappeared, while an oxygen-bearing methine signal was
observed at d 4.01 .d, Jˆ9.5 Hz), suggesting the existence
of a hydroxyl group at the C-9 position. As a result, the
signal of H-10a .d 5.06, dd, Jˆ6.0, 12.0 Hz) shifted up®eld
.d 4.75, d, Jˆ9.5 Hz). This was con®rmed by the signal of
droxy-2a,5a,10b-triacetoxy-4.20),
11-taxadiene
3,
6a,10b-dihydroxy-2a,5a,14b-triacetoxy-4.20), 11-taxa-
diene 4, 6a,9a,10b-trihydroxy-2a,5a,14b-triacetoxy-
4.20), 11-taxadiene 5, 9a,10b-O-.propane-2,2-diyl)-
2a,5a,14b-triacetoxy-4.20), 11-taxadiene 6, 9a-hydroxy-
2a,5a,10b,14b-tetra-acetoxy-4.20), 11-taxadiene formate
7 and 10b-hydroxy-2a,5a,9a,14b-tetra-acetoxy-4.20), 11-
taxadiene formate 8, respectively. Among them, the two
major products .1 and 2) were previously reported. Their
yields .measured by HPLC) were determined to be 70 and
12%, respectively. The other six minor products except 421
were new taxanes obtained by this approach in low yields.
The biotransformation reaction is illustrated in Scheme 2.
C-9 signi®cantly shifting to lower ®eld at d 79.3 in the 13
C
NMR spectrum. Therefore the structure of 5 was identi®ed
as 6a,9a,10b-trihydroxy-2a,5a,14b-triacetoxy-4.20), 11-
1
taxadiene, and all the H and 13C NMR signals of 5 were
assigned according to its 1H±1H COSY and HMQC spectra
.see Tables 1 and 2).
The FAB mass spectrum of 6 showed a quasi molecular ion
peak [M1Na]1 at m/z 541, consistent with the molecular
formula of C29H42O8. The 1H NMR spectrum of 6 was simi-
lar to that of 1, but one additional methyl group signal
occurred. The 13C NMR spectrum of 6 was also similar to
that of 1, except that an additional quaternary carbon signal
appeared at d 107.0. According to the literature,23±25 an
acetonide bridge might be introduced at the C-9 and C-10
position. This deduction was supported by 13C NMR, 1H±1H
The FAB mass spectrum of 3 showed a quasi molecular ion
peak [M1Na]1 at m/z 501, consistent with the molecular
formula of C26H38O8. The 1H NMR spectrum of 3 was simi-
lar to that of 1 except that the signal of H-14a .d 4.97, dd,
Jˆ4.5, 9.5 Hz) had disappeared, and one oxygen-bearing
methine signal appeared at d 4.09 .dd, Jˆ5.0, 9.0 Hz),
suggesting the presence of a hydroxyl group rather than
an acetoxy group at C-14. This conclusion was supported
by the appearance of a carbon resonance at d 67.5 instead of
d 70.0 in its 13C NMR spectrum. Therefore, compound 3
was elucidated to be the 14-deacetyl derivative of
COSY, HMQC and HMBC spectra of 6. All the 1H and 13
C
NMR signals of 6 .see Tables 1 and 2) were assigned
according to its DEPT, 1H±1H COSY, HMQC and
HMBC spectra. Interestingly, taxane with an acetonide
bridge has not been previously found in Taxus spp. This
type of reaction has never been reported in the ®eld of
compound 1. The H and 13C NMR spectral data of 3
were summarized in Tables 1 and 2.
1

5662
J. Dai et al. / Tetrahedron 58 52002) 5659±5668
Table 2. 13C NMR spectral data of compounds 3, 5±8 .CD3Cl, 125 MHz; dc)
Carbon
3
5
6
7
8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
63.2 d
70.6 d
44.0 d
142.3s
79.0 d
28.4 t
26.4 t
44.7 s
76.0 d
76.2 d
137.5 s
133.3 s
42.3 t
67.5 d
37.5 s
58.7 d
69.8 d
43.6 d
138.2s
82.0 d
69.8 t
34.2 t
44.1 s
79.3 d
72.1 d
136.7 s
134.8 s
37.4 t
70.2 d
39.5 s
31.8 q
26.0 q
21.6 q
21.6 q
121.1 t
59.4 d
69.9 d
42.9 d
141.2s
78.7 d
28.4 t
26.3 t
40.7 s
82.0 d
74.4 d
138.1 s
133.6 s
40.0 t
70.5 d
37.7 s
24.9 q
31.9 q
21.7 q
17.3 q
117.3 t
58.1 d
69.3 d
43.7 d
140.9s
77.7 d
27.7 t
26.7 t
43.8 s
77.0 d
71.6 d
137.2 s
132.0 s
39.0 t
69.6 d
36.5 s
25.4 q
31.0 q
20.4 q
16.9 q
17.7 t
58.1 d
69.3 d
43.7 d
140.9s
77.7 d
27.7 t
26.7 t
43.8 s
77.0 d
71.6 d
137.2 s
132.0 s
39.0 t
69.6 d
36.5 s
25.2 q
30.9 q
20.4 q
16.8 q
17.6 t
25.6 q
31.2 q
21.9 q
17.4 q
117.4 t
OCOCH3
170.5 s, 169.7
s, 169.6 s,
21.1 q, 21.3 q,
21.5 q
169.8 s, 169.4
s, 170.0 s,
20.8 q, 21.3 q,
21.3 q
169.9 s, 169.9 s, 169.7
s, 20.6 q, 21.3 q, 21.5 q
169.3 s, 169.2 s, 169.1 s,
169.0 s, 20.4 q, 20.7 q, 20.7 q,
21.2 q
169.3 s, 169.2 s, 169.1 s,
169.0 s, 20.4 q, 20.7 q, 20.7 q,
21.2 q
OCHO
159.4 d
159.4 d
107.0 s, 27.2 q, 26.8 q
biotransformations. However, it often occurs in chemical
synthesis to protect o-dihydroxy with acetone under acidic
conditions. It was therefore suggested that 6 might not be an
enzymatic product, since acetone was introduced in the
course of isolation.
showed the NOE relation between the proton signal at d
8.12 and the C-19 methyl group, so the formate group
should be attached to C-9 rather than C-10. Therefore, 7
was identi®ed as the C-9 formate derivative of 1, and all
the 1H and 13C NMR signals of 7 were assigned according to
its DEPT, 1H±1H COSY, HMQC and HMBC spectral data.
.see Tables 1 and 2).
The FAB mass spectrum of 7 showed a quasi molecular ion
peak [M1Na]1 at m/z 571, consistent with the molecular
formula of C29H40O10. The H NMR spectrum of 7 was
1
The FAB mass spectrum of 8 showed a quasi molecular ion
peak [M1Na]1 at m/z 571, consistent with the molecular
formula of C29H40O10. The H NMR spectrum of 8 was
similar to that of 1, except that one proton signal appeared
at lower ®eld .d 8.12) and the signals corresponding to H-9
and H-10 both shifted to lower ®eld at d 5.88 .d, Jˆ10.5 Hz)
and d 6.03 .d, Jˆ10.5 Hz), respectively. The 13C NMR
spectrum showed the appearance of a formate group signal,
which was determined to be a quaternary carbon by a DEPT
experiment at d 159.4. All of the above analysis suggested
the presence of a formate group at C-9 or C-10 position. The
HMBC .Scheme 3) spectrum showed that the proton signal
at d 8.12 was correlated to C-9, and NOESY .Scheme 3)
1
similar to that of 7, except that the signal corresponding to
the formate group shifted up®eld to d 8.05, and the signal
corresponding to formate methine proton shifted to lower
®eld at d 6.14, suggesting that the formate group was
present at C-10 position rather than at C-9 position. This
conclusion was supported by the HMBC and NOESY
experiments, since the proton signal at d 8.05 corresponding
to the formate group correlated to C-10 in the HMBC
Scheme 3. HMBC and NOESY correlations of compound 7.

J. Dai et al. / Tetrahedron 58 52002) 5659±5668
5663
Scheme 4. HMBC and NOESY correlations of compound 8.
spectrum, and no NOE relation between the formate and
C-19 methyl group was observed by NOESY. Therefore,
8 was identi®ed as the 10-formate derivative of 1, and all
the 1H and 13C NMR signals of 8 were assigned according to
its DEPT, 1H±1H COSY, HMQC and HMBC spectral data.
.see Tables 1 and 2, Scheme 4).
which was displayed in the course of the growth period of
Ginkgo cells;.3) selective formylation at C-9 and C-10
positions;and .4) selective acetylation of C-9 hydroxyl
group. In other words, there are several types of enzymes
involved in the bioprocess. Product 1 was subsequently used
for further biotransformation by Ginkgo cell cultures and 2
was identi®ed as the major product. Therefore 2 might be
biosynthesized from 1 by a deacetylation process. Though
no more experimental data were available on the order in
which these reactions took place, a comparison of the struc-
tures of the substrate and biotransformed products leads to
the proposed biotransformation pathway as shown in
Scheme 5.
From the above results, the biotransformation of sinenxan A
by Ginkgo cell suspension cultures involved various
reactions, such as: .1) selective hydroxylation at C-9 and
C-6 positions;.2) selective deacetylation at C-10 and C-14
positions, which were proven to be enzymatic reactions
since sinenxan A is very stable in the pH range .4.0±6.0),
Scheme 5. The proposed biotransformation pathway of sinenxan A by Ginkgo cell suspension cultures.

5664
J. Dai et al. / Tetrahedron 58 52002) 5659±5668
Figure 1. The kinetics of the growth and pH value of Ginkgo cell suspension cultures.
The speci®c hydroxylation at the unactivated C-9 position
of sinenxan A by Ginkgo cells constitutes an important step
in the semisynthesis from sinenxan A to paclitaxel and other
bioactive taxoids. In an endeavor to optimize the bio-
transformation conditions to increase the yields of the two
major products .1 and 2), the effects of the addition stage
and concentration of sinenxan A, also the kinetics of the
biotransformation reaction were investigated.
different sensitivities of cultured cells to exogenous
substrates.26,27
In order to clarify the optimal concentration of sinenxan A
addition for further investigation, the effects of concentra-
tion of sinenxan A .15, 30, 45, 60 and 75 mg/L, calculated
for the resulting solution concentration) on the bioconver-
sion were investigated. The results .Fig. 3) showed that the
optimal concentration for addition of sinenxan A was
60 mg/L. At this concentration, sinenxan A was ef®ciently
converted and the residual sinenxan A could hardly be
detected, meanwhile, the two major products reached their
highest yields, about 40 and 13 mg/L for 1 and 2, respec-
tively. Generally, exogenous substrates can be toxic for
cultured plant cells, however, biotranformation such as
hydroxylation and glucosylation are considered to be
detoxi®cation reactions. Hydrophobic materials such as
terpenoids and sterols may disturb the membranes of cells
and organelles when the molecules are incorporated and
phenolics may cause generation of active oxygen in the
cells. Therefore, if the substrates are not toxic to cells, the
cells may not respond to them. For example, Glycyrrhiza
cultured cells were sensitive to papaverine HCl at concen-
trations above 250 mg/L and yielded several biotransforma-
tion products. No product and no toxic effect were detected
in Saponaria of®cinalis, although more than one half the
amount of papaverine was taken into the cell when
papaverine HCl was added at 1250 mg/L.28 The results of
Firstly, the kinetics of Ginkgo cell growth and pH value
.Fig. 1), as well as the amount of residual sinenxan A and
the yields of 1 and 2 corresponding to different stages of
sinenxan A addition .Fig. 2) were investigated. The Ginkgo
cells grew very fast under the culture conditions and the
whole growth period lasted for 21 days and involved three
phases: .1) lag phase .0±9th day), .2) logarithmic phase .9±
18th day) and .3) stationary phase .18±21st day). The pH
values remained relatively stable ranging from 4.0 to 6.0.
The results also disclosed that the optimal stage for sinenxan
A addition was at the logarithmic-phase .9±18th day) of the
cell growth period. Substrate added at this phase, especially
on the 18th dayÐthe late logarithmic phase was converted
into the two major products ef®ciently. The substrate was
almost completely converted and the yields reached their
highest levels, about 16.7 and 5.4 mg/L for 1 and 2, respec-
tively .HPLC, approximately 70 and 12% of the amount of
added substrate). This may result from different activities of
the responsible enzymes in different cultural stages and
Figure 2. The effects of the timing of sinenxan A addition on its biotrans-
formation by Ginkgo cell suspension cultures.
Figure 3. The effects of the concentrations of sinenxan A added on its
biotransformation by Ginkgo suspension cultured cells.

J. Dai et al. / Tetrahedron 58 52002) 5659±5668
5665
be extracted, puri®ed and immobilized for large-scale
production. In addition, the genes encoding to the enzymes
could be cloned and transferred into a microorganism to
yield products industrially. With this view, a series of
experiments .Table 3) were designed to characterize the
enzymes responsible for the biocatalytic conversion of
sinenxan A to products 1 and 2. The results of treatment 2
in which the added sinenxan A was metabolized and
compound 1 and 2 were produced, suggest that the enzymes
are constitutive and extracellular. The results of other
treatments con®rmed this conclusion. Most probably, the
conversion of sinenxan A to 1 is cytochrome P-450-
dependent hydroxylation.29,30
Figure 4. The kinetics of sinenxan A biotransformation and compound 1
and 2 production by Ginkgo cell suspension cultures.
3. Discussion
this experiment also suggested that the concentration of
exogenous substrate addition affects the substrate to product
bioprocess.
Biotransformation refers to the technique that converts
various substrates to more useful products using freely
suspended, immobilized plant .or microorganisms) cells or
enzymes derived from those organisms. Biotransformations
employing plant cell cultures cover a wide range of
reactions, such as glucosylation, glucosyl esteri®cation,
hydroxylation, oxidoreductions, methylation and demethyl-
ation, hydrolysis, etc. To a certain extent, plant cells act as a
poly-enzyme system that can biocatalyze an exogenous
substrate to several products through various types of
reactions. In this paper, regio- and stereo-selective
hydroxylation .C-9, C-6), deacetylation .C-10, C-14) or
demethylation .acetoxyl group at C-9 and C-10) were
displayed by employing the same Ginkgo cell line, showing
the reaction diversity of sinenxan A catalyzed by Ginkgo
cells. However, the activities or amounts of enzymes
responsible for these reactions exhibited great differences,
which were suggested by signi®cant different yields of the
products. Obviously, the activities or amounts of enzymes
responsible for the yields of 1 and 2 were superior to those
of others, because the yields of 1 and 2 were much higher.
Thus, the major biotransformation of sinenxan A by Ginkgo
cells might be considered as C-9 hydroxylation. On the
other hand, the generation of a product could be controlled
depending upon identi®cation of the enzymes reponsible for
the process.
Based on the above results, the kinetics of the substrate to
product bioconversion were also investigated. 60 mg/L of
sinenxan A was added to the 18-day-old cultures and
incubated for 24, 48 and 72 h, respectively. The HPLC
analysis results .Fig. 4) revealed that: .1) the biotransforma-
tion rates of sinenxan A were about 40, 80 and 100%;.2) the
yield of 1 was about 26, 60 and 59.5%;and .3) the yield of 2
was about 13, 18 and 21.6% for the above mentioned three
incubation periods. These results suggested that the
metabolism of sinenxan A was consistent with production
of two major products within the ®rst 48 h, then production
of other minor products. Therefore, the optimal incubation
time should be 48 h as far as the production of 1 and 2 are
concerned.
Plant cells in vitro could convert the exogenous substrates to
new products. In essence, it is the enzymes produced by
plant cells that play the role. If the enzymes' properties
and additional information were well investigated, the
biotransformation condition could be greatly optimized,
and the yields of desired products could be improved by
modulating the enzymes' activities, or the enzymes could
Table 3. The basic properties of enzymes responsible for substrate biotransformation to compound 1 and 2
Treatments
Amount of
residual
substrate
.mg/L)
Yield of compound 1 .mg/L)
Yield of compound 2 .mg/L)
Havested
cell
cultures
.g/L, dry
weight)
1a
2b
3c
4d
5e
31.6^1.40
25.4^2.10
0.78^0.12
2.65^0.32
1.32^0.15
0
0
0
4.32^0.36
17.8^2.50
21.5^2.11
40.6^3.22
0.88^0.11
7.5^0.56
9.6^1.12
16.8^1.64
16.4^0.75
21.6^1.12
18.5^0.86
21.1^0.75
Each value was the mean of three repeated tests ^SE, the inoculum was 5 g/L of cell cultures .dry weight), and the reaction was quenched on day 21.
0.5 mL of substrate solution was added to the ¯ask without cell cultures inoculated.
a
b
0.5 mL of substrate solution was added on the 15th day to the ¯ask of which the cell cultures were ®ltered out.
0.5 mL of substrate solution was added on the 15th day to the ¯ask with cell cultures.
0.5 mL of substrate solution was added on the 15th day to the ¯ask, but on the 18th day, the cell cultures were ®ltered out, then an additional 0.5 mL substrate
solution was added to the same ¯ask.
0.5 mL of substrate solution was added on the 15th day to the ¯ask, and on the 18th day, an additional 0.5 mL of substrate solution was added to the same
¯ask.
c
d
e

5666
J. Dai et al. / Tetrahedron 58 52002) 5659±5668
In the long run, it would be desirable to produce paclitaxel
and analogs thereof by a process that does not depend on the
extraction of plant materials, and that can be carried out
under controlled conditions. Such a biotechnological
process may be based on plant cell culture fermentation or
a microbial process with a genetically engineered organism
in combination with some chemical/enzymatic steps. The
development of such a process would be greatly bene®ted
by an understanding of biosynthesis of paclitaxel in plants,
and by characterization of enzymes catalyzing various reac-
tions and identi®cation of genes encoding the key enzymes.
Identi®cation of rate-limiting enzymes would be essential to
increase paclitaxel synthesis. Some remarkable research
results on paclitaxel biosynthesis have been reported, and
have suggested that cytochrome P-450 enzymes are
responsible for the oxygenation steps.31±35 However
biotransformation of taxanes employing plant or microbial
cells may biomimic some steps of taxoid biosynthesis, such
as extensive oxidation of the taxane skeleton.
cell cultures and 0.5 mL EtOH alone to one additional ¯ask
as control. After an additional six days of incubation, the
cell cultures were ®ltered out under vacuum and washed
three times with distilled water. The ®ltrate was collected
and extracted three times and the dried cultures extracted
once by sonication with an equivalent volume of EtOAc,
and all the extracted solutions were concentrated under
vacuum at 408C. The residues were dissolved in acetone
and analyzed by TLC developed by the acetone±petroleum
ether .60±908C) .1:2.5) and detected by spraying with 10%
H2SO4 .in EtOH) followed by heating at 1058C. The TLC
results showed that some new spots appeared in the chro-
matogram of the medium extract of the treatment compared
with those of the cultures of treatment and the controls. For
preparative biotransformation experiments, the Ginkgo
suspension cell cultures were cultivated on a large scale
using 1000 mL ¯asks with 330 mL of medium. On the
day 15, 1.0 g of substrate was dissolved in 100 mL EtOH
and distributed into 50 ¯asks. After an additional six days of
incubation, all the media were collected, extracted and
concentrated as described above. The obtained residue
.1.5 g) was separated by silica gel chromatography .silica
gel H, 5±40 mesh) eluting with acetone±petroleum ether
.1:5±1:1) to yield compound 1 .500 mg), 2 .100 mg), 6
.5 mg), a mixture of 7 and 8 .6 mg), a mixture of 3 and 4
.10 mg), and a mixture of 4 and 5 .12 mg). Then the mixture
of 3±5 was further puri®ed by semi-preparative HPLC
.mobile phase: methanol±acetonitrile±water, 40:15:55,
v/v/v) to afford 3 .3 mg), 4 .5 mg), and 5 .6 mg).
In conclusion, we report a powerful method for preparation of
variety of derivatives by Ginkgo cells from sinenxan A. These
may not only supply new bioactive taxoids or intermediates
for synthesis of other new bioactive taxoids, but also give a
new alternative approach to study paclitaxel biosynthesis.
4. Experimental
4.1. General methods
4.3.1. 9a,14b-Dihydroxy-2a,5a,10b-triacetoxy-4ꢀ20), 11-
25
Optical rotations were measured with a Perkin±Elmer 243 B
polarimeter. IR spectra were obtained on a Perkin±Elmer
983 spectrophotometer .KBr). NMR spectra .1H, 13C NMR,
taxadiene 3. Colorless needles;[ a]D ˆ142.8 .c 0.0049,
MeOH);IR nmax .KBr): 3432, 2940, 1726, 1636, 1373,
1236, 1023 cm21;FABMS .NBA) m/z: 501 .M1Na, 15),
441 .1), 419 .2), 401 .3), 341 .4), 299 .6), 154 .77), 136
.100);HREIMS for C 26H38O8 requires: 478.2567;found:
1
DEPT, H±1H-COSY, HMQC, HMBC and NOESY) were
recorded in CDCl3 with Bruker DRX-500 or Varian
INOVA-500 instrument .1H NMR, 500 MHz; 13C NMR,
125 MHz) and chemical shifts were recorded in d using
TMS as internal standard. FABMS spectra were measured
on a KYKY-ZHP-5# mass spectrometer in positive mode;
and HREIMS analyses were performed on a JEOL JMX
HX-110 spectrometer .The Instrumental Analysis Center
for Chemistry, Tohoku University, Japan). The Ginkgo
suspension cells were shaked on the rotary shaker at
110 rpm at 258C in the darkness at the inoculum of 5 g/L
of cell cultures .dry weight). The pH values were adjusted to
5.8 before autoclaving for 20 min at 1218C and the amount
of sucrose added to the medium was 30 g/L.36 All chemicals
were obtained from Beijing Chemical Factory.
1
478.2566. The H and 13C NMR spectral data are sum-
marized in Tables 1 and 2.
4.3.2. 6a,10b-Dihydroxy-2a,5a,14b-triacetoxy-4ꢀ20), 11-
25
taxadiene 4. White powder;[ a]D ˆ147.18 .c 0.0072,
MeOH);other data matched Ref. 21.
4.3.3. 6a,9a,10b-Trihydroxy-2a,5a,14b-triacetoxy-4ꢀ20),
25
11-taxadiene 5. Colorless needles;[ a]D ˆ151.2 .c
0.0054, MeOH);IR nmax .KBr): 3432, 2940, 1726, 1636,
1373, 1236, 1023 cm21;FABMS .NBA) m/z: 517 .M1Na,
7), 154 .100), 136 .82);HREIMS for C 26H38O9 requires:
494.2516;found: 494.2520. The 1H and 13C NMR spectral
data are summarized in Tables 1 and 2.
4.2. Substrate
4.3.4. 9a,10b-O-ꢀPropane-2,2-diyl)-2a,5a,14b-triace-
25
2a,5a,10b,14b-Tetra-acetoxy-4.20), 11-taxadiene was isolated
from callus cultures of T. yunnanensis, and identi®ed by chemi-
cal and spectral methods.16 The substrate was dissolved in
EtOH and diluted to 10.0 mg/mL before use as stock solution.
toxy-4ꢀ20), 11-taxadiene 6. White powder;[ a]D ˆ136.7
.c 0.0062, MeOH);IR nmax .KBr): 3432, 2940, 1726, 1636,
1373, 1236, 1023 cm21;FABMS .NBA) m/z: 541 .M1Na,
18), 401 .3), 341 .6), 299 .10), 281 .16), 154 .40), 135 .66),
91 .100);HREIMS for C 29H42O8 requires: 518.2880;found:
518.2884. The 1H and 13C NMR spectral data are summar-
ized in Tables 1 and 2.
4.3. Biotransformation of sinenxan A by Ginkgo cell
suspension cultures
4.3.5. 9a-Hydroxy-2a,5a,10b,14b-tetra-acetoxy-4ꢀ20),
11-taxadiene formate 7 and 10b-hydroxy-2a,5a,9a,14b-
The suspension cells were cultured in 500 mL ¯ask with
150 mL MS liquid medium. 0.5 mL of the prepared
substrate solution was added to one ¯ask with suspension
tetra-acetoxy-4ꢀ20), 11-taxadiene formate 8. As
a

J. Dai et al. / Tetrahedron 58 52002) 5659±5668
5667
mixture, white powder;FABMS .NAB) m/z: 571 .M1Na,
18), 341 .5), 299 .11), 281 .30), 135 .100);HREIMS for
amounts of residual substrate, compound 1 and 2 measured
by HPLC. The results were averaged .in triplicate).
1
C29H40O10 requires: 548.2622;found: 548.2629. The H and
13C NMR spectral data are summarized in Tables 1 and 2.
Acknowledgements
4.4. Effect of substrate addition time on the
biotransformation
We thank The National Outstanding Youth Foundation by
NSF of China and Trans-Century Training Program Foun-
dation for the Talents by the Ministry of Education for
®nancial support. We also thank Dr Masayoshi Ando,
Faculty of Engineering, Niigata University, Japan, for the
help of HREIMS analyses.
On the 0, 3rd, 6th, 9th, 12th, 15th, 18th day during the cell
culture growth period, 35 mg/L of sinenxan A was added to
each 500 mL ¯ask with 150 mL of medium .in triplicates).
On the 21st day, the medium in each ¯ask was collected,
extracted and concentrated as described above. The residues
were dissolved in the HPLC mobile phase and diluted with
the same solution to give 2.0 mL and the amounts of
residual substrate, compound 1 and 2 were determined by
HPLC. HPLC analyses were performed by using a Zorbax
C18 column .25 cm£4.6 mm I.D., 5 mm) eluting with
methanol±acetonitrile±water .50:15:35, v/v/v) at the ¯ow
rate of 1.0 mL/min and detecting at 227 nm. The regression
References
1. Wani, M. C.;Taylor, H. L.;Wall, M. E.;Coggon, P.;McPhail,
A. T. J. Am. Chem. Soc. 1971, 93, 2325±2327.
2. Suffness, M. Ann. Rep. Med. Chem. 1993, 28, 305±315.
3. Rowinsky, E. K.;Cazenave, L. A.;Donehower, R. C. J. Natl
Cancer Inst. 1990, 82, 1247±1253.
equations of sinenxan A,
1
and
2
were:
Yˆ625,889X1154,506 .rˆ0.9997), Yˆ839,985X125,292
.rˆ0.9993), Yˆ683,770X246,645 .rˆ0.9999), respec-
tively, where Y refers to peak area, X the injection amount
.in mg), and r the correlation coef®cient.
4. McGuire, W. P.;Rowinsky, E. K.;Rosenshenein, N. B.;
Grumbine, F. C.;Ettinger, D. S.;Armstrong, D. K.;
Donehower, R. C. Ann. Intern. Med. 1989, 111, 273±276.
5. Holmes, F. A.;Kudelka, A. P.;Kavanagh, J. J.;Huber, M. H;
Ajani, J. A.;Valero, V. In Taxane Anticancer Agents: Basic
Science and Current Status;Georg, G. I., Chen, T. T., Ojima,
I., Vyas, D. M., Eds.; ACS Symposium series 583;American
Chemical Society: Washington, DC, 1995;pp 31±57.
6. Scniff, P. B.;Fant, J.;Horwitz, S. B. Nature 1979, 227, 9479±
9490.
4.5. Effect of substrate concentration on the
biotransformation
On the 18th day of culturing, the different concentrations of
substrate .15, 30, 45, 60, 75 mg/L) were added .in tripli-
cates). And on the 21st day, the media were collected,
extracted and concentrated, and the amounts of residual
substrate, compound 1 and 2 were determined by HPLC.
7. Manfredi, J. J.;Horwitz, S. B. Pharmacol. Ther. 1984, 25, 83±
89.
8. Nicolaou, K. C.;Yang, Z.;Liu, J. J.;Veno, H.;Nanterment,
P. G. Nature 1994, 367, 630±634.
4.6. Biotransformation kinetics
9. Georg, G. I.;Boge, T. C.;Cheruvallath, Z. S.;Clowers, J. S.;
Harriman, G. C. B.;Hepperle, M.;Park, H. In Taxol: Science
and Applications;Suffness, M., Ed.;CRC: Boca Raton, FL,
1995;pp 317±375.
On the day 18, 60 mg/L of sinenxan A was added .in tripli-
cates) and after every 24 h of incubation, the media were
collected, extracted and concentrated, and the amounts of
residual substrate, compound 1 and 2 were determined by
HPLC.
10. Kingston, D. J. I. Trends Biotechnol. 1994, 12, 222±228.
11. Patel, R. N. Annu. Rev. Microbiol. 1998, 361±398.
12. Bosch, I.;Croop, J. Biochim. Biophys. Acta 1996, 1288, 37±
54.
4.7. Basic properties of the enzymes responsible for the
production of compound 1 and 2
13. Ford, J. M.;Hait, W. N. Pharmacol. Rev. 1990, 42, 155±160.
14. Gottesman, M. M.;Pastan, I. Annu. Rev. Biochem. 1993, 62,
385±396.
In this experiment, a 500 mL ¯ask with 150 mL medium
was used for cell culturing. Treatment 1: 0.5 mL of substrate
stock solution was added to the ¯ask without cell cultures
inoculated. Treatment 2: 0.5 mL of substrate stock solution
was added to the ¯ask just after the cell cultures were
®ltered out on the 15th day. Treatment 3: 0.5 mL of
substrate stock solution was added to the ¯ask with 15-
day-old cell cultures. Treatment 4: 0.5 mL of substrate
stock solution was added to the ¯ask with 15-day-old cell
cultures, but on the 18th day of culturing, the cell cultures
were ®ltered out under vacuum, then additional 0.5 mL of
substrate solution was added to the ¯ask without cell
cultures. Treatment 5: 0.5 mL of substrate stock solution
was added to the ¯ask with 15-day-old cell cultures, and
on the 18th day of culturing, additional 0.5 mL of substrate
solution was added to the same ¯ask. On the 21st day, all the
media were collected, extracted and concentrated, and the
15. Ojima, I.;Bounaud, P. Y.;Bernacki, R. J. Chemtech June
1998, 31±35.
16. Cheng, K. D.;Chen, W. M.;Zhu, W. H.;Fang, Q. C. PCT Int.
Appl. WO9406, 740 .Cl. C07C35/37), 31 Mar 1994, JP Appl.
92/249, 047, 18 Sep 1992.
17. Wu, Y.;Zhu, W.;Lu, J.;Hu, Q.;Li, X. Chin. Pharm. J. 1998,
33, 15±18.
18. Huang, G.;Guo, J.;Liang, X. Acta Pharm. Sin. 1998, 33, 576±
586.
19. Hu, S. H.;Tian, X. F.;Zhu, W. H.;Fang, Q. C. Chin. Chem.
Lett. 1996, 6, 543±544.
20. Hu, S. H.;Tian, X. F.;Zhu, W. H.;Fang, Q. C. J. Nat. Prod.
1996, 59, 1006±1009.
21. Hu, S. H.;Tian, X. F.;Zhu, W. H.;Fang, Q. C. Tetrahedron
1996, 26, 8739±8746.
22. Dai, J.;Guo, H.;Lu, D.;Zhu, W.;Zhang, D.;Zheng, J.;Guo,
D. Tetrahedron Lett. 2001, 42, 4677±4679.

5668
J. Dai et al. / Tetrahedron 58 52002) 5659±5668
23. Kingston, D. G. I.;Chordia, M. D.;Jagtap, P. G.;Liang, J.;
Shen, Y.;Long, B. H.;Fairchild, C. R.;Johnston, K. A. J. Org.
Chem. 1999, 64, 1814±1822.
30. Petersen, M.;Alfermann, A. W. Z. Naturforsch. 1988, 43C,
501±504.
31. Koepp, A. E.;Hezari, M.;Zajicek, J.;Stofer;Vogel, B.;
LaFever, R. E.;Lewis, N. G.;Croteau, R. J. Biol. Chem.
1995, 270, 8686±8690.
24. Wiegerinck, P. H. G.;Fluks, L.;Hammink, J. B.;Mulders,
S. J. E.;de Groot, F. M. H.;van Rozendaal, H. L. M.;
Scheeren, H. W. J. Org. Chem. 1996, 61, 7092±7100.
25. Cheng, Q.;Oritani, T.;Honguchi, T.;Yamada, T.;Mong, Y.
Bioorg. Med. Chem. Lett. 2000, 10, 517±521.
32. Floss, H. G.;Mocek, U. In Taxol: Science and Applications;
Suffness, M., Ed.;CRC: Boca Raton, FL, 1995;pp 191±208.
33. Kingston, G. D. I. In Taxol: Science and Applications;
Suffness, M., Ed.;CRC: Boca Raton, FL, 1995;pp 287±315.
34. Lin, X.;Hezari, M.;Koepp, A. E.;Floss, H. G.;Croteau, R.
Biochemistry 1996, 35, 2968±2977.
26. Langebartels, C.;Harms, H. Z. P¯anzenphysiol. 1984, 113,
201±204.
27. Umetani, Y.;Tanaka, S.;Tabata, M. In Proceedings of the 5th
International Congress on Plant Tissue Cell Culture;
Fujiwara, M., Ed.;Manuzen: Tokyo, 1982;pp 383±385.
35. Schoendorf, A.;Rithner, C. D.;Williams, R. M.;Croteau,
R. B. Proc. Natl Acad. Sci. USA 2001, 98, 1501±1506.
36. Dai, J.;Zhu, W.;Wu, Y.;Hu, Q.;Zhang, D. Acta Pharm. Sin.
2000, 2, 151±156.
28. Inomata, S.;Yokoyama, M.;Seto, S.;Yanggi, M.
Microbiol. Biotechnol. 1991, 36, 315±318.
Appl.
29. Park, S. H.;Chang, Y. J.;Kim, K. Y.;Kim, S. U. J. Microbiol.
Biotechnol. 1999, 9, 147±149.