Regioselective acylation of 3-O-angeloylingenol by Candida antarctica Lipase B

Article abstract of DOI:10.1016/j.fitote.2009.02.001

Acylation of 3-O-angeloylingenol (1) with vinyl acetate, vinyl decanoate and vinyl cinnamate, catalyzed by Candida antarctica Lipase B, was investigated. In each case, compound 1 was quantitatively and regioselectively acylated to afford a single product,

Full text of DOI:10.1016/j.fitote.2009.02.001

Fitoterapia 80 (2009) 233–236
Contents lists available at ScienceDirect
Fitoterapia
journal homepage: www.elsevier.com/locate/fitote
Regioselective acylation of 3-O-angeloylingenol by Candida antarctica
Lipase B
a,
b,1
c
c
d
d
a
R.W. Teng ⁎, D. McManus , J. Aylward , S. Ogbourne , J. Johns , P. Parsons , A. Bacic
a
CRC for Bioproducts, School of Botany, The University of Melbourne, Victoria 3010, Australia
CRC for Bioproducts, Tridan-Albright Ltd. and Wilson (Aust) Ltd. Partnership, Victoria 3013, Australia
Peplin Ltd., 1 Breakfast Creek Rd, Newstead, Queensland 4006, Australia
Queensland Institute of Medical Research, Herston, Queensland 4006, Australia
b
c
d
a r t i c l e i n f o
a b s t r a c t
Article history:
Acylation of 3-O-angeloylingenol (1) with vinyl acetate, vinyl decanoate and vinyl cinnamate,
catalyzed by Candida antarctica Lipase B, was investigated. In each case, compound 1 was
quantitatively and regioselectively acylated to afford a single product, 3-O-angeloyl-20-O-
acetylingenol (1a), 3-O-angeloyl-20-O-decanoylingenol (1b) and 3-O-angeloyl-20-O-
cinnamoylingenol (1c), respectively. The structures of the novel compounds 1b–1c were
determined by MS and NMR, and product 1a by comparison of RP-HPLC and TLC with a
standard. Compounds 1b–1c induced a bipolar morphology of MM96L melanoma cells at a
similar concentration as compound 1, as well as having activity in inhibiting the growth of
MM96L melanoma cells.
Received 4 November 2008
Received in revised form 30 January 2009
Accepted 1 February 2009
Available online 8 February 2009
Keywords:
Candida antarctica Lipase B
Acylation
3-O-angeloylingenol
©
2009 Elsevier B.V. All rights reserved.
1
. Introduction
melanoma cells [8] and leukemia cells [9]. This compound
showed potential as a novel topical chemotherapeutic agent for
the treatment of skin cancer [10] and is being developed as a
topical therapy for actinic keratosis and superficial basal cell
carcinoma (www.peplin.com).
In this study, 3-O-angeloylingenol (1) was acylated with
vinyl acetate, vinyl decanoate and vinyl cinnamate in the
presence of C. antarctica Lipase B. The acylated products 1a–1c
were isolated, purified and their structures determined by MS
and NMR (Fig.1). The activity of compounds 1b–1c in inhibiting
the growth and inducing a bipolar morphology of MM96L
melanoma cells was assessed.
Candida antarctica Lipase B (commercial name: Novozyme
35) is a lipase which has proven to be a highly regio- and
4
stereo-selective enzyme in both hydrolysis [1] and synthesis
2]. This enzyme has been used to acylate a variety of sub-
[
strates, including many classes of natural products [3]. We
successfully used C. antarctica Lipase B for the regio- and
stereo-selective acylation of ginsenosides [4], iridoid glyco-
sides, a flavanoid glycoside and simple phenolic compounds
[5].
3
-O-angeloylingenol (1) was isolated from Euphorbia anti-
quorum and E. peplus, the extracts of which have been used in
the treatment of a number of conditions, such as warts, corns,
waxy growth and skin cancer [6]. First characterized as a skin
irritant and a toxic Euphorbia factor, An1 [7], 3-O-angeloylin-
genol (1) was thereafter shown to induce in vitro apoptosis of
2. Experimental
2.1. General
Vinyl acetate, vinyl decanoate, vinyl cinnamate and 2-
methylbutan-2-ol were purchased from Sigma. C. antarctica
Lipase B (commercial name: Novozyme 435), immobilized on
acrylic resin, was a kind gift from Novozymes Australia Pty
Ltd. RPMI 1640 media were purchased from Life Technologies,
USA, fetal calf serum and 96-well tissue culture plates from
CSL Biosciences, Australia.
⁎
Corresponding author. Current address: School of Biochemistry and
Molecular Biology, Australian National University, ACT 0200, Australia.
Tel.: +61 2 6125 0640; fax: +61 2 6125 0313.
E-mail address: tengrongwei@hotmail.com (R.W. Teng).
Current address: Nutrifina Pty Ltd., 103 Albert St, Port Melbourne,
1
Victoria 3207, Australia.
0367-326X/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.fitote.2009.02.001

234
R.W. Teng et al. / Fitoterapia 80 (2009) 233–236
mixture and analysed by TLC and RP-HPLC to monitor reaction
process. After substrate 1 was completely consumed (7 h for
reaction with vinyl acetate, 1 d for vinyl decanoate and 7 d for
vinyl cinnamate), the reaction was terminated by filtering off
the enzyme through a filter paper. The filtrate was dried in a
spin-vacuum prior to analysis.
To scale up, compound 1 (3 mg) was dissolved in 2-
methylbutan-2-ol (0.5 ml). To each of two separate solutions,
was added C. antarctica Lipase B (15 mg) and the acyl donors
(
180 µl), either vinyl decanoate or vinyl cinnamate, respec-
tively. The reactions were terminated after 1 d for reaction
with vinyl decanoate and after 7 d with vinyl cinnamate.
2.3. Isolation of acylated products
The dry residues were dissolved in MeOH (~1 ml) and then
purified by preparative RP-HPLC to afford compounds 1b and
1c. The isolated yields of the acylated products from the
scaled-up reactions were N90% for both compounds 1b and 1c.
Fig. 1. Structures of 3-O-angeloylingenol (1) and its acylated derivatives
1a–1c).
(
2.4. MS and NMR data of acylated products
Standards of 3-O-angeloylingenol and 3-O-angeloyl-20-O-
acetylingenol were from Peplin Ltd., Australia. Pre-coated thin
layer chromatogram (TLC) plates (silica gel 60 F254, 0.25 mm)
were purchased from Merck.
3-O-angeloyl-20-O-acetylingenol (1a). MALDI-TOF-MS
(C27
H O ; 474.59): m/z 473.11 [M+H] , 495.09 [M+Na] .
38 7
+
+
3-O-angeloyl-20-O-decanoylingenol (1b). MALDI-TOF-MS
+
1
3
-O-angeloylingenol (1) and its acylated products were
(C35
H
52
O
7
; 584.37): m/z 607.36 [M+Na] . H NMR (400 MHz,
separated on silica gel TLC using toluene:acetone (4:1, v/v) as
a solvent system. After the TLC plate was dried, TLC plates
were dipped into 5% (v/v) H SO (in ethanol) followed by
2 4
heating using a hair drier. 3-O-angeloylingenol (1) and its
acylated products appeared as yellow coloured spots.
δ in CDCl
3
, J in Hz): δ 6.16 (1H, qq, J=1.6, 7.2 Hz, H-3′), 6.12 (1H,
brd, J=4.0 Hz, H-7), 6.05 (1H, q, J=1.6 Hz, H-1), 5.56 (1H, s, H-3),
4.79 (1H, d, J=12.8 Hz, H-20a), 4.48 (1H, d, J=12.8 Hz, H-20b),
4.12 (1H, dd, J=4.4, 11.6 Hz, H-8), 3.88 (1H, s, H-5), 2.51 (1H, m,
2 7 2
H-11), 2.30 (2H, t, J=7.6, Me(CH ) CH CO), 2.26 (1H, ddd,
RP-HPLC was carried out on a Beckman Gold-126 (Beck-
man Coulter, Inc., USA) equipped with a variable wavelength
UV monitor. An analytical Luna C18(2) HPLC column (5 µm,
J=2.8, 8.8, 15.6 Hz, H-12β), 2.03 (3H, dq, J=7.2, 1.6 Hz, H-4′),
1.93 (3H, d, J=1.6 Hz, H-5′), 1.81 (3H, d, J=1.2, H-19), 1.76
(1H, ddd, J=6.0, 11.6, 15.6 Hz, H=12α), 1.25–1.29 (14H, m,
1
50×4.6 mm) and a semi-preparative Luna C18(2) HPLC
Me(CH
d, J=6.8 Hz, H-18), 0.94 (1H, t, J=6.8 Hz, H-14), 0.88 (3H, t,
J=6.8 Hz, Me(CH CH CO), 0.70 (1H, dt, J=6.4, 8.4 Hz, H-13).
3-O-angeloyl-20-O-cinnamoylingenol (1c), MALDI-TOF-
2 7 2
) CH CO),1.09 (3H, s, H-17),1.06 (3H, s, H-16), 0.98 (3H,
column (10 µm, 250×10 mm), purchased from Phenomenex
Australia Pty Ltd., were used. RP-HPLC was performed with
binary mixtures of 70% (v/v) MeOH in Milli-Q H
)
2 7
2
2
O (A) and
+
1
1
3
0
00% MeOH (B) with a gradient: 0–10 min 0–50% (v/v) B, 10–
5 min 50–83% B, then 35–40 min 83–100% B at a flow rate of
.5 and 2.8 ml/min for analytical and preparative scale,
MS (C34
(400 MHz, δ in CDCl
H
40
O
7
; 560.68): m/z 583.56 [M+Na] . H NMR
, J in Hz): δ 7.69 (1H, d, J=16.0 Hz, H-
3
7″), 7.37–7.55 (5H, m, H-2″ to H-6″), 6.45 (1H, d, J=16.0 Hz,
H-8″), 6.16 (1H, qq, J=1.6, 7.2 Hz, H-3′), 6.18 (1H, brd,
J=5.2 Hz, H-7), 6.06 (1H, q, J=1.2 Hz, H-1), 5.59 (1H, s, H-3),
4.95 (1H, d, J=12.8 Hz, H-20a), 4.62 (1H, d, J=12.8 Hz, H-
20b), 4.14 (1H, dd, J=4.0, 12.0 Hz, H-8), 3.94 (1H, s, H-5),
2.53 (1H, m, H-11), 2.28 (1H, ddd, J=2.8, 8.8, 16.0 Hz, H-12β),
2.03 (3H, dq, J=7.2, 1.6 Hz, H-4′), 1.93 (3H, brs, H-5′), 1.81
(3H, d, J=1.2, H-19), 1.77 (1H, ddd, J=6.0, 11.6, 16.0 Hz,
H=12α), 1.09 (3H, s, H-17), 1.06 (3H, s, H-16), 0.98 (3H, d,
J=7.2 Hz, H-18), 0.94 (1H, t, J=7.6 Hz, H-14), 0.70 (1H, dt,
J=6.4, 7.6 Hz, H-13).
respectively. Elution was monitored with UV detection at 210
and 230 nm.
1
H NMR were performed at room temp on a Varian
1
Unityplus-400, at 400 MHz ( H); δ in ppm relative to solvent
signal of CDCl (δ 7.24). MALDI-TOF-MS was performed on a
3 H
Voyager-DE STR biospectrometry workstation (Perspective
Biosystems) using α-cyano-4-hydroxycinnamic acid as a
matrix as previously described [5].
2
.2. Acylation of compound 1 with C. antarctica Lipase B
In small-scale experiments, compound 1 (approx. 0.3–
.6 mg) was dissolved in 2-methylbutan-2-ol (80–150 µl) and
2.5. Bioassay
0
then C. antarctica Lipase B (3 mg) and acyl donors (50 µl) were
added. Three separate incubations were performed, one for
each acyl donor (vinyl acetate, vinyl decanoate or vinyl
cinnamate). Blank controls were carried out for each acyl
donor with solutions lacking C. antarctica Lipase B. The resulting
mixtures were then incubated at 37 °C on a gyratory shaker at
Compounds 1, 1b and 1c were collected from RP-HPLC
elutes when the 10 µl reaction solutions withdrawn during
the small-scale acylation experiments were analysed by RP-
HPLC. Approximately equal amounts were dried by vacuum
centrifugation and dissolved in 50 µl of acetone. Control
experiment was carried out using 50 µl acetone. The solutions
were diluted at a series of 10-fold by complete cell culture
2
50 rpm. A 10 µl of reaction solution was taken from the

R.W. Teng et al. / Fitoterapia 80 (2009) 233–236
235
medium (RPMI 1640, supplemented with 10% (v/v) fetal calf
serum.
One μL of the solution was added to a single well of a 96-
well tissue culture plate containing 99 µl of complete cell
culture medium and approximately 3000 MM96L melanoma
cells which had been growing for 24 h . Assays for each
dilution were conducted in duplicate.
Two days after treatment, the cells were observed for
induction of a bipolar morphology, which indicated differentia-
tion activity, in comparison to the poly-dentritic control. Five
days after treatment, the cells were washed with phosphate-
buffered saline, fixed in ethanol and then the total protein, an
indicator of growth inhibition activity, was determined using
sulforhodamine B [11].
3
. Results and discussion
-O-angeloylingenol (1) was reacted with either vinyl
Fig. 2. Acylation kinetics of the conversion of 3-O-angeloylingenol (1) into
compound 1c using vinyl cinnamate in the presence of C. antarctica Lipase B.
3
acetate, vinyl decanoate or vinyl cinnamate in the presence of
C. antarctica Lipase B. The reactions were monitored by TLC
and HPLC. Small scale experiments indicated that 3-O-
angeloylingenol (1) was quantitatively acylated into a single
product in the presence of each of the three acyl donors. To
generate sufficient amounts of products for structural
determination, the reactions were scaled-up and the acylated
products were then purified by preparative RP-HPLC. The
structures of acylated products were determined by MS and
hydroxyl group, –OH(C20), is the most reactive hydroxyl
group in compound 1 (Fig. 1) and that the acylated products,
1b and 1c, should be the –OH(C20) acylated derivatives of
3-O-angeloylingenol (1).
In order to confirm the structures of compounds 1b and
1c, reactions were scaled up to generate sufficient quantities
of 1b–1c. The MALDI-TOF MS spectrum of compounds 1b and
1
+
H NMR. The results showed that acylations occurred only at
1c showed quasi-molecular ions at m/z 607.36 [M+Na] and
+
the primary hydroxyl –OH(C-20) to afford 3-O-angeoyl-20-O-
acetylingenol (1a), 3-O-angeloyl-20-O-decanoylingenol (1b)
and 3-O-angeloyl-20-O-cinnamoylingenol (1c), respectively.
The structures of compounds 1b and 1c have not been
previously reported.
In the control reactions, no product formation was
detected by either TLC or RP-HPLC. When reacted with vinyl
acetate in the presence of C. antarctica Lipase B, 3-O-
angeloylingenol (1) was completely transformed into a single
product (compound 1a) after only a 6 h incubation, as
indicated by the disappearance of the spot of 3-O-angeloylin-
genol (1) (relative shift (Rf): 0.2) and appearance of a new
spot (Rf: 0.6) on TLC. The result was further confirmed by RP-
HPLC where a new peak for compound 1a (retention time
583.56 [M+Na] , respectively, which were 154 amu and
130 amu greater than the substrate 3-O-angeloylingenol (1)
[10]. These increased molecular weights corresponded to an
additional decanoyl and cinnamoyl group, respectively, which
1
was confirmed by H NMR of compounds 1b and 1c showing
the additional 1H signals for the decanoyl and cinnamoyl
1
groups, respectively [4,5]. Moreover, compared with H NMR
data of compound 1 [12], the H-20 resonance of 1b and 1c
were shifted downfield by 0.45 and 0.80 ppm, respectively,
typical of acylation induced shifts [4,5]. Therefore, the
structures of compounds 1b and 1c were deduced to be 3-
O-angeloyl-20-O-decanoylingenol (1b) and 3-O-angeloyl-20-
O-cinnamoylingenol (1c), respectively (Fig. 1).
Acylation of compound 1, catalyzed by C. antarctica Lipase
B, afforded a quantitatively single mono-acylated product. In
addition, acylation occurred only at –OH (C20) while the
other two hydroxyl groups were un-reacted. This degree of
regioselectivity would be technically challenging for a non-
enzymatically based acylation reaction.
(
Rt): 29 min) appeared and the substrate, 3-O-angeloylin-
genol (1) (Rt: 27 min) disappeared.
Reaction of 3-O-angeloylingenol (1) with vinyl decanoate in
the presence of C. antarctica Lipase B was also rapid.
Approximately 90% of 3-O-angeloylingenol (1) was consumed
in 6 h and completely converted into compound 1b in 1 d (1b,
Rt: 48 min in RP-HPLC). Reaction of 3-O-angeloylingenol (1)
with vinyl cinnamate was much slower than with vinyl acetate
or vinyl decanoate (Fig. 2), presumably due primarily to the
steric hindrance of the bulky phenyl group. The results of RP-
HPLC fractionation indicated that 45% of 3-O-angeloylingenol
Compounds 1 and 1b–1c were dissolved in acetone and
diluted in complete cell culture medium by a series of 10-fold
dilutions for bioassay. All three compounds, in addition to
compound 1a assayed previously by Peplin Ltd. (data not
published), showed similar levels of activity at inducing a
bipolar morphology of the MM96L melanoma cells at approxi-
5
(1) was transformed into compound 1c (Rt: 37 min) in a 1 d
mately 1 ng/ml (a dilution of 10 ). All four compounds 1 and
reaction, 90% after a 3 d reaction, and complete conversion
after 7 d.
1a–1c, also showed growth inhibitory activity against the
MM96L melanoma cells. Whilst the results were not quantified,
these data are interesting as differentiation and growth
inhibition of mammalian cells are commonly associated with
protein kinase C (PKC) activity [13,14] and acylation of ingenol
esters at C20 would be expected to afford inactivity. These
compounds could therefore prove useful in the elucidation of
The product 1a was purified by preparative RP-HPLC and
its structure was determined as 3-O-angeoyl-20-O-acety-
lingenol (1a), an –OH(C20) acetylated product, by co-
elution on TLC and RP-HPLC with a 3-O-angeoyl-20-O-
acetylingenol standard. This suggested that the primary

236
R.W. Teng et al. / Fitoterapia 80 (2009) 233–236
[
3] (a) Oosterom MWV, Rantwijk FV, Sheldon RA. Biotechnol Bioeng
996;49:328.
PKC binding and activation kinetics. However, it is possible that
cell mediated hydrolysis of compounds 1a–1c yielded com-
pound 1, resulting in the observed positive biological assay
results. Given the potential of compound 1 as an anti-cancer
agent, compounds 1a–1c may be interesting novel targets for
product development. The bioassays indicate that acylation of
compound 1 appear not to improve its biological efficacy.
1
(b) Danieli B, Bertario A, Carrea G, Redigolo B, Secundo F, Riva S. Helv
Chim Acta 1996;76:2981.
(c) Danieli B, Luisetti M, Steurer S, Michelitsch A, Likussar W, Riva R, et al.
J Nat Prod 1996;62:670.
[
4] (a) Teng RW, Ang CS, McManus D, Armstrong D, Mau SL, Bacic A.
Tetrahedron Lett 2004;44:5661.
(b) Teng RW, Ang CS, McManus D, Armstrong D, Mau S, Bacic A. Helv
Chim Acta 2004;87:1860.
[
[
5] Teng RW, Bui TKA, McManus D, Armstrong D, Mau SL, Bacic A. Biocatal
Biotransform 2005;23:109.
6] Rizk AM, Hammouda FM, El-Missiry MM, Radwan HM, Evans FJ.
Phytochemistry 1985;24:1605.
7] Adolf W, Chanai S, Hecker EJ. J Sci Soc Thai 1983;9:81.
8] Gillespie SK, Zhang XD, Hersey P. Cancer Res 2004;3:1651.
Acknowledgements
This work was funded by a grant from the Australian
Government to the CRC for Bioproducts. We wish to thank
Novozymes Australia Pty Ltd. for providing Novozyme 435,
and Drs. T. K. Lim and Frances Separovic, School of Chemistry,
The University of Melbourne, Australia, for access to the NMR
spectrometer. We also thank Dr Ming-Long Liao, CRC for
Bioproducts, The University of Melbourne, Australia, for
critical comments.
[
[
[9] Hampson P, Chahal H, Khanim F, Hayden R, Mulder A, Assi LK, et al.
Blood 2005;106:1362.
10] (a) Kedei N, Lundberg DJ, Toth A, Welburn P, Garfield SH, Blumberg PM.
Cancer Res 2004;64:3243.
[
[
(b) Ogbourne SM, Suhrbier A, Jones B, Cozzi SJ, Boyle GM, Morris M,
et al. Cancer Res 2004;64:2833.
11] Rubinstein LV, Shoemaker RH, Paull KD, Simon RM, Tosini S, Skehan P,
et al. J Natl Cancer Inst 1990;82:1113.
[
[
12] Marco JA, Sanz-Cervera JF, Yuste A. Phytochemistry 1997;45:563.
13] Dempsey EC, Newton AC, Mochly-Rosen D, Fields AP, Reyland ME, Insel
PA, et al. Am J Physiol Lung Cell Mol Physiol 2000;279:L429.
14] Mellor H, Parker P. Biochem J 1998;332:281.
References
[
[1] (a) Brunet C, Zarevucka M, Zdenek W, Legoy MD. Enzyme Microb
Technol 2002;31:609.
(b) Jacobsen EE, Hellemond EV, Riise Moen A, Vazquez Prado LC,
Anthonsen T. Tetrahedron Lett 2002;44:8453.
[2] (a) Enaud E, Humeau C, Piffaut B, Girardin M. J Mol Catal B 2004;27:1.
(b) Skupinska KA, McEzchern EJ, Baird IR, Skerli RT, Bridger GJ. J Org
Chem 2004;68:3546.