Short synthesis of the optically active E-ring portion of (S)-camptothecin

Article abstract of DOI:10.1039/b109071m

The short synthesis of the optically active E-ring portion of (S)-camptothecin was presented. (5S)-5-Benzyloxy-5-ethyl-6-oxo-5,6-dihydro-2H-pyran-3-carboxylic acid, the protected E-ring moiety of (S)-camptothecin, were rapidly prepared in enantiomerically

Full text of DOI:10.1039/b109071m

View Article Online / Journal Homepage / Table of Contents for this issue
Short synthesis of the optically active E-ring portion of
(S)-camptothecin
S. Leue, W. Miao, A. Kanazawa, Y. Génisson, S. Garçon and A. E. Greene*
1
Université Joseph Fourier, LEDSS, BP 53X, 38041 Grenoble Cedex, France
Received (in Cambridge, UK) 5th October 2001, Accepted 11th October 2001
First published as an Advance Article on the web 23rd October 2001
(5S)-5-Benzyloxy-5-ethyl-6-oxo-5,6-dihydro-2H-pyran-3-
carboxylic acid, the protected E-ring moiety of (S)-campto-
thecin, has been rapidly prepared in enantiomerically enriched
form (98% ee) through enolate conjugate addition to a
ꢀ-bromo methacrylate derivative, followed by enzymatic
resolution with PLE.
The synthesis of this densely functionalized key component is
the subject of this paper.
At the outset we believed that the size of this molecule might
belie the synthetic challenge that lay ahead, and this indeed
turned out to be correct.3 First efforts were directed toward the
preparation of lactone ester VI, which, it was hoped, might be
transformed into the desired lactone acid V [eqn. (1)]. Toward
Camptothecin (1a) was isolated by Wall et al.1 in 1966 from
Camptotheca acuminata Decne (Nyssaceae) and immediately
generated interest because of significant antitumor activity
revealed in murine test systems. At NCI (National Cancer
Institute) misleading clinical trials carried out on the more
water-soluble seco-acid sodium salt of this alkaloid evidenced
severe toxicity, though, and testing was halted. Several sub-
sequently prepared semisynthetic or totally synthetic congeners,
such as 9-aminocamptothecin, topotecan, and irinotecan (1b–
d), however, have demonstrated clinically useful efficacy in the
treatment of certain cancers. Camptothecin and, presumably,
its analogues act as poisons of topoisomerase I by trapping the
cleavable DNA–topoisomerase I complex.2
(1)
this end, known4 4-oxo-δ-valerolactone (ternatolide) was pre-
pared from furfuryl alcohol as previously described5 for related
lactones; however, all attempts to convert this material into
the lactone ester VI were unsuccessful. Treatment with tris-
(methylthio)methyllithium yielded the γ-lactone from rapid
translactonization of the initial adduct, while attempted
palladium-catalyzed carbomethoxylation of the enol triflate †
derivative resulted in degradation, most likely through lactone
cleavage. In addition, the corresponding cyanohydrin could not
be productively advanced. Several alternative approaches to V,
such as through application of Rousseau and coworkers’ three-
component coupling6 [eqn. (2)] or by modification of com-
mercially available methyl coumalate, also proved unfruitful.
(2)
To develop a route to camptothecin and derivatives based on
novel bond connections, we have been studying the convergent
approach shown retrosynthetically in Scheme 1. N4–C3 C-ring
Fortunately, however, a convergent synthesis was eventually
developed starting from methacrylate 2a, easily available by a
Baylis–Hillman7 or Wittig–Horner8 reaction, and 2-bromo-
butyric acid (Scheme 2). It was recognized that to reach V it was
Scheme 1
closure by amide nitrogen addition to a double or triple bond,
followed by photolytic or conjugate-addition driven C14–C15
D-ring formation, oxidation, and removal of the R and/or
XЈ groups could be expected to allow access to a variety of
camptothecinoids from II. This intermediate should be access-
ible by Sonogashira or Stille coupling from halide III, which in
turn would be available from amine IV and the enantio-
merically enriched E-ring component lactone acid V (RЈ = H).
Scheme 2 Reagents and conditions: i, CH2O, K2CO3, H2O, 20 ЊC, 2 h;
ii, Ac2O, H2SO4(cat), Et2O, 20 ЊC, 3 h; iii, Br2, CH2Cl2, reflux, 53% from
TEPA; iv, TBAF, HMPA, 20 ЊC, 24 h, 55–68%; v, LDA,
C2H5CH(OBn)CO2Me (7), THF, Ϫ78 ЊC, 15 min, then 4, Ϫ78 ЊC, 12 h,
69%; vi, LiOH, THF–H2O, 20 ЊC, 12 h, 79%.
DOI: 10.1039/b109071m
J. Chem. Soc., Perkin Trans. 1, 2001, 2903–2905
This journal is © The Royal Society of Chemistry 2001
2903

View Article Online
only necessary to append, E-stereoselectively, a protected 2-
hydroxybutyrate to the β-carbon of adduct 2a. Therefore,
methacrylate 2a, in our hands best secured from triethyl
phosphonoacetate (TEPA) and formaldehyde in the presence of
potassium carbonate, was converted into its acetate 2b. Com-
pound 2b was then treated with bromine in dichloromethane to
provide the expected9 dibromide 3 in 53% overall yield from
TEPA. Pure (E)-β-bromo acrylate 4 was next secured stereo-
selectively with tetrabutylammonium fluoride in HMPA9 in
up to 68% yield. Other solvent–base combinations gave inferior
yields and generated variable amounts of the corresponding
Z-isomer in addition to various by-products, which complicated
the purification.
Methyl 2-benzyloxybutyrate (7) could be readily prepared
from 2-bromobutyric acid through reaction with benzyl alcohol
in the presence of sodium hydride, followed by Fischer esterifi-
cation (77% yield). After some experimentation, it was found
that the conjugate addition–elimination reaction was best per-
formed by using equimolar amounts of 4 and 7 and LDA as the
base in THF at Ϫ78 ЊC. Under these conditions the desired
adduct ( )-5 to the exclusion of the allylic bromide was formed;
this was fully expected on the basis of the substantially greater
nucleofugal character of the bromide in comparison with the
acetoxy group (69% yield). Again, only the E-derivative was
produced. The desired lactone acid 6, in racemic form, was next
obtained in good yield from the triester through treatment with
lithium hydroxide in aqueous THF.10
Experimental
( )-(E )-2-Acetoxymethyl-4-benzyloxy-4-ethylpent-2-ene-
1,5-dioic acid 1-ethyl 5-methyl diester ( )-5
A solution of diisopropylamine (0.720 cm3, 520 mg, 5.14 mmol)
in THF (10 cm3) at Ϫ30 ЊC under argon was treated with n-
BuLi in hexanes (2.3 M, 2.16 cm3, 5.0 mmol) and then stirred at
0 ЊC for 15 min, whereupon it was cooled to Ϫ78 ЊC, and ester 7
(936 mg, 4.49 mmol) dissolved in THF (4 cm3) was added. The
resulting solution was stirred for 15 min and then treated with a
solution of bromide 4 (1.13 g, 4.50 mmol) in THF (6 cm3).
After being stirred at Ϫ78 ЊC for 12 h, the reaction mixture was
processed in the usual manner and the crude product was puri-
fied by SiO2 column chromatography with 30% Et2O in pentane
to provide ( )-5 (1.17 g, 69%) as an oil. νmax/cmϪ1 1744, 1719,
1651, 1234; δH (300 MHz, CDCl3) 0.98 (3H, t, J 7.5), 1.33 (3H,
t, J 7.2), 1.90 (3H, s), 2.13 (2H, m), 3.82 (3H, s), 4.28 (2H, q,
J 7.2), 4.48 (2H, ABq, JAB 11.0, δA Ϫ δB = 0.05), 5.11 (2H, s),
7.36 (6H, m); δC (75.5 MHz, CDCl3) 8.3, 14.5, 20.8, 32.4, 52.9,
58.4, 61.6, 67.6, 84.0, 127.5, 127.9, 128.6, 131.6, 138.1, 143.7,
166.2, 170.4, 172.1 [Found: (M ϩ H)ϩ, 379.1783. C20H26O7 ϩ H
requires M, 379.1757].
( )-5-Benzyloxy-5-ethyl-6-oxo-5,6-dihydro-2H-pyran-3-carb-
oxylic acid ( )-6
A solution of triester ( )-5 (252 mg, 0.67 mmol) and LiOHؒ
H2O (185 mg, 4.41 mmol) in 20% aqueous THF–H2O (20 cm3)
was stirred at 20 ЊC for 12 h. Usual treatment of the reaction
mixture provided ( )-6 (145 mg, 79%), mp 134 ЊC. νmax/cmϪ1
3483, 1751, 1713, 1668; δH (200 MHz, CDCl3) 0.99 (3H, t,
J 7.5), 1.96 (2H, m), 4.45 (2H, ABq, JAB 10.6, δA Ϫ δB = 0.12),
Triester ( )-5 also proved a suitable precursor for the
required (S)-lactone acid (Scheme 3). From ( )-5, with pig liver
5.10 (2H, AB of ABX, JAB 17.5, JAX 1.0, JBX 2.1, δA Ϫ δB
=
0.16), 7.11 (1H, X of ABX), 7.31 (5H, m), 10.91 (1H, s); δC
(50.3 MHz, CDCl3) 7.6, 31.8, 67.1, 68.7, 76.6, 127.8, 127.9,
128.3, 128.7, 137.1, 141.6, 167.1, 169.1; m/z (CI) 294 (100) (M ϩ
NH4)ϩ, 277 (M ϩ H)ϩ [Found: C, 65.3; H, 5.75. C15H16O5
requires C, 65.2; H, 5.8%].
Scheme 3 Reagents and conditions: i, PLE, pH 7 phosphate buffer,
MeCN, 25 ЊC, 48 h, separation; ii, mixture of acids, LiOH, THF–H2O,
20 ЊC, 12 h, recrystallisation, 98% ee, 34% of theory.
(S)-(؉)-5-Benzyloxy-5-ethyl-6-oxo-5,6-dihydro-2H-pyran-
3-carboxylic acid (؉)-6
To a solution of triester ( )-5 (943 mg, 2.49 mmol) in MeCN
(7.0 cm3) was added pH 7.0 phosphate buffer solution (35 cm3)
and pig liver esterase (1.35 cm3, suspension in 3.2 M (NH4)2SO4
solution, ca. 1800 units, Fluka). The reaction mixture was
stirred vigorously at 25 ЊC for 48 h, whereupon volatiles were
removed under vacuum and 10% NaOH was added. Et2O
extraction provided, following filtration of the crude material
over SiO2 with CH2Cl , (R)-(ϩ)-5 (420 mg) [α]2D5 ϩ9.0 (c 1.15,
CHCl3). The aqueous2phase was then acidified with 5% HCl
and extracted with Et2O to give a mixture of acids (350 mg).
A 348 mg sample of this mixture was treated with LiOH as
described above to afford the lactone acid as a solid. The
mother liquor from recrystallization (CH2Cl2–C6H12) of this
material yielded (ϩ)-6 (l18 mg, 34% of theory): mp 81–83 ЊC;
[α]2D5 ϩ79 (c 2.26, CHCl3).
esterase (PLE)11 in pH 7.0 phosphate buffer and acetonitrile at
25 ЊC for 48 h (55% conversion), a mixture of partially hydro-
lyzed products together with recovered (R)-(ϩ)-triester was
produced. The latter was converted into the (R)-(Ϫ)-lactone
acid (75% ee‡). The former on saponification and recrystalliz-
ation gave the required (S)-(ϩ)-lactone acid (98% ee,‡ 34% of
theory, overall).
The assignment of configuration in the lactone acids was
made by conversion of the recovered (R)-(ϩ)-triester to the
amide ester (Ϫ)-9 of known12 absolute stereochemistry
(Scheme 4).
Acknowledgements
We thank Professor Y. Vallée, Dr P. Léon, and Dr B. Hartmann
for their interest in our work, Professor K. J. Hale for some
unpublished information, and Dr A. Tatibouët for some
preliminary experiments. Financial support from the CNRS
(UMR 5616) and Aventis Pharma and a fellowship award (to
S. G.) from Aventis CropScience are gratefully acknowledged.
Scheme 4 Reagents and conditions: i, O3, CH2Cl2, Ϫ78 ЊC, then Me2S,
Ϫ78 ЊC
20 ЊC, 12 h, 55%; ii, NaClO2, Me2C᎐CHMe, NaH2PO4, t-
᎐
BuOH–H2O, 20 ЊC, 12 h, 86%; iii, 2-chloro-1-methylpyridinium iodide,
Et3N, Et2NH, CH2Cl2, 20 ЊC, 24 h, 73%; iv, H2, 10% Pd/C, EtOH, 20 ЊC,
6 h, 84%; v, MeOCH2Br, i-Pr2EtN, CH2Cl2, 20 ЊC, 48 h, 92%.
Notes and references
In summary, an effective synthesis of lactone acid (ϩ)-6
in highly enantio-enriched form has been realized as part of
a novel approach to the camptothecinoids. Amides of the
general structure II, readily prepared from this key compound,
are at present being subjected to N4–C3, C14–C15 bicycliz-
ation tactics, the results of which will be disclosed at a later
date.
† The IUPAC name for triflate is trifluoromethanesulfonate.
‡ The enantiomeric excess was determined by HPLC of the correspond-
ing ethyl ester: Chiralcel OD-H column, 5 mm, propan-2-ol–hexane =
10 : 90, 0.5 mL minϪ1, T r (R) 16.51 min, T r (S) 18.09 min.
1 M. E. Wall, M. C. Wani, C. E. Cook, K. H. Palmer, A. T. McPhail
and G. A. Sim, J. Am. Chem. Soc., 1966, 88, 3888.
2904
J. Chem. Soc., Perkin Trans. 1, 2001, 2903–2905

View Article Online
2 For recent reviews, see: Camptothecins: New Anticancer Agents,
eds. M. Potmesil and H. Pinedo, CRC Press, Boca Raton, 1995;
S. Sawada, T. Yokokura and T. Miyasaka, Curr. Pharm. Des., 1995,
1, 113; M. Torck and M. Pinkas, J. Pharm. Belg., 1996, 51, 200.
3 For a discussion of the various approaches, see: S. Leue, Ph.D.
Thesis, Université J. Fourier de Grenoble, 2000 .
4 A. Padwa, A. T. Price and L. Zhi, J. Org. Chem., 1996, 61, 2283.
5 M. P. Georgeadis, S. A. Haroutounian and C. D. Apostolopoulos,
Synthesis, 1991, 379.
7 H.-S. Byun, K. C. Reddy and R. Bittman, Tetrahedron Lett., 1994,
35, 1371.
8 J. Villiéras and M. Rambaud, Synthesis, 1982, 924.
9 T. Ben Ayed, H. Amri and M. M. El Gaied, Tetrahedron, 1991, 47,
9621.
10 For a preparation of racemic V (R = RЈ = H), see: M. R. Peel and
D. D. Sternbach, Bioorg. Med. Chem. Lett., 1994, 4, 2753.
11 For reviews on the use of PLE in organic chemistry, see: M. Ohno
and M. Otsuka, Org. React., 1989, 37, 1; C.-H. Wong and G. M.
Whitesides, Enzymes in Synthetic Chemistry, Pergamon, Oxford,
1994, pp. 66–68.
6 A. Quendo and G. Rousseau, Tetrahedron Lett., 1988, 29, 6443;
A. Quendo, S. M. Ali and G. Rousseau, J. Org. Chem., 1992, 57,
6890.
12 M. A. Ciufolini and F. Roschangar, Tetrahedron, 1997, 53, 11049.
J. Chem. Soc., Perkin Trans. 1, 2001, 2903–2905
2905