Enantioselective total synthesis and structure revision of spirodihydrobenzofuranlactam 1. Total synthesis of stachybotrylactam

Article abstract of DOI:10.1021/ol030039j

(Matrix presented) The enantioselective total synthesis and structure revision of spirodihydrobenzofuranlactam 1 and of its regioisomer 25 are presented. Optically pure (+)-Wieland-Miescher ketone was utilized to construct the AB bicyclic core in 10 steps

Full text of DOI:10.1021/ol030039j

ORGANIC
LETTERS
2003
Vol. 5, No. 10
1785-1788
Enantioselective Total Synthesis and
Structure Revision of
Spirodihydrobenzofuranlactam 1. Total
Synthesis of Stachybotrylactam
Andrew S. Kende,* Wei-Ping Deng, Min Zhong, and Xiao-Chuan Guo
Department of Chemistry, UniVersity of Rochester, Rochester, New York 14627
kende@chem.rochester.edu
Received March 13, 2003
ABSTRACT
The enantioselective total synthesis and structure revision of spirodihydrobenzofuranlactam 1 and of its regioisomer 25 are presented. Optically
pure (+)-Wieland−Miescher ketone was utilized to construct the AB bicyclic core in 10 steps. Introduction of the resorcylate D-ring unit was
achieved by use of our tert-butyl ester metalation sequence. Subsequent stereoselective spirocyclization to form the C-ring was followed by
regioselective ring cyanation and lactam formation to produce the pentacyclic structure 1 and its regioisomer 25.
A series of spirodihydrobenzofuranlactams, active as an-
tagonists of endothelin and as inhibitors of HIV-1 protease,
were isolated from the cultures of two different Stachybotrys
species by Roggo et al. in 19961a,b and assigned structures
1-5 on the basis of extensive NMR studies. The pseudo-
symmetric “dimer” 5 is the most potent representative of
this series. These novel spirodihydrobenzofuranlactams were
reported to contain the unique axial hydroxyl group at the
C-19 position and the benzofuranlactam moiety. Their
undetermined absolute configuration as well as their phar-
macological activity1a encouraged us to develop the first
synthetic access to these new compounds. We describe herein
an enantioselective total synthesis of structure 1 and its
regioisomer 25 as well as their comparisons with natural
material.
Our approach (Scheme 1) commenced with the preparation
Miescher ketone. Ketone 6 was then converted to the enol
of the AB ring segment. trans-Decalone 6 was prepared by
trimethylsilyl ether, which was reacted with methyl iodide
a published procedure4-6 from enantiopure (+)-Wieland-
and benzyltrimethylammonium fluoride (BTAF)4 to give the
R-methyl derivative 7 in 82% yield. Ketone 7 was treated
with hydrazine in ethanol at reflux to afford the hydrazone
(1) (a) Roggo, B. E.; Petersen, F.; Sills, M.; Roesel, J. L.; Moerker, T.;
Peter, H. H. J. Antibiot. 1996, 49, 13. (b) Roggo, B. E.; Hug, P.; Moss, S.;
Sta¨mpfli, A.; Kriemler, H.-P.; Peter, H. H. J. Antibiot. 1996, 49, 374.
8 in 84% yield, followed by treatment with iodine and DBU
10.1021/ol030039j CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/19/2003

Scheme 1. Construction of AB-Ring Segment 10
Scheme 2. Attempted Spiroannulation of C- and D-Rings
in anhydrous ethyl ether and then with DBU in benzene to
produce the vinyl iodide 9 in 86% yield.7 At this stage,
lithiation of 9 was followed by treatment with methyl
4-formyl-3,5-dibenzyloxybenzoate,3 but no reaction was
observed, presumably due to the steric repulsion between
the angular methyl and two benzyloxy groups on the benzene
ring. Alternatively, a formyl group was installed on the AB
ring by treatment of lithiated compound 9 with DMF in THF
at -78 °C to furnish the R,â-unsaturated aldehyde 10 in 83%
yield.
Bromine-lithium3 exchange of tert-butyl-4-bromo-3,5-
dibenzyloxybenzoate 118 in THF at -78 °C, followed by
quenching with R,â-unsaturated aldehyde 10, gave rise to
the carbinol 12 (Scheme 2). Removal of the secondary
benzylic hydroxyl group to afford compound 13 was
achieved using the NaBH3CN and ZnI2 system.9 Subsequent
treatment of 13 with thionyl chloride in excess methanol
produced the methyl ester 14 in 89% yield. Hydrogenation
of compound 14 by employing either 10% Pd/C or Pd(OH)2
resulted in mixtures due to cleavage of one to three benzyl
groups and reduction of the alkene. In contrast, Raney nickel9
selectively removed both benzyl groups on the benzene ring
to give compound 15 in 94% yield. Compound 15 was then
subjected to acid-catalyzed spiro-heteroannulation. Unfor-
tunately, either no reaction or complicated mixtures were
obtained by treatment of compound 15 using Corey’s method
(THF/ethylene glycol, 2 N HCl),10 McMurry’s method
(Amberlyst 15/DCM),11 or other acidic conditions (TFA/
CHCl3; Hg(OCOCF3)2/THF; PTS/benzene).
From the X-ray crystal structure of 15 (Scheme 3), two
benzylic side chains were found to lie on the same side of
the AB-bicyclic core, which indicates possible steric hin-
drance between these two benzylic side chains upon the
spiroannulation. We considered that the “remote” axial
benzyloxy group on the A-ring could be a factor in the failed
spiroannulation. With this in mind, compound 16 was then
prepared in quantitative yield by cleavage of the benzyl group
in 15 using 10% Pd/C.
Spiroannulation now proceeded smoothly (Scheme 3)
when compound 16 was treated with Amberlyst 15 in
dichloromethane at room temperature, giving the chromato-
graphically separable isomers benzofuran 17 and benzopyran
18 in a ratio of 4:5 (90% overall yield). A series of reaction
conditions were screened including variation of solvents and
temperatures to optimize the reaction. Finally, Amberlyst 15
with dichloromethane as solvent at 0 °C was found to afford
benzofuran and benzopyran in a ratio of 1.7:1 (yield of
benzofuran 60%), comparable to Corey’s10 and McMurry’s11
results in related systems. Protic solvents and dipolar aprotic
solvents gave benzofuran as the major product (up to 3:1),
but in low conversion (up to 30%); extension of the reaction
time in these solvents led to gradual decomposition without
increase of conversion. Control experiments showed the
generation of 17 and 18 is irreversible under our optimum
reaction conditions.
(2) Buchschacher, P.; Fu¨rst, A.; Gutzwiller, J. Organic Syntheses;
Wiley: New York, 1990; Collect. Vol. VII, p 368.
(3) Kende, A. S.; Zhong, M. Synth. Commun. 1999, 29, 3401.
(4) Smith, A. B., III; Mewshaw, R. J. Org. Chem. 1984, 49, 3685.
(5) (a) Kirk, D. N.; Petrow, V. J. Chem. Soc. 1962, 1091. (b) Shimizu,
T.; Hiranuma, S.; Hayashibe, S.; Yoshioka, H. Synlett 1991, 833.
(6) Yusui, K.; Kawada, K.; Kagawa, K.; Tokura, K.; Kitadokoro, K.;
Ikenishi, Y. Chem. Pharm. Bull. 1993, 41, 1698.
(7) Di Grandi, M. J.; Jung, D. K.; Krol, W. J.; Danishefsky, S. J. J. Org.
Chem. 1993, 58, 4989.
(8) Lampe, J. W.; Hughes, P. F.; Biggers, C. K.; Smith, S. H.; Hu, H. J.
Org. Chem. 1996, 41, 4572.
1786
Org. Lett., Vol. 5, No. 10, 2003

Scheme 3. Construction of C-Ring
Scheme 4. Construction of E-Ring
at C-5 or C-10. Therefore, we decided to test the possibility
that the natural spirolactam reported by Roggo actually
possessed structure 25 instead of 1. Structure 25 corresponds
to stachybotrylactam, a compound first isolated in 1995 by
Jarvis et al.14,15 With benzofuran 17 in hand, we could prepare
the regioisomer 25 by redirecting the ring bromination so
as to obtain the regioisomer 23 and then following the same
sequence as was employed to make compound 1. Since the
Regioselective aromatic bromination of benzofuran 17
using NBS in dichloromethane provided 19 in 82% yield,
together with 5-10% of chromatographically unseparable
dibromo compound. The structure of 19 was confirmed by
single-crystal X-ray analysis of the derived compound 20,
which was chemoselectively obtained in 95% yield by
treatment of 19 with benzyl bromide and K2CO3 in aceto-
nitrile. Reaction of bromide 20 with cuprous cyanide (10
equiv) in DMF at 100 °C10 gave nitrile 21 in 92% yield.
Spirodihydrobenzofuranlactam 1 was then obtained in 80%
yield by hydrogenation of nitrile 21 using PtO2/EtOH/
Scheme 5. Synthesis of Regioisomer 25
12
CHCl3 and subsequent lactamization with base (Scheme
4), and the structure 1 was confirmed by single-crystal X-ray
analysis.
1
To our surprise, the H NMR and 13C NMR spectra of
our synthetic compound 1 were not identical to those reported
for the natural spirodihydrobenzofuranlactam 1.13 A careful
reading of the assignment of the structure of natural product
1 reported by Roggo1b revealed that they had placed the sole
aromatic proton at C-9 based only on the absence of an NOE
from that aromatic proton to the methylene protons at C-5
or C-10. Actually, the unique regioisomer 25 also would lack
an NOE from that aromatic proton to the methylene protons
(9) Barrero, A. F.; Alvarez-Manzaneda, E. J.; Chahboun, R. Tetrahedron
1998, 54, 5635.
(10) Corey, E. J.; Das, J. J. Am. Chem. Soc. 1982, 104, 5551.
(11) McMurry, J. E.; Erion, M. D. J. Am. Chem. Soc. 1985, 107, 2712.
(12) Taishi, T.; Takechi, S.; Mori, S. Tetrahedron Lett. 1998, 39, 4347.
(b) Wrobel, J.; Dietrich, A.; Gorham, B. J.; Sestanj, K. J. Org. Chem. 1990,
25, 2694.
Org. Lett., Vol. 5, No. 10, 2003
1787

phenol group in 17 directed exclusively to ortho-bromination,
a bulky protecting group on the phenol might alter bromi-
nation regioselectivity. By using the bulky TBDPS ether 22,
we found that NBS and silica gel (excess) in DCM gave a
1:1 mixture of the two bromo compounds, which were easily
separable by chromatography. In the absence of silica gel,
or using other solvents or bromination reagents, inferior para
regioselectivity was observed. Subsequent desilylation and
benzyl ether formation after isomer separation gave 23.
Cyanation followed by nitrile reduction led to the lactam
25.16 Comparisons of 1H NMR and 13C NMR spectra of our
synthetic 25 established the structural identity of the latter
with the natural spirolactam isolated by Roggo et al.1 reported
to have structure 1 and with stachybotrylactam isolated by
Jarvis.15 The absolute configuration of the natural lactam
could not be established because neither the Roggo group
nor the Jarvis laboratories had reported its optical rotation.
However, a very closely related spirobenzofuranlactam
N-ethanol derivative reported by Jarvis has the negative sign
of optical rotation ([R]20 ) -16, c 0.1, MeOH), which is
D
very similar to that of our synthetic lactam 25 ([R]20
)
D
-21.3, c 1.10, MeOH). Thus, the absolute configuration of
natural stachybotrylactam is most likely correctly represented
by stereoformula 25.
Acknowledgment. We thank Drs. John Huffman (Indiana
University) and Rene Lachicotte (University of Rochester)
for X-ray analyses and Prof. B. Jarvis (University of
Maryland) and Dr. S. Roggo (Ciba-Geigy Ltd.) for kindly
1
sending us copies of H and 13C NMR spectra of natural
stachybotrylactam.
1
Supporting Information Available: Data including H
NMR and 13C NMR spectra for 1, 10, 13, 15, 17, 20, 21,
and 23-25. ORTEP plots of 1, 15, and 20. This material is
available free of charge via the Internet at http://pubs.acs.org.
(13) Data for 1: colorless crystals; [R]20 ) 0.5 (c 1.01, MeOH); mp
D
255 °C dec (MeOH/EtOAc); IR (cm-1) 3441, 3264, 2937, 1671; 1H NMR
(400 MHz, CD3OD) δ 0.74 (3H, d, J ) 6.4 Hz), 0.90 (3H, s), 1.00 (3H, s),
1.05 (3H, s), 1.05-1.09 (1H, m), 1.52-1.61 (5H, m), 1.83-1.97 (3H, m),
2.13-2.16 (1H, m), 2.86 (1H, AB, J ) 17.2 Hz), 3.25 (1H, AB, J ) 17.2
Hz), 3.35 (1H, br), 4.27 (2H, s), 6.66 (1H, s); 13C NMR (100 MHz, CD3OD)
δ 14.55, 15.08, 20.68, 21.50, 23.91, 24.57, 27.51, 30.80, 31.36, 37.16 (2C),
39.93, 42.04, 42.88, 75.00, 93.92, 96.97, 117.80, 122.73, 132.78, 148.16,
163.08, 172.81; APCI 386 ([MH]+, 100). Anal. Calcd for C23H31NO4‚
1/2CH3OH (shown by X-ray): C, 70.30; H, 8.28. Found: C, 70.26; H, 8.49.
(14) Direct comparison of Jarvis’s spectroscopic data with Roggo’s data
was impossible due to the different solvent systems used in measuring the
NMR spectra. Roggo et al. have recently independently concluded that their
lactam 1 is structurally identical to the stachybotrylactam reported by
Jarvis.15 We have been unable to obtain actual reference samples of either
the Roggo spirolactam or the Jarvis stachybotrylactam.
OL030039J
(16) Data for 25: amorphous solid; [R]20 ) -21.3 (c 1.10, MeOH);
D
mp 210 °C dec (EtOAc); IR (cm-1) 3460, 3264, 2935, 1681; 1H NMR
(400 MHz, DMSO-d6) 0.64 (3H, d, J ) 6.4 Hz), 0.80 (3H, s), 0.88 (3H, s),
0.95 (3H, s), 0.88-0.95 (1H, m), 1.35-1.52 (5H, m), 1.69-1.85 (3H, m),
2.01 (1H, m), 2.75 (1H, AB, J ) 16.8 Hz), 3.10 (1H, AB, J ) 16.8 Hz),
3.18 (1H, br), 4.09 (1H, AB, J ) 16.8 Hz), 4.10 (1H, br), 4.20 (1H, AB,
J ) 16.8 Hz), 6.55 (1H, s), 8.35 (1H, br), 9.70 (1H, br); 13C NMR (100
MHz, DMSO-d6) 15.91, 16.016, 20.83, 22.78, 24.22, 25.29, 29.05, 31.15,
32.313, 36.85, 37.70, 39.62, 42.18, 42.41, 73.84, 98.08, 101.23, 114.38,
116.89, 134.61, 154.38, 156.49, 170.61; APCI 386 ([MH]+, 100). Anal.
Calcd: C, 71.66; H, 8.11. Found: C, 71.50; H, 7.92.
(15) Jarvis, B. B.; Salemme, J.; Morais, A. Nat. Toxins 1995, 3, 10.
1788
Org. Lett., Vol. 5, No. 10, 2003