Synthesis of ent-alantrypinone

Article abstract of DOI:10.1021/ja010066+

This paper presents a synthesis of ent-alantrypinone (ent-6), the enantiomer of a natural product produced by the fungus Penicillium thymicola. The synthesis revolves around the Li[Me₃AlSPh]-promoted isomerization of iminobenzoxazine 33 to quin

Full text of DOI:10.1021/ja010066+

5892
J. Am. Chem. Soc. 2001, 123, 5892-5899
Synthesis of ent-Alantrypinone†
David J. Hart* and Nabi A. Magomedov
Contribution from the Department of Chemistry, The Ohio State UniVersity, 100 West 18th AVenue,
Columbus, Ohio 43210
ReceiVed January 8, 2001
Abstract: This paper presents a synthesis of ent-alantrypinone (ent-6), the enantiomer of a natural product
produced by the fungus Penicillium thymicola. The synthesis revolves around the Li[Me3AlSPh]-promoted
isomerization of iminobenzoxazine 33 to quinazolinone 34, an N-acyliminium ion cyclization that converts
enamide 9 to bridged indole 35, and rearrangement of 35 to oxindole ent-6. Ancillary chemistry that involves
thermal fragmentation of an iminobenzoxazine to a nitrile ylide and Me2AlSPh-mediated cyclization of oxime
ether-ester 22 to pyrrolidinone 23 is also described.
Introduction
[2,1-b]quinazoline substructure of fumiquinazoline F. Spiro-
quinazoline (4) is a structurally unique alkaloid, isolated from
extracts of Aspergillus flaVipes.7 We imagine that Nature might
produce spiroquinazoline by an oxidative cyclization of fumi-
quinazoline F via the formal equivalent of an N-acyliminium
ion of type 5. Finally, alantrypinone (6) is a structurally simpler
relative of spiroquinazoline, recently isolated from Penicillium
thymicola, that clearly could be produced biosynthetically via
a similar oxidative cyclization.8
Fungi belonging to the genera Aspergillus and Penicillium
serve as a rich source of alkaloids clearly derived from amino
acids.1 A number of these alkaloids incorporate anthranilic acid
and tryptophan as amino acid components. Some classical
examples include the tryptoquivalines,2 aszonalenins,3 asper-
licins,4 and ardeemins.5 More recently a series of alkaloids
known as the fumiquinazolines have been reported.6 These vary
in structural complexity, but all seem to be biosynthetically
derivable from fumiquinazoline F (1) and its C3-epimer fumi-
quinazoline G. For example fumiquinazoline B (2) has an
alanine appended to the indole nitrogen, and oxidative cycliza-
tion has occurred onto the indole ring. This structural feature
also appears in the tryptoquivalines. One can imagine that
fumiquinazoline C (3) is derived from 2 via an oxidative
cyclization resulting in bridging of C3 and C14 in the pyrazino-
† This paper is dedicated to Professor David A. Evans on the occasion
of his 60th birthday.
(1) Yamamoto, Y.; Arai, K. In The Alkaloids; Brossi, A., Ed.; Academic
Press: Orlando, 1986; Vol. 29, p 185.
(2) Clardy, J.; Springer, J. P.; Buchi, G.; Matsuo, K.; Wightman, R. J.
Am. Chem. Soc. 1975, 97, 663. Buchi, G.; Luk, K. C.; Kobbe, B.; Townsend,
J. M. J. Org. Chem. 1977, 42, 244. Yamazaki, M.; Fujimoto, H.; Okuyama,
E. Tetrahedron Lett. 1976, 2861. Yamazaki, M.; Fujimoto, H.; Okuyama,
E. Chem. Pharm. Bull. 1977, 25, 2554. Yamazaki, M.; Fujimoto, H.;
Okuyama, E. Chem. Pharm. Bull. 1978, 26, 111. Yamazaki, M.; Okuyama,
E.; Maebayashi, Y. Chem. Pharm. Bull. 1979, 27, 1611. Fujimoto, H.;
Negishi, E.; Yamaguchi, K.; Nishi, N.; Yamazaki, M. Chem. Pharm. Bull.
1996, 44, 1843.
(3) Kimura, Y.; Hamasaki, T.; Nakajima, H.; Isogai, A. Tetrahedron Lett.
1982, 23, 225.
(4) Chang, R. S. L.; Lotti, V. J.; Monaghan, R. L.; Birnbaum, J.; Stapley,
E. O.; Goetz, M. A.; Albers-Schonberg, G.; Patchett, A. A.; Liesch, J. M.;
Hemsens, O. D.; Springer, J. P. Science 1985, 230, 177. Goetz, M. A.;
Lopez, M.; Monaghan, R. L.; Chang, R. S. L.; Lotti, V. J.; Chen, T. B. J.
Antibiot. 1985, 38, 1633. Liesch, J. M.; Hensens, O. D.; Springer, J. P.;
Chang, R. S. L.; Lotti, V. J. J. Antibiot. 1985, 38, 1638. Goetz, M. A.;
Monaghan, R. L.; Chang, R. S. L.; Ondeyka, J.; Chen, T. B.; Lotti, V. J. J.
Antibiot. 1988, 41, 875. Liesch, J. M.; Hensens, O. D.; Zink, D. L.; Goetz,
M. A. J. Antibiot. 1988, 41, 878.
(5) Karwoski, J. P.; Jackson, M.; Rasmussen, R. R.; Humphrey, P. E.;
Poddig, J. B.; Kohl, W. L.; Scherr, M. H.; Kadam, S.; McAlpine, J. B. J.
Antibiotics 1993, 46, 374. Hochlowski, J. E.; Mullally, M. M.; Spanton, S.
G.; Whittern, D. N.; Hill, P.; McAlpine, J. B. J. Antibiot. 1993, 46, 380.
(6) Numata, A.; Takahashi, C.; Matsushita, T.; Miyamoto, T.; Kawai,
K.; Usami, Y.; Matsumura, E.; Inoue, M.; Ohishi, H.; Shingu, T.
Tetrahedron Lett. 1992, 33, 1621. Takahashi, C.; Matsushita, T.; Doi, M.;
Minoura, K.; Shingu, T.; Kumeda, Y.; Numata, A. T. Chem. Soc., Perkin
Trans. 1995, 2345.
In addition to their interesting structures, a number of the
aforementioned compounds are biologically active. For example,
spiroquinazoline (4) is a competitive inhibitor of substance P
binding at the human NK-1 receptor and thus a potential lead
(7) Barrow, C. J.; Sun, H. H. J. Nat. Prod. 1994, 57, 471.
(8) Larsen, T. O.; Frydenvang, K.; Frisvad, J. C.; Christophersen, C. J.
Nat. Prod. 1998, 61, 1154.
10.1021/ja010066+ CCC: $20.00 © 2001 American Chemical Society
Published on Web 05/30/2001

Synthesis of ent-Alantrypinone
J. Am. Chem. Soc., Vol. 123, No. 25, 2001 5893
Scheme 1
Scheme 2
compound for development of analgesics.7,9 We have initiated
a synthetic program with the goal of developing efficient
methods for the preparation of spiroquinazoline and related
substances.10,11 Although we have yet to reach spiroquinazoline,
a synthesis of the enantiomer of alantrypinone is in hand.11 This
paper presents the details of this synthesis and some of the
observations made along the way.
From the start, our studies were guided by the biosynthetic
hypotheses described above and thus one of our objectives was
to generate N-acyliminium ions of type 5 to observe their
behavior. Although several routes to such ions were studied,
this paper will describe the chemistry of ions derived from
protonation of enamides of type 9, where R represents a
hydrogen or an appropriate nitrogen protecting group (eq 1).
We initially hoped to prepare 9 via cyclization of quinazolinones
7 and 8, and thus, these compounds were set as initial targets
for synthesis.
anhydride gave amide 11 in 97% yield.12 Treatment of 11 with
the aluminum amide derived from trimethylaluminum and
allylamine provided 12 in 87% yield.13 Acylation of both 11
and 12 with pyruvoyl chloride provided 13 (81%) and 14
(81%).14 Treatment of 13 with O-benzylhydroxylamine gave
15 in quantitative yield.
The cyclodehydration of anthranilic acid derivatives 14 and
15 to quinazolinones 7 and 8 was problematic, but interesting.
On the basis of methodology reported by Ganesan during the
course of a synthesis of fumiquinazoline G (C3-epi-1), it was
hoped that treatment of these amides with triphenylphosphine
(Ph3P)-iodine (I2) in the presence of Hunig’s base (DIPEA)
would effect the desired transformation.15 In the event, treatment
of 14 (C24H24N4O4) with this cyclodehydration system provided
a product (C24H22N4O3), initially suspected to be quinazolinone
7, in 56% yield (Scheme 2). This proved not to be the case.
We were initially surprised that this product showed no
propensity for intramolecular cyclization of the N-allylamide
nitrogen onto the carbonyl group of the methyl ketone. In an
attempt to effect cyclodehydration of this product (suspected
to be 7) to an enamide of type 9 (R ) allyl), this substance was
eventually warmed at 110-115 °C in DMSO. A rather clean
isomerization was observed in 80% yield. It was difficult to
deduce the structure of this rearrangement product solely on
the basis of spectral data. Crystallization of the isomerization
Preliminary Studies. It was hoped that 14 and 15 would be
the penultimate intermediates en route to 7 and 8, respectively.
It was imagined that cyclodehydration of these intermediates,
using known methodology, would provide the desired quinazoli-
nones. Syntheses of 14 and 15 are described in Scheme 1. Thus,
reaction of the methyl ester of (S)-tryptophan with isatoic
(9) Cascieri, M. A.; Macleod, A. M.; Underwood, D.; Shiao, L.-L.; Ber,
E.; Sadowski, S.; Yu, H.; Merchant, K. J.; Swain, C. J.; Strader, C. D.;
Fong, T. M. J. Biol. Chem. 1994, 269, 6587. Snider, R. M.; Constantine, J.
W.; Lowe, J. A., III; Longo, K. P.; Lebel, W. S.; Woody, H. A.; Drozda,
S. E.; Desai, M. C.; Vinick, F. J.; Spencer, R. W.; Hess, H.-J. Science 1991,
251, 435.
(10) Hart, D. J.; Magomedov, N. A. J. Org. Chem. 1999, 64, 2990. Freed,
J.; Hart, D. J.; Magomedov, N. A. J. Org. Chem. 2001, 66, 839.
(11) Hart, D. J.; Magomedov, N. A. Tetrahedron Lett. 1999, 40, 5429.
(12) Bhat, B.; Harrison, D. M. Tetrahedron 1993, 49, 10655.
(13) Basha, A.; Lipton, M.; Weinreb, S. M. Tetrahedron Lett. 1977, 18,
4171.
(14) Ottenheijm, H. C. J.; de Man, J. H. M. Synthesis 1975, 163.
(15) Wang, H.; Ganesan, A. J. Org. Chem. 1998, 63, 2432.

5894 J. Am. Chem. Soc., Vol. 123, No. 25, 2001
Hart and MagomedoV
Scheme 3
derived methyl group, explaining the abnormally upfield chemi-
cal shift of the methyl group. In addition, 8 crystallizes in a
conformation in which the ester carbonyl projects away from
the nitrogen of the oxime ether. This could explain the reluctance
of 8 to undergo cyclodehydration to a structure of type 9 (vide
infra).
The results described above suggest that, once again, the
original cyclodehydration conditions (Ph3P-I2-DIPEA) had
transformed 15 into iminobenzoxazine 20, rather than quinazoli-
none 8, and that the observed isomerization was actually the
conversion of 20 into 8.18 It was imagined that this might have
occurred by Lewis acid assisted nucleophilic opening of 20 to
an intermediate such as 21, followed by cyclization to provide
8. Whereas the mechanism of this transformation is merely
conjecture, the generality of this rearrangement process has been
established (vide infra).
Some Comments and Diversions. At this point in our
studies, a literature search revealed that the cyclodehydration
of N-acylanthranilamides to iminobenzoxazines using Ph3P-
Br2-Et3N had been reported by Mazurkiewicz, as well as the
acid-promoted rearrangement of these iminobenzoxazines to
quinazolinones.19 This report, and the observations noted in
Schemes 2 and 3, suggested that the cyclodehydration results
reported in Ganesan’s elegant fumiquinazoline G synthesis were
suspect.15 Indeed, this was nicely shown to be the case by the
Snider group and was later recognized by the authors of the
fumiquinazoline G synthesis.20,21
product from acetonitrile, however, provided a suitable single
crystal for X-ray crystallographic analysis, which unambiguously
determined its structure as 16.16 With the structure of 16
established, it was possible to postulate a mechanism for its
formation. Given that 16 was racemic, it was felt that the
oxazoline substructure might be derived from an intramolecular
[3 + 2]-cycloaddition between the nitrile ylide and ketone
groups in achiral intermediate 19. It seemed unlikely that 19
could be derived from 7 but reasonable that it be formed from
iminobenzoxazine 17. For example, fragmentation of 17 would
provide 18 and a subsequent proton transfer would then provide
19. This suggested that the original cyclodehydration of 14 had
not provided quinazolinone 7 but had in fact provided imi-
nobenzoxazine 17.
We also note that attempts to convert 8 to 9 (R ) OBn) were
unsuccessful. For example, we were able to develop a nucleo-
philic addition-ring closure procedure for the cyclization of
ethyl levulinate derivative 22 to 23 in 55% yield (eq 2) using
Concurrent with the studies shown in Scheme 2, we were
examining the cyclodehydration of 15 with the intention of
preparing quinazolinone 8 (Scheme 3). Thus, treatment of 15
(C29H26N4O5) with Ph3P-I2-DIPEA gave a 90% yield of a
product (C29H24N4O4) originally suspected to be quinazolinone
8. Once again this was not the case. This product had all the
spectroscopic features one might expect to see in 8 but exhibited
unexpected chemical behavior. It was sensitive to aqueous acid
and could not be induced to undergo cyclodehydration to an
enamide of type 9 (R ) OBn). In one attempt to accomplish a
nucleophile-induced cyclization, this product (suspected to be
8) was treated with 3 equiv of lithium trimethyl(phenylsulfido)-
aluminate (Li[Me3AlSPh]) for 20 h at room temperature.17 This
reaction provided an isomeric material in 90% yield. The
spectroscopic features of this isomerization product were also
remarkably consistent with what one would expect for 8, with
the exception that the pyruvate-derived methyl group in this
isomer (δ 1.19 in C6D6) was significantly more shielded than
in the starting isomer (δ 2.02 in C6D6). X-ray crystallographic
analysis of the rearrangement product unambiguously revealed
that the rearrangement product was 8 and, thus, the original
cyclodehydration product was something else!16 The crystal
structure of 8 also revealed that this compound crystallizes in
a conformation with the indolyl group folded over the pyruvate-
dimethylaluminum thiophenoxide (Me2AlSPh).22 Treatment of
8 with Me2AlSPh under identical conditions, however, returned
only starting material. One explanation for the lack of reactivity
of 8 evolves from analysis of its crystal structure.16 The crystal
structure of 8 reveals that the dihedral angle between the plane
of the quinazolinone ring and the plane of the CdN π-bond is
about 45°. This suggests that there is no through-conjugation
between the oxime oxygen and the quinazolinone unit. With
this in mind, a possible explanation for the failure of Me2AlSPh
to transform 8 to 9 might involve steric congestion inherent in
(18) Mass spectrometry provided another clear method for distinguishing
between tryptophan-derived iminobenzoxazines and quinazolinones. For
example, quinazolinone 8 undergoes a McLafferty rearrangement and
exhibits a peak due to a cation-radical at m/z 201, which loses a methoxy
radical to give a strong peak at m/z 170. These peaks are absent in
iminobenzoxazine 20. Both 8 and 20 show large peaks due to the
3-indolylmethyl cation at m/z 130.
(16) We thank Dr. Judith C. Gallucci for performing X-ray crystal-
lographic analyses of 8 and 16 at the Ohio State University Department of
Chemistry X-ray Crystallography Facility. Details are provided in the
Supporting Information.
(17) Itoh, A.; Ozawa, S.; Oshima, K.; Nozaki, H. Tetrahedron Lett. 1980,
21, 361.
(19) Mazurkiewicz, R. Monatsh. Chem. 1989, 120, 973.
(20) He, F.; Snider, B. B. J. Org. Chem. 1999, 64, 1397.
(21) Wang, H.; Ganesan, A. J. Org. Chem. 2000, 65, 1022.
(22) Prepared by analogy with the preparation of Et2AlSPh: Yasuda,
A.; Takahashi, M.; Takoya, H. Tetrahedron Lett. 1981, 22, 2413.

Synthesis of ent-Alantrypinone
J. Am. Chem. Soc., Vol. 123, No. 25, 2001 5895
2,3-disubstituted quinazolinones. One can imagine that addition
of an external nucleophile across the C9-N3 double bond would
increase steric congestion in the system upon going from sp2
to sp3 hybridization at C9.23
The rearrangement of iminobenzoxazine 17 to quinazolinone
7 was also examined. Treatment of 17 with Li[Me3AlSPh]
provided a complex mixture of products that was not possible
to characterize.24 Whatever the actual composition of this
complex mixture, upon treatment with trifluoroacetic acid in
chloroform (1:10) at reflux, 24 was obtained in approximately
50% yield (eq 3). The structure of 24 was assigned principally
The results described above were both encouraging and
discouraging. Whereas we had not been able to develop a
reaction sequence that provided isolable enamides of type 9, a
reliable procedure for the synthesis of highly functionalized
quinazolinones was in hand, and the observations presented in
eq 3 suggested that generation and cyclization of the desired
N-acyliminium ion was feasible. Thus, we abandoned pyruvic
acid derived precursors of enamide 9 and focused on precursors
where the oxidation state needed for generation of the N-
acyliminium ion was displaced by one carbon. Thus, sulfide
28 was set as the next target for synthesis in anticipation that it
could be converted to 9 (R ) H) using elimination chemistry
(eq 5).
1
on the basis of H NMR spectroscopy and eventually by a
ent-Alantrypinone (ent-6). The synthesis of 28 required
preparation of cysteine-derived acid chloride 31. This was
accomplished in two steps from commercially available S-
methyl-L-cysteine (29) using the standard Carpino protocol.25
Acylation of 29 with FmocCl in aqueous dioxane provided
urethane 30 in 98% yield. Reaction of 30 with excess thionyl
comparison of its 1H NMR spectrum with that of bridged indole
35 (vide infra). We imagine that 24 results from formation of
7, cyclization to a variety of compounds that could serve as
precursors to the carbocation ion derived from enamide 9 (R )
allyl),24 acid-mediated generation of such a carbocation, and
an intramolecular electrophilic aromatic substitution reaction.
This process, however, was messy and thus a longer, but more
controlled reaction sequence was sought.
One option that was explored involved conversion of 14 to
oxime ether 25, followed by cyclodehydration to afford ben-
zoxazine 26 in 92% overall yield (eq 4). Application of the
chloride furnished a 91% yield of crystalline acid chloride 31.
Acylation of anthranilic acid derivative 11 (vide supra) with
31 under Schotten-Baumann conditions provided a 96% yield
of 32 (Scheme 4). Cyclodehydration of 32 with Ph3P-I2-
DIPEA in dichloromethane at room temperature gave an 80%
yield of iminobenzoxazine 33. Treatment of 33 with 10 equiv
of Li[Me3AlSPh] in tetrahydrofuran at -78 °C, then at -10
°C for 12 h, and finally at room temperature for 8 h provided
a 46% yield of 28. A better conversion of 33 to 28 was achieved
under more selective conditions for the rearrangement. Thus,
treatment of 33 with 5 equiv of the aluminum complex at -78
°C for 30 min and then at -10 °C for 36 h provided a 76%
yield of quinazolinone 34 in addition to 8% of 28. Quinazolinone
34 was cleanly transformed into 28 using standard Fmoc
removal conditions (piperidine in tetrahydrofuran at 0 °C for
90 min).26 Oxidation of 28 with m-chloroperoxybenzoic acid
provided a 3:2 mixture of diastereomeric sulfoxides, and gentle
reflux of a suspension of these sulfoxides in benzene containing
triphenylphosphine provided enamide 9 (R ) H) in 79% yield
from 28.27 Triphenylphosphine was required to intercept meth-
benzoxazine-quinazolinone rearrangement then provided 27 in
93% yield. To our surprise, we were unable to induce 27 to
cyclize, even under such drastic conditions as reflux in
chloroform-trifluoroacetic acid (10:1). After such treatment,
unchanged starting material could be completely recovered.
(24) The anticipated quinazolinone (7) has keto and amide groups in
close proximity and might exist as a pair of diasteromeric cyclols. The 1H
NMR of the rearrangement mixture showed that some of the products had
incorporated a thiophenoxy group. Such compounds could be derived from
the cyclols by exchange of a hydroxyl group with a thiophenoxy group.
(25) Carpino, L. A.; Cohen, B. J.; Stephens Jr., K. E.; Sadat-Aalaee, S.
Y.; Tien, J.-H.; Langridge, D. C. J. Org. Chem. 1986, 51, 3732.
(26) Carpino, L. A. Acc. Chem. Res. 1987, 20, 401.
(23) As noted above 8 was uneffected by Li[Me3AlSPh] and also by
trimethylsilylthiophenoxide in the presence of catalytic amounts of tri-
methylsilyl triflate. It was possible to convert 8 to the corresponding
carboxylic acid using either lithium hydroxide in aqueous tetrahydro-
furan (42% with 44% recovery of starting material) or molten imidazole at
160-170 °C (90%). Attempts to convert this acid to 9 by activation of the
carboxylic acid were also unsuccessful.
(27) Rich, D. H.; Tam, J. P. J. Org. Chem. 1977, 42, 3815.

5896 J. Am. Chem. Soc., Vol. 123, No. 25, 2001
Hart and MagomedoV
Scheme 4
Scheme 5
resulted in enhancement of the signals due to H19 and H2 (δ
9.55). The downfield chemical shift of the C3 methyl group was
surprising but can be explained by its orientation relative to
the deshielding region of the indole nucleus. We note that the
1H NMR spectrum of 24 (vide supra) was a close match with
the spectrum of 35, including the chemical shift of the C3 methyl
group.
anesulfenic acid, and omission of this reagent reduced the yield
of 9 by 20-30%.
A stereochemical aspect of the chemistry presented in Scheme
4 deserves comment. Although Li[Me3AlSPh] is presumably a
nonbasic reagent, the presence of a potentially epimerizable
tryptophan-derived methine in 33 led us to look carefully for
epimerization products related to 28. In one experiment,
conducted on a large scale, the C3 epimer of 28 was isolated in
To complete the synthesis of alantrypinone, it was necessary
to conduct oxidative rearrangement of the indole to an oxindole.
A number of procedures for conducting this transformation are
known. We settled on N-bromosuccinimide (NBS) in aqueous
acid as a medium for accomplishing this task.32 Reaction of 35
with 1.1 equiv of NBS in a mixture of water, acetic acid, and
tetrahydrofuran (1:1:1) did not give the desired oxindole but
rather provided a 3:2 mixture of diastereomeric bromoindolines
36. Treatment of the mixture of bromoindolines with strong
acids (HBr in DME or TFA in THF) at 0 °C resulted in reversal
of the transformation to give starting 35. Eventually it was found
that treatment of 35 with five 1-equiv portions of NBS in a
mixture of tetrahydrofuran-trifluoracetic acid-water (5:4:4) at
0 °C accomplished the desired oxidative rearrangement. This
reaction, however, was accompanied by polybromination of the
indolinone (principally at C23).33 Thus, the crude reaction
product was hydrogenolyzed over platinum on carbon to give,
after separation by column chromatography, ent-alantrypinone
(ent-6) and ent-17-epialantrypinone (37) in 30% and 44% yields,
respectively.34,35 The relative configuration of the spiro center
(C17) in these stereoisomers was readily assigned on the basis
1
4% yield. H NMR spectroscopy could be used to distinguish
the two diastereomers. Compound 28 probably exists in a
boatlike conformation with the indole unit folded over the
pseudoaxial CH2SMe substituent.28 As a result, one of the
protons of the methylene R to sulfur is highly shielded and
appears at δ 0.32 in CDCl3 (the other appears at δ 2.60). In the
C3 epimer of 28, neither of the protons of the methylene under
consideration is shielded (δ 2.37 and 2.67). To establish that
the minor product was epimeric to 28 at C3 rather than C14,
this sulfide was oxidized to a mixture of sulfoxides which were
converted to 9 in 75% yield upon refluxing in toluene. The
optical rotation of this material matched that of 9 derived from
sulfide 28, both in sign and magnitude, providing clear evidence
that the epimerization had occurred at C3.29,30
The synthesis of ent-alantrypinone was completed as shown
in Scheme 5. Treatment of 9 (R ) H) with trifluoroacetic acid
in chloroform (1:20) under reflux for 2 h effected intramolecular
electrophilic attack of a presumed intermediate N-acyliminium
ion onto the indole to provide 35 in 89% yield. The orientation
of the indole nucleus was proven by difference NOE experi-
ments.31 Irradiation of H19 (δ 11.20 in DMSO-d 6) gave
enhancements of signals due to H21 (δ 7.37) and the C3 methyl
group (δ 2.12). In addition, irradiation of the C3 methyl group
1
of their H NMR spectra. In the spectrum of ent-17-epialant-
rypinone (37), H23 and H24 appeared at abnormally high field
(δ 6.66 and 5.87, respectively, in DMSO-d 6), a result of the
(31) It is possible that this cyclization proceeds mechanistically via initial
attack of the electrophile at C3 of the indole, followed by a Wagner-
Meerwein shift. For related cyclizations in substituted diketopiperizines
see: Ottenheijm, H. C. J.; Plate, R.; Noordik, J. H.; Herscheid, J. D. M. J.
Org. Chem. 1982, 47, 2147.
(32) Pellegrini, C.; Strassler, C.; Weber, M.; Borschberg, H.-J. Tetra-
hedron: Asymmetry 1994, 5, 1979. Stahl, R.; Borschberg, H.-J.; Acklin, P.
HelV. Chim. Acta 1996, 79, 1361.
(28) For related observations of shielding in diketopiperazines see:
Anteunis, M. J. O. Bull. Soc. Chim. Belg. 1978, 87, 627.
(29) We note that a variation of the conditions used to convert
iminobenzoxazines to quinazolinones by others were problematic when
applied to 33. For example, treatment of 33 with a 10% solution of piperidine
in acetonitrile at room temperature for 1.5 h provided a crude mixture of
materials. Heating these materials in acetonitrile at reflux for 18 h provided
enamide 9 (R ) H) in 40% yield (from 33). These conditions, however,
resulted in significant racemization as the optical rotation of this material
(9) was one-fifth of that recorded for material prepared using Li[Me3AlSPh].
(30) A synthesis of 9 (R ) H) that uses a different approach has been
reported: He, F.; Snider, B. B. Synlett 1997, 483.
(33) When only 2 equiv of NBS was used, under much the same
conditions, bromination of the aromatic ring was limited largely to C23.
However, small amounts of unreacted 35 interfered with isolation of pure
ent-6 (after debromination), complicating optical rotation measurements due
to the large positive specific rotation of 35. Thus, the polybromination-
debromination protocol was preferred.
(34) The hydrogenolysis was also conducted using Pd/C, but the rate of
the reaction was slower than with Pt/C.
(35) Over a series of experiments, the ratios of 37:ent-6 ranged from
50:50 to 60:40.

Synthesis of ent-Alantrypinone
J. Am. Chem. Soc., Vol. 123, No. 25, 2001 5897
stereochemical relationship of these protons to the π-system of
the quinazolinone substructure. In addition, the mobility of 37
on silica gel was much lower than that of ent-6, presumably a
reflection of the greater steric accessibility of the oxindole
carbonyl group (in 37) for binding to the solid support. Finally,
we note that the spectral and physical properties of ent-6 were
identical to those reported for the natural product with the
exception of the sign of the specific rotation.8,36 This confirms
the absolute configuration of the natural product as 3R,14R as
shown in structure 6.37 This also means that the biosynthesis of
alantrypinone either starts with the less common D-tryptophan
or that L-tryptophan is used and an isomerization occurs along
the biosynthetic pathway.38
we are currently focusing on the conversion of alantrypinone
or 35 to spiroquinazoline.
Conclusions
This paper describes a 10-step synthesis of ent-alantrypinone
(ent-6) that proceeds in 12% yield and confirms the absolute
configuration of the natural product. The studies describe several
observations of interest to the field of heterocyclic chemistry
including a new method for rearrangement of iminobenzoxazines
to quinazolinones that has some advantages over existing
methodology, fragmentation of an iminobenzoxazine to a nitrile
ylide, and an example of a nucleophile-induced cyclization of
an oxime-ester to a highly functionalized lactam.
As alluded to in the Introduction, one objective of this
research yet to be accomplished is a synthesis of spiroquinazo-
line (4). With this objective in mind, enamide 9 (R ) H) was
converted to N-acylindole 38 in 25-34% yield upon treatment
Experimental Section40
(()-1-(Benzyloxy)-5-methyl-5-(phenylthio)-2-pyrrolidinone (23).
To a stirred solution of 115 µL of PhSH in 1 mL of dry CH2Cl2 was
added dropwise 0.56 mL of a 2 M solution of AlMe3 in toluene over
a period of 3 min at room temperature under an argon atmosphere.
The solution was stirred at room temperature for 10 min, then cooled
to -78 °C, and transferred via cannula into a cooled (-78 °C) solution
of 200 mg (0.80 mmol) of 22 in 1 mL of dry CH2Cl2. The cooling
bath was removed and the reaction mixture was stirred for 1 h 20 min.
After that time, the reaction mixture was quenched with 1 mL of 1 N
aqueous HCl and partitioned between 100 mL of CH2Cl2 and 10 mL
of water. The organic layer was washed with 10 mL of water, dried
(MgSO4), and concentrated in vacuo. The residue was chromatographed
over 20 g of flash silica gel (gradient elution with Et2O:hexanes, 1:3,
then 1:1, then neat Et2O) to yield 138 mg (55%) of pyrrolidone 23 as
a thick colorless oil: IR (neat) 3060, 2975, 2938, 2880, 1726, 1453,
1374, 1062 cm-1; H NMR (acetone-d6, 300 MHz) δ 1.27 (ddd, J )
1
17.0 Hz, 10.0 Hz, 9.0 Hz, 1H, CHHCS), 1.63 (s, 3H, CH3), 1.97 (ddd,
J ) 17.0 Hz, 9.8 Hz, 2.1 Hz, 1H, CHHCS), 2.21 (ddd, J ) 13.6 Hz,
9.8 Hz, 9.0 Hz, 1H, COCHH), 2.43 (ddd, J ) 13.6 Hz, 10.0 Hz, 2.1
Hz, 1H, COCHH), 5.05 (d, J ) 9.9 Hz, 1H, CHHO), 5.29 (d, J ) 9.9
Hz, 1H, CHHO), 7.34-7.48 (m, 6H, ArH), 7.52-7.55 (m, 2H, ArH),
7.58-7.61 (m, 2H, ArH); 13C NMR (acetone-d6, 75.5 MHz) δ 26.7
(t), 27.2 (q), 32.9 (t), 75.4 (s), 78.9 (t), 129.3 (d), 129.3 (d), 129.5 (d),
130.0 (d), 130.0 (d), 130.2 (d), 130.2 (d), 130.4 (d), 131.9 (s), 136.6
(s), 138.1 (d), 138.1 (d), 170.6 (s); mass spectrum (EI), m/z (relative
intensity) 314 (M + 1+, 0.03), 204 (14), 179 (10), 110 (25), 109 (11),
92 (10), 91 (100), 77 (14), 65 (13), 51 (8), 43 (11); exact mass calcd
for C18H20NO2S (M + 1+) m/z 314.1215, found m/z 314.1261.
with p-nitrophenyl N-Cbz-glycinate in the presence of KF, 18-
C-6, and DIPEA in acetonitrile.39 We have not yet been able,
however, to accomplish the conversion of 38 to spiroquinazoline.
For example, treatment of 38 with either TFA-chloroform (1:
10) or p-toluenesulfonic acid in benzene, at reflux, provided
complex mixtures of products. These mixtures were subjected
to catalytic hydrogenolysis (Pd/C) to remove Cbz groups, but
a comparison of the 1H NMR spectra of the crude products with
the reported spectrum of 4 indicated no formation of the target
compound. In addition, treatment of enamide 9 with Amberlyst-
15 in aqueous acetonitrile, with the hope of intercepting the
spiroindoline 39 (expected to be an intermediate en route from
9 to 35) in a Ritter-type reaction, provided exclusively 35. Thus,
N-[N-[(R)-2-Carboxyamino-3-(methylthio)propionyl]anthraniloyl]-
L-tryptophan, N-(Fluoren-9-ylmethyl) Methyl Ester (32). To a
vigorously stirred two-phase system consisting of a solution of 432
mg (1.28 mmol) of aniline 11 in 12 mL of CH2Cl2 and 12 mL of 10%
aqueous NaHCO3 was added dropwise a solution of 529 mg (1.41
mmol) of acid chloride 31 in 10 mL of CH2Cl2 over a period of 5 min
at room temperature. The resulting mixture was stirred for 15 min,
after which time TLC (silica, EtOAc/Hex, 2:1) indicated complete
consumption of starting material. The organic layer was separated and
the aqueous layer was extracted with 50 mL of CH2Cl2. The combined
organic solutions were dried (MgSO4) and concentrated in vacuo to
give about 1 g of light-yellow foam which was fairly clean by 1H NMR.
Flash chromatography over 65 g of silica gel (gradient elution with
EtOAc/Hex 1:1, then 2:1) gave 887 mg of tripeptide 32 as a white
(36) It is interesting that the major fragmentation pathway in the mass
spectra of both ent-6 and 37 involves a retro-Diels-Alder reaction (loss of
3-methylideneoxindole) to give an ion at m/z 227, suggesting alternate
approaches to this family of natural products.
foam: [R]17 +16.2 (c 0.585, CHCl3); IR (KBr) 3370, 3055, 2951,
D
1732, 1714, 1696, 1683, 1651, 1644, 1600, 1587, 1537, 1519, 1503,
1447 cm-1; 1H NMR (CDCl3, 300 MHz, 60 °C) δ 2.14 (s, 3H, CH3S),
2.99-3.11 (m, 2H, CH2SCH3), 3.31 (dd, J ) 15.2 Hz, 6.0 Hz, 1H,
CHHCHCOO), 3.39 (dd, J ) 15.2 Hz, 5.5 Hz, 1H, CHHCHCOO),
3.71 (s, 3H, CH3O), 4.29 (t, J ) 7.2 Hz, 1H, CHCH2O), 4.42 (dd, J )
10.6 Hz, 7.2 Hz, 1H, OCHH), 4.50 (dd, J ) 10.7 Hz, 7.2 Hz, 1H,
OCHH), 4.49-4.57 (m, 1H, CHCH2S), 5.05 (ddd, J ) 7.6 Hz, 6.0 Hz,
(37) We note that Larsen had assigned absolute stereochemistry to 6 on
the basis of X-ray crystallographic studies and this research supports that
assignment.8
(38) We note the absolute configuration of fumiquinazoline G also
requires that its biosynthesis employ D-tryptophan or involve an isomer-
ization along the way.15,30
(39) Nakagawa, M.; Ito, M.; Hasegawa, Y.; Akashi, S.; Hino, T.
Tetrahedron Lett. 1984, 25, 3865. Klausner, Y. S.; Chorev, M. J. Chem.
Soc., Perkin Trans. 1 1977, 627.
(40) See spectra in supporting information for numbering scheme used
in H NMR assignments.
1

5898 J. Am. Chem. Soc., Vol. 123, No. 25, 2001
Hart and MagomedoV
5.5 Hz, 1H, CHCOOCH3), 5.72 (bd, J ) 7.0 Hz, 1H, NHCOO), 6.65
mL of 0.5 N aqueous HCl (caution: brisk gas evolution during the
acid-base reaction) and 200 mL of EtOAc, the organic layer was
separated, and the aqueous layer was extracted with two 80 mL portions
of EtOAc. The combined organic extracts were washed with 50 mL of
water, dried (MgSO4), and concentrated in vacuo to give a thick yellow
oil. The oil was flash chromatographed over 120 g of silica gel (gradient
elution with EtOAc:Hex, 1:3, 1:2, 1:1, 2:1, neat EtOAc) to give 1.0 g
(76%) of quinazolinone 34 as a yellow foam and 67 mg (8%) of 28 as
(bd, J ) 7.6 Hz, 1H, NHCHCOO), 6.94 (d, J ) 1.8 Hz, 1H, Nind
-
HCH), 7.00 (tm, J ) 7.8 Hz, 1H, ArH), 7.05 (tm, J ) 7.9 Hz, 1H,
ArH), 7.16 (tm, J ) 7.8 Hz, 1H, ArH), 7.26-7.32 (m, 4H, ArH), 7.35-
7.41 (m, 2H, ArH), 7.47 (t, J ) 7.8 Hz, 2H, ArH), 7.61 (d, J ) 7.4
Hz, 1H, ArH), 7.66 (d, J ) 7.4 Hz, 1H, ArH), 7.75 (d, J ) 7.4 Hz,
2H, ArH), 8.04 (bs, 1H, NindH), 8.59 (d, J ) 8.3 Hz, 1H, ArH), 11.6
(bs, 1H, NHCOCH); 13C NMR (CDCl3, 75.5 MHz, 60 °C) δ 16.0 (q),
27.6 (t), 37.1 (t), 47.6 (d), 52.5 (q), 53.6 (d), 55.8 (d), 67.7 (t), 109.9
(s), 111.5 (d), 118.5 (d), 119.9 (d), 120.1 (d), 120.8 (s), 121.7 (d), 122.4
(d), 123.0 (d), 123.5 (d), 125.4 (d), 127.0 (d), 127.2 (d), 127.8 (d),
132.8 (d), 136.5 (s), 139.1 (s), 141.5 (s), 144.0 (s), 144.3 (s), 156.2
(s), 168.4 (s), 169.4 (s), 172.2 (s); mass spectrum (EI), m/z (relative
intensity) 605 (1), 432 (2), 303 (3), 201 (71), 170 (11), 131 (12), 130
(100), 47 (10); mass spectrum (FAB), m/z (relative intensity) 677 (M
+ 1+). Anal. Calcd. for C38H36N4O6S: C, 67.49; H, 5.37. Found: C,
67.22; H, 5.28.
a white amorphous solid. Quinazolinone 34: [R]18 -295.8 (c 0.36,
D
EtOAc); IR (KBr) 3395, 3064, 2950, 2917, 1743, 1718, 1679, 1596,
1
1508, 1262, 1229 cm-1; H NMR (Me2CO-d6, 300 MHz, 57 °C) δ
1.59 (s, 3H, SCH3), 2.15 (dd, J ) 14.1 Hz, 4.7 Hz, 1H, CHHSCH3),
2.85 (dd, J ) 14.1 Hz, 9.4 Hz, 1H, CHHSCH3), 3.58 (s, 3H, OCH3),
3.76-3.86 (m, 2H, CH2CHCO), 4.21 (dd, J ) 7.4 Hz, 6.9 Hz, 1H,
OCH2CH), 4.31 (dd, J ) 10.3 Hz, 6.9 Hz, 1H, OCHH), 4.39 (dd, J )
10.3 Hz, 7.4 Hz, 1H, OCHH), 4.52 (ddd, J ) 9.7 Hz, 9.4 Hz, 4.7 Hz,
1H, SCH2CH), 5.67 (bm, 1H, NCHCO), 6.71 (bd, J ) 9.7 Hz, 1H,
CHNHCO), 6.82 (tm, J ) 7.8 Hz, 1H, ArH), 6.95 (d, J ) 1.9 Hz, 1H,
CHNindH), 7.04 (ddd, J ) 8.0 Hz, 7.2 Hz, 1.1 Hz, 1H, ArH), 7.25-
7.30 (m, 2H, ArH), 7.34-7.40 (m, 4H, ArH), 7.53-7.58 (m, 2H, ArH),
7.69 (d, J ) 7.5 Hz, 2H, ArH), 7.77-7.82 (m, 3H, ArH), 8.27-8.30
(m, 1H, Ha), 9.86 (bs, 1H, NindH); 13C NMR (Me2CO-d6, 100 MHz) δ
15.3 (q), 24.7 (t), 37.9 (t), 47.9 (d), 51.5 (d), 52.6 (q), 60.5 (d), 67.9
(t), 111.1 (s), 112.4 (d), 118.8 (d), 120.1 (d), 120.8 (d), 122.1 (s), 122.5
(d), 124.8 (d), 126.3 (d), 127.3 (d), 128.0 (d), 128.2 (d), 128.3 (d),
128.6 (d), 135.4 (d), 137.6 (s), 142.1 (s), 142.1 (s), 144.9 (s), 147.5
(s), 156.1 (s), 157.4 (s), 162.5 (s), 170.2 (s); mass spectrum (EI), m/z
(relative intensity) 658 (M+, 0.01), 629 (2), 458 (3), 356 (5), 214 (6),
201 (21), 196 (8), 179 (11), 178 (60), 176 (13), 170 (17), 166 (25),
165 (35), 131 (11), 130 (100), 76 (10), 43 (10); exact mass calcd for
C38H34N4O5S m/z 658.2250, found m/z 658.2296. Anal. Calcd for
C38H34N4O5S: C, 69.34; H, 5.21. Found: C, 69.01, 69.07; H, 5.21,
5.26.
N-[2-[(R)-1-Carboxyamino-2-(methylthio)ethyl]-4H-3,1-benzox-
azin-4-ylidene]-L-tryptophan, N-(Fluoren-9-ylmethyl) Methyl Ester
(33). To a solution of 5.11 g (19.50 mmol) of triphenylphosphine and
4.95 g (19.50 mmol) of iodine in 100 mL of dry CH2Cl2 was added,
via cannula, a solution of 2.63 g (3.90 mmol) of tripeptide 32 in 20
mL of CH2Cl2 (plus a 10 mL portion of CH2Cl2 to rinse), followed by
addition of 6.83 mL of Hunig’s base in one portion. The reaction
mixture was stirred for 5 h at room temperature under an argon
atmosphere, after which time TLC (silica, EtOAc/Hex, 2:1) indicated
complete consumption of starting material. The reaction mixture was
directly deposited to the top of a 250-g flash silica gel column. Gradient
elution with EtOAc/Hex 1:1, then with 2:1, and finally with neat EtOAc
(eluent contained 2% of triethylamine) provided 2.05 g (80%) of
iminobenzoxazine 33 as a yellow foam: [R]16D -124.3 (c 0.53, EtOAc);
1
IR (KBr) 3392, 3055, 2951, 2919, 1718, 1684, 1648 cm-1; H NMR
(MeCN-d3, 300 MHz) δ 2.02 (s, 3H, SCH3), 2.75 (dd, J ) 14.1 Hz,
7.7 Hz, 1H, CHHSCH3), 2.87 (dd, J ) 14.1 Hz, 5.9 Hz, 1H,
CHHSCH3), 3.26 (dd, J ) 14.4 Hz, 7.5 Hz, 1H, CHHCHCO), 3.41
(dd, J ) 14.4 Hz, 5.6 Hz, 1H, CHHCHCO), 3.64 (s, 3H, OCH3), 4.23
(t, J ) 6.5 Hz, 1H, OCH2CH), 4.43 (d, J ) 6.5 Hz, 2H, OCH2), 4.43-
4.51 (m, 1H, CHCH2S) 4.92 (dd, J ) 7.5 Hz, 5.6 Hz, 1H, NCHCO),
5.81 (bs, 1H, NHCO), 7.00 (ddd, J ) 7.9 Hz, 7.0 Hz, 1.0 Hz, 1H,
ArH), 7.09 (ddd, J ) 8.2 Hz, 7.0 Hz, 1.2 Hz, 1H, ArH), 7.12 (d,
J ) 2.3 Hz, 1H, NindHCH), 7.23-7.41 (m, 6H, ArH), 7.44 (ddd,
J ) 7.7 Hz, 7.7 Hz, 1.2 Hz, 1H, ArH), 7.59-7.65 (m, 4H, ArH),
7.80 (d, J ) 7.5 Hz, 2H, ArH), 8.09 (dd, J ) 7.9 Hz, 1.5 Hz, 1H, Ha),
8.88 (bs, 1H, NindH); 13C NMR (MeCN-d3, 75.5 MHz) δ 16.4 (q), 30.6
(t), 37.3 (t), 48.6 (d), 52.7 (q), 54.8 (d), 61.3 (d), 67.9 (t), 112.5 (d),
112.9 (s), 120.2 (d), 120.7 (s), 121.2 (d), 122.7 (d), 124.8 (d), 126.4
(d), 126.4 (d), 127.2 (d), 127.5 (d), 128.4 (d), 129.0 (d), 129.2 (s),
129.8 (d), 134.8 (d), 137.8 (s), 142.6 (s), 145.4 (s), 145.5 (s), 148.5
(s), 157.0 (s), 159.3 (s), 173.6 (s); mass spectrum (EI), m/z (rela-
tive intensity) 655 (0.4), 375 (3), 346 (19), 286 (8), 179 (14), 178
(100), 177 (11), 176 (12), 160 (6), 152 (9), 130 (42), 90 (14), no
molecular ion peak was detected; mass spectrum (FAB), m/z 659 (M
+ 1+). Anal. Calcd. for C38H34N4O5S: C, 69.34; H, 5.21. Found: C,
69.14; H, 5.10.
(1R,4S)-4-(Indol-3-ylmethyl)-1-[(methylthio)methyl]-2H-pyrazino-
[2,1-b]quinazoline-3,6(1H,4H)-dione (28). To a solution of 2.07 g
(3.146 mmol) of 34 in 100 mL of THF, cooled to 0 °C, was added 15
mL of piperidine in one portion. The reaction mixture was stirred at
ice bath temperature for 2 h 30 min. To the reaction mixture was added
100 mL of toluene, and the solution was concentrated on a rotary
evaporator to an approximate volume of 100 mL. The resulting solution
was partitioned between 30 mL of 1 N aqueous HCl and 200 mL of
EtOAc. The organic layer was sequentially washed with two 50 mL
portions of 0.5 N aqueous HCl and two 30 mL portions of water and
concentrated in vacuo. The residue was flash chromatographed over
65 g of silica gel (EtOAc) to provide 1.20 g (94%) of 28 as a light
yellow foam: [R]18D +465.0 (c 0.795, EtOAc); IR (KBr) 3339, 3061,
2918, 1681, 1610, 1593, 1568, 1473, 1405, 1334 cm-1 1H NMR
;
(CDCl3, 300 MHz) δ 0.32 (dd, J ) 13.8 Hz, 11.7 Hz, 1H, CHHSCH3),
1.75 (s, 3H, CH3), 2.60 (dd, J ) 13.8 Hz, 3.0 Hz, 1H, CHHSCH3),
3.74 (dd, J ) 15.1 Hz, 3.2 Hz, 1H, COCHCHH), 3.83 (dd, J ) 15.1
Hz, 4.7 Hz, 1H, COCHCHH), 4.24 (ddd, J ) 11.7 Hz, 3.0 Hz, 2.4 Hz,
1H, CHCH2S), 5.56 (dd, J ) 4.7 Hz, 3.2 Hz, 1H, NCHCO), 6.60 (bs,
1H, NHCO), 6.67 (d, J ) 2.4 Hz, 1H, NindHCH), 6.85 (ddd, J ) 8.0
Hz, 7.0 Hz, 0.9 Hz, 1H, ArH), 7.09 (ddd, J ) 8.0 Hz, 7.0 Hz, 1.0 Hz,
1H, ArH), 7.24 (dm, J ) 8.0 Hz, 1H, ArH), 7.28 (dm, J ) 8.0 Hz, 1H,
ArH), 7.52-7.57 (m, 2H, ArH), 7.76-7.81 (m, 1H, ArH), 8.12 (bs,
1H, NH), 8.40 (dd, J ) 8.2 Hz, 1.4 Hz, 1H, Ha); 13C NMR (Me2CO-
d6, 75.5 MHz) δ 15.2 (q), 27.1 (t), 41.8 (t), 55.6 (d), 57.9 (d), 109.5
(s), 112.3 (d), 119.1 (d), 120.0 (d), 121.0 (s), 122.5 (d), 125.2 (d), 127.2
(d), 127.4 (d), 128.7 (s), 135.4 (d), 136.9 (s), 147.9 (s), 151.1 (s), 161.4
(s), 167.5 (s), one doublet was not seen due to overlap with other peaks;
mass spectrum (EI), m/z (relative intensity) 404 (M+, 4), 356 (6), 170
(15), 143 (9), 131 (17), 130 (100), 129 (5), 115 (5), 103 (12), 102 (5),
77 (10), 48 (11), 47 (13), 45 (8); exact mass calcd for C22H20N4O2S
m/z 404.1307, found m/z 404.1304. Anal. Calcd for C22H20N4O2S: C,
65.39; H, 4.99. Found C, 64.75; H, 5.13.
(S)-2-[(R)-1-Carboxyamino)-2-(methylthio)ethyl]-r-(indol-3-yl-
methyl)-4-oxo-3(4H)-quinazolineacetic Acid, N-(Fluoren-9-ylmethyl)
Methyl Ester (34). To a stirred solution of 1.01 mL (9.88 mmol) of
thiophenol in 50 mL of dry THF, cooled to -78 °C, was added
dropwise 6.17 mL of a 1.6 M solution of n-BuLi in hexanes over a
period of 10 min under an argon atmosphere. The solution was stirred
at -78 °C for 5 min, the mixture was warmed to 0 °C, and AlMe3 (4.9
mL of 2 M solution in toluene, 9.88 mmol) was added dropwise over
a period of 2 min. To the resulting solution of the lithium trimethyl-
(phenylsulfido)aluminate, cooled to -78 °C, was added by cannula a
solution of 1.30 g (1.98 mmol) of iminobenzoxazine 33 in 10 mL of
THF (plus two 5 mL portions of THF to rinse). The reaction mixture
was stirred for 12 h at -25 °C. TLC (silica, EtOAc/Hex, 1:1) indicated
that reaction was slow at this temperature and that the reaction mixture
contained more starting material than the presumed product. The
reaction temperature was increased to -10 °C and the mixture was
stirred for additional 36 h. The mixture was partitioned between 50
(S)-4-(Indol-3-ylmethyl)-1-methylene-2H-pyrazino[2,1-b]quinazo-
line-3,6(1H,4H)-dione (9, R ) H). To a solution of 150 mg (0.37
mmol) of sulfide 28 in 10 mL of EtOAc was added a solution of 97
mg (0.48 mmol, 80% content) of m-CPBA in 1 mL of EtOAc in one
portion at -78 °C. The reaction mixture was stirred at -78 °C for 1

Synthesis of ent-Alantrypinone
J. Am. Chem. Soc., Vol. 123, No. 25, 2001 5899
h, after which time TLC (silica gel, EtOAc) indicated complete
consumption of starting material. The cold reaction mixture was
partitioned between 150 mL of EtOAc and 10 mL of a 1:1 mixture of
water and saturated aqueous Na2SO3. The organic layer was sequentially
washed with two 15 mL portions of sodium carbonate and three 30
mL portions of water and concentrated in vacuo without drying. During
one experiment, attempted drying over MgSO4 resulted in substantial
loss of the product, which appeared to have very strong affinity for
the surface of the drying agent. The residual moisture was removed by
addition of 30 mL of benzene to the mixture of sulfoxides followed by
(-)-Alantrypinone (ent-6); (1S,3′R,4S)-1-Methylspiro[1,4-ethano-
2H-pyrazino[2,1-b]quinazoline-13,3′-indoline]-2′,3,6(1H,4H)-tri-
one (37). To a stirred solution of 132 mg (0.371 mmol) of 35 in 13
mL of a mixture of THF:TFA:H2O (5:4:4), cooled to 0 °C, was added
66 mg (0.371 mmol) of solid NBS in one portion. After 1 h of stirring
at ice bath temperature, an additional 66 mg (0.371 mmol) of solid
NBS was added to the reaction mixture. A total of 5 equiv of NBS
was added to the reaction mixture over a period of 8 h. The reaction
mixture was partitioned between 200 mL of EtOAc and 50 mL of
saturated aqueous NaHCO3. The organic layer was washed with two
30 mL portions of water and concentrated in vacuo. The residue,
representing a fairly complex mixture of polybrominated spiro oxindoles
(1H NMR), was dissolved in 50 mL of MeOH. The solution was charged
with 200 mg of 5% Pt/C, 400 mg of NaOAc, and 0.80 mL of AcOH
and then hydrogenolyzed for 30 min under 1 atm of H2 at room
temperature. The reaction mixture was filtered through a 1 cm pad of
Celite followed by concentration in vacuo. The residue was dissolved
in 200 mL of EtOAc and washed with six 30 mL portions of water to
remove sodium acetate. Concentration in vacuo provided a fairly clean
mixture of ent-6 and 37, which was separated by preparative TLC (silica
gel, EtOAc) to give 42 mg (30%) of ent-6 and 61 mg (44%) of 37,
both as white amorphous solids. (-)-Alantrypinone (ent-6): mp >300
1
concentration in vacuo, the procedure being repeated twice. The H
NMR spectrum of the crude product (143 mg, 92%) showed the
presence of a mixture of diastereomeric sulfoxides (3:2) in addition to
trace amounts of 9. The mixture of sulfoxides was suspended in 100
mL of benzene. Triphenylphosphine (205 mg, 0.78 mmol) was added,
and the resulting mixture was gently refluxed for 18 h under an argon
atmosphere. The reaction mixture was cooled to room temperature and
concentrated in vacuo, and the residue was chromatographed over 12
g of flash silica gel (Et2O) to give 104 mg (79% from 28) of 9 as a
white amorphous solid: [R]18 +472.3 (c 0.419, EtOAc); IR (KBr)
D
3409, 3224, 2926, 1702, 1686, 1658, 1642, 1579, 1561, 1470, 1458,
1
1398, 1332 cm-1; H NMR (Me2CO-d6, 300 MHz) δ 3.49-3.61 (m,
2H, CHCH2), 4.36 (s, 1H, Hf), 5.24 (s, 1H, He), 5.57 (dd, J ) 4.6 Hz,
3.7 Hz, 1H, NCHCO), 6.70 (d, J ) 2.5 Hz, 1H, NindHCH), 6.73 (ddd,
J ) 8.1 Hz, 7.1 Hz, 1.0 Hz, 1H, ArH), 6.97 (ddd, J ) 8.1 Hz, 7.1 Hz,
1.1 Hz, 1H, ArH), 7.22 (dm, J ) 8.1 Hz, 1H, ArH), 7.27 (dm, J ) 8.1
Hz, 1H, ArH), 7.54-7.59 (m, 2H, ArH), 7.82 (m, 1H, Hc), 8.29 (ddd,
J ) 8.4 Hz, 1.6 Hz, 0.5 Hz, 1H, Ha), 9.43 (bs, 1H, NHCO), 10.03 (bs,
1H, NindH); 13C NMR (DMSO-d6, 75.5 MHz) δ 27.3 (t), 56.3 (d), 98.3
(t), 106.6 (s), 111.1 (d), 117.5 (d), 118.4 (d), 120.0 (s), 120.9 (d), 124.5
(d), 126.2 (d), 126.9 (d), 127.2 (d), 127.5 (s), 134.0 (s), 134.7 (d), 136.0
(s), 144.0 (s), 146.7 (s), 159.9 (s), 164.9 (s); mass spectrum (EI), m/z
(relative intensity) 356 (M+, 10), 341 (5), 227 (9), 170 (8), 143 (11),
131 (11), 130 (100), 129 (19), 103 (12), 102 (16), 77 (10); exact mass
calcd for C21H16N4O2 m/z 356.1273, found m/z 356.1273.
°C; [R]18 -40.3 (c 0.265, EtOH); IR (KBr) 3434, 3248, 1708, 1661,
D
1620, 1470, 1382, 1334 cm-1; 1H NMR (DMSO-d6, 300 MHz) δ 1.19
(s, 3H, CH3), 2.36-2.47 (m, 2H, H15), 5.59 (m, 1H, H14), 6.93 (d, J )
7.6 Hz, 1H, ArH), 7.11 (t, J ) 7.6 Hz, 1H, ArH), 7.20 (d, J ) 7.2 Hz,
1H, ArH), 7.32 (ddd, J ) 7.6 Hz, 7.6 Hz, 1.2 Hz, 1H, ArH), 7.60 (tm,
J ) 8.0 Hz, 1H, ArH), 7.70 (d, J ) 8.0 Hz, 1H, ArH), 7.86 (ddd, J )
8.2 Hz, 7.4 Hz, 1.3 Hz, 1H, ArH), 8.21 (dd, J ) 8.0 Hz, 1.1 Hz, 1H,
H10), 9.55 (d, J ) 1.7 Hz, 1H, H2), 10.63 (s, 1H, H19); 13C NMR
(DMSO-d6, 75.5 MHz) δ 13.4 (q), 35.9 (t), 52.0 (d), 54.7 (s), 61.8 (s),
109.9 (d), 120.0 (s), 122.3 (d), 123.8 (d), 126.3 (d), 127.2 (d), 127.7
(d), 129.1 (d), 129.9 (s), 134.7 (d), 142.6 (s), 146.8 (s), 152.9 (s), 158.3
(s), 169.7 (s), 176.6 (s); mass spectrum (EI), m/z (relative intensity)
372 (M+, 6), 228 (14), 227 (100), 199 (33), 198 (8), 170 (6), 145 (16),
144 (5), 130 (5), 117 (13), 103 (7), 90 (11), 89 (6); exact mass calcd
for C21H16N4O3 m/z 372.1222, found m/z 372.1211.
(6S,14S)-14,15-Dihydro-6-methyl-6,14-(iminomethano)-5H-indolo-
[2′,3′:4,5]azepino[2,1-b]quinazoline-12,15(6H)-dione (35). To a solu-
tion of 130 mg (0.37 mmol) of 9 (R ) H) in 50 mL of chloroform was
added 2.6 mL of TFA in one portion at room temperature. The resulting
mixture was refluxed for 1 h, cooled to room temperature, and con-
centrated in vacuo, and the residue was flash chromatographed over
12 g of silica gel (Et2O) to give 116 mg (89%) of bridged indole 35 as
Compound 37: mp >300 °C; [R]18D +84.3 (c 0.115, THF); IR (KBr)
3435, 3196, 2926, 1723, 1684, 1622, 1608, 1470, 1383, 1332 cm-1
;
1H NMR (THF-d8, 300 MHz) δ 1.35 (s, 3H, CH3), 2.29 (dd, J ) 14.0
Hz, 3.9 Hz, 1H, H15), 2.66 (dd, J ) 14.0 Hz, 1.9 Hz, 1H, H15), 5.67
(ddd, J ) 3.9 Hz, 3.2 Hz, 1.9 Hz, 1H, H14), 5.95 (dm, J ) 7.6 Hz, 1H,
H24), 6.62 (ddd, J ) 7.6 Hz, 7.6 Hz, 1.0 Hz, 1H, H23), 6.82 (dm, J )
7.6 Hz, 1H, H21), 7.10 (ddd, J ) 7.6 Hz, 7.6 Hz, 1.2 Hz, 1H, H22),
7.53 (ddd, J ) 7.9 Hz, 7.1 Hz, 1.2 Hz, 1H, H9), 7.62 (ddd, J ) 8.2 Hz,
1.2 Hz, 0.5 Hz, 1H, H7), 7.76 (ddd, J ) 8.2 Hz, 7.1 Hz, 1.6 Hz, 1H,
H8), 8.28 (bs, 1H, H2), 8.31 (ddd, J ) 7.9 Hz, 1.6 Hz, 0.5 Hz, 1H,
H10), 9.64 (bs, 1H, H19); 13C NMR (DMSO-d6, 75.5 MHz) δ 13.3 (q),
35.2 (t), 52.2 (d), 52.6 (s), 61.9 (s), 109.5 (d), 120.1 (s), 121.6 (d),
123.3 (d), 126.5 (d), 127.6 (d), 127.8 (d), 128.9 (d), 129.2 (s), 134.9
(d), 142.4 (s), 146.3 (s), 152.4 (s), 158.1 (s), 168.9 (s), 177.1 (s); mass
spectrum (EI), m/z (relative intensity) 372 (M+, 4), 317 (2), 267 (2),
228 (14), 227 (100), 199 (43), 198 (11), 172 (5), 171 (7), 170 (7), 146
(6), 145 (33), 144 (8), 130 (9), 117 (28), 116 (6), 103 (11), 90 (24), 89
(15), 63 (6); exact mass calcd for C21H16N4O3 m/z 372.1222, found
m/z 372.1208.
a white amorphous solid: [R]18 +267.1 (c 0.243, EtOAc); IR (KBr)
D
3343, 3269, 2903, 1692, 1672, 1609, 1569, 1470, 1452, 1385, 1332,
1305, 1242, 1188, 1160 cm-1; 1H NMR (DMSO-d6, 300 MHz) δ 2.12
(s, 3H, CH3), 3.23 (dd, J ) 17.4 Hz, 4.5 Hz, 1H, CHCHH), 3.43 (dd,
J ) 17.4 Hz, 2.8 Hz, 1H, CHCHH), 5.70-5.72 (m, 1H, CHCH2), 6.99
(ddd, J ) 7.9 Hz, 7.1 Hz, 0.9 Hz, 1H, ArH), 7.12 (ddd, J ) 8.1 Hz,
7.1 Hz, 1.1 Hz, 1H, ArH), 7.36-7.42 (m, 2H, ArH), 7.53 (tm, J ) 8.0
Hz, 1H, ArH), 7.64 (d, J ) 7.9 Hz, 1H, ArH), 7.81 (ddd, J ) 8.3 Hz,
7.3 Hz, 1.4 Hz, 1H, ArH), 8.14 (dd, J ) 8.0 Hz, 1.4 Hz, 1H, Ha), 9.55
(s, 1H, NHCO), 11.20 (s, 1H, NindH); 1H NMR (THF-d8, 300 MHz) δ
2.18 (s, 3H, CH3), 3.29 (dd, J ) 17.2 Hz, 4.6 Hz, 1H, CHCHH), 3.46
(dd, J ) 17.2 Hz, 2.8 Hz, 1H, CHCHH), 5.87 (ddd, J ) 4.6 Hz, 2.7
Hz, 1.5 Hz, 1H, CHCH2), 6.96 (ddd, J ) 8.0 Hz, 7.1 Hz, 1.1 Hz, 1H,
ArH), 7.06 (ddd, J ) 8.2 Hz, 7.1 Hz, 1.2 Hz, 1H, ArH), 7.27 (dm, J
) 8.2 Hz, 1H, ArH), 7.38 (d, J ) 7.7 Hz, 1H, ArH), 7.41 (ddd, J )
8.0 Hz, 7.1 Hz, 1.2 Hz, 1H, Hb), 7.57 (ddd, J ) 8.2 Hz, 1.2 Hz, 0.5
Hz, 1H, Hd), 7.67 (ddd, J ) 8.2 Hz, 7.1 Hz, 1.6 Hz, 1H, Hc), 8.19
(ddd, J ) 8.0 Hz, 1.6 Hz, 0.5 Hz, 1H, Ha), 8.48 (bs, 1H, NHCO),
10.20 (bs, 1H, NindH); 13C NMR (DMSO-d6, 75.5 MHz) δ 18.3 (q),
25.6 (t), 54.1 (d), 54.6 (s), 105.5 (s), 111.7 (d), 118.0 (d), 119.3 (d),
120.1 (s), 122.2 (d), 126.3 (d), 127.1 (d), 127.3 (d), 127.4 (s), 134.0
(s), 134.6 (d), 134.8 (s), 146.6 (s), 154.3 (s), 159.3 (s), 169.2 (s); mass
spectrum (EI), m/z (relative intensity) 357 (12), 356 (M+, 52), 342 (23),
341 (100), 317 (7), 313 (17), 312 (9), 299 (12), 286 (23), 267 (10),
212 (9), 211 (28), 183 (12), 163 (11), 162 (21), 151 (13), 113 (17), 70
(11), 51 (59); exact mass calcd for C21H16N4O2 m/z 356.1273, found
m/z 356.1287.
Acknowledgment. We thank the National Institutes of
Health for support of this research and the Campus Chemical
Instrumentation Center at the Ohio State University for technical
support.
Supporting Information Available: 1H and 13C NMR
spectra for most compounds, crystallographic data for 8 and
16, experimental procedures and spectral data for all compounds
not presented in the Experimental Section. This material is
available free of charge via the Internet at http://pubs.acs.org.
JA010066+