Asymmetric total synthesis of ent-(-)-roseophilin: Assignment of absolute configuration

Article abstract of DOI:10.1021/ja011271s

An asymmetric total synthesis of ent-(-)-roseophilin (1), the unnatural enantiomer of a novel naturally occurring antitumor antibiotic, is described. The approach enlists a room temperature heterocyclic azadiene inverse electron demand Diels-Alder reaction of dimethyl 1,2,4,5-tetrazine-3,6-dicarboxylate (7) with the optically active enol ether 6 bearing the C23 chiral center followed by a reductive ring contraction reaction for formation of an appropriately functionalized pyrrole ring in a key 1,2,4,5-tetrazine → 1,2-diazine → pyrrole reaction sequence. A Grubbs' ring closing metathesis reaction was utilized to close the unusual 13-membered macrocycle prior to a subsequent 5-exo-trig acyl radical-alkene cyclization that was used to introduce the fused cyclopentanone and complete the preparation of the tricylic ansa-bridged azafulvene core 32. Condensation of 32 with 33 under the modified conditions of Tius and Harrington followed by final deprotection provided (22S,23S)-1. Comparison of synthetic (22S,23S)-1 ([α]²⁵D, CD) with natural 1 established that they were enantiomers and enabled the assignment of the absolute stereochemistry of the natural product as 22R,23R. Surprisingly, ent-(-)-1 was found to be 2-10-fold more potent than natural (+)-1 in cytotoxic assays, providing ah unusually rewarding culmination to synthetic efforts that provided the unnatural enantiomer.

Full text of DOI:10.1021/ja011271s

J. Am. Chem. Soc. 2001, 123, 8515-8519
8515
Asymmetric Total Synthesis of ent-(-)-Roseophilin: Assignment of
Absolute Configuration
Dale L. Boger* and Jiyong Hong
Contribution from the Department of Chemistry and The Skaggs Institute for Chemical Biology,
The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, California 92037
ReceiVed May 23, 2001
Abstract: An asymmetric total synthesis of ent-(-)-roseophilin (1), the unnatural enantiomer of a novel naturally
occurring antitumor antibiotic, is described. The approach enlists a room temperature heterocyclic azadiene
inverse electron demand Diels-Alder reaction of dimethyl 1,2,4,5-tetrazine-3,6-dicarboxylate (7) with the
optically active enol ether 6 bearing the C23 chiral center followed by a reductive ring contraction reaction for
formation of an appropriately functionalized pyrrole ring in a key 1,2,4,5-tetrazine f 1,2-diazine f pyrrole
reaction sequence. A Grubbs’ ring closing metathesis reaction was utilized to close the unusual 13-membered
macrocycle prior to a subsequent 5-exo-trig acyl radical-alkene cyclization that was used to introduce the
fused cyclopentanone and complete the preparation of the tricylic ansa-bridged azafulvene core 32. Condensation
of 32 with 33 under the modified conditions of Tius and Harrington followed by final deprotection provided
(22S,23S)-1. Comparison of synthetic (22S,23S)-1 ([R]25D, CD) with natural 1 established that they were
enantiomers and enabled the assignment of the absolute stereochemistry of the natural product as 22R,23R.
Surprisingly, ent-(-)-1 was found to be 2-10-fold more potent than natural (+)-1 in cytotoxic assays, providing
an unusually rewarding culmination to synthetic efforts that provided the unnatural enantiomer.
Roseophilin (1, Figure 1), isolated and characterized by
Hayakawa and Seto et al. in 19921 from the culture broth of an
actinomycete identified as Streptomyces griseoViridis, is a novel
antitumor antibiotic that possesses a topologically unique
pentacyclic skeleton. It was found to consist of a potentially
strained macrocycle incorporated in an ansa-bridged azafulvene
linked to a characteristic conjugated heterocyclic ring system
containing a substituted furan and pyrrole. The structure 1 was
assigned by spectroscopy, largely NMR,1 and ultimately con-
firmed through total synthesis. While the relative stereochemistry
of 1 was assigned by NMR spectroscopy, the absolute stereo-
chemistry was not established. It was reported to exhibit
promising cytotoxic activity in the submicromolar range (IC50:
0.88 µM against KB human epidermoid carcinoma cells and
0.34 µM against K562 human erythroid leukemia cells).1 Its
unusual structure and this promising biological activity have
stimulated substantial synthetic work, culminating in a racemic
total synthesis by Fu¨rstner and several subsequent racemic and
optically active syntheses of the tricyclic core constituting formal
total syntheses of the natural product.2 Despite these efforts,
none have progressed to the stage of establishing the absolute
configuration of the natural product. Herein, and in conjunction
with the accompanying disclosure of Tius and Harrington,3 we
report an asymmetric total synthesis of ent-(-)-roseophilin and
the resulting assignment of the absolute configuration of the
natural product.
Roseophilin is structurally related to a larger class of impor-
tant pyrrolylpyrromethene red pigments, of which prodigiosin
(2) is the parent member, that also exhibit potent antitumor and
antibiotic activity. Although their site and mechanism of action
have not been established, prodigiosin has been shown to effect
double-strand DNA cleavage in the presence of Cu(II) and O2.4
The disclosure that members of the prodigiosin class of natural
products are also immunosuppressive, acting through a novel
mechanism, has generated additional interest in their use in
organ transplantation or in the treatment of autoimmune
diseases.5 Undecylprodigiosin5 and metacycloprodigiosin,6 as
well as prodigiosin,7 inhibit T-cell proliferation and exert their
effects by a mechanism distinct from that of cyclosporin A,
FK506, or rapamycin. They appear to inhibit IL-2 signal
transduction by inhibiting phosphorylation and activation of
JAK-3 at the IL-2 γ-chain.4 This complementary mechanism
(1) Hayakawa, Y.; Kawakami, K.; Seto, H.; Furihata, K. Tetrahedron
Lett. 1992, 33, 2701.
(2) (a) Nakatani, S.; Kirihara, M.; Yamada, K.; Terashima, S. Tetrahedron
Lett. 1995, 36, 8461. (b) Kim, S. H.; Fuchs, P. L. Tetrahedron Lett. 1996,
37, 2545. (c) Fu¨rstner, A.; Weintritt, H. J. Am. Chem. Soc. 1997, 119, 2944.
(d) Kim, S. H.; Figuerosa, I.; Fuchs, P. L. Tetrahedron Lett. 1997, 38, 2601.
(e) Fu¨rstner, A.; Weintritt, H. J. Am. Chem. Soc. 1998, 120, 2817. (f) Luker,
T.; Koot, W.-J.; Hiemstra, H.; Speckamp, W. N. J. Org. Chem. 1998, 63,
220. (g) Mochizuki, T.; Itoh, E.; Shibata, N.; Nakatani, S.; Katoh, T.;
Terashima, S. Tetrahedron Lett. 1998, 39, 6911. (h) Fu¨rstner, A.; Gastner,
T.; Weintritt, H. J. Org. Chem. 1999, 64, 2361. (i) Robertson, J.; Hatley,
R. J. D. Chem. Commun. 1999, 1455. (j) Harrington, P. E.; Tius, M. A.
Org. Lett. 1999, 1, 649. (k) Fagan, M. A.; Knight, D. W. Tetrahedron Lett.
1999, 40, 6117. (l) Bamford, S. J.; Luker, T.; Speckamp, W. N.; Hiemstra,
H. Org. Lett. 2000, 2, 1157. (m) Trost, B. M.; Doherty, G. A. J. Am. Chem.
Soc. 2000, 122, 3801. (n) Robertson, J.; Hatley, R. J. D.; Watkin, D. J. J.
Chem. Soc., Perkin Trans. 1 2000, 3389.
(3) Harrington, P. E.; Tius, M. A. J. Am. Chem. Soc. 2001, 123, 8509-8514.
(4) Melvin, M. S.; Tomlinson, J. T.; Saluta, G. R.; Kucera, G. L.;
Lindquist, N.; Manderville, R. A. J. Am. Chem. Soc. 2000, 122, 6333.
Melvin, M. S.; Ferguson, D. C.; Lindquist, N.; Manderville, R. A. J. Org.
Chem. 1999, 64, 6861.
(5) D’Alessio, R.; Bargiotti, A.; Carlini, O.; Colotta, F.; Ferrari, M.;
Gnocchi, P.; Isetta, A.; Mongelli, N.; Motta, P.; Rossi, A.; Rossi, M.; Tibolla,
M.; Vanotti, E. J. Med. Chem. 2000, 43, 2557. Tsuji, R. F.; Yamamoto,
M.; Nakamura, A.; Kataoka, T.; Magae, J.; Nagai, K.; Yamasaki, M. J.
Antibiot. 1990, 43, 1293. Lee, M. H.; Yamashita, M.; Tsuji, R. F.; Yamasaki,
M.; Kataoka, T.; Magae, J.; Nagai, K. J. Antibiot. 1998, 51, 92.
(6) Nakamura, A.; Nagai, K.; Ando, K.; Tamura, G. J. Antibiot. 1986,
39, 1155.
(7) Han, S. B.; Kim, H. M.; Kim, Y. H.; Lee, C. W.; Jang, E.-S.; Son,
K. H.; Kim, S. U.; Kim, Y. K. Int. J. Immunopharmacol. 1998, 20, 1.
10.1021/ja011271s CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/10/2001

8516 J. Am. Chem. Soc., Vol. 123, No. 35, 2001
Boger and Hong
Scheme 1
not only constitute a formal total synthesis of roseophilin but
also permit the absolute configuration assignment of 1 upon
condensation with 33. Thus, inverse electron demand Diels-
Alder reaction of the electron-deficient dimethyl 1,2,4,5-
tetrazine-3,6-dicarboxylate (7) with the optically active electron-
rich enol ether 6 was anticipated to provide 8 ideally
functionalized for elaboration to the pyrrole 9 for incorporation
into the tricyclic core. Formation of the triene 26 followed by
Grubbs’ ring closing metathesis (RCM)12 would serve to
introduce the 13-membered macrocycle on a nonstrained
precursor prior to subsequent 5-exo-trig acyl radical-alkene
cyclization that was anticipated to provide the strained cyclo-
pentanone, completing the preparation of the tricyclic core.
The starting, optically active electron-rich enol ether 6 was
prepared by LiAlH4 reduction of 3, derived from an Evans aldol
reaction of a chelated Ti(IV) enolate with s-trioxane14 followed
by O-benzyl ether formation, to provide 4 in 54% (Scheme 1).
TPAP oxidation15 of 4 (100%) and Wittig reaction of the
aldehyde 5 with Ph3PdCHOMe provided 6.
The key inverse electron demand Diels-Alder reaction of
the electron-deficient methyl 1,2,4,5-tetrazine-3,6-dicarboxylate
(7) with 6 proceeded effectively at room temperature to provide
the optically active 1,2-diazine 8 in excellent yield (Scheme
2). Moreover, the preparation of 8 was found to be most
convenient to conduct without purification of the intermediate
enol ether 6 and, following the Diels-Alder reaction, provided
8 in yields as high as 91% for the two steps. Thus, the
complementary match of the electron-rich dienophile 6 and the
electron-deficient 1,2,4,5-tetrazine 7 imparted by the substituents
provided a Diels-Alder reaction that proceeds effectively even
at room temperature. Reductive ring contraction of 8 effected
by treatment with Zn-TFA (25 °C, 1 h) gave the pyrrole 9 in
Figure 1.
of action to existing therapies suggests they could be used
synergistically in combination with existing drugs as well as in
an alternative immunosuppressive therapy on their own and has
provided further interest in developing synthetic routes to such
compounds.8 Because of our past studies on prodigiosin and
structurally related analogues,9 roseophilin emerged as a natural
and attractive synthetic target for us.
Our approach to roseophilin features both a 1,2,4,5-tetrazine
f 1,2-diazine f pyrrole Diels-Alder strategy10 for construction
of an appropriately functionalized pyrrole ring and a 5-exo-trig
acyl radical-alkene cyclization11 for formation of the cyclo-
pentanone found in the tricyclic core structure of 1 and is
summarized in Figure 1. Following disconnection of 1 into this
tricyclic core 32 and the heterocyclic side chain 33 defined in
the work of Fu¨rstner et al.,2e asymmetric synthesis of 32 would
(10) (a) Reviews: Boger, D. L. Tetrahedron 1983, 39, 2869. Boger, D.
L. Chem. ReV. 1986, 86, 781. Boger, D. L. Bull. Soc. Chim., Belg. 1990,
99, 599. Boger, D. L. Chemtracts: Org. Chem. 1996, 9, 149. Boger, D. L.;
Weinreb, S. M. Hetero Diels-Alder Methodology in Organic Synthesis;
Academic: San Diego, 1987. (b) Boger, D. L.; Coleman, R. S.; Panek, J.
S.; Yohannes, D. J. Org. Chem. 1984, 49, 4405. Boger, D. L.; Patel, M. J.
Org. Chem. 1988, 53, 1405. Boger, D. L.; Baldino, C. M. J. Org. Chem.
1991, 56, 6942. Boger, D. L.; Baldino, C. M. J. Am. Chem. Soc. 1993,
115, 11418. Boger, D. L.; Boyce, C. W.; Labroli, M. A.; Sehon, C. A.; Jin,
Q. J. Am. Chem. Soc. 1999, 121, 54. Boger, D. L.; Soenen, D. R.; Boyce,
C. W.; Hedrick, M. P.; Jin, Q. J. Org. Chem. 2000, 65, 2479.
(11) (a) Boger, D. L. Isr. J. Chem. 1997, 37, 119. (b) Boger, D. L.;
Mathvink, R. J. J. Org. Chem. 1988, 53, 3377. (c) Boger, D. L.; Mathvink,
R. J. J. Org. Chem. 1989, 54, 1777. (d) Boger, D. L.; Mathvink, R. J. J.
Am. Chem. Soc. 1990, 112, 4003. (e) Boger, D. L.; Mathvink, R. J. J. Am.
Chem. Soc. 1990, 112, 4008. (f) Boger, D. L.; Mathvink, R. J. J. Org. Chem.
1990, 55, 5442. (g) Boger, D. L.; Mathvink, R. J. J. Org. Chem. 1992, 57,
1429. (h) Chatgilialoglu, C.; Crich, D.; Komatsu, M.; Ryu, I. Chem. ReV.
1999, 99, 1991.
(8) The prodigiosins also act as H+/Cl- symporters, see: Sato, T.; Konno,
H.; Tanaka, Y.; Kataoka, T.; Nagai, K.; Wasserman, H. H.; Ohkuma, S. J.
Biol. Chem. 1998, 273, 21455.
(9) (a) Wasserman, H. H.; McKeon, J. E.; Smith, L.; Forgione, P. J.
Am. Chem. Soc. 1960, 82, 506. Wasserman, H. H.; Keith, D. D.; Nadelson,
J. J. Am. Chem. Soc. 1969, 91, 1264. (b) Rapoport, H.; Holden, K. G. J.
Am. Chem. Soc. 1962, 84, 635. (c) Boger, D. L.; Patel. M. J. Org. Chem.
1988, 53, 1405. Boger, D. L.; Patel, M. Tetrahedron Lett. 1987, 28, 2499.
(d) Wasserman, H. H.; Lombardo, L. J. Tetrahedron Lett. 1989, 30, 1725.
(e) D’Alessio, R.; Rossi, A. Synlett 1996, 513. (f) Fu¨rstner, A.; Szillat, H.;
Gabor, B.; Mynott, R. J. Am. Chem. Soc. 1998, 120, 8305. (g) Wasserman,
H. H.; Petersen, A. K.; Xia, M.; Wang, J. Tetrahedron Lett. 1999, 40, 7587.
(h) Fu¨rstner, A.; Grabowski, J.; Lehmann, C. W. J. Org. Chem. 1999, 64,
8275. (i) Fu¨rstner, A.; Krause, H. J. Org. Chem. 1999, 64, 8281 and
references cited therein.
(12) Grubbs, R. H.; Chang, S. Tetrahedron 1998, 54, 4413.
(13) Boger, D. L.; Panek, J. S.; Coleman, R. S.; Sauer, J.; Huber, F. X.
J. Org. Chem. 1985, 50, 5377. Boger, D. L.; Panek, J. S.; Patel, M. Org.
Synth. 1991, 70, 79.
(14) Evans, D. A.; Urpi, F.; Somers, T. C.; Clark, J. S.; Bilodeau, M. T.
J. Am. Chem. Soc. 1990, 112, 8215.
(15) Griffith, W. P.; Ley, S. V. Aldrichimica Acta 1990, 23, 13.

Total Synthesis of ent-(-)-Roseophilin
J. Am. Chem. Soc., Vol. 123, No. 35, 2001 8517
Scheme 2
methyl ester. However, direct dealkylative methyl ester hy-
drolysis (LiI, DMF, 130 °C) provided the desired carboxylic
acid 14 (74%) along with minor amounts of the corresponding
SEM-deprotected carboxylic acid (14%). Treatment of 14 with
ClCO2Et and Et3N followed by immediate NaBH4 reduction of
the mixed anhydride gave the corresponding alcohol 15 in 90%
yield.16 Treatment of 14 with BH3‚THF (6 equiv, THF, 25 °C,
6 h) also provided 15, but in lower conversions (50%). MnO2
oxidation of the benzylic alcohol 15 and reaction of the aldehyde
16 with the Wittig reagent derived from 1717 gave 18 (E:Z )
1:5, 96% for two steps). Treatment of 18 with H2/Pd-C served
to reduce the double bond and deprotect the benzyl alcohol to
give 19 (97%). TPAP oxidation and subsequent Wittig reaction
of 20 with Ph3PdCH2 provided 21 in 67-85% overall yield,
completing the first side chain introduction.
Hydrolysis of the lactone (LiOH), TMSCHN2 esterification,
TPAP oxidation, and reaction of the aldehyde 24 with the Wittig
reagent derived from 2518 completed the introduction of the
second side chain and gave the key triene 26 in 91% overall
yield for the four steps (Z:E >20:1). Ring closing metathesis
of 26 with 20 mol % of Grubbs’ catalyst (27, 0.25 mM, CH2Cl2,
40 °C, 72 h) gave 28 as a 1:1 mixture of E and Z olefin isomers
in yields as high as 88% (72-88%), completing the introduction
of ansa-bridged macrocycle. This olefin metathesis closure of
the ansa macrocycle prior to formation of the fused cyclopen-
tanone, itself introducing a large part of the strain associated
with the tricyclic core, does not require the use of specialized
substrates incorporating conformational constraints favoring
cyclization as required in the Fuchs and Speckamp RCM
approach. In this regard, our observations are analogous but
complementary to those of Fu¨rstner2h and especially Tius2j,3 who
employed substrates that contained acyclic precursors to both
the pyrrole and cyclopentanone ring systems or that lacked the
pyrrole vs cyclopentanone, respectively.
Because of the hindered nature of the methyl ester in 28, its
conversion to the corresponding carboxylic acid 29 proved
challenging. Only NaOH in refluxing EtOH-H2O (49%) was
successful in providing the carboxylic acid 29,19 whereas a range
of more standard conditions failed to react with 28. The
conversion of 29 to the key tricyclic core required formation of
the cyclopentanone and removal of the ansa chain internal
double bond. This was accomplished through use of an acyl
radical-alkene 5-exo-trig cyclization for closure of the five-
good yield and set the stage for the differentiation of the two
methyl esters. The reductive ring contraction reaction proceeded
with intermediate generation of the isolable 1,4-dihydro-1,2-
diazine 10 which in turn could be converted to 9 upon
reexposure to treatment with Zn-TFA (eq 1). Notably, this
reductive ring contraction reaction was much slower and less
effective when conducted in HOAc vs TFA, and attempts to
conduct the reaction at a later stage following formation of the
fused lactone were not as successful although this was not
investigated in detail.9c
Debenzylation of 9 with H2/Pd-C, acid-catalyzed lactoniza-
tion of 11 (70-83% for two steps) differentiating the two methyl
esters, and subsequent protection of 12 with SEMCl gave 13
(92%). This set the stage for the introduction of the first of two
pyrrole side chains terminating in a double bond suitable for
RCM. Initial attempts to selectively hydrolyze the methyl ester
in 13 (1 N LiOH, THF/MeOH/H2O) resulted in preferential
opening of the lactone to give the corresponding hydroxy acid
(16) Soai, K.; Yokoyama, S.; Mochida, K. Synthesis 1987, 647.
(17) Saunders, J.; Tipney, D. C.; Robins, P. Tetrahedron Lett. 1982, 23,
4147.
(18) Winter, M.; Naf, F.; Furrer, A.; Pickenhagen, W.; Giersch, W.;
Meister, A.; Willhalm, B.; Thommen, W.; Ohloff, G. HelV. Chim. Acta
1979, 62, 135.
(19) Lehmann, T. E.; Greenberg, W. A.; Liberles, D. A.; Wada, C. K.;
Dervan, P. B. HelV. Chim. Acta 1997, 80, 2002.

8518 J. Am. Chem. Soc., Vol. 123, No. 35, 2001
Boger and Hong
Scheme 3
Figure 2. CD spectrum of (+)-1‚HCl and ent-(-)-1‚HCl in MeOH.
with 32 failed to provide 36, leading instead to recovered starting
materials. Considerable effort over several months was expended
to ensure that the quality and dispersion of our CeCl3 preparation
were satisfactory including drying the reagent according to the
procedure of Imamoto22 and the removal of residual water with
t-BuLi.23 Despite these efforts, not even a trace of the desired
adduct 36 was detected in our reaction mixtures. However,
following the modified protocol detailed in the accompanying
article of Tius and Harrington,3 when this reaction was run with
the CeCl3 transmetalation conducted at -55 °C (2 h) before
recooling the solution to -78 °C for addition of 32, the elusive
labile adduct 36 was obtained in good conversions, Scheme 3.
Subsequent removal of the SEM and TIPS protecting groups
(Bu4NF) followed by brief treatment with aqueous HCl provided
roseophilin hydrochloride as a red solid. The remarkable optical
membered ring followed by hydrogenation of the remaining
double bond. Thus, treatment of 29 with (EtO)2P(O)Cl and Et3N
followed by addition of PhSeNa, generated in situ from NaH
and PhSeH, provided the phenyl selenoester 30 (83%), the key
acyl radical precursor.11g Notably, the conversions of 29 to 30
were low unless a respectable excess of reagents was employed
and alternative methods for formation of phenyl selenoesters
(N-phenylselenophthalimide, n-Bu3P; (PhSe)2, n-Bu3P; (COCl)2,
NaSePh) failed to react with this hindered substrate. 5-exo-trig
acyl radical-alkene cyclization (Bu3SnH, AIBN, C6H6, 90 °C,
3 h)11 smoothly and cleanly formed the strained cyclopentanone
and provided 31 as a single diastereomer (trans) in 83% yield.
Not only is this the thermodynamically most stable isomer, but
its preferential kinetic formation would be consistent with
cyclization through a chairlike allylic H-eclipsed conformation
with the olefin occupying an equatorial orientation20 and the
relative stereochemistry of 31 was confirmed by 1H NMR. The
coupling pattern of H-23 (δ 2.56, d, J ) 6.5 Hz) in the 1H NMR
spectrum showed no coupling to H-22. This same pattern is
observed in roseophilin1 and the tricyclic core2e and confirmed
the trans stereochemical relationship between the two chiral
centers. Catalytic hydrogenation (H2/PtO2, EtOAc, 2 h) of the
remaining double bond in 312l provided the tricyclic core 32,
identical in all aspects with that reported for authentic 32.2e,m,3
The heterocyclic side chain 33 was prepared by the method
of Kim and Fuchs,21 lithiated upon treatment with n-BuLi (1
equiv, THF, -57 °C, 5 h), and condensed with the model pyrrole
substrate 34, providing 35 in good yield (52%) comparable with
the report of Fu¨rstner,2h Scheme 3. However, all attempts to
follow the protocol described in detail by Fu¨rstner for conversion
of 33 to the corresponding cerium reagent and condensation
rotation of our synthetic (22S,23S)-1 ([R]25 -5100 (c 0.3 ×
D
10-4, CH3OH)) was nearly identical, but of an opposite sign to
that of natural roseophilin ([R]25 +5500 (c 0.8 × 10-4
,
D
CH3OH)), indicating that we prepared the unnatural enantiomer
and that natural 1 possesses the 22R,23R absolute stereochem-
istry. Similarly, the CD spectra of our synthetic ent-1 and natural
1 exhibited identical but opposite curves, further establishing
that they were enantiomers, Figure 2.
By testing both natural and ent-1 alongside prodigiosin (2),
we were able to establish that neither enantiomer of 1 cleaves
DNA effectively under conditions that have been disclosed for
2 and related compounds (Cu(II), O2).4,24 Although other
structural features may also contribute to these distinctions, the
comparisons suggest the methoxypyrrole ring of 2, vs the
methoxyfuran found in 1, is central to this activity, presumably
affecting metal chelation and the subsequent steps leading to
DNA cleavage. More importantly, and although the prodigiosins
may derive their cytotoxic activity through this metal mediated
ss or ds DNA cleavage, roseophilin most likely expresses its
activity through other mechanisms.
Remarkably, a side-by-side comparison of 35 and natural and
ent-(-)-roseophilin in a range of cytotoxic assays revealed not
only that natural roseophilin was more potent than the simplified
(22) Takeda, N.; Imamoto, T. Org. Synth. 1998, 76, 228.
(23) Paquette, L. A. In Encyclopedia of Reagents for Organic Synthesis;
Paquette, L. A., Ed.; John Wiley & Sons: New York, 1995; Vol. 2, p 1035.
(24) Trace amounts of DNA cleavage were observed with natural (+)-
roseophilin (<1% at 25 °C, 8% at 37 °C, 90 min) versus prodigiosin (44%
at 25 °C, 80% at 37 °C, 90 min). Synthetic ent-(-)-roseophilin (10% at 25
°C, 16% at 37 °C, 90 min) was more effective than the natural enantiomer,
but still less effective than prodigiosin (2).
(20) Beckwith, A. L. J.; Easton, C. J.; Serelis, A. K. J. Chem. Soc., Chem.
Commun. 1980, 482. Beckwith, A. L. J.; Lawrence, T.; Serelis, A. K. J.
Chem. Soc., Chem. Commun. 1980, 484.
(21) Kim, S. H. Ph.D. Thesis, Purdue University, 1998.

Total Synthesis of ent-(-)-Roseophilin
Table 1. In Vitro Cytotoxic Activity
J. Am. Chem. Soc., Vol. 123, No. 35, 2001 8519
As such, this unanticipated behavior of ent-(-)-1 provided an
unusually rewarding culmination to synthetic efforts that by
chance provided the unnatural enantiomer.
IC50 (µM)a
CCRF-CEM
compound
L1210
Experimental Section
ent-(-)-1
(+)-1
35
0.1
0.2
2.5
0.1
1.5
3.3
Full experimental details and compound characterizations are
provided in the Supporting Information.
a Average of triplicate determination.
Acknowledgment. We gratefully acknowledge the financial
support of the National Institutes of Health (CA42056) and the
award of a Bristol-Myers Squibb Fellowship to J.H. (2000-
2001). We are especially grateful to Professor M. A. Tius for
sharing his results prior to publication, especially those required
for condensation of 32 with 33 allowing our joint publication,
to Winston Tse for assistance with the DNA cleavage assays,
to Michael P. Hedrick for conducting the cytotoxic assays, to
Alan Saghatelian for the CD measurements, to Professor
Fu¨rstner for helpful discussions, and to Professor Y. Hayakawa
for a sample of authentic, naturally occurring roseophilin.
model 35 as expected from the work of Terashima2a but also
that ent-(-)-roseophilin was approximately 2-10-fold more
potent than the natural enantiomer (Table 1). Although we are
aware of instances where the unnatural enantiomer of a naturally
occurring antitumor agent is comparably or equally potent,25
we are not aware of an example where the unnatural enantiomer
exceeds the natural enantiomer potency by as much as 10-fold.
(25) Natural and ent-fredericamycin A: Boger, D. L.; Hu¨ter, O.; Mbiya,
K.; Zhang, M. J. Am. Chem. Soc. 1995, 117, 11839. Natural (+)- and ent-
(-)-CC-1065: Kelly, R. C.; Gebhard, I.; Wicnienski, N.; Aristoff, P. A.;
Johnson, P. D.; Martin, D. G. J. Am. Chem. Soc. 1987, 109, 6837. Boger,
D. L.; Coleman, R. S. J. Am. Chem. Soc. 1988, 110, 4796. Natural (+)-
and ent-(-)-duocarmycin SA and its derivative analogues: Boger, D. L.;
Johnson, D. S.; Yun, W. J. Am. Chem. Soc. 1994, 116, 1635. Natural and
ent-mitomycin C: Gargiulo, D.; Musser, S. S.; Yang, L.; Fukuyama, T.;
Tomasz, M. J. Am. Chem. Soc. 1995, 117, 9388.
Supporting Information Available: Full experimental
details and compound characterizations. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
JA011271S