Syntheses of the eastern halves of ritterazines B, F, G, and H, leading to reassignment of the 5,5-spiroketal stereochemistry of ritterazines B and F

Article abstract of DOI:10.1021/ja0705487

The ritterazine class of natural products comprises 26 compounds - all of which are spiroketal-containing steroidal heterodimers - that inhibit the proliferation of cultured human cancer cell lines with IC₅₀ values in the low nanomolar range. L

Full text of DOI:10.1021/ja0705487

Published on Web 05/01/2007
Syntheses of the Eastern Halves of Ritterazines B, F, G, and
H, Leading to Reassignment of the 5,5-Spiroketal
Stereochemistry of Ritterazines B and F
Scott T. Phillips and Matthew D. Shair*
Contribution from the Department of Chemistry & Chemical Biology, HarVard UniVersity,
Cambridge, Massachusetts 02138
Received January 29, 2007; E-mail: shair@chemistry.harvard.edu
Abstract: The ritterazine class of natural products comprises 26 compoundssall of which are spiroketal-
containing steroidal heterodimerssthat inhibit the proliferation of cultured human cancer cell lines
with IC50 values in the low nanomolar range. Little is known about their chemistry, cellular target(s), or
mechanism(s) of growth inhibition, due primarily to the small amount of material available from
natural sources. In this paper we report syntheses of the eastern halves of ritterazines B, F, G, and H and
address the energetic and mechanistic aspects of spiroketal equilibration for each. These studies have
led to reassignment of the 5,5-spiroketal stereochemistry of ritterazines B and F, and they have
enabled us to propose a quantitative description of the natural distribution of these ritterazine
compounds.
Introduction
be true of ritterazine B. The lack of correlation between the
cytotoxicity patterns of ritterazine B or cephalostatin 1 with
Ritterazine B, a marine natural product reported by Fusetani
et al. in 1995,1 is the most potent inhibitor of cultured human
cancer cell lines of the 26 known ritterazine compounds.1-4 Its
average GI50 value is 3.2 nM in the National Cancer Institute’s
60-cell-line screen,5 and in fact, ritterazine B is among the most
potent growth inhibitors ever tested by the National Cancer
Institute (NCI).5 However, the effects of ritterazine B on cells
and its potential as an anticancer treatment have not been studied
in depth due to lack of material.6 The structurally related natural
product cephalostatin 17,8 has a pattern of cytotoxicity similar
to that of ritterazine B in 10 of the NCI cell lines (NCI-10)
(with a Pearson correlation coefficient of 0.93),6,9 suggesting
that both compounds may have similar or identical cellular
target(s) and mechanism(s). Since early studies have shown
interesting effects of cephalostatin 1 on cells,10 the same may
molecules of known mechanism suggests that these natural
products also may have unique mechanisms.5,7,9-13
We have become interested in answering the following
questions surrounding the cephalostatins and the ritterazines:
(1) Do these compounds show some selective cytotoxic or
cytostatic effects for tumor cells versus nontransformed cells,
and does this activity translate to in vivo effects? (2) What are
the cellular target(s) and mechanism(s) of these molecules, and
are they related? (3) Might these compounds reveal new
anticancer cellular targets? (4) Which portions of these large
heterodimers are responsible for their biological activity? (5)
Why are certain ritterazine family members found in nature,
while other expected derivatives are not? Several of these
compelling questions could be answered if sufficient quantities
of the natural products were available (which they are not),14,15
but others only can be addressed by preparing derivatives by
synthesis. Therefore, efficient syntheses capable of accessing
hundreds of milligrams of these natural products and their
derivatives are needed.
(1) Fukuzawa, S.; Matsunaga, S.; Fusetani, N. J. Org. Chem. 1995, 60, 608-
614.
(2) Fukuzawa, S.; Matsunaga, S.; Fusetani, N. J. Org. Chem. 1994, 59, 6164-
6166.
(3) Fukuzawa, S.; Matsunaga, S.; Fusetani, N. Tetrahedron 1995, 51, 6707-
6716.
(4) Fukuzawa, S.; Matsunaga, S.; Fusetani, N. J. Org. Chem. 1997, 62, 4484-
We have initiated a program to study the chemistry and
biology of the most active ritterazines (shown in Chart 1) and
cephalostatin 1. With respect to the ritterazines, the western
4491.
(5) Flessner, T.; Jautelat, R.; Scholz, U.; Winterfeldt, E. Fortschritte der Chemie
Organischer Naturstoffe; Springer-Verlag: New York, 2004.
(6) Komiya, T.; Fusetani, N.; Matsunaga, S.; Kubo, A.; Kaye, F. J.; Kelley,
M. J.; Tamura, K.; Yoshida, M.; Fukuoka, M.; Nakagawa, K. Cancer
Chemother. Pharmacol. 2003, 51, 202-208.
(7) Pettit, G. R.; Inoue, M.; Kamano, Y.; Herald, D. L.; Arm, C.; Dufresne,
C.; Christie, N. D.; Schmidt, J. M.; Doubek, D. L.; Krupa, T. S. J. Am.
Chem. Soc. 1988, 110, 2006-2007.
(11) Rabow, A. A.; Shoemaker, R. H.; Sausville, E. A.; Covell, D. G. J. Med.
Chem. 2002, 45, 818-840.
(12) Boyd, M. R. The NCI In Vitro Anticancer Drug DiscoVery Screen; Humana
Press Inc.: Totowa, NJ, 1985-1995.
(8) Mu¨ller, I. M.; Dirsch, V. M.; Rudy, A.; Lo´pez-Anto´n, N.; Pettit, G. R.;
Vollmar, A. M. Mol. Pharm. 2005, 67, 1684-1689.
(9) Paull, K. D.; Shoemaker, R. H.; Hodes, L.; Monks, A.; Scudiero, D. A.;
Rubinstein, L.; Plowman, J.; Boyd, M. R. J. Natl. Cancer Inst. 1989, 81,
1088-1092.
(13) Boyd, M. R.; Paull, K. D. Drug DeV. Res. 1995, 34, 91-109.
(14) A 13.4 mg sample of ritterazine B was isolated from 5.5 kg of Ritterella
tokioka, and 139 mg of cephalostatin 1 was obtained from 166 kg of
Cephalodiscus gilchristi (wet mass).
(10) Lo´pez-Anto´n, N.; Rudy, A.; Barth, N.; Schmitz, L. M.; Pettit, G. R.;
Schulze-Osthoff, K.; Dirsch, V. M.; Vollmar, A. M. J. Biol. Chem. 2006,
281, 33078-33086.
(15) LaCour, T. G.; Guo, C.; Ma, S.; Jeong, J. U.; Boyd, M. R.; Matsunaga,
S.; Fusetani, N.; Fuchs, P. L. Bioorg. Med. Chem. Lett. 1999, 9, 2587-
2592.
9
10.1021/ja0705487 CCC: $37.00 © 2007 American Chemical Society
J. AM. CHEM. SOC. 2007, 129, 6589-6598
6589

A R T I C L E S
Phillips and Shair
Chart 1. Structures and Biological Activities for Ritterazines B, F, G, and H and Cephalostatin 1
halves are conserved, while minor structural changes in the
eastern halves cause substantial variations in growth inhibitory
activity. Since the activity of these compounds is dependent on
the eastern halves, we focused on preparing and studying these
portions as an important step toward studying the entire class
of compounds. In this paper we report syntheses of the eastern
halves of ritterazines B, F, G, and H, as well as the spiroketal
epimers of ritterazines G and H. Structural analyses of these
halves have led us to reassign the ritterazine B and F spiroketal
stereochemistry (C22) (and therefore those of six other rittera-
zines). Also, kinetics studies have been used to explain the
natural distribution of these ritterazine spiroketals and to predict
other, as of yet, undiscovered ritterazine natural products.
Less progress has been made on the ritterazines, though Fuchs
and co-workers have made important contributions here as well
by developing routes to ritterazines K21 and M,26 by reassigning
ritterazine M,27 and by synthesizing the eastern halves of
ritterazines G24 and I.28 Chemistry for related compounds also
has been developed by Winterfeldt,29-38 Heathcock,39,40 Tietze,41,42
Sua´rez,43,44 and others.16,17
Fuchs has shown that carrying the ritterazine G eastern half
through many steps causes epimerization of the spiroketal,
(22) Jeong, J. U.; Sutton, S. C.; Kim, S.; Fuchs, P. L. J. Am. Chem. Soc. 1995,
117, 10157-10158.
(23) Kim, S.; Sutton, S. C.; Guo, C.; LaCour, T. G.; Fuchs, P. L. J. Am. Chem.
Soc. 1999, 121, 2056-2070 and references therein.
(24) LaCour, T. G.; Guo, C.; Bhandaru, S.; Boyd, M. R.; Fuchs, P. L. J. Am.
Chem. Soc. 1998, 120, 692-707 and references therein.
(25) Li, W.; LaCour, T. G.; Fuchs, P. L. J. Am. Chem. Soc. 2002, 124, 4548-
4549.
Results and Discussion
Synthesis of the Ritterazine G Eastern Half. To study the
eastern halves of ritterazines B, F, G, and H, we required an
efficient and scalable synthesis of the ritterazine G eastern
half: from it, we planned to access the other eastern halves in
a few additional steps. A great deal of chemistry for building
these bis-steroidal natural products has been developed by Fuchs
and co-workers, particularly with respect to the cephalostatins.16-25
(26) Lee, S.; Fuchs, P. L. Org. Lett. 2002, 4, 317-318.
(27) Lee, S.; LaCour, T. G.; Lantrip, D.; Fuchs, P. L. Org. Lett. 2002, 4, 313-
316.
(28) LaCour, T. G.; Fuchs, P. L. Tetrahedron Lett. 1999, 40, 4655-4658.
(29) Ba¨sler, S.; Brunck, A.; Jautelat, R.; Winterfeldt, E. HelV. Chim. Acta 2000,
83, 1854-1880.
(30) Flessner, T.; Ludwig, V.; Siebeneicher, H.; Winterfeldt, E. Synthesis 2002,
1373-1378.
(31) Haak, E.; Winterfeldt, E. Synlett 2004, 1414-1418.
(32) Jautelat, R.; Muller-Fahrnow, A.; Winterfeldt, E. Chem.sEur. J. 1999, 5,
1226-1233.
(16) Ganesan, A. Angew. Chem. Int. Ed. Engl. 1996, 35, 611-615.
(17) Gryszkiewicz-Wojtkielewicz, A.; Jastrzebska, I.; Morzycki, J. W.; Ro-
manowska, D. B. Curr. Org. Chem. 2003, 7, 1257-1277.
(18) Guo, C.; Bhandaru, S.; Fuchs, P. L. J. Am. Chem. Soc. 1996, 118, 10672-
10673.
(19) Lee, J. S.; Fuchs, P. L. Org. Lett. 2003, 5, 2247-2250.
(20) Lee, J. S.; Fuchs, P. L. J. Am. Chem. Soc. 2005, 127, 13122-13123.
(21) Jeong, J. U.; Guo, C.; Fuchs, P. L. J. Am. Chem. Soc. 1999, 121, 2071-
2084 and references therein.
(33) Jautelat, R.; Winterfeldt, E.; MullerFahrnow, A. J. Prak. Chem./Chem.-
Ztg. 1996, 338, 695-701.
(34) Dro¨gemu¨ller, M.; Flessner, T.; Jautelat, R.; Scholz, U.; Winterfeldt, E. Eur.
J. Org. Chem. 1998, 2811-2831.
(35) Nawasreh, M.; Winterfeldt, E. Curr. Org. Chem. 2003, 7, 649-658.
(36) Scholz, U.; Winterfeldt, E. Nat. Prod. Rep. 2000, 17, 349-366.
(37) Winterfeldt, E. Pure Appl. Chem. 1999, 71, 1095-1099.
(38) Yunus, U.; Iqbal, R.; Winterfeldt, E. J. Heterocycl. Chem. 2005, 42, 1079-
1084.
9
6590 J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007

Eastern Half Syntheses of Ritterazines B, F, G, and H
A R T I C L E S
Scheme 1. Synthesis of the Ritterazine G Eastern Half and Its Spiroketal Diastereomera
a Reagents and conditions: (a) hν, dioxane; (b) BF3‚OEt2, PhMe, 83% (two steps); (c) Jones oxidation; (d) NaBH4, -78 °C; (e) Ac2O, DMAP, 91%
(three steps); (f) NaCNBH3, AcOH, 93%; (g) o-NO2PhSeCN, PBu3; (h) H2O2, 0 °C, 67% (two steps); (i) Hg(OAc)2, THF/H2O; NaBH4, 88%; (j) PhI(OAc)2,
I2, 97% (7:8 ) 2.7:1); (k) K2CO3, MeOH, 100%.
necessitating chromatographic separations. We decided to
simplify the experimental difficulties in handling this spiroketal
by performing a late-stage spiroketalization under conditions
that were nonequilibrating, but that would allow us to prepare
the two spiroketal epimers for each ritterazine half (meaning
we needed routes for the preparation of ritterazines G and H
and their spiroketal epimers and of ritterazines B and F). The
reaction we chose to form the spiroketal was a Sua´rez oxidative
spiroketalization (see Scheme 1).44
reaction. Reproducible, albeit somewhat lower, yields are
obtained using purified alcohol and by changing the solvent to
pyridine.45
Oxymercuration-demercuration of alkene 5 provides tertiary
carbinol 6 in 88% yield, along with 12% recovered starting
material. The demercuration step in this reaction is known to
be accelerated by hydroxide,46 but in our case, the presence of
hydroxide significantly increased the amount of starting material
recovered, even though the mercuration step was nearly
quantitative. However, when sodium borohydride (without
sodium hydroxide) is added directly to the reaction mixture,
the demercuration step proceeds with a minimal amount of
elimination.46
In the last step of our synthesis, a Sua´rez iodine(III)
oxidation44 is used to form a mixture of two epimeric 5,5-
spiroketals in nearly quantitative yield, where the ritterazine G
diastereomer (7) is the favored product. The two diastereomers
can be separated by silica gel chromatography, providing access
to both spiroketals for further studies. Alternatively, the crude
mixture of spiroketals can be equilibrated over MgSO4 during
workup to form the ritterazine G diastereomer almost exclusively
(see eq 1).
Our synthesis of the ritterazine G eastern half (Scheme 1)
was accomplished in 40% yield over 10 steps from hecogenin
acetate (1), a convenient and commercially available steroid
building block. The first step, based on the work of Winterfeldt,
is a Norrish type I photolytic cleavage of the C12-C13 bond,
affording aldehyde 2.34 BF3‚OEt2-catalyzed ene reaction on the
unpurified aldehyde (compound 2) proceeds to form a six-
membered ring with the C12 secondary carbinol in the R-con-
figuration (compound 3).34 Ritterazine G has a â-configured
secondary carbinol at this position, so an efficient sequence
of oxidation and reduction steps is used to invert the
alcohol.24 Reductive opening29 of spiroketal 4 affords a
primary alcohol that is then converted to terminal olefin 5 using
the selenation/oxidation protocol developed by Grieco.45 Ini-
tially, we encountered variable yields (30-87%) when the
selenation step was performed in THF. The variation appeared
inversely dependent on the purity of the starting material
(higher yields were obtained using the unpurified alcohol),
which suggested the possibility that residual NaHCO3 from
the previous workup had a beneficial effect on the Grieco
Syntheses of the Reported Ritterazine B and Ritterazine
F Eastern Halves. The reported ritterazine F eastern half was
prepared in one step from the ritterazine G eastern half (7) using
a Pt/C-catalyzed hydrogenation (Scheme 2). At the outset we
expected hydrogenation to occur from the less hindered R-face
of the olefin,47 but in this case, the allylic ether seems to direct
(39) Heathcock, C. H.; Smith, S. C. J. Org. Chem. 1994, 59, 6828-6839.
(40) Heathcock, C. H.; Smith, S. C. J. Org. Chem. 1995, 60, 6641-6641.
(41) Tietze, L. F.; Krahnert, W. F. Synlett 2001, 560-562.
(42) Tietze, L. F.; Krahnert, W. R. Chem.sEur. J. 2002, 8, 2116-2125.
(43) Betancor, C.; Freire, R.; Pe´rez-Martin, I.; Prange´, T.; Sua´rez, E. Org. Lett.
2002, 4, 1295-1297.
(44) Betancor, C.; Freire, R.; Pe´rez-Mart´ın, I.; Prange´, T.; Sua´rez, E. Tetrahedron
2005, 61, 2803-2814.
(45) Grieco, P. A.; Gilman, S.; Nishizawa, M. J. Org. Chem. 1976, 41, 1485-
1486.
(46) Brown, H. C.; Geoghegan, P. J., Jr. J. Org. Chem. 1970, 35, 1844-1850.
(47) Abad, A.; Agullo, C.; Arno, M.; Domingo, L. R.; Zaragoza, R. J. J. Org.
Chem. 1990, 55, 2369-2373.
9
J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007 6591

A R T I C L E S
Phillips and Shair
Scheme 2. Synthesis of the Reported Ritterazine F Eastern Halfa
Scheme 3. Synthesis of the Reported Ritterazine B Eastern Halfa
a Reagents and conditions: (a) Pt/C, H2, EtOH; (b) K2CO3, MeOH, 100%.
the heterogeneous catalyst to the hindered â-face,48 providing
ritterazine F in 82% yield. This procedure represents a particu-
larly effective route to spiroketal-substituted cis-hydrindanes,
which normally require several steps for their preparation
from the corresponding trans-hydrindane steroidal precur-
sors.49 This hydrogenation also is striking in that other reagents
(Pd(OH)2/C, PtO2 in EtOH, Crabtree’s catalyst, Wilkinson’s
catalyst, hydroboration/protonolysis, and 5% rhodium on alu-
mina) were ineffective and only provided starting material.
A similar hydrogenation of the C22 epimer of ritterazine G
(compound 8, Scheme 1) provided little of the desired ritterazine
B eastern half. Whereas the conversion of ritterazine G to
ritterazine F involved the hydrogenation of one thermodynamic
spiroketal to form another, the spiroketal in the non thermo-
dynamically favored configuration (due to one anomeric effect)
(compound 8) presented unexpected problems. In this case,
compound 8 equilibrated in ethanol at a rate competitive with
that for hydrogenation, leading to a mixture of products favoring
ritterazine G, ritterazine F, and a trans-hydrindane impurity.50
Alternatively, we accessed the structure reported to be the
eastern half of ritterazine B through a more circuitous path
(Scheme 3). Hydrogenation of the eastern half of ritterazine G
(7), rather than its non thermodynamically favored C22 epimer,
in acetic acid (instead of ethanol) delivers the cis-hydrindane
and reduces the spiroketal, affording 15 in 63% yield. Sua´rez
spiroketalization44 of 15 then forms a separable mixture of the
eastern halves of ritterazines B (16) and F (12) in quantitative
yield, where the eastern half of ritterazine B (16) is the less
abundant C22 spiroketal epimer.
a Reagents and conditions: (a) Pt/C, H2, AcOH; (b) PhI(OAc)2, I2, 100%
(12:16 ) 3:1); (c) K2CO3, MeOH, 100%.
diastereomer can be obtained almost exclusively from the Sua´rez
reaction if the workup includes MgSO4 instead of Na2SO4 as
the drying reagent, though such an equilibration is unnecessary
since ritterazine F can be accessed more efficiently as outlined
in Scheme 2. Overall, the ritterazine F eastern half can be
prepared in 11 steps and 33% yield from hecogenin acetate.
The ritterazine B eastern half is accessed less efficiently in 6%
yield over 12 steps from hecogenin acetate; the major loss of
material in this sequence occurs in the final Sua´rez reaction,
where the more stable (S)-C22 spiroketal (12) predominates over
the ritterazine B eastern half (Scheme 3).
Like the eastern halves of ritterazine G and its diastereomer,
the spiroketals of the eastern halves of ritterazines B and F
interconvert under slightly Lewis acidic conditions, with the
equilibrium favoring ritterazine F (see eq 2). The ritterazine F
(48) Hoveyda, A. H.; Evans, D. A.; Fu, G. C. Chem. ReV. 1993, 93, 1307-
1370.
(49) Jung, M. E.; Johnson, T. W. Tetrahedron 2001, 57, 1449-1481.
(50) Each product was synthesized through an independent route as shown in
Schemes 1-3. Chemical shift data for all compounds are tabulated in the
Supporting Information.
9
6592 J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007

Eastern Half Syntheses of Ritterazines B, F, G, and H
A R T I C L E S
Scheme 4. Synthesis of the Ritterazine H Eastern Half and Its
Spiroketal Epimer
Synthesis of the Ritterazine H Eastern Half. The eastern
half of ritterazine H and its C22 spiroketal epimer were prepared
by oxidation of ritterazines B and F, respectively (Scheme 4).
Not surprisingly, under equilibrating conditions the spiroketal
of the eastern half of ritterazine H is converted into the
thermodynamically more stable C22 epimer (19) (see eq 3).
Compound 19 is stabilized by two anomeric effects, whereas
compound 18 is only stabilized by one anomeric effect. The
ease of this equilibration indicates the importance of avoiding
both Brønsted and Lewis acids when handling this compound.
Overall, the ritterazine H eastern half is prepared in 14 steps
and 6% yield from hecogenin acetate, or in 98% yield from the
ritterazine B eastern half.
Structural Analysis of the Synthetic Spiroketal Dia-
stereomers. The stereochemistry of all six 5,5-spiroketals
prepared was verified by a pattern of nuclear Overhauser effects
(NOEs) (Figure 1) and, in the case of the ritterazine G and F
eastern halves, by X-ray crystallography (Figure 2).
For ritterazine G (7, Figure 1), both the NOEs and the X-ray
crystal structure reveal a conformation stabilized by two
anomeric effects, where the spiroketal carbon-oxygen bond
lengths are 1.42 and 1.43 Å, consistent with values expected
for anomeric carbon-oxygen bonds in 5,5-spiroketals.24 In
comparison, the diastereomer of ritterazine G (8) interconverts
rapidly (on the NMR time scale) between conformations having
one and two anomeric effects. The strongest NOEs are consistent
with the extended conformation (Figure 1b), which possesses
only one anomeric effect, but we also observed a weak NOE
consistent with a bent structure that has two anomeric effects
(see Figure 3); the number and intensities of the NOEs suggest
that the extended conformation is preferred despite its dimin-
ished anomeric stabilization. Consequently, the diastereomer of
ritterazine G may provide a particularly convenient model for
studying anomeric effects in 5,5-spiroketals, which are less well
understood than in 6,6-spiroketals.51,52
Figure 1. Representative NOEs for the ritterazine eastern halves. The NOEs
are separated into their backbone and spiroketal components. The intensities
of the NOEs are indicated by color: red, strong; black, medium; blue, weak.
show solution conformations favoring single anomeric effects;
no NOEs were found that support a conformation with two
anomeric effects. This result implies that the cis-hydrindane
induces a greater steric bias toward an extended spiroketal
conformation than does the olefin in the ritterazine G dia-
stereomer (8).
Comparison of Synthetic Ritterazines B, F, G, and H with
Their Natural Analogues. Because the ritterazines are linear
dimers of two steroidal units and because the two halves are
separated by a large distance (the molecules are ∼30 Å long),
In comparison, the eastern halves of ritterazines B and H (17
and 18), in which the C22 spiroketal is in the R configuration,
(51) Perron, F.; Albizati, K. F. Chem. ReV. 1989, 89, 1617-1661.
(52) Aho, J. E.; Pihko, P. M.; Rissa, T. K. Chem. ReV. 2005, 105, 4406-4440.
9
J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007 6593

A R T I C L E S
Phillips and Shair
in only one natural product. Our reassignment reverses the
trend: now ritterazine B and six other ritterazine natural products
have the thermodynamically favored spiroketal, whereas ritt-
erazine F has the spiroketal in the nonanomeric configuration;
this result is intellectually satisfying, particularly if the spiro-
ketals equilibrate in ocean water. However, while such an
equilibration seems reasonable, there are additional data that
contradict this idea: ritterazine H possesses the nonanomeric
spiroketal, but the corresponding thermodynamically favored
spiroketal has not been isolated. This issue of whether ritt-
erazines equilibrate in nature will be discussed in more detail
below.
The reassignment of ritterazines B and F (and consequently
ritterazines C, P, Q, R, S, and Y) is further supported by a trend
of downfield chemical shifts for protons in proximity to the
terminal spiroketal oxygen (Figure 5). For example, when the
terminal oxygen of the C22 spiroketal is R-configured (ther-
modynamically favored configuration), then the two steroid
protons (H16 and H17) in close proximity are shifted ca. 0.2-
0.3 ppm downfield relative to when the spiroketal oxygen is â
(nonanomeric configuration). Similarly, the proton (H20) in
proximity to the â-oxygen is shifted 0.2-0.3 ppm downfield
as compared to the R-diastereomer.
Figure 2. X-ray crystal structures for the ritterazine G and F eastern halves.
The crystals are of the p-nitrobenzoyl derivatives of ritterazines G and F,
which were prepared according to the procedures outlined in the Supporting
Information. The p-nitrobenzoate group was removed from carbon 12 for
clarity.
These downfield changes in chemical shifts are due to the
deshielding effect of the nearby electronegative oxygen atom.53
A similar phenomenon has been observed in 2-endo-norborneol,
where the alcohol lone pairs interact through space with nearby
protons and perturb their chemical shifts in an angle-dependent
manner.53-55 Not surprisingly, this effect is also distance-
dependent. In our systems, a rough correlation exists between
larger ∆δ values and smaller distances between the C16 proton
and the terminal spiroketal oxygen (Figure 5): ∆δ ) 0.23 ppm
for 9 versus 10, with a hydrogen-oxygen distance of 2.94 Å,56
∆δ ) 0.29 ppm for 19 versus 18, with a hydrogen-oxygen
distance of 2.88 Å,57 and ∆δ ) 0.32 ppm for 13 versus 17,
with a hydrogen-oxygen distance of 2.84 Å.58
Equilibration of Ritterazines in Ocean Water. Since we
had access to all spiroketal epimers of the eastern halves of
ritterazines B, F, G, and H, we decided to investigate why there
is an uneven distribution of the C22 spiroketal epimers in nature.
This required that we determine whether the ritterazines
equilibrate in ocean water, and if they did, then we needed to
study the kinetics and mechanism(s) for this process so that
we could understand the relationship between structure and
rates of equilibration. The results of these studies are presented
below.
Figure 3. NOEs for the diastereomer of ritterazine G showing the
conformational equilibrium in C6D6 at 20 °C.
the presence or absence of one half should have little impact
on the chemical shifts of protons near the ends of the other half
(i.e., near the C22 spiroketal units). This hypothesis is supported
by the nearly identical chemical shifts for the synthetic ritterazine
G eastern half and for those of the natural product (Figure 4a).1,3
In Figure 4a, select proton chemical shifts for natural ritterazine
G are shown on the left, with deviations from these values for
each synthetic spiroketal epimer shown to the right. Positive
values for ∆δ correspond to chemical shifts that are further
upfield from those for the natural product, whereas negative
values are for protons further downfield. Comparison of the
values in Figure 4a confirms that the spiroketal stereochemistry
in synthetic diastereomer 9 is the same as that originally
assigned. Similar correlations, however, were not found for
ritterazines B and F (see b and c of Figure 4, respectively). For
these systems, the chemical shifts for natural ritterazine B match
diastereomer 13 (the thermodynamically favored spiroketal), not
diastereomer 17 as assigned in the isolation paper.1 Likewise,
ritterazine F matches the spiroketal in the nonanomeric con-
figuration (17), not the thermodynamically favored spiroketal
as originally proposed. This reassignment does not conflict with
the data obtained by the authors of the isolation papers; in their
case, the spiroketals were assigned on the basis of a single and
somewhat ambiguous NOE; other correlations were absent
because of poor NOE buildup due to the unfavorable correlation
times of the full-length natural product.1 However, the reduced
size of our synthetic halves has allowed a complete NOE
analysis.
We are uncertain about the exact environment within the
tunicates (or possible symbiotic organisms) that produce the
ritterazines (salinity, acidity, etc.),59 but we reason that ocean
water should reflect the conditions under which the ritterazines
were found and should serve as an appropriate medium to
(53) Seidl, P. R.; Carneiro, J. W. de M.; Tostes, J. G. R.; Koch, A.; Kleinpeter,
E. J. Phys. Chem. A 2005, 109, 802-806.
(54) Seidl, P. R.; Tostes, J. G. R.; Carneiro, J. W. de M.; Taft, C. A.; Dias, J.
F. J. Mol. Struct. 2001, 539, 163-169.
(55) Abraham, R. J.; Barlow, A. P.; Rowan, A. E. Magn. Reson. Chem. 1989,
27, 1074-1084.
This reassignment is consistent with the distribution of the
ritterazine natural products: the ritterazine B eastern half was
originally assigned as the spiroketal in the nonanomeric con-
figuration, but this spiroketal is conserved among seven natural
products. Ritterazine F, on the other hand, was assigned as the
thermodynamically favored spiroketal, but its spiroketal appears
(56) Distances were obtained from the X-ray crystal structure of the eastern
half of ritterazine G, shown in Figure 2.
(57) Distances were calculated from the minimum energy conformation found
from a Monte Carlo conformational search using the MM2* force field.
(58) Distances were obtained from the X-ray crystal structure of the ritterazine
B eastern half, shown in Figure 2.
(59) The ritterazines were isolated from the tunicate R. tokioka in waters with
an average temperature of 21 °C: Fukuzawa, S. Personal communication.
9
6594 J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007

Eastern Half Syntheses of Ritterazines B, F, G, and H
A R T I C L E S
Figure 4. Comparison of proton chemical shifts for ritterazines B, F, G, and H with those of their synthetic counterparts. Proton chemical shifts were
measured at 22 °C in pyridine-d5. Values for the synthetic derivatives are shown as differences between the natural product chemical shift1,3 and the synthetic
chemical shift: ∆δ ) δnatural - δsynthetic
.
5,5-spiroketals equilibrate in ocean water. These results dem-
onstrate that ocean water does catalyze the equilibration of the
spiroketals, albeit at vastly different rates for each ritterazine
spiroketal. For example, the C22 diastereomer of the eastern
half of ritterazine G (10) equilibrates nearly 50% after 4 days
at 40 °C, and after 33 days only 6% of the starting spiroketal
epimer remains; this result explains why ritterazine G was
isolated, but its C22 diastereomer was not. However, given the
equilibrium ratio between ritterazine G and its diastereomer (line
3, Table 1), we would expect a small quantity of the diastere-
omer to exist in nature. Using the equilibrium ratio as a guide,
we estimate that ca. 132 µg of ritterazine G’s diastereomer
should have been present along with the 2.2 mg of ritterazine
G found in the 8.2 kg of tunicates used in the isolation process.3
In contrast to the rapid equilibration of ritterazine G’s epimer,
only 4% of ritterazine F equilibrates after 4 days at 40 °C, and
50% conversion is not reached until 24 days.62 The differences
in equilibration rates are even more pronounced with ritterazine
H, where only 1% epimerizes after 4 days at 40 °C. This slow
conversion explains why this spiroketal with the nonanomeric
configuration persists in nature.
These results raise the question of what catalyzes the equil-
ibration in ocean water; magnesium(II) clearly catalyzes the
equilibration in organic solvent, but the catalytic species in ocean
water is less obvious. Lines 4-6 in Table 1 outline our efforts
to determine the identity of this catalyst. For example, the
diastereomer of ritterazine G equilibrates slowly in carbonate-
Figure 5. Effect of spiroketal stereochemistry on proton chemical shifts
(pyridine-d5, 22 °C).
measure whether these spiroketals equilibrate naturally. Shown
in Table 1 are data indicating the extent to which ritterazine
(60) Pusztai, A. Biochem. J. 1966, 99, 93-101.
(62) The equilibration of ritterazine F may or may not be relevant to its native
environment, depending on the lifetime of the ritterazines within the
tunicate.
(61) Cohn, E. J.; Conant, J. B. Proc. Natl. Acad. Sci. U.S.A. 1926, 12, 433-
438.
9
J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007 6595

A R T I C L E S
Phillips and Shair
Table 1. Equilibration of the Nonanomeric Spiroketals in Ocean Water
a Aqueous solvents were mixed with acetone in a 1:1 ratio to ensure solubility of the compounds. b Ocean water was collected from Crane Beach, Ipswitch,
MA, and organics were removed prior to use.60,61 c Ratios were determined by 1H NMR analysis. d ∆G° values were calculated at 298 K using the equation
∆G° ) -RT ln K.
buffered water, but the equilibration is faster when MgCl2 is
added to mimic the magnesium(II) dissolved in ocean water
(on average, ocean water has 53.1 mM magnesium(II)63).
However, the pH within the tunicate also may be relevant to
the rate of spiroketal equilibration64 because even 10-fold more
acidic water equilibrates the diastereomer of ritterazine G more
quickly (line 6, Table 1).
on ∆Gq, which explains why seemingly small values for ∆∆Gq
correspond to large differences in reaction rates. For example,
the ∆Gq value for the ritterazine H eastern half is only 1.8
kcal‚mol-1 higher than that for the diastereomer of ritterazine
G, but the half-life for ritterazine H is nearly 35 times longer.
The ritterazine F eastern half, on the other hand, equilibrates
only 2-fold faster than ritterazine H, due to the mere 0.6
kcal‚mol-1 difference in transition-state free energies. Clearly,
small structural differences between the ritterazine eastern halves
have large ramifications on the equilibration rates by influencing
the free energy of the transition states.
Mechanistic Studies on Ritterazine Equilibration. Ritt-
erazines F and H and the diastereomer of ritterazine G are
structurally very similar, but as shown in the kinetics studies
above, the free energies of the transition states are different,
which leads to their unique rates of equilibration. To eliminate
the possibility that the differences in transition-state free energies
are the result of differences in mechanism, we performed two
experiments to determine which oxonium ion is operative in
the transition state.
The first experiment involved trapping the oxonium ion with
a hydride source (Scheme 5); we assumed that both possible
oxonium ion intermediates would reduce at equal rates and that
the rate of reduction would not be limiting, both of which are
plausible in cases where a large excess of the reductant is used.
A single reduction product was observed for each reaction
shown in Scheme 5, indicating that the three compounds
equilibrate by the same mechanism, which involves opening of
the terminal spiroketal ring. In each example, the identity of
the product was determined using a combination of 1H chemical
Kinetics Studies on Ritterazine Equilibration. The various
rates of spiroketal equilibration in ocean water are consistent
with the natural distribution of the ritterazine 5,5-spiroketals,
but it is not clear why these structurally similar systems have
such different behavior. To address this question, we performed
a series of kinetics and mechanistic experiments on p-nitro-
benzoylated derivatives of the ritterazine halves. These studies
were performed in organic solvents, where the equilibration rates
are faster and, therefore, more convenient to monitor relative
to the rates in ocean water. Also, we measured the equilibration
rates for each ritterazine half at six different temperatures, which
allowed us to calculate and compare the free energy values for
the transition states (∆Gq) (Table 2).
The magnitudes of the transition-state free energies (∆Gq)
shown in Table 2 are low, which is consistent with the rapid
equilibration of these spiroketals at room temperature, but it is
the ∆∆Gq values (∆Gq
- ∆Gq
and ∆Gq
- ∆Gq
)
rittG
rittH
rittB
rittH
that highlight the differences in reaction rates between spiro-
ketals. The relationship between reaction rate and ∆Gq can be
described quantitatively in terms of half-lives by modification
of the Eyring equation, as follows:
-1
kBT
h
kBT
h
q
q
k ) κ
e-∆G /RT w T1/2 ) ln 2 κ
e-∆G /RT
(
)
(63) Akhtar, W.; Zaidi, I. A. S. S. H.; Jilani, S. Water, Air, Soil Pollut. 1997,
94, 99-107.
(64) Doney, S. C. Sci. Am. 2006, 294, 58-65.
This equation reveals the exponential dependence of half-lives
9
6596 J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007

Eastern Half Syntheses of Ritterazines B, F, G, and H
A R T I C L E S
Table 2. Equilibration of the Ritterazine F, G, and H Nonanomeric Spiroketals in Toluenea
a Reactions were run at six temperatures using a 500 µM concentration of the compound in toluene and 1000 equiv of MgSO4. Equilibrations were
performed in triplicate, taking eight time points per temperature; diastereomer ratios were determined by integration of HPLC peak areas. Temperatures were
chosen to cover a 40-fold change in reaction rate for each spiroketal. b The preparation of these labeled derivatives is outlined in the Supporting Information.
c ∆G° values were calculated at 298 K from the equation ∆G° ) -RT ln Keq, where values for Keq were measured 3-8 times at 23 °C. d ∆Gq values are
reported at 298 K and were calculated from the following equation: ∆Gq ) ∆Hq - T∆Sq. e ∆Hq and ∆Sq values were calculated from a plot of ln(kT-1
versus T-1 and are reported as the average of three independent measurements.
)
Scheme 5. Mechanism of Equilibration for Ritterazines B, G,
and Ha
a The reduction reactions were quantitative, as determined by integration
of the H NMR spectrum.
Figure 6. Transition-state models from molecular mechanics simulations.
The low-energy transition states were calculated from Monte Carlo searches
using the MMFF force field68 provided in the Spartan modeling package.
Calculated atom distances are included as a measure of unfavorable
nonbonding interactions: methyl-methyl distances are measured between
carbons, while methyl-hydrogen distances are measured between the methyl
carbon and the hydrogen of interest.
1
shift and NOE data, where representative NOEs are shown only
for the diastereomer of ritterazine G. The structure of this
product was further confirmed by comparison of its deprotected
derivative (27) with the acetate-deprotected version of
intermediate 6 that was prepared by an independent route
(Scheme 1).
The second experiment involved the measurement of entropy
contributions to the transition-state free energies. In this case,
the mechanism suggested by the trapping experiments is
supported by the nearly equivalent ∆Sq values for all three
compounds (Table 2). The other possible transition state would
have two additional rotatable bonds, which would increase the
∆Sq term by ca. 14 cal‚mol-1, a value that would be distinguish-
able among the ∆Sq values shown in Table 2.65 Our experimental
values, however, are all approximately zero, which is consistent
with a common mechanism and is frequently observed in cases
where solvent becomes more ordered around the charged
transition states.66,67
Since ritterazines B, G, and H equilibrate through the same
mechanism, the different ∆Gq values, and consequently the
different equilibration rates, must arise from minor structural
(66) Carpenter, B. K. Determination of Organic Reaction Mechanisms; John
Wiley & Sons: New York, 1984.
(67) Zuman, P.; Patel, R. C. Techniques in Organic Reaction Kinetics; John
Wiley & Sons: New York, 1984.
(65) Searle, M. S.; Williams, D. H. J. Am. Chem. Soc. 1992, 114, 10690-10697.
9
J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007 6597

A R T I C L E S
Phillips and Shair
variations in the transition state. Monte Carlo conformational
searches for all three transition states68 reveal increasing A(1,2)
strain and syn-pentane interactions moving from ritterazine G
to B to H, which accounts for the higher ∆Gq values found in
the latter (Figure 6). These unfavorable nonbonding interactions
are both distant- and angle-dependent; the distances are shown
in Figure 6 to highlight the increasing strain between the
transition states. Ultimately, the natural distribution of the
ritterazine 5,5-spiroketals is more a function of the transition-
state free energies (∆Gq) than of the ground-state free energies
(∆G°), and even small deviations in the steroid backbone affect
the energy of the transition states and, therefore, the rate of
spiroketal equilibration.
the cellular target of these molecules and of determining which
elements of their structures are important for biological activity.
Finally, equilibration of these ritterazine 5,5-spiroketals in
toluene and in ocean water has led to the observation that minor
and remote structural features on the steroid backbone can
impact the free energy of the transition states for spiroketal
equilibration; in fact, the transition-state free energies dictate
the natural distribution of the ritterazine spiroketals much more
than the ground-state free energies.
Acknowledgment. This work was supported by Novartis
Institutes of Biomedical Research. S.T.P. is a Damon Runyon
Fellow supported by the Damon Runyon Cancer Research
Foundation (Grant DRG-1805-04). We also thank Neil Vasan
and Amy L. Wojciechowski for their contributions at the outset
of the program. The crystal structures for compounds 21 and
23 were solved by Dr. Richard J. Staples of the X-ray
Crystallographic Laboratory at Harvard University.
Conclusion
We have developed short, efficient, and scalable routes to
the eastern halves of ritterazines B, F, G, and H and to the
spiroketal epimers of ritterazines G and H, and in doing so, we
have reassigned the spiroketal stereochemistry for the eastern
halves of ritterazines B and F. Our syntheses also demonstrate
a convenient strategy for preparing spiroketal-substituted cis-
hydrindanes. Consequently, we now have a route that could
produce large quantities of 10 ritterazine eastern halves, which
is an important step toward our ultimate goal of uncovering
Supporting Information Available: Experimental procedures
and characterization for the compounds described above,
tabulation of kinetics data, and CIF files for compounds 21 and
23. This material is available free of charge via the Internet at
http://pubs.acs.org.
(68) Gundertofte, K.; Liljefors, T.; Norrby, P.; Pettersson, I. J. Comput. Chem.
1996, 17, 429-449.
JA0705487
9
6598 J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007