Enantioselective total synthesis and confirmation of the absolute and relative stereochemistry of streptorubin B

Article abstract of DOI:10.1021/ja109165f

The enantioselective total synthesis of the pyrrolophane natural product streptorubin B is described. Key steps in the concise route include the application of a one-pot enantioselective aldol cyclization/Wittig reaction and an anionic oxy-Cope rearrangement to forge the crucial 10-membered ring. Comparisons between CD spectra of synthetic and natural samples of streptorubin B coupled with X-ray crystallography allowed for the determination of the absolute stereochemistry of this natural product for the first time. These studies also provided unambiguous proof of the relative configuration between the butyl side chain and the bispyrrole subunit. Additional studies revealed a novel atropstereoselective Paal-Knorr pyrrole condensation and provided fundamental experimental insight into the barrier for atropisomerization of the natural product.

Full text of DOI:10.1021/ja109165f

ARTICLE
pubs.acs.org/JACS
Enantioselective Total Synthesis and Confirmation of the Absolute
and Relative Stereochemistry of Streptorubin B
Dennis X. Hu, Michael D. Clift, Kiel E. Lazarski, and Regan J. Thomson*
Department of Chemistry, Northwestern University, Evanston, Illinois 60208, United States
S Supporting Information
b
ABSTRACT: The enantioselective total synthesis of the pyrrolophane natural prod-
uct streptorubin B is described. Key steps in the concise route include the application
of a one-pot enantioselective aldol cyclization/Wittig reaction and an anionic oxy-
Cope rearrangement to forge the crucial 10-membered ring. Comparisons between
CD spectra of synthetic and natural samples of streptorubin B coupled with X-ray
crystallography allowed for the determination of the absolute stereochemistry of this
natural product for the first time. These studies also provided unambiguous proof
of the relative configuration between the butyl side chain and the bispyrrole subunit. Additional studies revealed a novel atrop-
stereoselective Paal-Knorr pyrrole condensation and provided fundamental experimental insight into the barrier for atropisomeri-
zation of the natural product.
’ INTRODUCTION
the basis of comparisons to the data collected by Floss, stated that
4 was in fact a natural product.11 A second synthesis of 4 was
reported in 2007 by Reeves, who came to the same conclusion as
F€urstner.12 These conclusions are confusing, given Floss’ report
that his data for the “pink pigment”,9 to which he assigned the
structure 4, “closely matched” that reported by Gerber, even
though at that time she had already revised her assignment in
favor of 1. Furthermore, in 2008, Challis and co-workers reported
that S. coelicolor, the organism from which Floss isolated his “pink
pigment”, produced 1; they found no evidence for production
of 4.5b
The structure of streptorubin B (1) and its identity as a natural
product are less questionable because of the full complement of
NMR spectroscopic analyses conducted Weyland and co-work-
ers in 19913 as well as those by Challis and workers in 2008.5b
Key to the structural assignment was the presence of an upfield
resonance at -1.55 ppm in the proton NMR spectrum that was
attributed to one of the diastereotopic C40 hydrogens (Weyland
numbering), which suffers anisotropic shielding by the pyrrole
nucleus as a result of the conformation of the strained 10-
membered ring. Metacycloprodigiosin (2)4 and prodigiosin R1
(5)6 have similar proton resonances. While the structures of
these two molecules have been confirmed through total syn-
thesis,11-13 there have been no reported syntheses of 1, although
the groups of F€urstner13c and Chang14 have each prepared the
pyrrole core. While the gross structure of 1 is not in doubt, several
stereochemical issues remain. Weyland and co-workers noted
that 1 possesses an element of planar stereochemistry;3 that is,
there are two potential atropdiastereomers of 1, depending on
the relative stereochemistry of the butyl side chain and the
The prodigiosin alkaloids have attracted widespread attention
from both chemists and biologists because of their unique
molecular architectures and the range of potentially useful
biological activities they display.1 For example, a prodigiosin-
inspired medicinal chemistry program at Gemin X led to the
development of a synthetic small-molecule Bcl inhibitor that is
currently in phase I and II clinical trials for the treatment of a
variety of cancers.2 From a structural standpoint, streptorubin B
(1)3 and metacycloprodigiosin (2)4 are especially notable be-
cause of their highly strained pyrrolophane cores, which are
formed by oxidative ring closure from a common precursor,
namely, the natural product undecylprodigiosin (3, Figure 1).5
The more recently isolated congener prodigiosin R1 (5)6 has a
structure similar to that of 2 and provides an interesting link
between these molecules and the related compound roseophilin
(6),7 while butylcycloheptylprodigiosin (4)8,9 is a constitutional
isomer of both 1 and 2. In fact, the structural similarity between 1
and 4 led to significant confusion regarding the actual structures
of these compounds. In 1975, Gerber and co-workers isolated a
red pigment from Streptomyces sp. Y-42 and Streptomyces abi-
koensis to which they assigned the structure given by 4.8 Likewise,
in 1985 the group of Floss assigned the structure 4 to a “pink
pigment” they isolated from a strain of Streptomyces coelicolor.9
Subsequent to these publications, in 1991 Weyland and co-
workers reported the isolation of a pigment from an actinomy-
cete (strain B 4358), which they showed to possess the pyrro-
lophane structure given by 1.3 They also went on to state that the
assignment of 4 put forth earlier by Gerber (and presumably
Floss also) should be revised in favor of 1. Curiously, Gerber
herself came to the same conclusion in 1978, although she did
not elaborate on the reasons for this reassignment.10 In 2005,
F€urstner and co-workers reported the first synthesis of 4 and, on
Received: October 12, 2010
Published: December 17, 2010
r
2010 American Chemical Society
1799
dx.doi.org/10.1021/ja109165f J. Am. Chem. Soc. 2011, 133, 1799–1804
|

Journal of the American Chemical Society
ARTICLE
Figure 3. Failed RCM approach to the streptorubin B core.
Scheme 1. Synthesis Plan for Streptorubin B (1)
we were unable to confirm the absolute configuration of these
natural products because of a lack of access to authentic samples.
Subsequent to our publication, we were able to confirm the
absolute configuration of natural 2 through comparison of the
CD spectra recorded for our synthetic material and a natural
sample isolated by Challis and co-workers.16 The synthetic
sample, which was R, matched the natural sample, thus establish-
ing that natural metacycloprodigiosin possesses the R configura-
tion for the side chain.
Figure 1. Some prodigiosin alkaloids and roseophilin.
Wishing to also establish the stereochemistry of streptorubin B
(1), we sought to apply the same strategy to its synthesis that had
proven successful for 2.13d Our total synthesis of 2 relied upon a
ring-closing metathesis (RCM) reaction of diene 7 to produce
the requisite 12-membered ring in 69% yield. In order to
synthesize 1, we prepared diene 8 by a route analogous to that
which we used to prepare 7. In contrast to what we found for 12-
membered ring formation, diene 8 failed to undergo the desired
ring closure (Figure 3). With the use of a variety of metathesis
catalysts, the only isolated products were mixtures of dimeric
species. Repositioning the alkene did not lead to any improve-
ment. The failure of 8 to undergo efficient RCM is most likely a
consequence of severe ring strain in the 10-membered product,
and we therefore devised an alternative strategy.
Our revised synthesis plan for 1 (Scheme 1) entailed initial
disconnection of the bispyrrole side arm of the natural product to
generate the pyrrolophane core (11). Simplification of 11 by the
Paal-Knorr pyrrole transform followed by some functional
group interconversions led back to cyclodecanone 12, which
contains the full retron for the anionic oxy-Cope rearrange-
ment.17 Thus, the synthesis was simplified to devising a route to
gain ready access to the functionalized cyclohexanol precursor
13. To this end, we wished to employ a proline-catalyzed
enantioselective desymmetrizing intramolecular aldol reaction
on dialdehyde 1518 followed by an in situ Wittig reaction to form
14, which could be easily converted to cyclohexanol 13.
Figure 2. Atropisomerism within streptorubin B (1).
bispyrrole side arm (Figure 2). This stereochemical relationship
has never been established. Additionally, the absolute configura-
tion of 1 has not been established. In fact, at the outset of our
work, the only compound of this family with known absolute
stereochemisty was the related compound roseophilin (6). The
independent synthetic efforts of the Tius and Boger groups15
established that 6 possesses the absolute stereochemistry indi-
cated in Figure 1. It was against this background that we initiated
a synthesis of 1 with the goal of clarifying the structural
ambiguities surrounding this fascinating molecule.
’ RESULTS AND DISCUSSION
Total Synthesis. In 2009, our research group completed the
first enantioselective total synthesis of metacycloprodigiosin (2)
and its relative, prodigiosin R1 (5).13d At the time of publication,
In 2003, List and co-workers reported the development
of a proline-catalyzed enantioselective exo-enol-6-exo-trig aldol
reaction as an effective means for desymmetrizing acyclic
1800
dx.doi.org/10.1021/ja109165f |J. Am. Chem. Soc. 2011, 133, 1799–1804

Journal of the American Chemical Society
ARTICLE
Scheme 2. Enantioselective Total Synthesis of Streptorubin B (1)
(as the HCl salt) was identical to the natural product as deter-
mined by 1H and 13C spectroscopy and mass spectrometry (MS).22
Thus, streptorubin B (1) was prepared in nine steps from 16 in
20% overall yield.
Relative Stereochemistry of Streptorubin B and Key Inter-
mediates. We were intrigued by the observation that the
initially formed species of the final condensation step in the
synthesis converted into the natural product upon standing. We
speculated that the initially formed species was one atropdiaste-
reomer of the natural product and that over time it isomerized
into the other, presumably lower-energy atropisomer, which has
the same structure as the natural product. Inspection of molec-
ular models of the two possible atropisomers indicated that the
lower-energy atropisomer should be anti-streptorubin B, in
which the severe eclipsing interactions present within syn-strep-
torubin B are avoided. Analysis of the two possible atropdiaster-
eomers, anti-1 and syn-1, revealed that anti-1 should display
NOESY cross-peaks for HA and HB as well as HC and HD but lack
HA-HC and HB-HD cross-peaks. The isomeric compound syn-
1 should display the opposite correlations. The NOESY data
obtained, as shown in Figure 4, were fully consistent with the
structure assigned as anti-1, where the two side arms effectively
project from opposite faces of the 10-membered ring. An
additional NOESY spectrum of the initially formed mixture
resulting from condensation between 11 and 21 showed a clear
HB-HD cross-peak, indicating that the initial adduct was indeed
syn-1. From 1H NMR spectroscopy, the equilibrium ratio of the
anti and syn isomers is approximately 10:1 at room temperature,
indicating that the anti isomer is thermodynamically favored by
approximately 1.4 kcal mol-1. By monitoring the change in the
syn-1/anti-1 ratio as a function of time at two temperatures, we
determined the activation barrier for the conversion of syn-1 into
Figure 4. NOESY analysis of streptorubin B isomers.
dialdehydes.18 In a footnote of the manuscript, the authors
reported the use of an in situ Horner-Wadsworth-Emmons
(HWE) reaction as a means of producing compounds with a
chromophore for easy determination of enantiomeric excess by
HPLC. We speculated that such a route employing a Wittig
reaction rather than the HWE reaction might provide rapid
access to the homoallylic alcohol (i.e., 14) that we needed for our
streptorubin B synthesis. Accordingly, we treated aldehyde 15,
which had been prepared in one step19 from commercially
available cycloheptene (16), with 10 mol % (S)-proline and
upon consumption of the dialdehyde added ylide 18 to the
reaction mixture (Scheme 2). In this way, the homoallylic alcohol
14 was obtained as the major diastereomer in 69% yield as a 98:2
mixture of enantiomers. Oxidation of 14 followed by addition of
the vinyl anion 1920 gave 13, the required precursor to the
anionic oxy-Cope rearrangement, with an er of 97:3. Exposure of
alcohol 13 to KHMDS and 18-crown-6 produced the desired 10-
membered ring 12 in 85% yield. The enantiopurity of 12 was
found to be 97:3 er, indicating a good level of stereochemical
transfer during the oxy-Cope rearrangement. With 12 in hand,
generation of the pyrrole core 11 proceeded smoothly following
alkene reduction with concomitant benzyl ether cleavage, oxida-
tion of the liberated alcohol to the aldehyde, and Paal-Knorr
pyrrole synthesis (67% yield over three steps). Completion of the
synthesis involved conducting an acid-promoted condensation
between pyrrole 11 and aldehyde 21,21 which was followed by
removal of the Boc group by basic methanolysis. Analysis of the
bright-red material thus obtained revealed an approximately 10:1
mixture of two compounds in which the major compound did
not match the natural product. Reexamination of this NMR
sample after 10 days, however, revealed that this mixture had
transformed almost completely to 1. The synthetic material
anti-1 to be approximately 20.5 kcal mol-1 23
. Thus, on the basis
of these NMR spectroscopy studies, we established the relative
stereochemistry of streptorubin B (1) for the first time. Unequi-
vocal evidence was later obtained through X-ray crystallography
(see below).
The question of why we initially observed a mixture of isomers
favoring the unnatural atropisomer, syn-streptorubin B, during
the final condensation step still remained unanswered, however.
We hypothesized that perhaps the pyrrole precursor 11 existed as
an atropisomer that favored the unnatural syn configuration,
which would have led to the initial generation of syn-streptorubin
1801
dx.doi.org/10.1021/ja109165f |J. Am. Chem. Soc. 2011, 133, 1799–1804

Journal of the American Chemical Society
ARTICLE
Scheme 3. Dissection of the Atropisomer Issue
Scheme 4. Atropselective Paal-Knorr Pyrrole Synthesis
temperature, anti-11 equilibrated to a 4:1 mixture favoring syn-
11. Furthermore, we showed that condensation of anti-11 with
bispyrrole 21 directly afforded 1 as the favored atropisomer.
In hindsight, it was fortunate that the “unnatural” pyrrole
isomer syn-11 could be converted into the natural product
through isomerization of the “unnatural” isomer of streptorubin
B (i.e., syn-1) into streptorubin B. Had this final isomerization
not occurred, however, completion of the total synthesis would
still have been possible by taking advantage of a kinetically
controlled Paal-Knorr pyrrole condensation of anti-22 to gen-
erate anti-11 and hence the correct atropisomer of 1.
B following condensation with bispyrrole 21 (Scheme 3). This
higher-energy species would then have undergone conversion
into the more stable anti isomer of the natural product. In fact,
F€urstner and co-workers had prepared racemic pyrrole 11 as part
of their work on 1 and noted in their manuscript13c that 11
displays an element of planar stereochemistry, although assign-
ment of this stereochemistry and conversion of 11 into the
natural product were not reported. They did, however, note that
coalescence of pyrrole 11 could not be reached, indicating a
barrier to rotation of greater than 17.5 kcal mol-1. The NOESY
experiment on pyrrole 11 conducted in our laboratories revealed
that it exists predominantly as the unnatural isomer syn-11 and
Absolute Configuration of Streptorubin B. As a final goal,
we wished to establish the absolute stereochemistry of 1, which
necessitated that we first establish the absolute stereochemistry
of our synthetic material. We had established the configuration of
the initial aldol/Wittig reaction product 14 through Mosher ester
analysis25 but were mindful of the fact that the stereochemistry of
the butyl side chain was set through the anionic oxy-Cope
rearrangement. While it seemed entirely reasonable to expect
the Cope product 12 to possess the stereochemistry indicated in
Scheme 2 by way of the minimized chairlike transition state (i.e.,
20), we required stronger evidence. Fortunately, the solution to
this issue came in the form of X-ray-quality crystals of the HCl
salt of synthetic 1 itself (Figure 5A). Thus, through anomalous
dispersion (Flack parameters of 0.04 for the indicated enantio-
mer and 0.87 for the opposite enantiomer)26 we were able to
unambiguously assign the configuration of our synthetic material
as R, as predicted in our synthetic scheme (see Scheme 2).
Additionally, this X-ray structure confirmed our NMR structural
assignment that 1 exists as the anti atropdiastereomer (Figure 4).
Comparisons between the CD spectra of a natural sample
and our synthetic streptorubin B were somewhat surprising
(Figure 5B). As shown in Figure 5B, the CD spectrum of the
natural sample (blue) is the mirror image of the CD spectrum of
synthetic (R)-streptorubin B (green), indicating that natural
streptorubin B possesses an S configuration for the side chain
(Figure 5C).27 This stereochemistry is the opposite of that of
metacycloprodigiosin (2) and is most likely a consequence of the
conformation adopted by the linear side chain within the
common biogenetic precursor, undecylprodigiosin (3), when it
undergoes enzyme-mediated oxidative cyclization to produce
either of the two natural products (i.e., streptorubin B or
metacycloprodigiosin). We further confirmed this stereochemi-
cal assignment through preparation of the S enantiomer of
streptorubin B (1) by using (R)-proline rather than (S)-proline
during the one-pot aldol/Wittig reaction. This material displayed
a CD spectrum (Figure 5B, red) identical in shape to that of the
natural material.
1
that by H NMR spectroscopy the ratio of syn-11 to anti-11 is
approximately 10:1 when 11 is first synthesized. When as-
synthesized 11 is left to stand over a period of 2 weeks or longer,
the equilibrium ratio appears to be approximately 4:1 in favor of
syn-11. Thus, a complete picture of this complex final condensa-
tion step has been revealed. When first synthesized, pyrrole
11 exists as a 10:1 mixture favoring the syn atropisomer, which
upon condensation with 21 affords a 10:1 mixture of adducts
favoring syn-1, which then isomerizes to the more stable com-
pound, anti-1.24
Further analysis of pyrrole 11 and its immediate precursor,
cyclodecanone syn-22, led us to speculate that perhaps the syn
atropisomer of 11 was formed kinetically during the Paal-Knorr
condensation (Scheme 4). If condensation of enamine 23 onto
the endocyclic ketone occurs faster than bond rotation to form
24, this might account for the selective (10:1) formation of syn-
11 (i.e., in a ratio greater than the equilibrium ratio, which we
determined to be 4:1). We hypothesized that if this were the case,
diastereomeric cyclodecanone anti-22 would afford anti-11 upon
Paal-Knorr pyrrole synthesis via enamine 24. Cyclodecanone
anti-22 was synthesized using a modified version of the route
used to prepare syn-22 (see the Supporting Information) and
exposed to the conditions for pyrrole synthesis. Analysis of the
pyrrole thus obtained revealed that the major isomer in this case
was anti-11, indicating that the cyclodecanone precursors do not
converge to the same intermediate upon condensation with
ammonium acetate and that the intermediate enamines 23 and
24 do not readily interconvert under the reaction conditions. We
did observe, however, that over the course of 10 days at room
1802
dx.doi.org/10.1021/ja109165f |J. Am. Chem. Soc. 2011, 133, 1799–1804

Journal of the American Chemical Society
ARTICLE
Figure 5. (A) X-ray structure of synthetic (R)-streptorubin B. (B) CD spectra. (C) Confirmed absolute and relative stereochemistry of streptorubin B.
’ CONCLUSIONS
’ REFERENCES
(1) For reviews of the chemistry and biology of the prodigiosins, see:
(a) F€urstner, A. Angew. Chem., Int. Ed. 2003, 42, 3582–3603. (b)
Williamson, N. R.; Fineran, P. C.; Leeper, F. J.; Salmond, G. P. C.
Nat. Rev. Microbiol. 2006, 4, 887–899. (c) Williamson, N. R.; Fineran,
P. C.; Gristwood, T.; Chawrai, S. R.; Leeper, F. J.; Salmond, G. P. C.
Future Microbiol. 2007, 2, 605–618.
(2) (a) Borthakur, G.; O’Brien, S.; Ravandi-Kashani, F.; Giles, F.;
Schimmer, A. D.; Viallet, J.; Kantarjian, H. Blood 2006, 108, 750. (b)
Nguyen, M.; et al. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 19512–19517.
(3) Laatsch, H.; Kellner, M.; Weyland, H. J. Antibiot. 1991, 44, 187–
191.
(4) (a) Wasserman, H. H.; Rodgers, G. C.; Keith, D. D. J. Am. Chem.
Soc. 1969, 91, 1263–1264. (b) Wasserman, H. H.; Keith, D. D.; Rodgers,
G. C. Tetrahedron 1976, 32, 1855–1861.
(5) (a) Cerde~no, A. M.; Bibb, M. J.; Challis, G. L. Chem. Biol. 2001, 8,
817–829. (b) Mo, S.; Sydor, P. K.; Corre, C.; Alhamadsheh, M. M.;
Stanley, A. E.; Haynes, S. W.; Song, L.; Reynolds, K. A.; Challis, G. L.
Chem. Biol. 2008, 15, 137–148. Also see ref 1.
The absolute and relative stereochemistry of streptorubin B
(1) has been established through enantioselective total synthesis,
providing definitive proof of the structure of this intriguing
compound after decades of speculation. The concise synthesis
made use of a novel one-pot enantioselective aldol/Wittig
reaction to form a key homoallylic alcohol with high levels of
enantiopurity. This powerful reaction, in combination with the
anionic oxy-Cope rearrangement, proved to be pivotal for the
construction of the strained 10-membered ring required for the
synthesis. This straightforward procedure should prove useful in
the enantioselective synthesis of other substituted 10-membered
rings prevalent in many bioactive natural products. The complex
question of atropstereoisomerism within 1 and its pyrrole pre-
cursor 11 was answered through a combination of NMR analysis
and X-ray crystallography, while the absolute stereochemistry of 1
was determined through comparisons of CD spectra. Alongside our
established enantioselective routes to the 12-membered pyrrolo-
phane prodigiosins (i.e., metacycloprodigiosin and prodigiosin
R1),13d the synthesis outlined herein for the 10-membered isomer
(i.e., streptorubin B) will enable the ready preparation of these
natural products (and analogues) for further biological evaluation.
(6) Kawasaki, T.; Sakurai, F.; Hayakawa, Y. J. Nat. Prod. 2008, 71,
1265–1267.
(7) Hayakawa, Y.; Kawakami, K.; Seto, H.; Furihata, K. Tetrahedron
Lett. 1992, 33, 2701–2704.
(8) (a) Gerber, N. N. J. Antibiot. 1975, 28, 194–199.(b) Gerber,
N. N.; Stahly, D. P. Appl. Microbiol. 1975, 30, 807–810.
(9) Tsao, S.-W.; Rudd, B. A. M.; He, X.-G.; Chang, C.-J.; Floss, H. G.
J. Antibiot. 1985, 38, 128–131.
(10) Gerber, N. N.; McInnes, A. G.; Smith, D. G.; Walter, J. A.;
Wright, J. L. C.; Vining, L. C. Can. J. Chem. 1978, 56, 1155–1163. In this
manuscript, Gerber and coworkers state that “Streptomyces and Strepto-
verticillium species also produce the related butylcycloheptylprodiginine
(1, R1R3 = 10,70-cyclo(CH2)6-CH(CH2)3CH3, R2 = R4 = H), which has
hitherto been incorrectly identified as a 2,3-cycloalkyl derivative.” In
other words, the compound called “the related butylcycloheptylprodi-
ginine” is actually streptorubin B.
’ ASSOCIATED CONTENT
S
Supporting Information. Complete ref 2b, detailed ex-
b
perimental procedures, spectral data for all compounds, and
X-ray structure data (CIF). This material is available free of
charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
r-thomson@northwestern.edu
’ ACKNOWLEDGMENT
This paper is dedicated to Professor David A. Evans. Support
for this work was provided by Northwestern University (NU)
and the NIH/NIGMS (1R01GM085322). D.X.H. is a 2010
Barry M. Goldwater Scholar and a 2010 ACS SURF recipient
and gratefully acknowledges support from NU in the form of an
Undergraduate Research Grant. We thank Professor Gregory
Challis (University of Warwick, U.K.) for a sample of natural
streptorubin B, the CD spectrum of natural metacycloprodigiosin,
and helpful discussions. Charlotte Stern (NU) and Dr. Amy
Sargeant (NU) are thanked for assistance with X-ray analysis.
(11) F€urstner, A.; Radkowski, K.; Peters, H. Angew. Chem., Int. Ed.
2005, 44, 2777–2781.
(12) Reeves, J. T. Org. Lett. 2007, 9, 1879–1881.
(13) For metacycloprodigiosin, see: (a) Wasserman, H. H.; Keith,
D. D.; Nadelson, J. J. Am. Chem. Soc. 1969, 91, 1264–1265. (b)
Wasserman, H. H.; Keith, D. D.; Nadelson, J. Tetrahedron 1976, 32,
1803
dx.doi.org/10.1021/ja109165f |J. Am. Chem. Soc. 2011, 133, 1799–1804

Journal of the American Chemical Society
ARTICLE
1867–1871. (c) F€urstner, A.; Szillat, H.; Gabor, B.; Mynott, R. J. Am.
Chem. Soc. 1998, 120, 8305–8314. (d) Clift, M. D.; Thomson, R. J. J. Am.
Chem. Soc. 2009, 131, 14579–14583. For prodigiosin R1, see ref 13d.
(14) Chang, M. Y.; Pai, C. L.; Chen, H. P. Tetrahedron Lett. 2005, 46,
7705–7709.
(15) (a) Harrington, P. E.; Tius, M. A. J. Am. Chem. Soc. 2001, 123,
8509–8514. (b) Boger, D. L.; Hong, J. Y. J. Am. Chem. Soc. 2001, 123,
8515–8519.
(16) See the Supporting Information for CD spectra of synthetic
material prepared as in ref 13d and the CD spectrum of natural
metacycloprodigiosin kindly provided by Prof. Gregory Challis.
(17) Evans, D. A.; Golob, A. M. J. Am. Chem. Soc. 1975, 97, 4765–
4766.
(18) Pidathala, C.; Hoang, L.; Vignola, N.; List, B. Angew. Chem., Int.
Ed. 2003, 42, 2785–2788.
(19) Hong, B.-C.; Tseng, H.-C.; Chen, S.-H. Tetrahedron 2007, 63,
2840–2850.
(20) 19 was prepared from the corresponding iodide through
lithium-halogen exchange (see the Supporting Information). The
iodide was prepared using the protocol of Huang and Negishi (see:
Huang, Z.; Negishi, E. Org. Lett. 2006, 8, 3675–3678).
(21) Aldrich, L. N.; Dawson, E. S.; Lindsley, C. W. Org. Lett. 2010,
12, 1048–1051.
(22) Key compound data for synthetic 1: IR (neat) 3436, 3371,
3019, 2937, 1764 cm-1. 1H NMR (500 MHz, CDCl3): δ 12.70 (s, 1H),
12.66 (s, 1H), 12.58 (s, 1H), 7.22 (s, 1H), 7.21 (s, 1H), 7.11 (s, 1H), 6.90
(s, 1H), 6.51 (s, 1H), 6.34 (t, 1H, J = 2.7 Hz), 6.09 (s, 1H), 4.02 (s, 3H),
3.33 (m, 1H), 3.10 (m, 1H), 2.53 (dt, 1H, J = 12.3, 5.4 Hz), 1.96-1.68
(m, 5H), 1.55 (m, 2H), 1.37 (m, 5H), 1.24 (m, 3H), 1.14 (m, 2H), 0.91
(m, 5H), 0.78 (m, 1H), -1.55 (dt, 1H, J = 14.1, 8.8 Hz). 13C NMR (125
MHz, CDCl3): δ 165.4, 154.4, 150.6, 147.0, 126.7, 125.0, 122.4, 120.3,
116.8, 116.6, 112.5, 111.5, 92.8, 58.7, 38.9, 37.5, 31.7, 30.9, 30.4, 30.0,
29.1, 27.7, 25.4, 22.9, 14.2. HRMS (ESI) m/z: exact mass calcd for
C25H33N3O [M þ Hþ], 392.2696; found, 392.2700.
(23) See the Supporting Information for full details regarding the
measurement and calculation of this activation energy.
(24) Because the isomers of pyrrole 11 do not interconvert rapidly at
room temperature, the Curtin-Hammett principle need not be con-
sidered in order to rationalize this outcome. If the barrier for inter-
conversion of the two pyrrole atropisomers were sufficiently low, one
might have predicted that anti-1 would be formed preferentially since
the rate of condensation of anti-11 with 21 is most likely significantly
faster than that of syn-11 with 21. Since this interconversion does not
occur rapidly under the reaction conditions, the product distribution
reflects the ground-state stabilities of the two starting atropisomers. For
further discussion of the Curtin-Hammett principle, see: Seeman, J. I.
Chem. Rev. 1983, 83, 83–134.
(25) (a) Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512–
519. (b) Hoye, T. R.; Jeffrey, C. S.; Shao, F. Nat. Protoc. 2007, 2, 2451–
2458. See the Supporting Information for details
(26) Flack, H. D. Acta Crystallogr., Sect. A 1983, 39, 876–881.
(27) Challis and co-workers came to the same conclusion on the
basis of their own studies of the biosynthesis of streptorubin B (private
communication).
1804
dx.doi.org/10.1021/ja109165f |J. Am. Chem. Soc. 2011, 133, 1799–1804