Evaluation of the biosynthetic proposal for the synthesis of marineosins A and B

Article abstract of DOI:10.1021/ol100034p

Chemical equation presented The first synthetic efforts toward marlneosins A and B, novel spiroaminals from a Streptomyces actinomycete, are described by evaluation of the proposed biosynthesis. The hypothesized biosynthetlc C1-C25 Diels-Alder substrate was prepared In 8 steps in 5.1% overall yield; however, the proposed biomimetic inverse-electron-demand hetero-Diels-Alder reaction failed to deliver the marineosin core. Molecular mechanics supports this observation.

Full text of DOI:10.1021/ol100034p

ORGANIC
LETTERS
2010
Vol. 12, No. 5
1048-1051
Evaluation of the Biosynthetic Proposal
for the Synthesis of Marineosins A and B
Leslie N. Aldrich, Eric S. Dawson, and Craig W. Lindsley*
Departments of Chemistry and Pharmacology, Vanderbilt Institute of Chemical
Biology, Vanderbilt UniVersity Medical Center, NashVille, Tennessee 37232
craig.lindsley@Vanderbilt.edu
Received January 7, 2010
ABSTRACT
The first synthetic efforts toward marineosins A and B, novel spiroaminals from a Streptomyces actinomycete, are described by evaluation
of the proposed biosynthesis. The hypothesized biosynthetic C1-C25 Diels-Alder substrate was prepared in 8 steps in 5.1% overall yield;
however, the proposed biomimetic inverse-electron-demand hetero-Diels-Alder reaction failed to deliver the marineosin core. Molecular mechanics
supports this observation.
Recently, Fenical and co-workers reported on the isolation,
characterization, and biological evaluation of two novel
spiroaminals, marineosins A (1) and B (2), from a marine-
derived Streptomyces-related actinomycete (Figure 1).1 Bio-
synthetically, 1 and 2 appear to be derived from previously
unknown modifications within the prodigiosin-like pigment
pathways,2 and both displayed inhibition of human colon
carcinoma (HCT-116 IC50’s of 0.5 and 46 µM, repsectively),
with 1 significantly more potent.1 Fenical also proposed a
possible biosynthesis of 1 and 2 via an inverse-electron-
demand hetero-Diels-Alder reaction on acyclic C1-C25
substrate 5 that would form the pyran ring and spiroaminal
in a single step (Figure 2).1 On the basis of the novel
Figure 1. Structures of marineosins A (1) and B (2).
structure, the biological activity, and the enticing biosynthetic
proposal for 1 and 2, efforts toward a biomimetic total
synthesis of 1 and 2 were pursued in order to test the
proposed biosynthesis.
Synthetic efforst focused on the biosynthesis proposed by
Fenical in which a condensation between the C1-C9 bis-
pyrrole 3 and the enone-containing C10-C24 pyrrole 4 was
envisioned to deliver the Diels-Alder substrate 5. An
intramolecular inverse-electron-demand hetero-DielssAlder
reaction would then provide in a single step the spiroaminal
pyranyl core 6. Subsequent reduction across C22-C23 and
C6-C7 would provide marineosins A (1) and B (2). In this
Letter, we describe the synthesis of 5, the acyclic C1-C25
biosynthetic intermediate, and attempts toward a biomimetic
synthesis of 1 and 2; unfortuately, the biosynthetic proposal
could not be validated in the laboratory.
(1) Boonlarppraadab, C.; Kauffman, C. A.; Jensen, P. R.; Fenical, W.
Org. Lett. 2008, 10, 5505–5508.
Synthetic efforts focused on the construction of the key
C1-C9 bis-pyrrole 3 and the enone-containing C10-C24
(2) Fuerstner, A. Angew. Chem., Int. Ed. 2003, 42, 3582–3603.
10.1021/ol100034p  2010 American Chemical Society
Published on Web 02/08/2010

Scheme 2. Retrosynthesis of C10-C24 Fragment 4
Figure 2. Proposed biosynthesis of marineosins A (1) and B (2)
or B(OH)2) 10 and 9-bromo-1-nonene 11. Both sp2-sp3
Suzuki couplings and SN2 reactions between 10 and 11 failed
to afford the desired product. Similarly, via pathway b,
wherein the R in 13 was a CH2Br or CH2OMs, SN2 chemistry
also failed with organometallic reagents 14. Ultimately,
addition of an octenyl Grignard to a protected pyrrole-2-
carboxaldehyde proved effective, but with quite unexpected
results.
from the condensation of 3 and 4 to deliver 5, followed by a inverse-
electron-demand Diels-Alder reaction to produce 6.
pyrrole 4. The synthesis of 3 proved to be starightforward
following literature methods utilized in the synthesis of
structurally similar tamajamines and the BcI inhibitor oba-
toclax.3,4 As shown in Scheme 1, a Vilsmeier-Haack
Scheme 1. Synthesis of C1-C9 Protected Bis-pyrrole 9
Scheme 3. Grignard Addition to Aldehyde 15
As shown in Scheme 3, treatment of Boc-protected
pyrrole-2-carboxaldehyde 15 with Gringard 16 provided the
unexpected oxazolidinone 17, resulting from 1,2-addition,
followed by cyclization, in 78% yield. Although this was
not a productive venue to access 4, it led us to hypothesize
a one-pot addition, rearrangement, deoxygenation, and
deprotection cascade based on the precendent set forth by
Muchowski.5 In 1985, Muchowski demonstrated that NaBH4
reduction of phenylsulfonamide-protected 2-carboxypyrroles
18 led directly to deoxygenated and deprotected congeners
23 (Scheme 4).5 Thus, we postulated that addition of
Gringard 16 to pyrrole 24, followed by NaBH4, may lead
directly to the desired 25 in a single pot. In the event, 25
was indeed produced in the one-pot cascade, but the major
product 26 was delivered in a 1:10 ratio (25:26) in 70% yield.
Ultimately, 25 could be delivered in high yield by a three-
step sequence involving Grignard 16 addition to pyrrole
aldehyde 24 to provide secondary alochol 26. Ley oxidation
haloformylation was performed on 4-methoxy-3-pyrrolin-2-
one 7 to provide bromoenamine 8 in 59% yield. Suzuki
coupling with Boc-1H-pyrrol-2yl boronic acid delivered the
Boc-protected analogue 9 of the C1-C9 3 in 48% yield.
Retrosynthesis of the enone-containing C10-C24 pyrrole
4 led to a number of reasonable approaches (Scheme 2). In
both disconnection routes a and b, the enone was envisioned
to be installed through a cross-metathasis with 12. Discon-
nection pathway a led to the 2-pyrrole organometallic (Li
(3) Dari, K.; Yao, Y.; Faley, M.; Tripathy, S.; Rioux, E.; Billot, X.;
Rabouin, D.; Gonzalez, G.; Lavalle, J. F.; Attardo, G. Org. Process Res.
DeV. 2007, 11, 1051–1057
.
(4) Pinkerton, D. M.; Banwell, M. G.; Willis, A. C. Org. Lett. 2007, 9,
(5) Greenhouse, R.; Ramirez, C.; Muchowski, J. M. J. Org. Chem. 1985,
50, 2961–2964.
5127–5130
.
Org. Lett., Vol. 12, No. 5, 2010
1049

Scheme 4
.
One-Pot Addition, Rearrangement, Deoxygenation,
and Elimination
Scheme 6. Synthesis of C10-C24 Fragment 4
ature allowed the cross-metathasis product 4 to be isolated
in 40% yield.
With the two biosynthetic fragments, 9, the C1-C9
Boc-protected 3, and the C10-C24 enone 4 in hand, an
acid-mediated condensation smoothly formed the C9-C10
bond, delivering the C1-C25 acyclic precursor 5 for the
proposed inverse-electron-demand hetero-Diels-Alder
reaction in 92% yield (Scheme 7). Isolation of this highly
affords the corresponding ketone 27, and application of
Muchowski’s one-pot cascade delivers 25 in 48% for the
three steps (Scheme 5).
Scheme 5. Synthesis of C10-C21 Fragment 25
Scheme 7. Synthesis of C1-C25 Diels-Alder Substrate 5
With 25 in hand, the stage was set for the cross-metathasis
reaction to assemble the C10-C24 fragment 4.6 Once again,
this “straightforward” approach provided unexpected results.
Under standard cross-metathasis conditions with catalytic
Grubbs II, 25, and 12 (Scheme 6), the desired 4 was produced
as a minor product along with conjugate addition products
28 and 29. To the best of our knowledge, this is the first
account of Grubbs II catalyzing 1,4-conjugate additions, and
the reaction is quite general with respect to both electron-
rich pyrroles and acyclic Michael acceptors.7 However,
increasing catalyst loading from 0.5 to 30 mol % concomitant
with lowering the temperature from 40 °C to room temper-
colored poly pyrrole was difficult and required significant
refinement to the isolation and purification steps, but 5
could ultimately be produced, purified via reverse phase
chromatography, and stored in gram quantities. It is known
that 5 and related prodigiosins exist as two stable isomers
about the C8-C9 bond in solution, and the equilibrium
distribution is dependent on both solvent and pH. For
clarity as a DA substrate, we depicted a single isomer 5.8
At this point, we had completed the synthesis of the key
biosynthetic precursors 3-5 prescribed by Fenical and
were ready to attempt the proposed inverse-electron-
demand hetero-Diels-Alder reaction.
As shown in Scheme 8, the proposed inverse-electron-
demand hetero-Diels-Alder reaction of 5 to deliver 6 proved
unsuccessful. Over months of study and hundreds of reaction
(6) Chatterge, A. K.; Morgan, J. P.; Scholl, M.; Grubbs, R. H. J. Am.
Chem. Soc. 2000, 122, 3783–3786.
(7) Manuscript in preparation.
1050
Org. Lett., Vol. 12, No. 5, 2010

Scheme 8. Attempted Inverse-Electron-Demand
Hetero-Diels-Alder Reaction of 5 To Afford 6
conditions (heat, microwave, photochemical, Lewis acid
catalysis, mineral acid catalysis, solvent, and additives), we
were unable to effect the conversion of 5 to 6 in anything
other than presumed trace amounts detected only by LCMS.9
On the basis of these results, we enlisted molecular modeling
in an attempt to understand why the Diels-Alder reaction
failed. Molecular mechanics sampling for 5 was conducted
starting from the hypothesized transition geometry using both
stochastic and systematic conformer searches and gradient
energy minimization with the Merck MMFF94 forcefield as
implemented in the MOE software package (Chemical
Computing Group).10 Analysis of the top 10,000 conformers
with the lowest relative energies (20 kcal from lowest energy
conformer) indicated a failure to identify favorable Diels-
Alder transition state geometries. Out of 10,000 systematic
search conformers generated for 5, less than 15% of the
structures sampled have a folded topology, and the key atoms
remained separated by almost 5 Å (Figure 3). Of those
conformers with a “folded” topology, 15% form intramo-
lecular hydrogen bonds.8 Thus, as a result of intramolecular
hydrogen bonds and a large degree of conformational
flexibility present in the long, alkyl linker moeity, the
intramolecular Diels-Alder mechanism is likely energeti-
cally disfavored. Moreover, attempts at intermolecular vari-
ants9 proved equally unsuccessful, suggesting the C8-C9
olefin of the extended poly pyrrole π-system is not a
competent dienophile.
Figure 3. Mist favored Merck MMFF94 minimized conformer of 5.
In short order, we prepared the proposed biosynthetic
building blocks 3 and 4 and synthesized the key C1-C25
inverse-electron-demand hetero-Diels-Alder substrate 5 in
8 steps (5.1% overall yield). Hundreds of reaction conditions
were explored, but the proposed biomimetic intramolecular
inverse-electron-demand hetero-Diels-Alder reaction was
not successful. Modeling studies supported the inability of
5 to affect this transformation. Thus, a fundamentally new
synthetic strategy is now underway to synthesize marin-
eosines A and B. In the course of this work, we also
discovered a novel Grubbs II catalyzed 1,4-conjugate addition
reaction. While synthetically we could not validate the
biosynthetic proposal, an enzyme-templated process in nature
may still align 5 in a manner conducive for the Diels-Alder
reaction to occur.
Acknowledgment. The authors acknowledge the Depart-
ment of Pharamcology for support of this research. L.N.A.
acknowledges the Vanderbilt Institute of Chemical Biology
(VICB) for a predoctoral fellowship.
In summary, we evaluated the biosynthetic proposal put
forth by Fenical and co-workers for marineosins A and B.
Supporting Information Available: Experimental pro-
cedures, characterization data, and 1H and 13C NMR spectra
for all new compounds 4, 5, 8, 9, 17, 25-28. This material
is available free of charge via the Internet at http://pubs.acs.org.
(8) Furstner, A.; Grabowski, J.; Lehman, C. W. J. Org. Chem. 1999,
64, 8275–8280.
(9) See Supporting Information.
(10) Vainio, M. J.; Johnson, M. S. J. Chem. Inf. Model. 2007, 47, 2462–
2474.
OL100034P
Org. Lett., Vol. 12, No. 5, 2010
1051