Total synthesis of RNA-polymerase inhibitor ripostatin B and 15-deoxyripostatin A

Article abstract of DOI:10.1002/anie.201108749

Keep me skipped: A highly convergent total synthesis of ripostatin B, an inhibitor of the bacterial RNA polymerase, is described. The key steps to construct and avoid isomerization of the skipped triene are a double Stille cross-coupling reaction and a ring-closing metathesis. Furthermore, 15-deoxyripostatin A, a stable and conformationally locked analogue of ripostatin A (see scheme, 15-OH group red), was prepared and tested in vivo. Copyright

Full text of DOI:10.1002/anie.201108749

Angewandte
Chemie
DOI: 10.1002/anie.201108749
Ripostatin B (2)
Total Synthesis of RNA-Polymerase Inhibitor Ripostatin B and
15-Deoxyripostatin A**
Wufeng Tang and Evgeny V. Prusov*
Natural products isolated from marine and terrestrial micro-
organisms, plants, and fungi play a crucial role in drug
discovery and development in important areas, such as
antibiotics research and cancer treatment.[1] The emergence
of multi-drug-resistant strains over the past decades is
a problem that plagues public health.[2] Consequently, the
continuous search for new drug candidates that use a novel
mode of action is essential for effective infection control
measures.
The ripostatins A (1) and B (2; Scheme 1) are secondary
metabolites isolated in 1994 by Hçfle et al. from the
fermentation broth of the gliding bacteria strain Sorangium
cellulosum So ce 377.[3a] They are polyketide macrolides and
characterized by a 14-membered macrolactone with an
attached side chain. The synthetic challenge of these mole-
cules is derived through three separated double bonds that
easily isomerize into conjugation under acidic or basic
conditions. In particular at the stage of synthetic intermedi-
ates, this problem generates a significant hurdle. Moreover,
À
the C2 C3 Z-configured double bond is prone to isomer-
ization into conjugation with the acid functionality.
The initial biological evaluation of ripostatins revealed
only modest activity against Staphylococcus aureus and
Escherichia coli with minimal inhibitory concentration
(MIC) values in the range of 1 mgmLÀ1.[3b] At the same
time, the ripostatins were found to have a cytostatic effect on
L-929 mouse fibroblast cells.
However, more recently Ebright and co-workers reported
the crystal structure of the bacterial RNA polymerase with
myxopyronin (3); in that crystal structure the secondary
metabolite is bound to the “switch region” of the poly-
merase.[4] Further studies with mutant bacterial strains
provided strong evidence that corallopyronin A (4) and
ripostatin A (1) bind to the same pocket. The fact that the
amino acid sequence of the RNA polymerase is highly
conserved among bacterial species but differs substantially
from that of the mammalian enzyme makes the polymerase
an appealing target for the development of new antibiotics.
Consequently, we became attracted to the ripostatins as
a promising lead structure in our efforts to discover and
develop novel antibiotics. Our aim was to provide a general
synthetic access to the ripostatin framework and answer
fundamental questions regarding its synthesis and structure–
activity relationship by providing 15-deoxyripostatin A (27)
and ripostatin B (2) as probes to answer the question whether
the open-chain isomer or the cyclic hemiacetal are respon-
sible for the biological activity of the ripostatins. Although
preliminary studies towards the total synthesis of ripostatins
were published by Kirschning and co-workers,[5] the con-
struction of the macrocyclic lactone was proven to be rather
difficult. The main challenge in the synthesis of ripostatins is
to establish and maintain a so-called “skipped polyene” motif
(C2–C9) within the macrocyclic ring; this motif is notorious
for its lability and tendency to isomerize into the conjugated
dienoate.[6]
Scheme 1. Natural products that are inhibitors of the bacterial RNA
polymerase.
[*] W. Tang, Dr. E. V. Prusov
Department of Medicinal Chemistry, Helmholtz Centre for Infection
Research, Inhoffenstrasse 7, 38124 Braunschweig (Germany)
E-mail: evgeny.prusov@helmholtz-hzi.de
Homepage: http://www.helmholtz-hzi.de/en/research/
research_groups/pharmaceutical_research/medici-
nal_chemistry/
[**] We thank Dr. R. Jansen, Dr. H. Irschik, and K. Schober for providing
us with the NMR spectroscopy data of authentic ripostatin B. We
are grateful to Dr. G. Drꢀger for assistance with high-resolution
mass spectra; Dr. J. Fohrer and C. Kakoschke are thanked for NMR
spectroscopy measurements. We are also indebted to Dr. F. Sasse
and B. Hinkelmann for performing in vivo assays. We thank Dr. V.
Wray for help with the manuscript preparation. Financial support
from DFG (grant PR 1328/1-1 to E.V.P.) is greatly acknowledged.
We envisioned a retrosynthetic strategy (Scheme 2),
which we hoped would overcome those difficulties by
introducing the “skipped polyene” at a late stage in the
synthesis and by performing the ring closure immediately
afterwards. We argued that a double Stille cross-coupling
directly followed by a ring-closing metathesis might be the
key transformation for the successful construction of ripo-
statin B. The required precursor 5 can be derived by attach-
ment of two allylic groups to the double vinyl iodide 6, which,
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201108749.
Angew. Chem. Int. Ed. 2012, 51, 3401 –3404
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3401

.
Angewandte
Communications
selective reduction of triple bond with Red-Al followed by
iodine quench and two consecutive oxidations with manga-
nese oxide and sodium chlorite.[9]
With these building blocks in hand, the assembly of
ripostatin B (2) was attempted (Scheme 4). At first, fragments
9 and 10 were connected using a Paterson aldol reaction[10]
Scheme 2. Retrosynthetic analysis of ripostatin B. TBS=tert-butyldime-
thylsilyl, RCM=ring-closing metathesis.
in turn, can be obtained through esterification of the
iodoacrylic acid 8 with the corresponding alcohol fragment.
Our synthesis takes advantage of three equally complex
building blocks and is outlined in Scheme 3. In the synthetic
direction, (Æ)-epichlorohydrin was converted into the (S)-7-
(trimethylsilyl)hept-1-en-6-yn-4-ol 12 in four steps according
to the published procedure.[7] After simple protecting-group
manipulations, the terminal vinyl iodide moiety was installed
by carboalumination with subsequent iodine quench. Dihy-
droxylation of the double bond using the Sharpless protocol
and subsequent periodate cleavage furnished aldehyde 10.
The intermediate for the Paterson aldol reaction 9 was readily
synthesized from 3-butyn-1-iodide (15) using a one-pot
carboalumination/cross-coupling reaction[8] with benzyl bro-
mide followed by halogen–metal exchange and acylation with
the Weinreb amide of acetic acid. The third segment, the
carboxylic acid 8, was prepared from TBS-protected 3-butyn-
1-ol in five steps, including one carbon elongation, cis-
Scheme 4. Completion of the synthesis. (+)-DIP-Cl=(+)-B-chloro-
diisopinocampheylborane, DIBAL-H=diisobutylaluminum hydride,
TCBT=2,4,6-trichlorobenzoyl chloride, DMP=Dess–Martin periodi-
nane, DMAP=4-dimethylaminopyridine, PPTS=pyridinium p-toluene-
sulfonate, Grubbs II=(1,3-bis(2,4,6-trimethylphenyl)-2-imidazolidinyli-
dene)dichloro(phenylmethylene)(tricyclohexylphosphine)ruthenium,
py=pyridine.
that proceeds in a good yield and diastereoselectivity. No
attempts were made to improve either the yield or selectivity
since both starting materials were readily available. Then,
hydroxyketone 7 was subjected to an anti 1,3-reduction
according to the Evans–Hoveyda[11] protocol to give the
acetate-protected diol 18. Its absolute configuration was
confirmed to be 15-R by Mosher ester analysis.[12] After
careful examination of different condensation protocols, the
esterification under Yamaguchi[13] conditions was found to be
the method of choice for linking the iodoacrylic acid fragment
8 to the alcohol 19. In the next step, the simultaneous
introduction of two terminal allylic groups was performed by
the Stille cross-coupling[14] with excess of allyltributylstan-
nane. The reaction proceeded without any difficulties at
elevated temperatures in a mixture of benzene and DMF in
the presence of 5 mol% tetrakis(triphenylphosphin)palla-
dium as catalyst. At this stage we realized that attachment of
the allylic group to the acrylate fragment by cross-coupling
reaction after the esterification step is crucial for the success
of the synthesis; otherwise a rapid shift of the double-bond
position under esterification conditions would occur.
Scheme 3. Synthesis of the main building blocks. nBuLi=n-butyl-
lithium, tBuLi=tert-butyllithium, TBSOTf=tert-butyldimethylsilyl tri-
fluoromethane sulfonate, Red-Al=sodium bis(2-methoxyethoxy)-
aluminium hydride, Cp2ZrCl2 =bis(cyclopentadienyl) zirconium dichlor-
ide, Pinnick ox.=NaClO2, NaH2PO4, 2-methyl-2-butene, tert-BuOH/
H2O.
In the last stage of the synthesis, the macrolactone ring
was formed by exposure of a 1 mm solution of the diene 5 in
dichloromethane to the second-generation Grubbs catalyst at
room temperature, whereas reaction at reflux temperature
3402 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 3401 –3404

Angewandte
Chemie
produced a substantial number of side products. Gratifyingly,
only the trans-alkene product was isolated from the meta-
thesis reaction.[15] Finally, the exocyclic carboxylic acid
functionality was secured by selective cleavage of the primary
TBS ether followed by two consecutive oxidations with Dess–
Martin periodinane and sodium hypochlorite, respectively. To
protect the rather electron rich “skipped polyene” motif from
the electrophilic chlorine species during the Pinnick oxida-
tion, it was imperative to add dimethyl sulfoxide (DMSO) to
the reaction mixture as a co-solvent and scavenger;[16]
otherwise, a number of unidentified by-products were pro-
duced. Finally, the global deprotection was achieved by
treatment of the penultimate intermediate with the Olah
reagent at À258C to give ripostatin B in 43% yield.[17]
Scheme 6. Synthesis of the tetrahydropyrane core by Prins cyclization.
To gain a deeper understanding of the binding mode of the
ripostatins, we decided to clarify whether the hemiketal or
ketone form of ripostatin A (Scheme 5) is responsible for its
Scheme 7. Synthesis of the 15-deoxyripostatin A. CSA=10-camphorsul-
fonic acid, DIAD=diisopropyl azodicarboxylate.
According to our plan, the benzyl side chain was to be
introduced through regioselective elaboration of the left-
hand triple bond in the presence of the other trimethylsilyl-
protected one. Unfortunately, despite numerous attempts, we
were unable to perform carboalumination and cross-coupling
reactions in a one-pot manner as it was used earlier for 3-
butyn-1-iodide (15). We propose that this failure may be
attributed to coordination effects of the vinylaluminum
species with the oxygen atom of the tetrahydropyrane.
Nevertheless, conversion into vinyl iodide 23 followed by
Negishi cross-coupling with benzylzinc bromide[8a] provided
us with the desired product. Except for removal of the
trimethylsilyl group from the left-hand triple bond and
Mitsunobu esterification, the remaining steps in the synthesis
parallel those described for ripostatin B.
The biological validation identified 15-deoxyripostatin A
as inactive against several bacterial strains. Among several
possible explanations, we rationalize that the keto form is
necessary for binding to the RNA polymerase. Alternatively,
a functional group with strong hydrogen-bond donor–
acceptor properties must be present near or within the
tetrahydropyrane cycle.
Scheme 5. Keto–hemiketal equilibrium in ripostatin A, decomposition
pathway, and proposed stabilized 15-deoxy-analogue.
biological activity. It is also known that ripostatin A is not
stable under slightly basic conditions and decomposes into
inactive ripostatin C by b-acetoxy elimination from the
ketone form.[3a] Therefore, the 15-deoxyripostatin A was
proposed as a stable structural analogue, in which the
tetrahydropyran moiety would represent the locked keto–
hemiketal equilibrium.
The tetrahydropyrane core of the ripostatin A analogue
27 was easily accessed through a Prins cyclization[18] between
the previously described homoallylic alcohol 12 and 4-
pentynal (Scheme 6). Under optimized conditions, this trans-
formation produced a mixture of two separable diastereomers
in a 2.2:1 ratio and 54% combined yield. As expected, the all-
syn diastereomer 21a was the major product. The poor
diastereoselectivity at this step was not an issue, because both
diastereomers were taken further in the synthesis and
converted later into ester 25 (Scheme 7) employing an
appropriate esterification method, either using the Yamagu-
chi protocol or Mitsunobu[19] conditions with concomitant
inversion at carbon atom C13.
In summary, a short and efficient total synthesis of
ripostatin B is described in which the longest linear sequence
includes 18 steps and 0.22% overall yield. Because of its
convergent nature the synthesis route enables the rapid
assembly of further analogues to improve the pharmacolog-
Angew. Chem. Int. Ed. 2012, 51, 3401 –3404
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3403

.
Angewandte
Communications
4684; b) I. Paterson, M. J. Coster, D. Y.-K. Chen, K. R. Gibson,
D. J. Wallace, Org. Biomol. Chem. 2005, 3, 2410 – 2419; c) C. J.
Cowden, I. Paterson, Org. React. 1997, 51, 1 – 200.
ical properties of the ripostatins required to advance this
promising lead into medicinal applications. The key steps to
avoid double-bond isomerization of the skipped triene are
a double Stille cross-coupling reaction and a ring closing
metathesis. Furthermore, a stable and conformationally
locked analogue of ripostatin A was prepared and tested
in vivo. Further work to produce optimized analogues is
currently in progress and will be reported in due course.[20]
[11] D. A. Evans, A. H. Hoveyda, J. Am. Chem. Soc. 1990, 112, 6447 –
6449.
[12] J. A. Dale, D. L. Dull, H. S. Mosher, J. Org. Chem. 1969, 34,
2543 – 2549.
[13] J. Inanaga, K. Hirata, H. Saeki, T. Katsuki, M. Yamaguchi, Bull.
Chem. Soc. Jpn. 1979, 52, 1989 – 1993.
[14] a) M. Abarbria, J.-L. Parrainb, A. DuchÞne, Tetrahedron Lett.
1995, 36, 2469 – 2472; b) A. Rivkin, J. T. Njardarson, K. Biswas,
T.-C. Chou, S. J. Danishefsky, J. Org. Chem. 2002, 67, 7737 –
7740; c) See also references [5] and [6].
Received: December 12, 2011
Revised: January 17, 2012
Published online: February 29, 2012
[15] a) A. Fꢁrstner, Angew. Chem. 2000, 112, 3140 – 3172; Angew.
Chem. Int. Ed. 2000, 39, 3012 – 3043; b) K. C. Nicolaou, P. G.
Bulger, D. Sarlah, Angew. Chem. 2005, 117, 4564 – 4601; Angew.
Chem. Int. Ed. 2005, 44, 4490 – 4527; c) A. Gradillas, J. Perez-
Castells, Angew. Chem. 2006, 118, 6232 – 6247; Angew. Chem.
Int. Ed. 2006, 45, 6086 – 6101.
[16] E. Dalcanale, F. Montanari, J. Org. Chem. 1986, 51, 567 – 569.
[17] Analytical data obtained for the synthetic ripostatin B were in
good agreement with those reported in the reference [3a]. The
specific optical rotation was found to be of exactly the same sign
and value as for the natural product. The 1H NMR spectrum
exhibits an almost perfect match with the authentic one and most
of the 13C resonances are within 1 ppm of those reported. A
somewhat higher deviation of two signals in the vicinity of the
carboxylic group may be explained by salt formation during the
basic workup.
Keywords: inhibitors · natural products · ripostatins ·
synthesis design · total synthesis
.
[1] a) G. M. Cragg, D. J. Newman, K. M. Snader, J. Nat. Prod. 1997,
60, 52 – 60; b) J. W.-H. Li, J. C. Vederas, Science 2009, 325, 161 –
165.
[2] C. A. Arias, B. E. Murray, N. Engl. J. Med. 2009, 360, 439 – 443.
[3] a) H. Augustiniak, G. Hçfle, H. Irschik, H. Reichenbach, Liebigs
Ann. 1996, 1657 – 1663; b) H. Irschik, H. Augustiniak, K. Gerth,
G. Hçfle, H. Reichenbach, J. Antibiot. 1995, 48, 787 – 792.
[4] J. Mukhopadhyay, K. Das, S. Ismail, D. Koppstein, M. Jang, B.
Hudson, S. Sarafianos, S. Tuske, J. Patel, R. Jansen, H. Irschik, E.
Arnold, R. H. Ebright, Cell 2008, 135, 295 – 307.
[5] C. Kujat, M. Bock, A. Kirschning, Synlett 2006, 419 – 422.
[6] Studien zur Totalsynthese von Ripostatin A und B.; C. Kujat.
Dissertation, Wilhelm Leibniz Universitꢀt Hannover, 2007.
(http://deposit.ddb.de/cgi-bin/dokserv?idn =
[18] a) V. V. Vintonyak, B. Kunze, F. Sasse, M. E. Maier, Chem. Eur.
J. 2008, 14, 11132 – 11140; b) P. A. Wender, B. A. DeChristopher,
A. J. Schrier, J. Am. Chem. Soc. 2008, 130, 6658 – 6659; For
recent reviews on Prins reaction, see: c) L. E. Overman, L. D.
Pennington, J. Org. Chem. 2003, 68, 7143 – 7157; d) I. M. Pastor,
M. Yus, Curr. Org. Chem. 2007, 11, 925 – 957.
984056041&dok_var= d1&dok_ext = pdf&filename =
984056041.pdf).
[7] S. A. Burova, F. E. McDonald, J. Am. Chem. Soc. 2004, 126,
2495 – 2500.
[8] a) E. I. Negishi, H. Matsushita, N. Okukado, Tetrahedron Lett.
1981, 22, 2715 – 2718; b) B. H. Lipshutz, T. Butler, A. Lower, J.
Servesko, Org. Lett. 2007, 9, 3737 – 3740.
[19] a) D. W. Custar, T. P. Zabawa, J. Hines, C. M. Crews, K. A.
Scheidt, J. Am. Chem. Soc. 2009, 131, 12406 – 12414; b) S. F.
Martin, J. A. Dodge, Tetrahedron Lett. 1991, 32, 3017 – 3020.
[20] Contemporaneous with our efforts an independent total syn-
thesis of ripostatin B has been achieved: P. Winter, W. Hiller, M.
Christmann, Angew. Chem. 2012, 124, 3452 – 3456; Angew. Chem
Int. Ed. 2012, 51, 3396 – 3400.
[9] M. Kanematsu, M. Shindo, M. Yoshida, K. Shishido, Synthesis
2009, 2893 – 2904.
[10] a) I. Paterson, J. M. Goodman, M. A. Lister, R. C. Schumann,
C. K. McClure, R. D. Norcross, Tetrahedron 1990, 46, 4663 –
3404 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 3401 –3404