Total Synthesis of Ripostatin B and Structure-Activity Relationship Studies on Ripostatin Analogs

Article abstract of DOI:10.1021/acs.joc.8b00193

Described is the total synthesis of the myxobacterial natural product ripostatin B and of a small number of analogs. Ripostatin B is a polyketide-derived 14-membered macrolide that acts as an inhibitor of bacterial RNA-polymerase, but is mechanistically distinct from rifamycin-derived RNA-polymerase inhibitors that are in use for tuberculosis treatment. The macrolactone ring of ripostatin B features two stereocenters and a synthetically challenging doubly skipped triene motif, with one of the double bonds being in conjugation with the ester carbonyl. Appended to the macrolactone core are an extended hydroxy-bearing phenylalkyl side chain at C13 and a carboxymethyl group at C3. The triene motif was established with high efficiency by ring-closing olefin metathesis, which proceeded in almost 80% yield. The side chain-bearing stereocenter α to the ester oxygen was formed in a Paterson aldol reaction between a methyl ketone and a β-chiral β-hydroxy aldehyde with excellent syn selectivity (dr >10:1). The total synthesis provided a blueprint for the synthesis of analogs with modifications in the C3 and C13 side chains. The C3-modified analogs showed good antibacterial activity against efflux-deficient Escherichia coli but, as ripostatin B, were inactive against Mycobacterium tuberculosis, in spite of significant in vitro inhibition of M. tuberculosis RNA-polymerase.

Full text of DOI:10.1021/acs.joc.8b00193

Subscriber access provided by - Access paid by the | UCSB Libraries
Total Synthesis of Ripostatin B and Structure-
Activity Relationship Studies on Ripostatin Analogs
Florian Glaus, Darija Dedic, Priyanka Tare, Valakunja Nagaraja, Liliana Rodrigues, Jose A. Ainsa,
Jens Kunze, Gisbert Schneider, Ruben C. Hartkoorn, Stewart T. Cole, and Karl-Heinz Altmann
J. Org. Chem., Just Accepted Manuscript • Publication Date (Web): 15 Mar 2018
Downloaded from http://pubs.acs.org on March 15, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Total Synthesis of Ripostatin B and Structure-
Activity Relationship Studies on Ripostatin Analogs
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
‡
‡
†
Florian Glaus, Darija Dedić, Priyanka Tare,
†
&
ˡˡ
‡
Valakunja Nagaraja, Liliana Rodrigues, José Antonio Aínsa, Jens Kunze,
‡
§
§
‡,
Gisbert Schneider, Ruben C. Hartkoorn, Stewart T. Cole, and Karl-Heinz Altmann, *
ORCID-ID (KHA): https://orcid.org/0000-0002-0747-9734
‡ETH Zürich, Department of Chemistry and Applied Biosciences, Institute of
Pharmaceutical Sciences, Zürich, Switzerland.
karl-heinz.altmann@pharma.ethz.ch
§École Polytechnique Fédérale de Lausanne, Global Health Institute, Lausanne,
Switzerland
†Department of Microbiology and Cell Biology, Indian Institute of Science,
Bangalore 560 012, India
ˡˡGrupo de Genética de Micobacterias, Departamento de Microbiología, Medicina
Preventiva y Salud Pública, Facultad de Medicina and BIFI, Universidad de Zaragoza,
Zaragoza, Spain and CIBER Enfermedades Respiratorias (CIBERES), Madrid, Spain
&Fundación Agencia Aragonesa para la Investigación y el Desarrollo (ARAID),
Zaragoza, Spain
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract
Described is the total synthesis of the myxobacterial natural product ripostatin B and of a
small number of analogs. Ripostatin B is a polyketide-derived 14-membered macrolide that
acts as an inhibitor of bacterial RNA-polymerase, but is mechanistically distinct from
rifamycin-derived RNA-polymerase inhibitors that are in use for tuberculosis treatment. The
macrolactone ring of ripostatin B features two stereocenters and a synthetically challenging
doubly skipped triene motif, with one of the double bonds being in conjugation with the
ester carbonyl. Appended to the macrolactone core are an extended hydroxy-bearing
phenylalkyl side chain at C13 and a carboxymethyl group at C3. The triene motif was
established with high efficiency by ring-closing olefin metathesis, which proceeded in almost
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
8
0% yield. The side chain-bearing stereocenter  to the ester oxygen was formed in a
Paterson aldol reaction between a methyl ketone and a -chiral -hydroxy aldehyde with
excellent syn selectivity (dr >10:1). The total synthesis provided a blueprint for the synthesis
of analogs with modifications in the C3 and C13 side chains. The C3-modified analogs
showed good antibacterial activity against efflux-deficient Escherichia coli but, as ripostatin
B, were inactive against Mycobacterium tuberculosis, in spite of significant in vitro inhibition
of M. tuberculosis RNA-polymerase.
ACS Paragon Plus Environment

Page 3 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Introduction
In 1995, Höfle, Reichenbach and co-workers described the isolation of two antibacterial
metabolites from the culture supernatant of Sorangium cellulosum (strain So ce377;
1
,2
obtained from a Kenyan soil sample in 1989). The two compounds were named ripostatin
A and B, respectively, which reflects their biological mechanism of action, the inhibition of
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ribonucleic acid polymerase (RNAP; vide infra). The isolation procedure delivered 30 mg of
ripostatin per liter of extract, with ripostatin A being the major constituent (A/B ca. 25:1).1
Ripostatins A (1) and B (2) are 14-membered polyketide-derived macrolides incorporating
an unusual 1,4,7-triene-framework, an exocyclic carboxylic acid functionality and an apolar
side chain terminating in a phenyl ring. The two compounds show a high degree of structural
homology, differing only in the oxidation state at C15.
Ripostatin A (1) exists as an equilibrium mixture of a hemiacetal and a keto form (Scheme
1
);2 the latter can undergo elimination to the biologically inactive ripostatin C (3), a process
that is triggered even by mildly basic conditions (36% elimination in methanolic buffer
solution pH 8 at 40 °C within 1 hour). In contrast, ripostatin B (2) is significantly more stable,
with saponification of the lactone functionality being observed only above pH 11.2
Ripostatins A (1) and B (2) were reported by the isolation group to be narrow spectrum
antibiotics, with notable activity (minimum inhibitory concentrations (MICs) between 0.63
and 2.5 g/mL) being observed only against Staphylococcus aureus and an Escherichia coli
1
"
mutant with altered outer membrane" (E. coli Tol C ), although the exact mutation in the
tolC gene was not specified. In addition, ripostatin B (2), but not ripostatin A (1), also showed
1
minor activity against yeasts and fungi. Based on the differential activity of ripostatins
against wild-type E. coli and a corresponding Tol C mutant the isolation group suggested that
the lack of broad-based antibacterial activity of ripostatins was related to poor penetration
through the bacterial cell wall rather than degradation of the compounds.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 1. Keto/hemiacetal equilibrium in ripostatin A (1) and elimination to ripostatin C
(
3).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Both ripostatin A (1) and B (2) appear to exhibit low toxicity towards mouse fibroblasts,
although it is not entirely clear how the numerical data provided in ref. 1 are to be
interpreted.
At the mechanistic level, ripostatin A (1) was demonstrated by the isolation group to be a
1
potent inhibitor of RNAP from E. coli with an IC50 of 0.1 M. At the same time, (eukaryotic)
wheat germ RNA polymerase II was insensitive to the compound; the RNAP-inhibitory
activity of ripostatin B (2) was not investigated in the context of the isolation work.
Intriguingly, ripostatin-resistant S. aureus mutants showed no cross-resistance with
rifampicin and vice versa, thus suggesting for the binding sites of the two compounds on
RNAP to be distinctly different.
The lack of cross-resistance between ripostatin A (1) and rifampicin in S. aureus was
3
subsequently confirmed for a whole series of mutant strains. More recently, both
ripostatins have also been demonstrated to inhibit the growth of the hypersensitive E. coli
4
strain D21f2tolC::Tn10 with MICs below 1 g/mL, thus confirming the earlier findings of the
isolation group with the unspecified E. coli Tol C mutant strain. The E. coli strain
D21f2tolC::Tn10 harbors two primary defects in the outer membrane, a “deep rough” (rfa)
5
mutation that causes loss of heptose from the outer-membrane lipopolysaccharides and a
6
Tn10 insertion mutation in the tolC gene, which disrupts the Tol C protein that serves as an
outer membrane channel for several efflux pumps. In combination, these defects lead to
greatly enhanced susceptibility of the bacteria towards hydrophobic, but not hydrophilic,
5
antibacterials. The activity of ripostatins against E. coli D21f2tolC::Tn10 re-enforces the
hypothesis that the intrinsic activity of the compounds against bacterial RNAP would be
ACS Paragon Plus Environment

Page 5 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
sufficient to elicit potent antibacterial effects, which, however, do not become manifest in
most cases due to a lack of efficient intracellular accumulation.
In 2008, Ebright and co-coworkers discovered that ripostatin A (1) showed a high degree of
7
8,9
cross-resistance with the myxopyronin /corallopyronin class of natural RNAP inhibitors,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
thus indicating that these antibiotics all shared the same binding site on RNAP, or at least
bound to largely overlapping sites.10
Analysis of the myxopyronin-resistance substitution cluster suggested this binding site to
be located in the RNAP switch region, which was subsequently confirmed by X-ray
crystallography of myxopyronin A in complex with RNAP from Thermus thermophilus.10
Binding of myxopyronin to the “switch region” of RNAP stabilizes a closed conformation of
the enzyme, which prevents access of the dsDNA substrate to the active site cleft, and the
same is likely to be true for ripostatin A (1). Importantly, this inhibition mode is
fundamentally different from that of rifamycin-type inhibitors;10,11 as a result, switch region
binders exhibit very limited cross-resistance with rifampicin.3,10
While virtually no work on ripostatins as targets for total synthesis had been reported in
the literature for more than 15 years after their discovery, with the notable exception of a
1
2
paper by Kirschning in 2006, three total syntheses of ripostatin B (2) were communicated
1
3
14
simultaneously in 2012 by Christmann and co-workers, Prusov and co-workers and our
own group.15,16 While all these syntheses were based on ring-closure by ring-closing olefin
metathesis (RCM), different approaches were pursued to prepare the requisite diene
13
precursor, especially in the case of the Christmann synthesis. More recently, the Prusov
1
7
group has also reported the total synthesis of ripostatin A (1) followed by a full paper on
the synthesis of both ripostatins and a limited number of analogs for structure-activity
relationship (SAR) studies.4
Our own interest in ripostatins as targets for total synthesis arose in the context of a
broader program directed at the discovery of new natural product-based antimicrobials in
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
general, and of antimycobacterial agents in particular. As demonstrated by the clinical
importance of several rifamycin derivatives for tuberculosis (TB) treatment, RNAP is a highly
validated target for TB drugs. At the same time, it appears that the potential of bacterial
RNAP as a drug target for TB is still underexplored, given the fact that there is only one
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
8
structural class of RNAP inhibitors in clinical use. While natural ripostatins are not suitable
drug candidates per se, primarily due to their inability to cross the bacterial cell wall
efficiently, this issue might be addressable by structural modifications. It was also not clear a
priori if the lack of antibacterial activity of ripostatins would necessarily extend to M.
19
tuberculosis (Mtb), given the unique nature of the mycobacterial cell wall. In order to
provide insight into the effects of ripostatins on the growth of Mtb and to establish the
chemical basis for SAR studies, we embarked on the stereoselective total synthesis of
ripostatin B (2) as the (chemically) more stable of the two ripostatin variants. At the time
when this work was initiated, it could only be inferred that ripostatin B (2), like ripostatin A
(1), was an inhibitor of the bacterial RNAP, as no experimental data on this question were
available in the literature. In the meantime, work by Prusov and Ebright has shown that 1
and 2 inhibit RNAP from E. coli with similar potency, while 2 is ca. 10-fold more potent than
1
against the enzyme from S. aureus.4
In this report, we present full details on the total synthesis of ripostatin B (2), the essential
elements of which we have already reported in previous communications.15,16 In addition,
we describe the synthesis of a series of new ripostatin analogs and their (partial) assessment
as inhibitors of RNAP from Mtb. While some of the compounds were found to inhibit this
enzyme with sub-M IC50s, none of them (including the natural product 2) showed any
measurable antibacterial activity against the pathogen. Like natural ripostatins, several
compounds showed significant activity against an efflux-impaired strain of E. coli, which has
allowed to deduce an SAR for their intrinsic antibacterial potency.
Results and Discussion
One of the key structural characteristics of the ripostatins is the doubly skipped triene
motif from C2 to C9, which we felt should be established only late in the synthesis under
mild conditions, in order to minimize the risk for migration of the C5–C6 double bond into
conjugation with the enoate system. A particularly suitable approach towards the
construction of this critical structural motif thus appeared to be macrocyclic ring-closure by
ACS Paragon Plus Environment

Page 7 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
ring-closing olefin metathesis (RCM), which is conducted under very mild conditions in the
2
0
absence of acids or bases. More specifically, the C5–C6 double bond was selected for
disconnection, as this led to a precursor (4) with two unencumbered double bonds (Scheme
2
).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 2. Retrosynthesis of ripostatin B (2).
The RCM substrate 4 was envisioned to be accessed from vinyl iodide 5 by Stille coupling
with allyltributylstannane. Vinyl iodide 5 can be further disconnected into acid 6 and the
partially protected triol 7. The latter was planned to be accessed through a stereoselective
Paterson aldol reaction between methyl ketone 8 and protected β-hydroxy aldehyde 9,
followed by stereoselective anti 1,3-reduction of the resulting aldol product and protecting
group manipulations. In contrast to most other methods, the Paterson aldol reaction is
known to furnish reasonable levels of stereoselectivity also for reactions involving methyl
ketones21,22 and, thus, was considered the most promising approach for our system. It
should also be noted that the particular order of steps projected for the assembly of 4 from
7
was chosen based on Kirschning’s finding that dienoic acids derived from 6 were
12
incompatible with various esterification protocols. Finally, aldehyde 9 was projected to be
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
derived from chiral epoxide 10 by epoxide ring opening with an appropriate carbon
nucleophile and subsequent elaboration of the diene moiety. Epoxide 10 was literature
known and could be obtained in 4 straightforward steps from D-aspartic acid as the ultimate
source of the C11 stereocenter.23
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
For building block 9, the specific implementation of the strategy outlined above involved
treatment of epoxide 10 with the anion of ethyl propiolate in the presence of BF
3
·OEt ,
2
followed by reaction of the ensuing secondary alcohol with TBSCl, to furnish TBS-ether 11 in
84% overall yield (Scheme 3).
Scheme 3. Synthesis of aldehyde 9.
The subsequent stereoselective transformation of the triple bond in 11 into the required
trisubstituted E-configured double bond was based on methodology that was originally
developed by Kobayashi and Mukaiyama24,25 and more recently has been applied to the
ACS Paragon Plus Environment

Page 9 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
26
synthesis of a closely related variant of enoate 13 by Takamura and co-workers. Thus, the
treatment of 11 with thiophenol/NaOMe afforded a ca. 8:1 mixture of Z and E thioenol
ethers, from which the desired Z isomer 12 could be isolated in yields of around 85% on
small scale. On multigram scale, the isomers could not be separated in a single
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
chromatographic run, but separation was performed at the stage of ester 13 or the
corresponding alcohol. Reaction of 12 with dimethylcuprate24,25 then provided enoate 13 in
≥
95% yield as a single isomer. The reaction appeared to be very sensitive to the exact
experimental conditions, such that the desired product was only obtained if the reaction
mixture, after addition of 12 to the suspension of dimethylcuprate at −78 °C, was allowed to
warm to −30 °C in the cooling bath over a period of 30 - 40 min. No conversion was observed
upon slow warming to room temperature or if 12 was added at −40 °C or 0 °C.
After reduction of ester 13 to the corresponding primary alcohol, the latter was treated
27
with CBr
4
and PPh
3
in the presence of 2,6-lutidine to effect formation of allylic bromide 14
(Scheme 3). The presence of 2,6-lutidine in this Appel-type reaction was crucial, in order to
avoid decomposition (presumably caused by HBr). The 1,4-diene unit was then installed by
2
8
Stille cross coupling of 14 with tributylvinylstannane and [Pd
2
(dba)
3
]/AsPh in DMF, which
3
gave diene 15 in 79% yield. It should be noted that washing with brine was normally avoided
during work-up of the bromination reaction, as this led to partial substitution of chlorine for
bromine; for all practical purposes, however, the halogen exchange was inconsequential, as
the use of mixtures of 14 and the corresponding allylic chloride afforded similar yields in the
subsequent Stille reaction as pure 14.
The final elaboration of 15 into aldehyde 9 then proved to be unexpectedly challenging. An
initial attempt to remove the PMB-group by standard oxidative cleavage with DDQ only led
to decomposition. This finding was not entirely surprising, as difficulties with the DDQ-
mediated cleavage of a PMB-ether in the presence of a 1,4-diene moiety have been reported
2
9
previously; however, a range of other methods investigated also failed to deliver the
2
9
desired primary alcohol. Thus, BF
3
·OEt
2
/thiophenol led to selective TBS deprotection,
0,31
32
33
BCl
3
·SMe 3
2
and SnCl
4
/thiophenol both induced decomposition, and AlMe
3
left the
·OEt
, which has been reported by Iwasaki and co-workers as a mild reagent for PMB-
molecule unchanged even at elevated temperature (40 °C). A combination of MgBr
and SMe
2
2
2
34
removal afforded 15 in scale-dependent yields, i. e., 74% on a 13 mg scale, but only 50%
(25% recovered starting material) on an 80 mg scale; thus, this method was also considered
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
unsuitable. Finally, we discovered that the PMB-group was most efficiently removed by
3
5
treatment of 9 with TMSI followed by cleavage of the ensuing TMS-ether with K
MeOH (Scheme 3) which afforded primary alcohol 16 in 92% yield. The reaction likely
proceeds via activation of the benzylic oxygen by TMS addition, followed by S 2-type attack
2
CO
3
in
N
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
of iodide (Scheme 3). Oxidation of 16 under Swern conditions then completed the synthesis
of building block 9.
The synthesis of ketone 8, comprising the C14–C24 unit of ripostatin B (2), departed from
phenylacetaldehyde (17) (Scheme 4). The latter was elaborated into unsaturated -
unsaturated ester 19 in a known two-step sequence that involved addition of
2
-propenylmagnesium bromide (18) and subsequent reaction of the resulting secondary
3
6
alcohol with triethyl orthoacetate and Johnson-Claisen rearrangement. Ester 19 was
obtained in 49% overall yield.
Scheme 4. Synthesis of methyl ketone 8.
37
Following a procedure developed by Williams and co-workers, 19 was then directly
transformed into methyl ketone 8 in 78% yield with MeNH(OMe) and an excess of MeMgBr.
A comparable yield was obtained if the intermediate Weinreb amide was isolated and
reacted with MeMgBr in a separate step.
With methyl ketone 8 and aldehyde 9 in hand, we began our investigations into the
feasibility of the critical Paterson aldol reaction.21,22 As alluded to above, we considered this
method to be best suited for our system, although it was not clear initially how the -
substituent on the aldehyde would affect the stereochemical outcome of the reaction. In the
2
2
event, the aldol reaction between the (+)-Ipc
2
BCl-derived Ipc B-enolate of 8 and aldehyde
2
ACS Paragon Plus Environment

Page 11 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
9
(Scheme 5) proceeded with high diastereoselectivity (>10:1 dr), to provide the desired
aldol product 20 in 62% yield with a dr of 97:3 after chromatographic purification.
Scheme 5. Building block assembly. Aldol reaction between ketone 8 and aldehyde 9.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The configuration of the newly formed stereocenter in 20 was verified explicitly at the
stage of partly protected triol 7 (vide infra); direct Mosher analysis38,39 of 20 was not feasible
due to a pronounced tendency of the Mosher esters to undergo β-elimination.
As Paterson aldol additions with methyl ketones, in the absence of additional elements of
stereocontrol, typically proceed with selectivities of 50%-80% ee,21,22 the high selectivity
observed for the reaction between 9 and 8 with (+)-Ipc Cl as the Lewis acid clearly points to
2
synergistic stereoinduction by the β-substituent of the aldehyde. This assumption is in line
with results obtained by Prusov and Tang as part of their work on the synthesis of ripostatin
17
BCl and aldehyde 224,17 to be completely
A (1), where they found the reaction of 8, (-)-Ipc
2
non-selective.
In contrast, and in agreement with our own observations, formation of the enolborinate of
with (+)-Ipc BCl and subsequent reaction with 22 gave the corresponding aldol product
8
2
17
with high diastereoselectivity (11-15:1 dr), thus clearly indicating the matched nature of
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the latter combination of reactants. 1,3-Syn induction has previously been reported for
additions of methyl ketone-derived enolborinates to aldehydes bearing a silyloxy group at a
β-stereocenter.40,41 Notably, this phenomenon cannot be accounted for by the Evans 1,3-
induction model, which predicts for a β-stereocenter of an aldehyde to favor formation of
the 1,3-anti addition product.42,43 Only recently, Yamamoto and co-workers have proposed a
transition state model that explains the 1,3-syn induction in the addition of methyl ketone-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
44
derived enolborinates to chiral β-supersilyloxy aldehydes (Scheme 6): DFT calculations for
the addition of the enolborinate derived from acetone and 9-BBNOTf to aldehyde M1
revealed a transition state such as A to be preferred, in spite of the C=O group being located
gauche to both the methyl and the silyloxy group. Transition state B, in which gauche
interactions are minimized, is about 1.5 kcal/mol higher in energy than A; this was suggested
.44
to be the result of steric interactions between the silyloxy moiety and the enolate CH
2
Scheme 6. Proposed transition state models for stereoinduction in the addition of the
enolborinate derived from acetone and 9-BBN to β-silyloxy aldehyde M1.44
In analogy, the silyloxy group was proposed to interfere with the C=O group in the case of
transition state C (Scheme 6), which was found to be the lowest energy transition state
4
4
leading to anti-products (about 1 kcal/mol higher in energy than A). Unfortunately, no
calculations have been performed for aldehydes carrying sterically less demanding -silyloxy
groups (such as OTES, OTBS, OTIPS, OTBDPS). It appears likely, however, that the above
considerations could also account for the 1,3-syn induction observed in the reaction of 9 and
2
2, respectively, with the enolborinate of 8 obtained with (+)-Ipc Cl.
2
ACS Paragon Plus Environment

Page 13 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
In order to establish the stereocenter at C15 (ripostatin numbering), aldol product 20 was
4
5
submitted to an Evans-Tishchenko reduction (Scheme 5). The reaction proceeded smoothly
and furnished the desired hydroxy ester 21 in excellent yield (93%) and with high selectivity
(
>20:1 dr). Hydroxy ester 21 was then elaborated into bis-TBS-protected triol 7 by reaction
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
with TBSCl and subsequent reduction of the ester moiety with DIBAL-H (Scheme 5); 7 was
obtained in 82% yield from 21.
As alluded to above, the configuration of the C13 stereocenter was verified by Mosher
ester analysis of 7 (Scheme 7).38,39 Furthermore, removal of the ester moiety in 21 followed
by ketalization with 2,2-dimethoxypropane furnished a ketal (25) suitable for stereochemical
analysis according to Rychnovsky, which also allowed unambiguous assignment of the
configuration of the stereocenter at C15 as S46,47 (Scheme 7).
Scheme 7. Configurational assignment of the C13 and C15 stereocenters.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
48
With 7 in hand, we turned our attention to its esterification with acid 6, which proved to
be far more difficult than anticipated. A series of esterification methods were investigated,
4
9
50
51
52
including the use of Mukaiyama’s salt, DCC (without and with DMAP), PyBOP or TBTU
as coupling agents, but none of these methods provided the desired ester in notable
quantities. Appreciable amounts of ester 5 (Scheme 8) (< 30% yield) were obtained only
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
upon preactivation of 6 with 2,4,6-trichlorobenzoyl chloride (TCBC) according to Yamaguchi
53
and co-workers und subsequent reaction of the mixed anhydride with 7. Closer
investigation of the reaction conditions revealed that immediate mixing of all components
(6, 7, TCBC, Et N, and DMAP), rather than pre-activating the carboxylic acid, led to a slight
3
increase in yield to 35% (with ca. 20% of 7 recovered). Interestingly, the reaction proved to
be very fast, reaching maximum conversion in less than 20 min at room temperature. This
observation led us to speculate that the incomplete conversion was caused by
decomposition of the anhydride intermediate and we started to investigate lower reaction
temperatures.
Scheme 8. Completion of building block assembly and macrocyclization.
ACS Paragon Plus Environment

Page 15 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
These experiments showed that mixing all components at −78 °C and then allowing the
reaction mixture to warm to −20 °C, gave 5 in isolated yields of around 60% (with 20-25% of
7 recovered). Finally, we discovered that the temperature ideally had to be kept between
−
40 and −50 °C (after mixing the reactants at −78 °C). In this temperature regime, maximum
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
conversion was reached within 2-3 h and the desired ester 5 could be isolated in 80% yield
(Scheme 8); in addition, ca. 8% of alcohol 7 were also recovered.
Having established a reliable and high-yielding esterification protocol, the second 1,4-diene
unit in 4 was installed in 86% yield by Stille coupling of 5 and tributylallylstannane (Scheme
8). The stage was thus set for the crucial ring-closure reaction. The efficiency of the RCM was
initially assessed in a series of small-scale experiments (ca. 2 mg of 4) with either Grubbs-II
2
0
or Hoveyda-Grubbs-II catalyst, where conversion was monitored by low-resolution ESI-MS;
after concentration, the reaction products were further analyzed by TLC and, in some cases,
1
also by H-NMR spectroscopy. With both catalysts, rapid consumption of starting material
was observed, regardless of solvent (DCM, 1,2-dichloroethane or toluene) and temperature
(room temperature to 80 °C). Unfortunately, three close spots were usually observed on TLC
(after concentration).
In light of the rapid reaction rates with the ruthenium-based second-generation catalysts,
it appeared that the less reactive first-generation catalysts might still show practically useful
conversion rates. In the event, the Grubbs-I catalyst gave reaction rates comparable to the
ones observed for second-generation catalysts. At the same time, the reaction was cleaner,
allowing isolation of macrocycle 26 in 77% yield as a single isomer (Scheme 8). The E
1
configuration of the newly formed C5–C6 double bond was confirmed by H-NMR
spectroscopy (J = 14.8 Hz). Best results were obtained if the reaction was not allowed to
reach completion, which could be achieved by using less than 15 mol-% of the catalyst.
Otherwise, formation of side products was observed. Interestingly, and in contrast to the
Grubbs-I catalyst, the Hoveyda-Grubbs-I catalyst left the starting material unaffected.
Noteworthy, we noticed decomposition of initial batches of macrocycle 26 upon storage
even at low temperature (about 50% decomposition within 72 hours at −20 °C), which we
presumed was caused by ruthenium-derived by-products. Following a procedure developed
54
by Georg and co-workers for the removal of such by-products, the RCM reaction was thus
quenched by the addition of DMSO (50 equivalents relative to the catalyst) and stirring was
continued for at least 12 hours, then the reaction mixture was concentrated and purified by
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
silica gel chromatography. If this procedure was applied, 26 could be stored for 48 hours at
−
20 °C without loss in purity. Nonetheless, as a rule, 26 was usually directly subjected to TBS
removal without storage.
Selective cleavage of the primary TBS-ether was best achieved with buffered HF·pyr
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Scheme 8), which was superior to TASF (decomposition), AcOH/THF/H
2
O 1:2:1 (slow
reaction) or NaIO
4
,55 and delivered the desired primary alcohol 27 in 81% yield. While NaIO
4
allowed deprotection of the primary TBS-ether with good selectivity on a scale smaller than
10 mg of 26 (ca. 70% yield), extended reaction times were required even on slightly larger
scale, which were also associated with over-deprotection (i. e. 50% yield on a 25 mg scale).
The final conversion of 27 into ripostatin B (2) in a first step required oxidation of the free
hydroxy group to the corresponding carboxyl functionality (Scheme 9). In order to
circumvent handling of the intermediate aldehyde, whose carbonyl group is part of a
vinylogous malonate-type system and which we thus presumed to be highly labile, we
initially focused on methods that should directly deliver acid 28. Unfortunately, however,
5
6
57
TEMPO/PhIOAc
material.
2
, TPAP/NMO·H O as well as PDC all led to decomposition of the starting
2
Scheme 9. Side chain oxidation and final deprotection.
In light of the failure of direct oxidation methods, we decided to investigate the feasibility
of two-step protocols. These experiments revealed that the Dess-Martin periodinane
5
8
(DMP) afforded very clean oxidation of 27 to the intermediate aldehyde within 5 min, as
demonstrated by conducting the reaction in an NMR tube (in THF-d ). After aqueous work-
8
ACS Paragon Plus Environment

Page 17 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
up (saturated aqueous NaHCO
3
/saturated aqueous Na
2
S
2
O ), however, numerous side
3
products were visible on TLC and subsequent Pinnick oxidation59,60 of this material furnished
28 in very low yields only. This finding confirmed our original suspicions about the labile
nature of the aldehyde derived from 27 and led us to perform the final oxidation of the
aldehyde without any work-up of the DMP oxidation by simply adding the necessary Pinnick
reagents to the reaction mixture. Gratifyingly, conducting the reactions in this fashion
enabled isolation of acid 28 in 81% yield (Scheme 9). Final cleavage of the secondary TBS-
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ethers was best accomplished with 5% HF in aqueous CH
88% yield (Scheme 9). Other protocols, such as the use of HF·pyr, AcOH/THF/H
NaIO 55
gave distinctly less clean reactions.
3
CN, furnishing ripostatin B (2) in
2
O 3:1:1, or
4
A comparison of the NMR data for synthetic 2 with the published data for the natural
1
product revealed certain deviations in a number of signals in both the H- (Δ > 0.1 ppm for
13
three protons) and the C-NMR (Δ > 0.5 ppm for five carbons) spectra, especially for the
1
C11–C15 region. However, our H-NMR spectrum fully matched with the one that has been
1
4
obtained for natural 2 by the group of Prusov, as well as with the ones for synthetic 2 from
1
4
13
both Prusov and Christmann. Furthermore, HPLC chromatograms of our synthetic 2 and
natural ripostatin B were superimposable (experiments carried out in the group of Prof. Rolf
Müller at the Helmholtz Institute for Pharmaceutical Research Saarbrücken, Germany, data
not shown).
As alluded to in the Introduction, so far, specific SAR data for ripostatins are available only
from the recent work of Prusov, Ebright and co-workers on a limited number of synthetic
1
7
analogs. While a few semisynthetic ripostatin derivatives were reported in the context of
2
the original isolation work, no numerical data were provided for these compounds; rather,
the methyl ester of 2, a C15-2-phenylbutyryloxy derivative of 2, and the methyl ketal form of
1
were simply classified as "inactive", while the C15-epimer of 2 was stated to be equally
active as natural 2, without specifying which type of activity these conclusions related to.
The investigations conducted by Prusov and Ebright suggest that the active form of
ripostatin A (1) is the bicyclic hemiketal, based on the assessment of RNAP inhibition (for E.
coli as well as S. aureus RNAP) and antibacterial activity against E. coli D21f2tolC and S.
aureus ATCC 12600. Prusov and Ebright also prepared the C14-difluoro analogs of 2 and 15-
epi-2, respectively, and found those two compounds to exhibit markedly different biological
activity, with C14-difluoro-2 being ca. 20-fold more potent than its C15-epimer. This finding
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
stands in contrast to the statement in the isolation paper about the similar activity of 2 and
C15-epi-2. Finally, Prusov and Ebright found both natural ripostatins to be inactive against
RNAP from Mtb (vide infra).
Our own work on ripostatin analogs has focused (a) on the replacement of the carboxylic
acid group by other polar functionalities, as we speculated that the charge associated with
this group at neutral pH might hamper cell wall and/or membrane penetration, an effect
that has already been noted decades ago for rifamycin B;61,62 and (b) on evaluating
modifications of the C13-side chain that would reduce the overall lipophilicity of the
corresponding analogs (relative to ripostatin B (2)), again hoping that this would improve cell
wall penetration.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The first analog prepared in the context of these SAR evaluations was triol 29 (Scheme 10),
as this was easily accessible from the bis-TBS-ether 27. Thus, treatment of 27 with 5% HF in
aqueous acetonitrile furnished analog 29 in 67% yield after HPLC purification (Scheme 10).
Scheme 10. Synthesis of C3-side chain-modified analogs.
ACS Paragon Plus Environment

Page 19 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
While not investigated, 29 should also have been accessible by global deprotection of tris-
TBS-ether 26, an approach that was later applied in the preparation of related analogs (vide
infra).
Advanced intermediate 27 was also elaborated into analog 32, bearing a C3-side chain
terminating in a methyl carbamate moiety (Scheme 10). This analog had emerged as a
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
plausible and attractive target from preliminary docking studies, based on the X-ray crystal
10
structure of RNAP from T. thermophilus with the bound natural product myxopyronin7
using GOLD.63 The docking suggested a binding mode of 2 where the carboxy group of the
C3-side chain would overlap with the methylcarbamate moiety in bound myxopyronin,
hence leading to the design of analog 32.
Intermediate 27 was first converted into primary amine 31 by Mitsunobu reaction with
phthalimide followed by treatment of the ensuing N-substituted phthalimide 30 with
ethanolic hydrazine (Scheme 10). Treatment of crude 31 (ca. 80% pure) with methyl
chloroformate in the presence of Et N followed by removal of the TBS-protecting groups
3
under the previously established conditions then furnished analog 32 in 21% overall yield
from 30 after HPLC purification.
As will be discussed below, analog 29 was found to be biologically more active than
ripostatin B (2), which also led us to investigate homologated analog 38 with an extended
C3-side chain, in order to assess the effect of side chain length on biological activity (Scheme
11). The synthesis of 38 followed the same overall route that we had established for the
synthesis of ripostatin B (2), with homologated acid 34 substituting for 6 in the synthesis of
the natural product. Acid 34 was obtained in 5 steps and 41% overall yield from 1-pentyn-5-
ol via known propargylic alcohol 33.64
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 11. Synthesis of C3-side chain-extended analog 38.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Our strategy towards ripostatin B analogs with reduced lipophilicity entailed truncation of
the phenyl-bearing C13-side chain to a hydroxy-substituted pentyl or isoheptyl substituent,
respectively (analogs 45a and 45b; Scheme 12). In light of the sustained biological activity of
2
9, analogs incorporating a hydroxyethyl group at C3, rather than the natural side chain
were targeted for an initial evaluation.
The synthesis of analogs 45a and 45b was conceptually analogous to the total synthesis of
ripostatin B (2), which served as an efficient blueprint for analog generation. Thus, the
addition of the enolborinates derived from commercially available ketones 39a or 39b by
reaction with (+)-Ipc
in 66% and 70% yield, respectively, and with high diastereoselectivities (>10:1 dr) (Scheme
2). Of note, in both cases the excess ketone could be readily removed by rotary
2
BCl to aldehyde 9 (Scheme 3) afforded β-hydroxy ketones 40a and 40b
1
evaporation, allowing more straightforward purification than for hydroxy ketone 20 (leading
to the natural product).
ACS Paragon Plus Environment

Page 21 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Scheme 12. Synthesis of ripostatin analogs with truncated C13-side chains.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The elaboration of 40a/b into 42a/b was uneventful for both series and so was the low-
temperature Yamaguchi esterification, which afforded high yields of both 43a and 43b (83%
and 76%, respectively). After Stille cross coupling, the RCM precursors 44a/b were treated
with Grubbs-I catalyst to effect macrocyclization. As discussed above, the reactions were not
allowed to go to completion and a work-up protocol involving the addition of DMSO to the
54
reaction mixture was applied in order to remove ruthenium by-products. Separation of the
macrocycle from unreacted starting material was not attempted, rather, after filtration
through a plug of silica gel, the mixture was immediately submitted to global deprotection.
After purification by preparative RP-HPLC, 45a and 45b were obtained in 29% and 57% yield,
respectively, for the two-step sequence from 44a/b.
In order to assess the intrinsic antibacterial activity of ripostatin analogs, all compounds
were tested against E. coli JW5503-1 (originating from the Keio collection, and supplied by
the Coli Genetic Stock Center, CGSC-11430), which (among other changes), lacks the Tol C
6
5
efflux channel (ΔtolC732::kan). As can be seen from the data summarized in Table 1,
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
analogs 29, 32, and 38, harboring modified C3-side chains, are more potent against E. coli
JW5503-1 than either the parent compound ripostatin B (2) or ripostatin A (1).
Table 1. Antibacterial activity of natural ripostatins and selected analogs against efflux-
deficient E. coli JW5503-1.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
E. coli JW5503-1
Compound
MIC (µg/mL)
Ripostatin B (2)
Ripostatin A (1)a
3.125 - 6.25
6.3 - 12.5
0.39 - 0.78
0.78 - 1.56
1.56 - 3.125
>100
2
3
3
9
8
2
45a
5b
4
50 - 100
aRipostatin A (1) was a kind gift of Prof. Rolf Müller (Helmholtz Institute for Pharmaceutical
Research (HIPS), Saarbrücken, Germany).
These data clearly suggest that the carboxylate group in ripostatin B (2) is not an essential
requirement for intrinsic antibiotic activity. In contrast, truncated analogs 45a and 45b
showed no meaningful activity (MICs >50 g/mL), thus indicating that a hydrophobic tail of
sufficient length is required for ripostatin-type analogs, in order to display antibacterial
activity. In light of their poor (or non-existent) intrinsic antibiotic activity, truncated analogs
4
5a and 45b were not further investigated.
Interestingly, as we discovered only very recently, a natural product with a similar
6
6
structure as ripostatin B (2) was reported already in 1988 by Singh et al. This compound,
named aerocavin (46), was isolated from a non-pigmented strain of Chromobacterium
violaceum and it displays the same relative configuration as ripostatin B (2).
While the absolute configuration of 46 has not been secured, its specific rotation has the
2
2
25
= +35.7 (c 1 in MeOH) for 22),
same sign as that of 2 ([]
D
= +25.1° (c 0.9 in MeOH) vs. []
D
which makes it very likely that the two compounds also have the same absolute
ACS Paragon Plus Environment

Page 23 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
configuration. Aerocavin (46), whose C13-side chain has the same length as that of ripostatin
B (2), has been reported to be moderately active against different strains of S. aureus, with
MICs similar to those that have been obtained for ripostatin B (2) (3.1 - 6. 3 g/mL). The
compound was found to be inactive against wild-type E. coli.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
As we were particularly interested in the potential of ripostatins to serve as lead structures
for the development of new antitubercular agents, ripostatin B (2) and analog 29 were also
assessed for activity against Mtb. Disappointingly, however, both 2 and 29 proved to be
inactive against Mtb H37Rv or Mycobacterium smegmatis (MICs >100 g/mL). These findings
4
are in line with the data reported recently by Prusov and Ebright. However, in contrast to
Prusov and Ebright, who have reported both ripostatins to be inactive against RNAP from Mtb,
we have found good activity of 2 and 29 against Mtb RNAP in a promoter non-specific
6
7
transcription assay using calf thymus DNA as a template; IC50 values were 2.8 M for 2 and
.44 M for 29 vs. 0.1 M for rifampicin. Using the same assay format, IC50 values obtained
0
against E. coli RNAP were 0.2 M for 2, and 0.11 M for rifampicin; the value for 2 is about
one order of magnitude higher than the one reported recently by Prusov and Ebright (IC50 of
4
2 against E. coli RNAP of 0.018 M). It should be noted, however, that literature IC50 values
for bacterial RNAP inhibition by ripostatins vary significantly, even if reported by the same
laboratory. Thus, the IC50 value of 1 against E. coli RNAP reported in ref. 10 is 0.6 M (i. e. >20-
fold higher than in ref. 4). For S. aureus RNAP the IC50s for 1 in ref. 4 and ref. 10 are 0.88 M
and 6 M, respectively.
The reasons for the discrepancy between our own data for Mtb RNAP inhibition by
ripostatin B (2) and those reported by Prusov and Ebright are unclear at this point but may
be related to different assay set-ups. If so, the question arises which of the assays employed
is the more relevant; while one might argue that the lack anti-Mtb activity of 2 indicates that
the compound is a poor inhibitor of Mtb RNAP indeed, this conclusion is not inevitable.
Obviously, the possible RNAP inhibitory activity of 2 could be obscured by poor penetration
into Mtb (due to rapid efflux68,69 or impaired influx or both), as is the case for E. coli. We
have tried to address this question by exposing Mtb H37Rv to ripostatin B (2) in the presence
of sub-inhibitory concentrations (one fourth of their MIC) of the efflux inhibitors carbonyl
cyanide m-chlorophenyl hydrazone (CCCP), chlorpromazine (CPZ) and phenylalanyl-arginine-
β-naphthylamide (PAβN); in no case, however, could any antibacterial activity be observed
70
(MIC >100 g/mL). In contrast, an MIC of 12.5 g/mL was obtained for inhibition of E. coli
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
71
(
ATCC 25932) in the presence PAβN, a competitive inhibitor of the AcrAB-TolC system (vs.
MIC >100 g/mL for wild-type E. coli). Interestingly, the addition of 0.01% sodium dodecyl
7
2
sulfate (SDS), a known chemical penetration enhancer, to the growth medium led to
detectable (but still high) MIC values for ripostatin B (2) and analog 29 against Mtb H37Rv
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
MIC ca. 50 g/mL), suggesting that limited cell wall permeability might at least partly
contribute to the lack of antimycobacterial activity of ripostatin B (2) and related structures.
Finally, preliminary stability experiments with analog 29 in Middlebrook 7H9 broth medium
(
used in this work for determination of Mtb susceptibility to ripostatins) have indicated that
the complete lack of activity of the compound against Mtb H37Rv is not caused by massive
degradation in the culture medium.
It remains to be investigated if compounds such as 29, 32, and 38, which lack an ionizable
group in the C3 side chain, exhibit more potent antibacterial activity than 2 against bacteria
with intact efflux systems.
Conclusions
In conclusion, we have developed a successful total synthesis of the myxobacterial natural
product ripostatin B (2), which provided the target molecule in 21 steps (longest linear
sequence, starting from D-aspartic acid) and 3.6% overall yield. The molecule's Achilles heel,
the doubly skipped C2–C9 triene motif, was installed at a late stage of the synthesis by ring-
closing olefin metathesis with Grubbs-I catalyst, which afforded a cleaner reaction than
second-generation catalysts and produced macrocycle 26 in 78% yield as the pure E isomer
at C5–C6.
A major hurdle to overcome was the esterification of acid 6 with alcohol 7, a reaction that
failed to produce synthetically useful yields under a plethora of conditions examined. This
problem could finally be solved by employing a low-temperature Yamaguchi protocol, which
delivered ester 5 in 80% yield. Other key steps in the assembly of the RCM precursor were a
Paterson aldol reaction to install the stereocenter at C13 and a reduction under Evans-
Tishchenko conditions to set the stereocenter at C15.
The total synthesis provided a conceptual blueprint for the synthesis of an analog with an
extended C3 side chain (38) and two ripostatin B variants with truncated C13-side chains.
While analogs 29, 32, and 38 were active against the efflux-deficient E. coli strain JW5503-1,
neither ripostatin B (2) nor 29, as the most potent inhibitor of E. coli JW5503-1 showed any
ACS Paragon Plus Environment

Page 25 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
activity against Mtb H37Rv. While the data available at this point may be interpreted such as
to suggest that 2 lacks sufficient inhibitory activity again Mtb RNAP, this conclusion is still
premature, as efflux phenomena in Mtb are highly complex68,69 and most likely have not
been fully addressed by our experiments. Testing of 29, 32, and 38 against other species of
bacteria has not been performed yet.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We continue to work on new ripostatin analogs, in order to identify compounds whose
intrinsic antibacterial activity, as displayed against efflux-deficient strains of E. coli, will also
become manifest in bacteria with intact efflux systems. The results of these efforts will be
reported in due course.
Experimental Section
General experimental methods. All non-aqueous reactions were performed under an argon
atmosphere using flame-dried glassware and standard syringe/septa techniques.
CH
(DCM from CaH
purchased as anhydrous grade from Fluka (puriss.; dried over molecular sieves; H
0.005%) and used without further purification unless otherwise stated. Solvents for
2
Cl
2
(DCM), THF and Et
2
O used for reactions were distilled under argon prior to use
2
, THF and Et
2
O from Na/benzophenone). All other absolute solvents were
2
O
<
extractions, flash column chromatography (FC) and thin layer chromatography (TLC) were
purchased as commercial grade and distilled prior to use. All other commercially available
reagents were used without further purification unless otherwise stated. Reactions were
magnetically stirred and monitored by TLC performed on Merck TLC aluminum sheets (silica
gel 60 F254). Spots were visualized with UV light (λ = 254 nm) or through staining with
Ce
2
(SO
4
)
3
/phosphomolybdic acid/H
2
SO
4
(CPS), vanillin/H
2
4
SO /EtOH or KMnO
4
/K
2
CO .
3
Chromatographic purification of products (FC) was performed using Fluka silica gel 60 for
preparative column chromatography (particle size 40-63 μm).
Melting points were obtained in open capillary tubes using a Büchi melting point
apparatus B-540 and are uncorrected.
1
13
H- and C-NMR spectra were recorded in CDCl
3
, CD
3
OD or C
6
6
D on a Bruker AV-400 400
MHz or on a Bruker AV-500 500 MHz spectrometer at room temperature. Chemical shifts (δ)
1
are reported in ppm and are referenced to chloroform (δ 7.26 ppm for H, δ 77.16 ppm for
13
1
13
1
C), MeOH (δ 3.31 ppm for H, δ 49.00 ppm for C) or benzene (δ 7.16 ppm for H, δ 128.06
1
3
13
ppm for C). All C-NMR spectra were measured with complete proton decoupling. Data for
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
NMR spectra are reported as follows: s = singlet, d = doublet, t = triplet, q = quartet, m =
multiplet, br = broad signal, J = coupling constant in Hz.
Infrared spectra (IR) were recorded on a Jasco FT/IR-6200 instrument. Resonance
-
1
frequencies are given as wavenumbers in cm .
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Optical rotations were measured on a Jasco P-1020 polarimeter at the sodium D line with
ퟐퟒ
a 10 or 100 mm path length cell and are reported as follows: [풂] : (concentration (g/100
퐃
mL), and solvent).
High resolution mass spectra (HRMS) were recorded on a Bruker maXis ESI-Qq-TOF-MS or
on a Waters Micromass AutoSpec Ultima EI-Sector-MS, respectively, by the ETH Zürich MS
service (Louis Bertschi, Rolf Häfliger and Oswald Greter under the direction of Dr. Xiangyang
Zhang).
For analytical HPLC the following combination of devices by VWR HITACHI was used:
column oven L-2350, PDA detector L-2455, autosampler L-2200, pump L-2130. A reversed
phase Waters Symmetry C18 column (3.5 μm, 4.6x100 mm) or chiral, normal phase Daicel
Chiralpak AD-H (6 μm, 4.6x150 mm) were used. For preparative HPLC a device by Gilson
equipped with a Waters SymmetryPrep C18 column (5 μm, 19x100 mm, room temperature)
was used.
Ethyl (R)-5-hydroxy-7-((4-methoxybenzyl)oxy)hept-2-ynoate (E-1): To a solution of ethyl
propiolate (5.1 mL, 49.5 mmol, 2.0 equiv.) in THF (200 mL) at –78 °C was added dropwise n-
BuLi (1.6m in hexane, 29.4 mL, 47.0 mmol, 1.9 equiv.). The resulting yellow solution was
stirred at –78 °C for 20 min, then a solution of epoxide 4 (5.15 g, 24.7 mmol, 1.0 equiv.) in
THF (50 mL) was added at a rate such that that the internal temperature did not rise above
−
70 °C, followed by neat BF
min at –78 °C, the orange solution was quenched by the addition of MeOH (30 mL). Sat. aq.
NaHCO (50 ml), water (50 mL) and EtOAc (150 mL) were next added (the pH should be >7).
The phases were separated and the aqueous layer was extracted with EtOAc (3x100 mL). The
combined organic layers were washed with brine (10 mL), dried over MgSO , filtered and
concentrated under reduced pressure. Purification by flash chromatography (hexane/EtOAc
3
·OEt
2
(7.6 mL, 61.8 mmol, 2.5 equiv.). After being stirred for 45
3
4
2
:1) afforded alcohol E-1 (6.51 g, 21.2 mmol, 86%) as a yellow oil.
1
TLC: R
f
= 0.30 (hexane/EtOAc 2:1, UV, CPS); H-NMR (400 MHz, CDCl
3
): δ 7.26-7.21 (m, 2H),
6
.90-6.85 (m, 2H), 4.45 (s, 2H), 4.20 (q, J = 7.1, 2H), 4.08-3.99 (m, 1H), 3.80 (s, 3H), 3.75-3.68
ACS Paragon Plus Environment

Page 27 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
(
m, 1H), 3.67-3.60 (m, 1H), 3.32 (d, J = 3.5, OH), 2.56 (dd, J = 17.2, 5.8, 1H), 2.49 (dd, J = 17.2,
1
3
6
.6, 1H), 1.94-1.78 (m, 2H), 1.29 (t, J = 7.1, 3H); C-NMR (100 MHz, CDCl ): δ 159.5, 153.7,
3
129.8, 129.5, 114.0, 85.8, 75.0, 73.2, 69.5, 68.4, 62.0, 55.4, 35.4, 27.3, 14.1; IR (thin film): 
3
1
3
454, 3010, 2980, 2938, 2910, 2864, 2837, 2234, 1705, 1612, 1513, 1464, 1366, 1301, 1244,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
173, 1070, 1032, 847; HRMS (ESI): calculated for C17
H26NO
5
[M+NH
]+: 324.1805, found
4
ퟐퟒ
24.1799; [풂] : ‒1.0 (c 0.5 in CHCl ).
3
퐃
TBS-ether 11: To a solution of alcohol E-1 (6.49 g, 21.2 mmol, 1.0 equiv.) in DCM (150 mL)
at 0 °C was added 2,6-lutidine (4.2 mL, 36.0 mmol, 1.7 equiv.) followed by TBSOTf (5.9 mL,
27.5 mmol, 1.3 equiv.). The pale yellow solution was stirred at 0 °C for 30 min. Water (50 mL)
was added, the phases were separated and the aqueous layer was extracted with DCM (3x50
mL). The combined organic layers were dried over MgSO , filtered and concentrated under
4
reduced pressure. Purification by flash chromatography (hexane/EtOAc 10:1) afforded
silylether 11 (8.78 g, 20.9 mmol, 98%) as a colorless oil.
1
TLC: R
f
= 0.73 (hexane/EtOAc 2:1, UV, CPS); H-NMR (400 MHz, CDCl
3
): δ 7.28-7.22 (m, 2H),
6.90-6.85 (m, 2H), 4.43 (d, J = 11.4, 1H), 4.38 (d, J = 11.4, 1H), 4.21 (q, J = 7.1, 2H), 4.09-4.00
m, 1H), 3.81 (s, 3H), 3.52 (dd, J = 6.8, 5.6, 2H), 2.52 (dd, J = 17.2, 6.2, 1H), 2.47 (dd, J = 17.1,
(
5
0
6
.7, 1H), 1.95-1.85 (m, 1H), 1.85-1.75 (m, 1H), 1.29 (t, J = 7.1, 3H), 0.88 (s, 9H), 0.09 (s, 3H),
13
.06 (s, 3H); C-NMR (100 MHz, CDCl
3
): δ 159.3, 153.8, 130.6, 129.4, 113.9, 86.5, 74.8, 72.8,
7.6, 66.2, 61.9, 55.4, 37.1, 28.1, 25.9, 18.1, 14.2, −4.5, −4.7; IR (thin film):  2953, 2930,
2893, 2856, 2238, 1710, 1613, 1586, 1513, 1472, 1464, 1366, 1302, 1245, 1173 1092, 1072,
+
1035, 1005, 917; HRMS (ESI): calculated for C23
H
37
O
5
Si [M+H] 421.2405, found 421.2418;
ퟐퟒ
[
풂] : ‒13.4 (c 0.7 in CHCl ).
3
퐃
Thioether 12: To a solution of alkyne 5 (8.78 g, 20.9 mmol, 1.0 equiv.) in MeOH (100 mL)
was added thiophenol (2.4 mL, 23.0 mmol, 1.1 equiv.) and sodium methoxide (56.4 mg, 1.04
mmol, 0.05 equiv.). The reaction was stirred at room temperature for 14 h before being
filtered through a plug of celite (eluting with EtOAc). The filtrate was concentrated under
reduced pressure. Purification by flash chromatography (hexane/EtOAc 10:1) gave thioether
1
2 in quantitative yield as an about 8:1 mixture of (Z/E)-isomers. On small scale, it was
possible to separate the isomers in a single chromatographic run, affording (Z)-12 in yields of
about 85%. Analytical data are for pure (Z)-12.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
TLC: R
f
= 0.36 (hexane/EtOAc 4:1, UV, CPS); H-NMR (400 MHz, CDCl
3
): δ 7.52-7.48 (m, 2H),
7
.38-7.28 (m, 3H), 7.24-7.18 (m, 2H), 6.90-6.85 (m, 2H), 5.90 (s, 1H), 4.28 (s, 2H), 4.27-4.14
(m, 2H), 3.81 (s, 3H), 3.77-3.69 (m, 1H), 3.22-3.12 (m, 2H), 2.40-2.28 (m, 2H), 1.66-1.56 (m,
1
3
1
H), 1.52-1.42 (m, 1H), 1.30 (t, J = 7.1, 3H), 0.80 (s, 9H), ‒0.08 (s, 3H), ‒0.15 (s, 3H); C-NMR
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
100 MHz, CDCl ): δ 166.0, 159.3, 157.0, 135.8, 131.4, 130.7, 129.4, 129.3, 129.2, 115.5,
3
1
2
1
13.9, 72.6, 68.3, 66.2, 60.0, 55.4, 44.8, 37.1, 25.9, 18.1, 14.6, ‒4.5, ‒4.7; IR (thin film): 
952, 2928, 2894, 2855, 1701, 1613, 1582, 1513, 1472, 1463, 1440, 1389, 1366, 1331 1302,
+
247, 1196, 1171, 1092, 1035, 1005, 921; HRMS (ESI): calculated for C29
H46NO
5
SSi [M+NH ] :
4
ퟐퟒ
548.2860, found 548.2861; [풂] : ‒11.4 (c 0.6 in CHCl ).
3
퐃
Ester 13: To a brown-grey suspension CuI (7.83 g, 41.1 mmol, 1.93 equiv.) in THF (120 mL)
at ‒78 °C was added dropwise MeMgBr (2.8m in Et O, 27.8 mL, 77.7 mmol, 3.7 equiv.) (the
2
internal temperature increased to −65 °C). The grey suspension was immersed in an ice bath
for about 5 min, giving a viscous green suspension. At ‒78 °C, a solution of thioether 12 (11.3
g, 21.3 mmol, 1.0 equiv.; mixture of the (Z)- and the (E)-isomer, vide supra) in THF (30 mL)
was added drop-by-drop. The remaining dry ice was removed from the cooling bath and the
yellow mixture was allowed to warm to ‒30 °C in the cooling bath over ca. 30-40 min. Sat.
aq. NH
phases were separated and the aqueous layer was extracted with EtOAc (3x100 mL). The
combined organic layers were washed with brine, dried over MgSO , filtered through a short
4
Cl (130 mL), 25% aq. NH
3
solution (40 mL) and EtOAc (100 mL) were added, the
4
plug of celite and concentrated under reduced pressure. Purification by flash
chromatography (hexane/EtOAc 10:1 → 7:3) afforded ester 13 (9.05 g, 20.7 mmol, 97%) as a
yellow oil (contains ca. 6% of the undesired (Z)-isomer derived from the minor isomer of the
thiophenol addition reaction). Note: If pure (Z)-12 was employed, (E)-13 was also obtained in
quantitative yields (≥95%). Sometimes, a yellow instead of a green suspension was obtained
upon addition of MeMgBr to the suspension of CuI and warming to 0 °C, leading to the same
outcome.
1
TLC: R
f
= 0.57 (hexane/EtOAc 4:1, UV, CPS); H-NMR (400 MHz, CDCl
3
): δ 7.27-7.22 (m, 2H),
6
=
1
7
.89-6.83 (m, 2H), 5.66 (q, J = 1.3, 1H), 4.44 (d, J = 11.5, 1H), 4.38 (d, J = 11.5, 1H), 4.14 (qd, J
7.1, 1.6, 2H), 4.06-3.97 (m, 1H), 3.80 (s, 3H), 3.50 (t, J = 6.5, 2H), 2.29 (ddd, J = 13.0, 6.4,
.0, 1H), 2.24 (ddd, J = 13.0, 6.2, 0.8, 1H), 2.16 (d, J = 1.3, 3H), 1.81-1.63 (m, 2H), 1.26 (t, J =
13
.1, 3H), 0.85 (s, 9H), 0.04 (s, 3H), 0.02 (s, 3H); C-NMR (100 MHz, CDCl ): δ 166.7, 159.3,
3
ACS Paragon Plus Environment

Page 29 of 65
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
1
56.6, 130.7, 129.4, 118.5, 113.9, 72.8, 68.1, 66.6, 59.6, 55.4, 49.4, 37.3, 26.0, 19.6, 18.2,
1
4.5, ‒4.4, ‒4.6; IR (thin film):  2952, 2929, 2856, 2362, 1714, 1646, 1613, 1587, 1513,
1472, 1463, 1442, 1366, 1302, 1247, 1221, 1172, 1148, 1094, 1037, 1005, 938; HRMS (ESI):
+
ퟐퟒ
calculated for C24
H44NO
5
Si [M+NH
4
] : 454.2983, found 454.2987; [풂] : +1.0 (c 0.5 in CHCl ).
3
퐃
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(R,E)-5-((t-butyldimethylsilyl)oxy)-7-((4-methoxybenzyl)oxy)-3-methylhept-2-en-1-ol (E-
2
): To a solution of 13 (9.05 g, 20.7 mmol, 1.0 equiv.; mixture of (E/Z)-isomers, vide supra) in
DCM (150 mL) was added dropwise DIBAL-H (1.2m in toluene, 37.1 mL, 44.6 mmol, 2.15
equiv.). The solution was stirred for at ‒78 °C for 30 min. Sat. aq. Rochelle salt (200 mL) was
added and stirring was continued at room temperature until two clear phases formed (ca. 3
h). The phases were separated and the aqueous layer was extracted with DCM (4x100 mL).
The combined organic layers were washed with brine, dried over MgSO
concentrated under reduced pressure. Purification by flash chromatography (hexane/EtOAc
:1 → 2:1 → 1:1) furnished allylic alcohol E-2 (7.1 g, 18.0 mmol, 87%, pure (E)-isomer) as a
4
, filtered and
3
pale yellow oil. Note: If pure (E)-13 was employed, (E)-E-2 was obtained in yields of between
86-94%.
1
TLC: R
f
= 0.13 (hexane : EtOAc 4:1, UV, CPS); H-NMR (400 MHz, CDCl
3
): δ 7.29-7.22 (m,
2
4
1
H), 6.90-6.84 (m, 2H), 5.41 (tq, J = 6.8, 1.3, 1H), 4.43 (d, J = 11.5, 1H), 4.38 (d, J = 11.5, 1H),
.15-4.09 (m, 2H), 3.99-3.91 (m, 1H), 3.80 (s, 3H), 3.51 (dd, J = 7.2, 6.0, 2H), 2.21 (dd, J =
3.3, 6.0, 1H), 2.13 (dd, J = 13.3, 6.9, 1H), 1.83-1.72 (m, 1H), 1.67 (m, 3H), 1.68-1.58 (m, 1H),
1
3
1.24 (t, J = 5.5, OH), 0.87 (s, 9H), 0.04 (s, 3H), 0.03 (s, 3H); C-NMR (100 MHz, CDCl
3
): δ
59.3, 136.6, 130.8, 129.4, 126.7, 113.9, 72.8, 68.2, 66.9, 59.5, 55.4, 48.2, 37.0, 26.0, 18.2,
7.1, ‒4.2, ‒4.6; IR (thin film):  3387, 2952, 2928, 2855, 1613, 1586, 1512, 1472, 1463,
442, 1361, 1302, 1247, 1173, 1089, 1036, 1004, 938; HRMS (ESI): calculated for C22 O Si
1
1
1
H
39
4
+
ퟐퟒ
[
M+H] : 395.2612, found 395.2623; [풂] : ‒7.3 (c 0.6 in CHCl ).
3
퐃
Allylic bromide 14: To a solution of E-2 (2.60 g, 6.59 mmol, 1.0 equiv.), 2,6-lutidine (1.2 mL,
10.6 mmol, 1.6 equiv.) and PPh (3.80 g, 14.5 mmol, 2.2 equiv.) in DCM (130 mL) at −78 °C
was added dropwise a solution of CBr (4.81 g, 14.5 mmol, 2.2 equiv.) in DCM (5 mL). After 2
3
4
h, the reaction was carefully quenched with water (50 mL). The phases were separated and
the aqueous layer was extracted with DCM (3x50 mL). The combined organic layers were
dried over MgSO
4
. (Note: Washing with brine leads to partial substitution of chloride for
ACS Paragon Plus Environment