Total Synthesis and Biological Evaluation of the Glycosylated Macrocyclic Antibiotic Mangrolide A

Article abstract of DOI:10.1002/anie.201805770

The macrocyclic antibiotic mangrolide A has been described to exhibit potent activity against a number of clinically important Gram-negative pathogens. Reported is the first enantioselective total synthesis of mangrolide A and derivatives. Salient features of this synthesis include a highly convergent macrocycle preparation, stereoselective synthesis of the disaccharide moiety, and two β-selective glycosylations. The synthesis of mangrolide A and its analogues enabled the re-examination of its activity against bacterial pathogens, and only minimal activity was observed.

Full text of DOI:10.1002/anie.201805770

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Total Synthesis and Biological Evaluation of the Glycosylated
Macrocyclic Antibiotic Mangrolide A
Authors: Hiromu Hattori, Joel Roesslein, Patrick Caspers, Katja Zerbe,
Hideki Miyatake-Ondozabal, Daniel Ritz, Georg Rueedi, and
Karl Gademann
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201805770
Angew. Chem. 10.1002/ange.201805770
Link to VoR: http://dx.doi.org/10.1002/anie.201805770
http://dx.doi.org/10.1002/ange.201805770

10.1002/anie.201805770
Angewandte Chemie International Edition
COMMUNICATION
Total Synthesis and Biological Evaluation of the Glycosylated
Macrocyclic Antibiotic Mangrolide A
Hiromu Hattori,[a] Joel Roesslein,[a] Patrick Caspers,[b] Katja Zerbe,[a] Hideki Miyatake-Ondozabal,[a] Daniel
Ritz,[b] Georg Rueedi[b] and Karl Gademann*[a]
Dedicated to Professor E. J. Corey on the occasion of his 90th birthday.
Abstract: The macrocyclic antibiotic mangrolide A has been reported to
exhibit potent activity against a number of clinically important Gram-
negative pathogens. Herein we report the first enantioselective total
synthesis of mangrolide A and derivatives. Salient features of our synthesis
Mangrolide A (1) was isolated from the SNA18 strain of
Actinoalloteichus sp. following phenotypic screening against
Burkholderia cepacia (Figure 1).[2] A preliminary biological study
revealed that this natural product is active against Gram-negative
bacteria including Acinetobacter baumannii (MIC = 0.25 µg/mL) and
Pseudomonas aeruginosa (MIC = 1.0 µg/mL). This antibacterial activity
has been suggested to involve interaction with the 30S subunit of the
ribosome leading to mistranslation.[4] Interestingly, fidaxomicin (2), a
clinically approved antibiotic against Gram-positive bacteria, shares an
almost identical 18-membered macrocyclic scaffold with mangrolide (1),
while the carbohydrate and resorcylate moieties are replaced. Most
surprisingly, the biological activity and target protein completely
changes with these structural alterations from activity against Gram-
positive bacteria by RNA polymerase inhibition (for 2) to activity
against Gram-negative species by ribosome inhibition (for 1).[4, 5] This
change of biological properties when switching the molecular decoration
of a macrocyclic antibiotic is highly unusual and warrants further studies.
The structure of mangrolide A (1) is characterized by a highly
unsaturated 18-membered macrocycle decorated with the unusual
carbohydrates, D-2,4-dimethyl quinovose and D-mycaminose,[3] both
connected via b-glycosidic bonds. Synthetic challenges of mangrolide A
(1) include (1) the stereoselective construction of the macrocyclic
macrocyclic core,[6] (2) the regioselective functionalization as well as the
stereoselective synthesis of both carbohydrates and, most importantly,
(3) the two b-selective glycosylation reactions. Despite notable
advances in carbohydrate chemistry, methods for the construction of b
glycosidic bonds of N-alkylamino sugars are rare,[7] although some
successful total syntheses of compounds containing this motif have been
published.[8] In addition, the b-selective glycosylation at the sterically
demanding C11 hydroxy group was expected to be challenging based on
our earlier experiences in the fidaxomicin synthesis.[6b] To date, one
synthetic study towards mangrolide A (1) has been documented in a
patent by De Brabander and co-workers.[4] However, to the best of our
knowledge, no total synthesis nor detailed antibiotic testing has been
reported, presumably due to the challenging glycosylation chemistry and
the complex macrocyclic features.[9] In this communication, we report
on the first total synthesis of mangrolide A (1) and on the biological
evaluation against bacterial pathogens as well as in vitro mechanism of
action studies.
include
a highly convergent macrocycle preparation, stereoselective
synthesis of the disaccharide moiety and two b-selective glycosylations. The
synthesis of mangrolide A and its analogues enabled the re-examination of
its activity against bacterial pathogens, and only minimal activity was
observed.
Antimicrobial resistance (AMR) is an increasing serious threat to global
public health and the annual loss of life is expected to reach 10 million
deaths by 2050 with an economic cost of $100 trillion.[1] Effectively
addressing AMR requires a multifaceted approach that facilitates
sustainable and appropriate use of existing antimicrobials, but also calls
for new antibiotics. The development of lead structures with new
molecular scaffolds and mechanisms of action thus constitutes an
important goal. Particularly, while several antibiotics against infections
caused by Gram-negative bacteria are in clinical development today,
many constitute modifications of existing antibiotic classes, as the
discovery of novel antibiotics that i) can penetrate the outer and inner
membrane, and ii) are not subject to extensive efflux remains very
challenging.
OH
7
20
OH
H
Me
D-Quinovose
H
10
1
Me
O
O
Me
11
H
Me
MeO
HO
β
H
H
H
O
O
Me2N
O
O
3´´ MeO
Me
β
3´
14
17
HO
D-Mycaminose
Me
Me
Me
D-Rhamnose
Mangrolide A (1)
O
Me
HO
Et
Cl
Cl
β
O
O
OH
OH
OH
H
Me
Me
OMe
H
Me
Me
O
O
O
O
OH
H
O
Me
Dichlorohomoorsellinate
H
H
H
O
O
OH
HO
β
Me
D-Noviose
Me
Me
Me
Fidaxomicin (2, Tiacumicin B, Lipiarmycin A3, Clostomicin B1)
Figure 1. Bis-b-glycosylated antibiotics, mangrolide A (1) and fidaxomicin (2).
Our retrosynthetic analysis of mangrolide A (1) is illustrated in Scheme
1. Mangrolide A was disconnected at the b-glycosidic bond expecting
that the glycosylation with azidodisaccharide 3 and the protected
macrocycle 4 would proceed with the aid of the neighbouring acyl
group.[10] We envisaged that the strategic late-stage N-dimethylation of
the azide intermediate not only would facilitate smooth glycosidation,
but also would provide a key azide analogue to investigate the biological
importance of the dimethylamino group (vide infra). The advanced
carbohydrate intermediate 3 was expected to be synthesized from D-
quinovose 5 and D-mycaminose precursor 6 by b-selective glycosylation
[a]
[b]
H. Hattori, J. Roesslein, Dr. H. Miyatake-Ondozabal, Dr. K. Zerbe,
Prof. Dr. K. Gademann
Department Chemistry
University of Zurich
Winterthurerstrasse 190, CH-8057 Zurich
E-mail: karl.gademann@uzh.ch
Dr. P. Caspers, Dr. D. Ritz, Dr. G. Rueedi
Idorsia Pharmaceuticals Ltd
Hegenheimermattweg 91, CH-4123 Allschwil
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/anie.201805770
Angewandte Chemie International Edition
COMMUNICATION
[20]
with the aid of the nitrile effect.[11] Quinovose 5 would be obtained from
glycoside 7 and the D-mycaminose precursor 6 from 2-acetylfuran (8) by
employing the protocol developed by O'Doherty.[12] The macrocyclic
part 4 was proposed to be synthesized by ring closing metathesis, Suzuki
coupling and Yamaguchi esterification as key steps.[6a, b]
trifluorophenylacetimidate 14[19] and Ph3CB(C6F5)4
in a n-PrCN/
CH2Cl2 mixture resulted in disaccharide b-15 in 81% yield with 4.3 to 1
selectivity in favor of the b diastereoisomer. This ratio and yield are
worth noting as the selectivity of nitrile assisted glycosylations are
generally low in the presence of a silyl group, especially when the C4
hydroxy group of another sugar is employed as the acceptor.[6b] With
careful control of temperature and stirring, this reaction was scaled to
gram level. We speculate that the very weak tendency of the bulky
tetrakis(pentafluorophenyl)borate anion towards coordination of the
oxocarbenium cation[21] contributes to the preferential formation of the
α-nitrile adduct, which is effectively attacked from the b side to form the
desired glycosidic bond. To the best of our knowledge, this catalyst
system is the first example of b-glycosylation using acetimidate donors,
although the activation of glycosyl fluorides and thioglycosides are well
studied by Mukaiyama and co-workers using the same catalyst.[18a, 22]
Next, epoxide opening[17] of disaccharide 15 using an azide nucleophile
was performed and the desired equatorial adduct 16 was isolated in 80%
yield after acetylation of the resultant C2´ alcohol. At this stage, the use
of the acetate as a protecting group was essential for the separation of
the two regioisomers. Deprotection of the PMB group and formation of
the imidate afforded donor 17 as a mixture of anomers (α/β 1:1) in
quantitative yield.
Yamaguchi
Esterification
TBSO
RCM
β-selective
glycosylation
Me
TBSO
O
O
Me
Me
MeO
TBSO
O
1
O
N3
O
Me
Me
Me Me
Me
MeO
β-selective
glycosylation
AcO
OLG
3
Suzuki
Coupling
OH
4
de novo
synthesis
Me
Me
HO
HO
HO
O
O
O
O
MeO
TBSO
HO
O
Me
MeO
LG
HO
O
OMe
OPMB
Methyl
D-glucoside (7)
2-acetylfuran (8)
5
6
Scheme 1. Retrosynthetic analysis. LG = leaving group, PMB = 4-methoxybenzyl,
TBS= tert-butyldimethylsilyl, RCM = ring-closing metathesis.
Ph
O
O
TBSO
I
3 steps[12]
73%
a) DIBAL
O
O
BnO
TBSO
7
b) PPh3, I2, imid.
76% (2 steps)
MeO
MeO
OMe
OMe
Me
9
10
Me
O
Me
HO
Me
O
O
MeO
TBSO
MeO
TBSO
O
e) MeOTf
2,6-(t-Bu)2Py
O
c) Pd(OH)2, H2
DIPEA
Me
Me
O
6
OPMB
O
O
MeO
MeO
MeO
TBSO
HO
TBSO
O
CF3
O
OPMB
a) Ph3CB(C6F5)4
–95 to –40 °C
β/α 4.3:1
88%
3``
d) Pd/C, H2
90% (2 steps)
MeO
β-15 (81%)
MeO
OMe
OMe
14
PhN
11
12
b) NaN3
c) Ac2O
80% (2 steps)
Me
Me
O
O
MeO
TBSO
g) CF3C(=NPh)Cl
K2CO3, 82%
f) H2SO4, Ac2O;
MeO
TBSO
Me
Me
O
Me
O
Me
MeO
K2CO3, MeOH
quant.
MeO
MeO
TBSO
MeO
TBSO
O
CF3
Ph
OH
O
O
O
O
N3
O
CF3
NPh
d) DDQ, quant.
e) CF3C(=NPh)Cl, quant.
13
14
N3
MeO
MeO
N
3`
OAc
AcO
OPMB
Scheme 2. Synthesis of the quinovose domain 14. Reagents and conditions: a) DIBAL,
PhMe, –20 to 10 °C; b) PPh3, I2, imid., CH2Cl2, reflux, 76% (2 steps); c) Pd(OH)2, H2
(balloon), DIPEA, MeOH; d) Pd/C, H2 (balloon), MeOH, 90% (2 steps), e) MeOTf, 2,6-
(t-Bu)2Py, DCE (1 M), 88%; f) H2SO4 (5 mol%), Ac2O, 0 °C; K2CO3, MeOH, quant.;
g) CF3C(=NPh)Cl, K2CO3, acetone, 0 °C to rt, 82%. DIBAL = diisobutylaluminium
17
16
Scheme 3. Synthesis of the azidodisaccharide donor 17. Reagents and conditions: a) 14
(1.2 eq), Ph3CB(C6F5)4 (5 mol%), MS 3Å, n-PrCN/CH2Cl2 = 1:5, –95 to –40 °C, 4 h,
β/α 4.3:1, 81% (β); b) NaN3, NH4Cl, EtOH, 80 °C; c) Ac2O, Et3N, DMAP, 80%
(2 steps); d) DDQ, CH2Cl2, pH7 phosphate buffer, quant.; e) CF3C(=NPh)Cl, K2CO3,
acetone, 0 °C to rt, quant. MS = molecular sieves, DMAP = 4-(dimethylamino)pyridine,
DDQ = 2,3-dichloro-5,6-dicyano-1,4-benzoquinone.
hydride, imid.
= imidazole, DIPEA = N,N-diisopropylethylamine, OTf =
trifluoromethanesulfonate, DCE = 1,2-dichloroethane, Py = pyridine.
The synthesis of the quinovose domain began by installing the C2´´
methyl group and C3´´ TBS ether onto methyl D-glucopyranoside (7)
(Scheme 2).[13] Notably, the sequence was completely selective to give
9 in 73% over 3 steps including a single chromatographic purification.
Selective cleavage of the benzylidene acetal at the C6´´ position[14]
followed by iodination[15] gave iodide 10, which was reductively
dehalogenated and deprotected to provide quinovose 11. Next, alcohol
11 was treated with a strong methylating reagent MeOTf[16] to provide
methyl ether 12. Retention of the TBS group at the C3´´ hydroxy group
was confirmed by transformation to the corresponding acetate (TBAF;
Ac2O). Acetal exchange under acidic conditions and hydrolysis of the
intermediate acetate followed by the imidate addition reaction afforded
acetimidate 14, suitable for glycosylation.
Next, we set out to investigate the b-selective glycosylation with
separately synthesized epoxy alcohol 6[12,
glycosyl donors and activation conditions were examined without
success (glycosylfluoride,[18a] thioglycoside,[18b] phosphoramidate,[18c]
phosphite
experimentation
Having synthesized the azidodisaccharide donor 17, we set about the
final assembly starting with macrocycle preparation (Scheme 4). Suzuki
coupling[6b, c] of previously synthesized iodide 18[6a, f] and separately
prepared boronate 19[23] provided the desired alcohol in excellent yield.
This synthetic intermediate was then converted into the corresponding
linear polyene 22 through Yamaguchi esterification with acid 20.[6a, f]
A
ring-closing metathesis reaction with second generation Grubbs catalyst
and subsequent hydrolysis of the nitrobenzoate ester at C11 afforded
alcohol 4 as a mixture of (E,E) and (E,Z) isomers. Fortunately, these
stereoisomers could be successfully separated after the TMS protection,
and the subsequent silyl deprotection gave access to the pure (E,E)
isomer. Next, azide donor 17 was reacted with macrolactone 4 using
TBSOTf as an activator. To our surprise, the reaction proceeded
smoothly to provide the desired b-isomer in 42% yield with complete
selectivity. This yield was striking given that the glycosylation of such
a hindered system could never be achieved on a related system in our
17]
(Scheme 3). Various
ester,[18d]
trichloroacetimidate[18e,f]).
Extensive
of
revealed
that the combination
This article is protected by copyright. All rights reserved.

10.1002/anie.201805770
Angewandte Chemie International Edition
COMMUNICATION
previous fidaxomicin synthesis.[6b, 6f] The key to this success is attributed
to higher stability[24] and less steric hindrance of the rationally designed
glycosyl donor. In addition, the use of the azide donor was essential, as
a
model study using various glycosyl donors with the N,N-
dimethylamino group and different activation conditions never resulted
in the desired product in acceptable yield.
OTBS
OTBS
5
OTBS
4
a) boronate 19
Pd(PPh3)4
quant.
TBSO
TBSO
O
TBSO
O
c) Grubbs II
E/Z = 2.7:1
Me
Me
Me
Me
1
CO2H
acid 20
O
O
Me
Me
Me
Me
Me
Me Me
Me
Me
Me Me
17
d) NaOMe
91% (2 steps)
11
I
b) acid 20
TCBC, 68%
Me
15
14
14
HO
Me
OPNB
18
OPNB
OH
Me
(E)-4 and (Z)-4
(E)-4 only
e) TMSCl, imid., 53%
f) AcOH, MeOH, quant.
22
PinB
boronate 19
Me
OR1
H
O
Me
O
MeO
TBSO
OR1
20
O
Me
7
H
O
CF3
NPh
N3
17
g) TBSOTf (15 mol%), rt, 42%
Me
MeO
O
O
j) TCEP;
aq. CH2O
NaBH3CN
25%
Me
AcO
11
MeO
R1O
H
Me
O
H
H
H
Mangrolide A (1)
O
N3
O
β
3´´
O
Me
MeO
3´
OR2
Me
Me
Me
h) NaOMe, MeOH
i) 3HF·NEt3, 80% (2 steps)
23: R1 = TBS, R2 = Ac
24: R1 = R2 = H
Scheme 4. Total synthesis of mangrolide A (1). Reagents and conditions: a) 19 (1.25 eq), Pd(PPh3)4 (10 mol%), TlOEt (1.5 eq), THF, H2O, 0 °C to rt, 30 min, quant.; b) 20 (2.5 eq),
TCBC, Et3N, THF; DMAP, PhMe, 68%; c) Grubbs II (20 mol%), PhMe, 100 °C, 2 h, E, E/E, Z = 2.7:1; d) NaOMe, MeOH, Et2O, 2 h, 91% (2 steps); e) TMSCl, imid., CH2Cl2, 0
°C, 15 min, 53%; f) AcOH, MeOH, 5 h, quant.; g) 17 (1.5 eq), TBSOTf (15 mol%), CH2Cl2, rt, 15 min, 42%; h) NaOMe, MeOH, Et2O, 14 h; i) 3HF•NEt3, THF, 50 °C, 24 h, 80%
(2 steps); j) TCEP, phosphate buffer, THF, MeOH, 2N NaOH, H2O, 2 days; 37% aq. CH2O, NaBH3CN, H2O, MeCN, 1 h, 25% (2 steps). Pin = pinacolato, THF = tetrahydrofuran,
TCBC = 2,4,6-trichlorobenzoyl chloride, TMSCl = chlorotrimethylsilane. TBSOTf = tert-butyldimethylsilyl trifluoromethanesulfonate, TCEP = tris(2-carboxyethyl)phosphine
hydrochloride.
Next, the acetyl and TBS groups were cleaved to give tetraol 24. Finally,
reduction of the azide group by TCEP followed by reductive amination
and purification by reversed-phase HPLC afforded fully synthetic
mangrolide (1). The NMR spectra of the formate adduct (2 equiv.) of the
synthetic material 1 matched those reported for the natural product.[2, 4]
With synthetic (+)-mangrolide A (1) and the derivative 24 in hand, we
began testing against a panel of bacteria (see also the Supporting
4]
Information).[2, As outlined in Table 1, while for natural 1 low
micromolar activity had been reported,[2, 4] synthetic 1 did not show any
activity against the reference isolates and clinical strains tested. There
was no activity against a laboratory-generated mutant of E. coli with
increased permeability (DrfaC) and decreased efflux (DtolC). The
spectrum and the activity of fidaxomicin (2) was essentially as
reported.[25]
This article is protected by copyright. All rights reserved.

10.1002/anie.201805770
Angewandte Chemie International Edition
COMMUNICATION
the macrocycle preparation from three key components, all of which are
accessed within 8 steps from commercially available starting materials.
Evaluation of mangrolide A (1) and its analogues against a panel of
pathogenic bacteria revealed no antibiotic activity, however, weak
inhibition of translation was observed in an in vitro enzyme assay system.
The detailed biological study of mangrolide A (1) and synthesis of
similar 18-membered natural and unnatural antibiotic candidates is
ongoing, and will be disclosed in due course.
Table 1. Minimal inhibitory concentrations of mangrolide A and derivatives against
Gram-positive and Gram-negative bacterial strains[a]
compound
S.a.[b]
E.f.[c]
C.d.[d]
E.c.[e]
E.c.[f]
A.b.[g]
P.a.[h]
Mangrolide A
24
>16
>16
8
>16
>16
4
>16
>16
>16
>16
>32
>16
>16
8
>16
>16
>32
>16
>16
>32
Fidaxomicin
≤0.06
Acknowledgements
[a] Bacterial isolates are from the Idorsia strain collection. Minimal inhibitory
concentrations (MICs, in µg/ml) were determined by the broth micro-dilution method
according to guidelines of the Clinical and Laboratory Standards Institute (CLSI).[26]
[b] S.a.; Staphylococcus aureus ATCC 29213. [c] E.f.; Enterococcus faecalis ATCC
29212. [d] C.d.; Clostridium difficile ATCC 70057. [e] E.c.; Escherichia coli ATCC
25922. [f] E.c.; Escherichia coli 2085 (DrfaC DtolC). [g] A.b.; Acinetobacter
baumannii A-1305. [h] P.a.; Pseudomonas aeruginosa ATCC 27853.
We gratefully acknowledge partial financial support by the SNSF. We
thank Angela Amsler and Cédric Lüthi for technical support, Simon Jurt
for assistance with NMR spectroscopy, and Myriam Gwerder, Jonathan
Delers, Hans H. Locher, and Daniela Sabato for antibacterial testing. We
further thank Stefan Diethelm and Simon Williams for stimulating
discussions.
As the bacterial strains used in the literature may be more susceptible
than the ones described herein, we decided to perform an enzyme assay
to investigate the proposed mechanism of inhibition and in vitro potency
(Table 2). Mangrolide A (1) and derivatives were tested in a coupled E.
coli S30 in vitro transcription/translation (IVTT) assay. Mangrolide A
(1) inhibited the translation with equal potency also when a phage
promoter was used, just as the erythromycin positive control did, but in
contrast to fidaxomicin (2), where the inhibition transcription was
dependent on a bacterial promoter. We therefore concluded that
mangrolide A (1) inhibits protein translation as previously described.
However, the IC50 value was 50 – 100-fold higher than the ones
determined for erythromycin or fidaxomicin, which may explain the
absence of antibacterial activity. Comparing the in vitro activity of 1 to
24 it also became apparent that the N,N-dimethylamino group plays a
key role for biological activity. The reason for the discrepancy of the
observed to the reported biological activity for compound 1 remains
unclear, contamination by a highly active minor constituent in the
natural product would present an explanation.[27]
Keywords: natural products • antibiotics • carbohydrate chemistry •
stereoselective synthesis • glycosylation
[1]
[2]
WHO 2017.12, Antibacterial agents in clinical development.
M. T. Jamison, PhD thesis, The University of Texas Southwestern Medical
Center at Dallas (US), 2013.
[3]
D-Mycaminose is found in macrolide antibiotics (e.g. tylosin or carbomycin).
Interestingly, the veterinary antibiotic drug, tylosin targets the 50S ribosomal
subunit and inhibits translation, which implies a different mechanism of action
to mangrolide A (1). a) J. M. McGuire, W. S. Boniece, C. E. Higgens, M. M.
Hoehn, W. M. Stark, J. Westhead, R. N. Wolfe, Antibiot. Chemother. 1961, 11,
320–327; b) R. B. Woodward, Angew. Chem. 1957, 69, 50–58; c) R. B.
Woodward, L. S. Weiler, P. C. Dutta, J. Am. Chem. Soc. 1965, 87, 4662–4663;
d) P. Constable, K. W. Hinchcliff, S. Done, W. Gruenberg, Veterinary Medicine
(Elsevier) 11th ed. 2017, pp. 153–174.
[4]
[5]
[6]
J. DeBrabander, PCT/US2015/030621, 2015.
I. Artsimovitch, J. Seddon, P. Sears, Clin. Infect. Dis. 2012, 55, 127–131.
a) H. Miyatake-Ondozabal, E. Kaufmann, K. Gademann, Angew. Chem. Int. Ed.
2015, 54, 1933–1936; Angew. Chem. 2015, 127, 1953–1956; b) E. Kaufmann,
H. Hattori, H. Miyatake-Ondozabal, K. Gademann, Org. Lett. 2015, 17, 3514–
3517; c) F. Glaus, K. H. Altmann, Angew. Chem. Int. Ed. 2015, 54, 1937–1940;
Angew. Chem. 2015, 127, 1957–1961; d) W. Erb, J. M. Grassot, D. Linder, L.
Neuville, J. Zhu, Angew. Chem. Int. Ed. 2015, 54, 1929–1932; Angew. Chem.
2015, 127, 1949–1952; e) L. Jeanne-Julien, G. Masson, E. Astier, G. Genta-
Jouve, V. Servajean, J. M. Beau, S. Norsikian, E. Roulland, Org. Lett. 2017, 19,
4006–4009; f) H. Hattori, E. Kaufmann, H. Miyatake-Ondozabal, R. Berg, K.
Gademann, J. Org. Chem. DOI 10.1021/acs.joc.8b00101; g) L. Jeanne-Julien,
G. Masson, E. Astier, G. Genta-Jouve, V. Servajean, J.-M. Beau, S. Norsikian,
E. Roulland, J. Org. Chem. 2018, 83, 921–929.
Table 2. Activities in E. coli cell-free transcription and translation assay (IVTT)[a]
compound
E. coli promoter
SP6 promoter
Mangrolide A
24
6.45
80.8
0.09
0.04
8.31
ND
Erythromycin
Fidaxomicin
0.05
76.6
[7]
a) K. Kitamura, Y. Maezawa, Y. Ando, T. Kusumi, T. Matsumoto, K. Suzuki,
Angew. Chem. Int. Ed. 2014, 53, 1262–1265; Angew. Chem. 2014, 126, 1286–
1289; b) I. Smith, E. N. Pitsinos, C. K. Hwang, Y. Mizuno, H. Saimoto, G. R.
Scarlato, T. Suzuki, K. C. Nicolaou, J. Am. Chem. Soc. 1993, 115, 7625–7635;
c) K. C. Nicolaou, Z. Lu, R. Li, J. R. Woods, T. I. Sohn, J. Am. Chem. Soc. 2015,
137, 8716–8719; d) S. A. Hitchcock, M. Y. Chu-Moyer, S. H. Boyer, S. H.
Olson, S. J. Danishefsky, J. Am. Chem. Soc. 1995, 117, 5750–5756.
a) A. G. Myers, J. Liang, M. Hammond, P. M. Harrington, Y. Wu, E. Y. Kuo, J.
Am. Chem. Soc. 1998, 120, 5319–5320; b) K. Tatsuta, Y. Amemiya, Y.
Kanemura, H. Takahashi, M. Kinoshita, Tetrahedron Lett. 1982, 23, 3375–
3378; c) S. L. Gholap, C. M. Woo, P. C. Ravikumar, S. B. Herzon, Org. Lett.
2009, 11, 4322–4325; d) H.-S. Oh, R. Xuan, H.-Y. Kang, Org. Biomol. Chem.
2009, 7, 4458–4463.
[a] Assay as described[28] with the following modifications to distinguish inhibitors
of transcription from inhibitors of translation. IC50 determined using an E. coli S30
extract either in presence of plasmid pBestluc (n E. coli promoter) or in presence
of plasmid pSP-luc+NF (phage SP6 promoter) with added phage SP6 RNA
polymerase. In experiments using plasmid pBestluc inhibitors of bacterial
transcription or translation interfere with the assay, whereas in experiments using
plasmid pSP-luc+NF only inhibitors of the bacterial translation can interfere. IVTT
IC50 [µM] are averages of two independent experiments. . ND: not determined.
[8]
[9]
In summary, we have developed a convergent and modular strategy for
the first total synthesis of mangrolide A (1) in 25 steps (longest linear
Reviews in antibiotic natural product synthesis: a) K. Toshima, Carbohydr. Res.
2006, 341, 1282–1297; b) S. E. Rossiter, M. H. Fletcher, W. M. Wuest, Chem.
Rev. 2017, 117, 12415–12474.
sequence). Salient features of this route include
a b-selective
glycosylation without neighbouring group assistance to form the
disaccharide (6 + 14 à 15) and another with the highly congested
acceptor alcohol (4 + 17 à 23). We also demonstrated the robustness of
[10] Given the incompatibility of azide with the macrocycle forming events (e.g.
Suzuki Coupling and RCM), we planned to introduce the disaccharide moiety
This article is protected by copyright. All rights reserved.

10.1002/anie.201805770
Angewandte Chemie International Edition
COMMUNICATION
after cyclizing the ring; an example of azidodisaccharide on macrocyclic natural
product synthesis: A. M. Szpilman, E. M. Carreira, Org. Lett. 2009, 11, 1305–
1307.
[11] K. K. -K. T. Mong, T. Nokami, N. T. T. Tran, P. B. Nhi in Selective
Glycosylations: Synthetic Methods and Catalysts. (Ed.: C. S. Bennett), Wiley-
VCH, Weinheim, 2017, pp. 59–77.
[12] H. Guo, G. A. O’Doherty, Angew. Chem. Int. Ed. 2007, 46, 5206–5208; Angew.
Chem. 2007, 119, 5298–5300.
[13] a) D. Icheln, B. Gehrcke, Y. Piprek, P. Mischnick, W. A. König, M. A. Dessoy,
A. F. Morel, Carbohydr. Res. 1996, 280, 237–250.
[14] N. Tanaka, I. Ogawa, S. Yoshigase, J. Nokami, Carbohydr. Res. 2008, 343,
2675–2679.
[15] P. J. Garegg, R. Johansson, B. Samuelsson, Synthesis, 1984, 189–195.
[16] The use of MeI, NaH, THF or Me2SO4, NaH, THF led to the migration of the
TBS group. Ag2O, MeI (neat), at reflux gave no reaction.
[17] M. P. Croatt, E. M. Carreira, Org. Lett. 2011, 13, 1390–1393.
[18] a) T. Mukaiyama, H. Jona, Proc. Japan Acad. Ser. B, Phys. Biol. Sci. 2002, 78,
73–83; b) L. Bohé, D. Crich in Selective Glycosylations: Synthetic Methods and
Catalysts. (Ed.: C. S. Bennett), Wiley-VCH, Weinheim, 2017, pp. 115–133; c)
S. Hashimoto, Y. Yanagiya, T. Honda, H. Harada, S. Ikegami, Tetrahedron Lett.
1992, 33, 3523–3526; d) T. Tsuda, R. Arihara, S. Sato, M. Koshiba, S.
Nakamura, S. Hashimoto, Tetrahedron 2005, 61, 10719–10733; e) J. Mlynarski,
J. Ruiz-Caro, A. Fürstner, Chem. Eur. J. 2004, 10, 2214–2222; f) P. Peng, R. R.
Schmidt, J. Am. Chem. Soc. 2015, 137, 12653–12659.
[19] B. Yu, J. Sun, Chem. Commun. 2010, 46, 4668–4679.
[20] K. Takeuchi, T. Mukaiyama, Chem. Lett. 1998, 555–556.
[21] I. Krossing, I. Raabe, Angew. Chem. Int. Ed. 2004, 43, 2066–2090.
[22] a) K. Takeuchi, T. Tamura, T. Mukaiyama, Chem. Lett. 2000, 29, 124–125; b)
S. I. Tanaka, M. Takashina, H. Tokimoto, Y. Fujimoto, K. Tanaka, K. Fukase,
Synlett 2005, 2325–2328.
[23] See supporting information for the synthesis. The stereogenic center was
constructed either by Jacobsen´s hydrolytic kinetic resolution (HKR) or using
(+)-valine. a) M. Tokunaga, J. F. Larrow, F. Kakiuchi, E. N. Jacobsen, Science
1997, 277, 936–938; b) T. Taguri, M. Yamamoto, T. Fujii, Y. Muraki, T. Ando,
Eur. J. Org. Chem. 2013, 6924–6933.
[24] The glycosyl bromide used in the fidaxomicin synthesis was very unstable and
there was a mismatch in reactivity between glycosyl donor and accepter.
[25] a) E. J. C. Goldstein, F. Babakhani, D. M. Citron, Clin. Infect. Dis. 2012, 55,
143–148; b) E. J. Goldstein, D. M. Citron, K. L. Tyrrell, C. V. Merriam.
Antimicrob Agents Chemother. 2013, 57, 4872–4876.
[26] CLSI document M07-A9, Clinical Laboratory Standards Institute, Wayne, PA,
USA; 2012.
[27] Recent example of ‘inactive’ natural products: a) M. T. Villaume, E. Sella, G.
Saul, R. M. Borzilleri, J. Fargnoli, K. A. Johnston, H. Zhang, M. P. Fereshteh,
T. G. M. Dhar, P. S. Baran, ACS Cent. Sci. 2016, 2, 27–31; b) A. Cernijenko, R.
Risgaard, P. S. Baran, J. Am. Chem. Soc. 2016, 138, 9425–9428; c) M. Moreno,
W. A. M. Elgaher, J. Herrmann, N. Schläger, M. M. Hamed, S. Baumann, R.
Müller, R. W. Hartmann, A. Kirschning, Synlett 2015, 26, 1175–1178.
[28] H. H. Locher, P. Caspers, T. Bruyère, S. Schroeder, P. Pfaff, A. Knezevic, W.
Keck, D. Ritz. Antimicrob Agents Chemother. 2014, 58, 901–908.
This article is protected by copyright. All rights reserved.

10.1002/anie.201805770
Angewandte Chemie International Edition
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
COMMUNICATION
Nitrile assisted
β-Selective Glycosylation
RCM
HO
The macrocyclic natural product,
mangrolide A was prepared in
synthetic form, via a rapid
construction of the rare 18-
membered macrocycle and two b-
selective glycosylation reactions.
The synthesis of truncated
derivatives and a detailed
biological evaluation of the
synthetic natural product is also
presented.
Hiromu Hattori, Joel Roesslein,
Patrick Caspers, Katja Zerbe, Hideki
Miyatake-Ondozabal, Daniel Ritz,
Georg Rueedi and Karl Gademann*
OH
H
Me
H
Me
O
Me
O
MeO
HO
H
Me
β
H
O
H
H
O
O
O
Me2N
Me
MeO
β
HO
Me
Me
Me
Page No. – Page No.
Neighbouring Group Assisted
β-Selective Glycosylation
Suzuki Coupling
Total Synthesis and
Biological Evaluation of
the Glycosylated
Macrocyclic Antibiotic
Mangrolide A
Mangrolide A
This article is protected by copyright. All rights reserved.