Synthesis of Norfijimycin A with Activity against Mycobacterium tuberculosis

Article abstract of DOI:10.1071/CH16559

The total synthesis of norfijimycin A, a simplified analogue of the marine natural product fijimycin A, is described. Fijimycin A is a cyclic depsipeptide that has been shown to possess activity against methicillin-resistant Staphylococcus aureus. The natural product contains a rare N,β-dimethyl leucine unit with unknown stereochemistry at the β-carbon. To evaluate the importance of the β-methyl group for antimicrobial activity, we introduced N-methyl leucine into the natural product scaffold. The resulting norfijimycin A was shown to possess significant activity against Mycobacterium tuberculosis, the etiological agent of tuberculosis.

Full text of DOI:10.1071/CH16559

RESEARCH FRONT
CSIRO PUBLISHING
Aust. J. Chem.
Communication
http://dx.doi.org/10.1071/CH16559
Synthesis of Norfijimycin A with Activity against
Mycobacterium tuberculosis
A
B
B
Alexander Stoye, Gayathri Nagalingam, Warwick J. Britton,
,
A C
and Richard J. Payne
A
School of Chemistry, The University of Sydney, Sydney, NSW 2006, Australia.
Tuberculosis Research Program, Centenary Institute, and Sydney Medical School,
The University of Sydney, Sydney, NSW 2006, Australia.
B
C
Corresponding author. Email: richard.payne@sydney.edu.au
The total synthesis of norfijimycin A, a simplified analogue of the marine natural product fijimycin A, is described.
Fijimycin A is a cyclic depsipeptide that has been shown to possess activity against methicillin-resistant Staphylococcus
aureus. The natural product contains a rare N,b-dimethyl leucine unit with unknown stereochemistry at the b-carbon. To
evaluate the importance of the b-methyl group for antimicrobial activity, we introduced N-methyl leucine into the natural
product scaffold. The resulting norfijimycin A was shown to possess significant activity against Mycobacterium
tuberculosis, the etiological agent of tuberculosis.
Manuscript received: 2 October 2016.
Manuscript accepted: 15 November 2016.
Published online: 9 December 2016.
Antimicrobial resistance has the potential to lead to a devas-
tating global health problem if new antimicrobials with novel
widespread resistance of Mtb to currently available drugs has led
to significant effort in discovering new therapeutic leads for
drug-resistant TB therapy. Natural products have been a rich
source of molecules for drug discovery, accounting for 34 % of
all Food and Drug Administration (FDA)-approved drugs in
1981–2010, and, as such, it is not surprising that the research
community continues to investigate natural products and ana-
[
1]
mechanisms of action are not developed. For example,
methicillin-resistant Staphylococcus aureus (MRSA) is an
enormous problem in hospitals and is responsible for increasing
[
2]
numbers of deaths over the past decade. Tuberculosis (TB) is
caused by the infection with the bacterium Mycobacterium
tuberculosis (Mtb), and according to the World Health Orga-
nization, was responsible for 1.5 million deaths and the emer-
[4]
logues thereof as new antibiotic drug leads.
Fijimycin A (1, Fig. 1) was isolated in 2011 from the
fermentation broth of strain CNS-575, a Streptomyces sp.
cultured from a marine sediment sample collected off Nasese,
[2a]
gence of 9.6 million new cases in 2014. The frontline drugs
that are currently used to treat TB are rifampicin, isoniazid,
ethambutol, and pyrazinamide. Although these therapies are
effective against drug-sensitive TB, they cannot effectively treat
multi-drug resistant (MDR) and extensively drug-resistant
[5]
Fiji. It was isolated together with its congeners fijimycin B,
[
6]
fijimycin C, and the diastereomer etamycin A (2). The latter
compound (also called virifogrisein I) had been shown (before
the fijimycin isolation report) to exhibit considerable activity
[2b,3]
(XDR) TB and require the use of second-line drugs.
The
HO
HO
OH
O
OH
O
O
O
N
O
N
H
Me
H
Me
N
N
N
N
HN
HN
O
HN
O
O
HN
O
O
O
O
O
O
H
H
N
N
O
Me HN
N
Me
Me HN
N
Me
∗
O
O
O
1: (R): Fijimycin A
3: Norfijimycin A
2: (S): Etamycin A
Fig. 1. Structures of fijimycin A (1), etamycin A (2), and the simplified target analogue of this study
norfijimycin A (3).
Journal compilation � CSIRO 2016
www.publish.csiro.au/journals/ajc

B
A. Stoye et al.
[
7]
against Gram-positive bacteria as well as Mtb. The cyclodep-
sipeptide fijimycin A (1) contains several unusual amino acid
residues as part of its structure, namely sarcosine, D-allo-
hydroxyproline, D-N-methylphenylglycine (D-a-phenylsarco-
sine), 3-hydroxypicolinic acid, and N,b-dimethyl leucine.
Though the absolute stereochemistry of most residues within
the natural product was determined from the isolated material,
the stereochemistry at the b-centre of N,b-dimethyl leucine was
not elucidated. To date, a single total synthesis of etamycin A
From here, Fmoc-Ala-OH could be coupled to the resin-bound
peptide, followed by Fmoc deprotection under standard condi-
tions. Unfortunately, under these conditions, we observed
exclusively a de-esterified resin-bound peptide product, pre-
sumably due to diketopiperazine formation caused by nucleo-
philic attack from the nitrogen atom onto the carboxyl group of
D-a-phenylsarcosine. Although disconnection B was initially
deemed less attractive, because of the potential risk of epimer-
ization during macrolactamization, we were able to synthesize
the linear peptide on resin implementing the on-resin esterifica-
tion as the last coupling step. Unfortunately, epimerization at the
a-phenylsarcosine during the on-resin esterification was again
observed and could not be prevented despite screening several
conditions. As a consequence, we synthesized a building block
(4, Scheme 1) containing a preformed ester linkage, and separa-
tion of the diastereomers followed.
[
9]
[
8]
has been reported; however, to our knowledge there have not
yet been any reports of analogues of the natural product family.
Herein, we report the total synthesis of norfijimycin A (3) using
solid-phase peptidesynthesis (SPPS). In this study, N,b-dimethyl
leucine was substituted with N-methyl leucine in order to
evaluate the importance of the b-methyl group for antimicrobial
activity, specifically against virulent Mtb.
Our initial strategy (disconnection A, Fig. 2) consisted of the
attachment of sarcosine to 2-chlorotrityl chloride (2-CTC) resin,
which after SPPS and cleavage from the resin could be used as
the C-terminal cyclization point in the final macrolactamization.
The clear advantage of this approach was the inhibition of
epimerization during ring closure due to the absence of a
stereogenic centre in sarcosine. A key step in this approach was
an on-resin esterification to form the depsipeptide ester linkage
between threonine and D-a-phenylsarcosine. Unfortunately, this
process led to complete epimerization of a-phenylsarcosine.
The synthesis of 4 began from L-threonine (5) which was
allyloxycarbonyl (Alloc)-protected followed by the formation
of the corresponding tert-butyl ester to afford 6 in 72 % yield
over the two steps (Scheme 1; see Supplementary Material for
synthesis details). Esterification with (R)-Fmoc-a-phenylsarco-
0
sine (7) using N-(3-dimethylaminopropyl)-N -ethylcarbodii-
mide hydrochloride (EDCꢀHCl) and catalytic amounts of
4-(dimethylamino)pyridine (DMAP) provided 8 as a mixture
of diastereomers (66 : 34 R : S at the a-centre of the a-phenyl-
sarcosine unit). Cleavage of the tert-butyl ester then furnished
orthogonally protected ester 4 together with diastereomer 9
(
dr 66 : 34 4 : 9), which were separated by reversed-phase HPLC
HO
to afford the pure diastereomers in excellent yield (60 % for 4
and 33 % for 9).
OH
O
O
N
Me
N
With diastereomerically pure building block 4 in hand,
Fmoc-Ala-OH was next loaded onto 2-CTC resin followed by
iterative assembly of resin-bound peptide 10 via Fmoc-strategy
SPPS. Extension of this peptide through incorporation of the
synthetic building block 4 then provided resin-bound 11
(Scheme 2; see Supplementary Material for synthesis details).
N
H
N
HN
O
HN
O
O
A
O
O
H
Me
Me HN
N
Me
O
Me
Treatment of 11 with Pd(PPh ) and phenylsilane successfully
3 4
O
removed the Alloc-protecting group from the threonine
residue to enable branching of the peptide chain. From here
O-allyl-protected 3-hydroxypicolinic acid 12 was coupled
to resin-bound peptide 11 using (7-azabenzotriazol-1-yloxy)
tripyrrolidinophosphonium hexafluorophosphate (PyAOP)
B
Fig. 2. Synthetic disconnections towards norfijimycin A (3).
EDC⋅HCl
DCM, DMAP (cat.)
3
h, 0ЊC
Me
N
O
1
. Allyl chloroformate,
sat. aq. NaHCO3/THF
. CuCl, DIC, t-BuOH
OH
O
OH
O
O
O
Fmoc
OH
2
7
t
t
Ph
quant.
BuO
AllocHN
O
OH
O Bu
72 % over 2 steps
N
8
NH2
NHAlloc
Me
Fmoc
5
6
dr 66:34 (R:S)
1
1
2
. TFA/DCM (1:1, v/v)
5 min, rt.
. prep. HPLC
O
O
O
HO
AllocHN
N
Me
Fmoc
4
9
: (R): 60 %
: (S): 33 %
0
Scheme 1. Synthesis of the ester building block 4. DIC, N,N -diisopropylcarbodiimide; cat., catalytic; quant., quantitative; prep.
HPLC, preparative HPLC; rt, room temperature.

Synthesis of Norfijimycin A
C
and N-methylmorpholine (NMM). Initially, we tried to remove
the Fmoc-protection group from D-a-phenylsarcosine,
followed by cleavage with 30 vol-% hexafluoroisopropanol
1-hydroxy-7-azabenzotriazole (HOAt) and NMM in a dilute
(5mM) solution of DMF/DCM (1: 1, v/v). After 24h, the starting
material was consumed and two diastereomers of the cyclic
peptide (epi-3 and 3) were formed in approximately equal
amounts and were isolated in 60% combined yield (33% epi-3,
27 % 3). It is presumed that in this case, the epimerization was
caused by the unwanted formation of an oxazoline formed by
intramolecular nucleophilic attack of the oxygen from N-methyl
leucine onto the activated C-terminus of alanine. Liquid chroma-
tography–mass spectrometry (LC/MS) analysis of several ali-
quots taken from the reaction mixture at several time points
(
HFIP) in CH Cl , yielding the side-chain-protected peptide.
2 2
Unfortunately, this led to epimerization of the a-centre of
a-phenylsarcosine. As a result, we instead decided to remove
the O-allyl-protecting group first, followed by full cleavage of
the peptide from the solid support with the N-terminal protec-
tion still intact. The epimerization-free removal of the Fmoc-
protecting group of the D-a-phenylsarcosine N-terminus was
then achieved by palladium-catalyzed hydrogenation, thereby
avoiding the base-promoted epimerization with piperidine and
indicated that one diastereomer (t ¼ 4.83 min) corresponded to
R
[
10]
providing crude depsipeptide 13.
epi-3, whereas the other diastereomer (t ¼ 5.05min) was norfi-
R
Following reversed-phase HPLC depsipeptide 13 was
afforded in 41 % yield (based on the initial resin loading) as a
single diastereoisomer. The purified linear peptide 13 was next
subjected to a solution-phase macrolactamization using O-(7-
jimycin A (3) (see Supplementary Material for details).
Having successfully prepared target natural product ana-
logue 3 and its epimer (epi-3), we next assessed their activities
against the virulent H37Rv strain of Mtb using a resazurin-
[11]
based assay. Compound epi-3 exhibited moderate activity
against Mtb with a minimum inhibitory concentration required
0
0
azabenzotriazol-1-yl)-N,N,N ,N -tetramethyluroniumhexafluoro-
phosphate (HATU) as a coupling reagent in combination with
O
O
O
O
O
HO
O
O
N
Fmoc-SPPS
NHAlloc
N
N
Me
N
Fmoc
N
Me
4
Me
H
H
AllocHN HN
H
N
H N
2
O
O
PyAOP, HOAt, NMM, DMF
O
O
Cl
-CTC
O
O
Me
O
N
2
O
HN
Me
HN
Me
O
Me
O
O
N
Fmoc
O
O
10
11
1
. Pd(Ph ) , PhSiH , CH Cl
3
4
3
2
2
HO
2.
OAll
OH
O
12
PyAOP, NMM
O
HN
N
Me
H
N
CO H
N
N
2
HATU, HOAt, DIPEA,
DMF/DCM (1:1), 5 mM
HN
H
O
O
3. Pd(Ph ) , PhSiH , CH Cl
3 4 3 2 2
O
O
4. TFA/TIS/H O (90:5:5)
24 h, rt
2
N
Me
5. Pd/C, H , MeOH, 12 h, rt
2
Me
O
HN
Me
NH
O
O
Me
OH
·
CF CO H
3
2
4
1 % after HPLC
13
(based on initial
resin loading)
HO
HO
OH
O
OH
O
H
H
O
HN
N
O
HN
N
Me
Me
N
N
N
N
HN
O
O
HN
H
O
O
O
O
O
O
H
N
ϩ
N
O
Me HN
N
Me
O
Me HN
N
Me
O
O
O
O
epi-Norfijimycin A
epi-3)
3 % isolated
Norfijimycin A (3)
7 % isolated
dr 1:1 (HPLC analysis of reaction mixture)
0 % combined yield
(
3
2
6
Scheme 2. Synthesis of norfijimycin A (3) and epi-norfijimycin A (epi-3). DIPEA, N,N-diisopropylethylamine; TIS, triisopropylsilane; MeOH,
methanol.

D
A. Stoye et al.
120
110
100
epi-3
Scripps Institute of Oceanography, University of California San Diego, for
providing NMR data for authentic fijimycin A.
3
90
80
70
60
50
40
30
20
10
0
References
[
1] J. M. A. Blair, M. A. Webber, A. J. Baylay, D. O. Ogbolu, L. J. V.
Piddock, Nat. Rev. Microbiol. 2015, 13, 42. doi:10.1038/
NRMICRO3380
[
2] (a) World Health Organization, Global Tuberculosis Report 2015
(
(
WHO Press: Geneva).
b) K. Dheda, T. Gumbo, N. R. Gandhi, M. Murray, G. Theron,
Z. Udwadia, G. B. Migliori, R. Warren, Lancet Respir. Med. 2014, 2,
21. doi:10.1016/S2213-2600(14)70031-1
c) N. R. Gandhi, P. Nunn, K. Dheda, H. S. Schaaf, M. Zignol,
3
0.1
1
10
100
1000
(
Concentration of drugs [μM]
D. van Soolingen, P. Jensen, J. Bayona, Lancet 2010, 375, 1830.
doi:10.1016/S0140-6736(10)60410-2
3] A. Koul, E. Arnoult, N. Lounis, J. Guillemont, K. Andries, Nature
Fig. 3. Inhibition curves of epi-3 and 3 against Mtb.
[
[
[
2
011, 469, 483. doi:10.1038/NATURE09657
4] A. L. Harvey, R. Edrada-Ebel, R. J. Quinn, Nat. Rev. Drug Discovery
015, 14, 111. doi:10.1038/NRD4510
2
to inhibit the growth of 50 % of organisms (MIC ) of 70 mM,
whereas norfijimycin A (3) possessed potent activity against
Mtb with an MIC50 of 5 mM (Fig. 3).
5
0
5] P. Sun, K. N. Maloney, S.-J. Nam, N. M. Haste, R. Raju,
W. Aalbersberg, P. R. Jensen, V. Nizet, M. E. Hensler, W. Fenical,
Bioorg. Med. Chem. 2011, 19, 6557. doi:10.1016/J.BMC.2011.06.053
6] (a) N. M. Haste, V. R. Perera, K. N. Maloney, D. N. Tran, P. Jensen,
W. Fenical, V. Nizet, M. E. Hensler, J. Antibiot. 2010, 63, 219.
doi:10.1038/JA.2010.22
In summary, we have developed a solid-phase-based syn-
thetic approach to access norfijimycin A (3) – a simplified
analogue of fijimycin A (1). This analogue exhibited significant
activity against Mtb and suggests that N-methyl leucine could
serve as a replacement for the rare and more synthetically
challenging N,b-dimethyl leucine residue in future analogue
studies. In addition, the synthetic route described here may
provide a viable route to the fijimycin and etamycin class of
cyclodepsipeptide natural products in the future.
[
(b) J. C. Sheehan, H. G. Zachau, W. B. Lawson, J. Am. Chem. Soc.
1958, 80, 3349. doi:10.1021/JA01546A039
(
c) R. B. Arnold, A. W. Johnson, A. B. Mauger, J. Chem. Soc. 1958,
466. doi:10.1039/JR9580004466
d) J. C. Sheehan, H. G. Zachau, W. B. Lawson, J. Am. Chem. Soc.
957, 79, 3933. doi:10.1021/JA01571A095
4
(
1
[
[
[
7] B. Heinemann, A. Gourevitch, J. Lein, D. L. Johnson, M. A. Kaplan,
D. Vanas, I. R. Hooper, Antibiot. Annu. 1955, 2, 728.
8] J. C. Sheehan, S. L. Ledis, J. Am. Chem. Soc. 1973, 95, 875.
doi:10.1021/JA00784A041
9] (a) M. C. Lai, E. M. Topp, J. Pharm. Sci. 1999, 88, 489. doi:10.1021/
JS980374E
Supplementary Material
13
H and C NMR spectra, analytical HPLC patterns, and high-
resolution mass spectra of all novel compounds are available on
the Journal’s website.
1
(
b) C. Goolcharran, R. T. Borchardt, J. Pharm. Sci. 1998, 87, 283.
doi:10.1021/JS970325M
10] T. Maegawa, Y. Fujiwara, T. Ikawa, H. Hisashi, Y. Monguchi,
Acknowledgements
[
The authors thank Dr Ian Luck and Dr Nick Proschogo for technical support
with NMR and mass spectrometry analyses, respectively, and Andrew M.
Giltrap for helpful discussions. We also thank Professor William Fenical,
H. Sajiki, Amino Acids 2009, 36, 493. doi:10.1007/S00726-008-0109-7
[11] N. K. Taneja, J. S. Tyagi, J. Antimicrob. Chemother. 2007, 60, 288.
doi:10.1093/JAC/DKM207