Total Synthesis of Teixobactin

Article abstract of DOI:10.1021/acs.orglett.6b01324

The first total synthesis of the cyclic depsipeptide natural product teixobactin is described. Synthesis was achieved by solid-phase peptide synthesis, incorporating the unusual l-allo-enduracididine as a suitably protected synthetic cassette and employing a key on-resin esterification and solution-phase macrolactamization. The synthetic natural product was shown to possess potent antibacterial activity against a range of Gram-positive pathogenic bacteria, including a virulent strain of Mycobacterium tuberculosis and methicillin-resistant Staphylococcus aureus (MRSA).

Full text of DOI:10.1021/acs.orglett.6b01324

Letter
pubs.acs.org/OrgLett
Total Synthesis of Teixobactin
Andrew M. Giltrap,†,⊥ Luke J. Dowman,†,⊥ Gayathri Nagalingam,‡ Jessica L. Ochoa,§
Roger G. Linington,§,∥ Warwick J. Britton,‡ and Richard J. Payne*,†
‡
†School of Chemistry and Tuberculosis Research Program, Centenary Institute, and Sydney Medical School The University of
Sydney, Sydney, NSW 2006, Australia
§Department of Chemistry and Biochemistry, University of California, Santa Cruz, California 95064, United States
∥Department of Chemistry, Simon Fraser University, Burnaby, British Columbia BC V5A 1S6, Canada
S
* Supporting Information
ABSTRACT: The first total synthesis of the cyclic depsipeptide natural product teixobactin is described. Synthesis was achieved
by solid-phase peptide synthesis, incorporating the unusual L-allo-enduracididine as a suitably protected synthetic cassette and
employing a key on-resin esterification and solution-phase macrolactamization. The synthetic natural product was shown to
possess potent antibacterial activity against a range of Gram-positive pathogenic bacteria, including a virulent strain of
Mycobacterium tuberculosis and methicillin-resistant Staphylococcus aureus (MRSA).
he emergence of drug resistant strains of pathogenic
bacteria has compromised the effectiveness of a growing
T
number of clinically employed antibiotics.1 Mycobacterium
tuberculosis (Mtb), the etiological agent of tuberculosis (TB), is
an example of a pathogen to which widespread resistance to
frontline antibiotic treatments has developed.2−4 Mtb is
estimated to latently infect one-third of the global population,
and of the 9.6 million new cases of TB in 2014, a significant
proportion were infected with drug-resistant strains of Mtb, thus
complicating treatment and compromising global efforts to
eradicate the disease.4 Unfortunately, the growing burden of
antibiotic resistance is coupled with decreased effort in the
development of new antibiotics.5 Indeed, of the antibiotics
currently in clinical trials, the majority are variations on current
drug architectures, e.g., rifapentine and delamanid for TB.6,7 It is
well established that nature provides a rich source of diverse
molecules with privileged antibacterial activity, highlighted by
the fact that numerous clinically approved antibiotics are natural
products or derivatives thereof.8,9 However, very few genuine
antibiotic leads have been discovered from natural sources over
the past two decades.
Figure 1. Structure of teixobactin (1).
drug-resistant clinical isolates of Mtb (MIC = 0.125 μg mL−1)
and methicillin-resistant Staphylococcus aureus (MRSA, MIC =
0.25 μg mL−1). Structurally, teixobactin is an undecadepsipeptide
with a cyclized C-terminus and a methylated N-terminus. The
natural product possesses four D-amino acids, the unusual amino
acid L-allo-enduracididine, and has some structural similarities to
other natural products, e.g. hypeptin and mannopeptimycin.11,12
Teixobactin was shown to exhibit its antibacterial action by
binding to lipid II and lipid III, two key enzymatic substrates in
the biosynthesis of peptidoglycan and teichoic acid, respec-
tively.10,13 Ling et al. attempted to generate teixobactin-resistant
In early 2015, Ling et al.10 reported the isolation and
characterization of a novel peptidic natural product called
teixobactin (1, Figure 1) from a previously “uncultivable” soil
bacterium Eleftheria terrae. Teixobactin was discovered using
iChip, a new technology that enabled the bacterium to be
cultured for the production of sufficient material for isolation and
structural and functional characterization. The natural product
was shown to exhibit potent antibacterial activity against a wide
range of Gram-positive bacteria including virulent (H37Rv) and
Received: May 6, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.6b01324
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
mutants in S. aureus and Mtb by treating the organisms with
sublethal doses of the natural product; however, no resistant
mutants could be generated.10 This striking result is thought to
be due to teixobactin binding to multiple enzymatic substrates
rather than to an enzyme; ultimately, gaining resistance by
mutating the substrate for an enzyme is inherently more difficult
for an organism than mutating amino acids within an enzyme.
This mechanism of action coupled with the potent antibacterial
activity against a range of clinically relevant pathogens has made
teixobactin a realistic antibiotic candidate for Gram-positive and
Mtb infections and an attractive target for total synthesis.
While no total synthesis of teixobactin has been reported, an
efficient synthesis of the unnatural amino acid L-allo-
enduracididine was published by Craig et al.14 In addition, Jad
et al.15 and Parmar et al.16 both reported the synthesis of an
analogue of the natural product in which the synthetically
challenging L-allo-enduracididine residue was replaced by a
simplified L-arginine residue. A second analogue was also
reported by Parmar et al.,16 whereby the D-configured amino
acids, except threonine, were replaced by L-configured residues.
Both analogues exhibited less potent inhibition of S. aureus,
revealing the importance of both the D-configured amino acids
and the L-allo-enduracididine residue for antibacterial activity of
the natural product. In this study, we sought to develop a robust
synthetic route to teixobactin (1) which would be efficient,
amenable to rapid analogue generation, and would ultimately
facilitate thorough profiling of the antibiotic activity.
the two steps. Triflation of the γ-alcohol in 6 under basic
conditions then afforded cyclic guanidine 7 in good yield. Finally,
acidic deprotection of the α-amine and α-carboxylate followed by
Fmoc protection furnished the target building block 8 [Fmoc-
End(Cbz)2-OH] in 57% yield over two steps.
With the suitably protected L-allo-enduracididine building
block 8 in hand, we next began assembly of the depsipeptide
chain of teixobactin. It was envisaged that the depsipeptidic core
of the natural product could be assembled using Fmoc-SPPS
including a key on-resin esterification step. It was proposed that,
following cleavage of the resin under weakly acidic conditions, a
solution-phase cyclization followed by global side-chain
deprotection (including of the Cbz protection on the
enduracididine residue) under strongly acidic conditions would
afford the natural product.
Initial efforts involved the loading of Fmoc-D-Thr-OH (with
an unprotected side chain) to 2-chlorotrityl chloride (2-CTC)
functionalized polystyrene resin followed by coupling of Fmoc-L-
Ser(tBu)-OH. At this point, the key on-resin esterification step
with protected L-Ile was attempted using a number of
esterification conditions. Surprisingly, we could not find an
effective set of conditions to facilitate this on-resin trans-
formation, with starting material remaining even after multiple
treatments. We reasoned that the steric bulk of the 2-CTC linker
adjacent to the side chain hydroxyl of the D-Thr side chain was
impeding esterification, and as such, we sought a less sterically
encumbered resin linker as an alternative. Toward this end, we
selected (4-(hydroxymethyl)-3-methoxyphenoxy)acetic acid
(HMPB) functionalized polyethylene glycol-based NovaPEG
resin, taking advantage of the decreased steric bulk surrounding
the loaded amino acid as well as the increased swelling properties
provided by the polyethylene glycol-based support.19 Fmoc-D-
Thr(TES)-OH (9) was loaded to the resin via the symmetrical
anhydride (generated by treatment with N,N′-diisopropylcarbo-
diimide (DIC) and 4-(dimethylamino)pyridine (DMAP) to
afford 10 (Scheme 2). Next, the Fmoc group was removed via
treatment with piperidine in DMF, followed by removal of the
triethylsilyl (TES) protecting group by double treatment with
acetic acid buffered tetrabutylammonium fluoride (TBAF).
Fmoc L-Ser(tBu)-OH was next coupled using (benzotriazol-1-
yloxy)tripyrrolidinophosphonium hexafluorophosphate
(PyBOP) as the coupling reagent and 4-methylmorpholine
(NMM) as the base in DMF to afford 11. The key on-resin
esterification to the D-Thr side chain was then attempted with
Alloc-L-Ile-OH using DIC and catalytic DMAP as the
esterification conditions. Gratifyingly, complete esterification
was achieved in one 16 h treatment to provide depsipeptide 12 as
judged by LC−MS analysis. Having successfully branched the
peptide chain, the remaining linear portion of the target peptide
was extended using conventional Fmoc-SPPS, including
incorporation of the L- and D-configured amino acids within
the natural product and the N-terminal N-methyl-Boc-D-Phe-
OH to afford resin-bound 13.
We began by preparing a suitably protected L-allo-
enduracididine building block that could be installed directly
into Fmoc-SPPS. The early steps in our synthesis took
inspiration from Rudolph et al.17 Specifically, nitromethane
addition to the free carboxylate side chain of Boc-L-Asp-OtBu (2)
provided nitroketone 3 (Scheme 1). Stereoselective reduction of
Scheme 1. Synthesis of Suitably Protected L-allo-
Enduracididine Building Block 8
Following the successful assembly of resin-bound 13, the
synthesis continued on the branched L-Ile residue (Scheme 3).
Specifically, on-resin Alloc deprotection was achieved by
treatment with Pd(PPh3)4 and PhSiH3. The next step involved
coupling of the suitably protected L-allo-enduracididine building
block Fmoc-End(Cbz)2-OH 8, which was smoothly effected
through the use of 1-[bis(dimethylamino)methylene]-1H-1,2,3-
triazolo[4,5-b]pyridinium 3-oxide hexafluorophosphate
(HATU) as the coupling reagent in combination with 1-
the ketone in 3 with L-Selectride provided a diastereomeric
mixture (dr: 5:1 (2S,4R):(2S,4S)) which was readily separable by
flash column chromatography to afford alcohol 4 as a single
diastereoisomer in 52% yield over two steps. From here,
hydrogenation of the δ-nitro moiety,18 followed by guanidiny-
lation of the resulting amine with a bis-Cbz-protected variant of
Goodman’s guanidinylating reagent 5, gave 6 in 72% yield over
hydroxy-7-azabenzotriazole (HOAt) as an additive and Hunig’s
̈
B
DOI: 10.1021/acs.orglett.6b01324
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 2. Synthesis of Resin-Bound Depsipeptide
Teixobactin Precursor 13
Scheme 3. Synthesis of Teixobactin (1)
base to afford resin-bound 14. At this point, 14 was subjected to
conventional Fmoc deprotection conditions (10 vol % piperidine
in DMF, 2 × 3 min). Unfortunately, this treatment led to the
formation of a de-esterified resin-bound peptide, presumably due
to unwanted diketopiperazine formation caused by the
nucleophilic cyclization of the α-amine of the deprotected L-
allo-enduracididine residue onto the α-carboxyl of L-isoleu-
cine.20,21 To overcome diketopiperazine formation, the resin was
treated with 10 vol % of piperidine in DMF for a shorter duration
(30 s) and washed rapidly with DMF and DCM before
immediately treating the resin with a preactivated solution of
Fmoc-L-Ala-OH, PyBOP, and NMM in DMF. This led to the
formation of resin-bound 15 with minimal diketopiperazine
formation. The Fmoc protecting group from the coupled L-Ala
residue was next removed under standard conditions followed by
cleavage of the linear side-chain-protected depsipeptide 16 from
the resin using 1% TFA in CH2Cl2.
With depsipeptide 16 in hand, and without purification, we
next performed the key macrolactamization step by treating 16
with 4-(4,6-dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpholi-
nium tetrafluoroborate (DMTMM·BF ) and Hunig’s base in a
̈
4
dilute (10 mM) solution of DMF, which provided the cyclized
side-chain protected depsipeptide after 16 h. All that remained
for the completion of the target was removal of the side-chain
protecting groups and purification. Using deprotection con-
ditions reported by Koide et al.22 for the 4-methoxybenzyl
protecting group, and later adopted by Hondal et al.,23 we were
able to remove all the protecting groups, including the Cbz
moieties, in one step using a mixture of 70:12:10:8 v/v/v/v TFA,
trifluoromethanesulfonic acid (TfOH), thioanisole, and m-
cresol. Subsequent purification by RP-HPLC and lyophylization
yielded teixobactin as a TFA salt. This was then lyophilized
multiple times in the presence of 5 mM HCl24 to yield
teixobactin as the bis-HCl salt in 3.3% yield (over 24 steps from
original resin loaded amino acid 10, average of 87% per step).
This conversion to the bis-HCl salt was carried out to enable
direct comparison with the isolated natural product, which was
characterized in this salt form. Gratifyingly, all spectroscopic data
were consistent with that reported for the isolated natural
product reported by Ling et al.10
We next assessed the antibacterial activity of 1 to further
characterize the synthetic natural product. Specifically, synthetic
teixobactin 1 was screened against the virulent H37Rv strain of
Mtb using a resazurin-based assay25 and against a range of Gram-
negative and Gram-positive strains including Bacillus subtilis 168,
C
DOI: 10.1021/acs.orglett.6b01324
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Staphylococcus aureus, methicillin-resistant S. aureus (MRSA),
Escherichia coli, Providencia alcalifaciens, Ochrobactrum anthropi,
Enterobacter aerogenes, Acinetobacter baumannii, Vibrio cholerae,
Salmonella typhimurium, Pseudomonas aeruginosa, and Yersinia
pseudotuberculosis using standard methods.26 The activity of
synthetic 1 against these organisms was consistent with that for
the natural product reported by Ling et al.,10 despite some
differences in the strains of the organisms used in this study
(Table 1). Specifically, 1 exhibited potent activity against Mtb
ACKNOWLEDGMENTS
■
We thank Dr. Ian Luck (The University of Sydney) for technical
support for NMR spectroscopy, Dr. Nick Proschogo (The
University of Sydney) for technical support for mass
spectrometry, and Jake Haeckl (Simon Fraser University) and
Walter Bray (University of California, Santa Cruz) for technical
support for antibacterial screening. We thank the Australian
Postgraduate Award for PhD funding (A.M.G. and L.J.D.), an
ARC Future Fellowship to R.J.P. (FT130100150), NSERC
Discovery support (R.G.L.), and a Cota-Robles fellowship to
J.L.O.
Table 1. Activity of Synthetic Teixobactin (1) against a Panel
a
of Pathogenic Bacteria
REFERENCES
■
teix MIC
vanc MIC
line MIC
cipro MIC
organism
(μM)
(μM)
(μM)
(μM)
(1) Blair, J. M. A.; Webber, M. A.; Baylay, A. J.; Ogbolu, D. O.; Piddock,
L. J. V. Nat. Rev. Microbiol. 2014, 13, 42.
S. aureus (MSSA)
S. aureus (MRSA)
E. coli
1.1
0.69
1.4
0.69
(2) Dheda, K.; Gumbo, T.; Gandhi, N. R.; Murray, M.; Theron, G.;
Udwadia, Z.; Migliori, G. B.; Warren, R. Lancet Respir. Med. 2014, 2, 321.
(3) Gandhi, N. R.; Nunn, P.; Dheda, K.; Schaaf, H. S.; Zignol, M.; van
Soolingen, D.; Jensen, P.; Bayona, J. Lancet 2010, 375, 1830.
(4) WHO. Global Tuberculosis Report 2015; WHO, 2015.
(5) Silver, L. L. Clin. Microbiol. Rev. 2011, 24, 71.
(6) Koul, A.; Arnoult, E.; Lounis, N.; Guillemont, J.; Andries, K. Nature
2011, 469, 483.
(7) Zumla, A. I.; Gillespie, S. H.; Hoelscher, M.; Philips, P. P. J.; Cole, S.
T.; Abubakar, I.; McHugh, T. D.; Schito, M.; Maeurer, M.; Nunn, A. J.
Lancet Infect. Dis. 2014, 14, 327.
(8) Harvey, A. L.; Edrada-Ebel, R.; Quinn, R. J. Nat. Rev. Drug Discovery
2015, 14, 111.
1.1
>27
0.21
>27
>27
>27
>27
>27
>27
>27
>27
0.87
>66
0.17
>66
>66
>66
>66
>66
>66
>66
>66
1.2
>66
0.22
>66
>66
>66
>66
>66
>66
>66
>66
>66
0.013
0.13
B. subtilis
P. alcalifaciens
O. anthropi
0.027
0.85
E. aerogenes
A. baumanii
V. cholerae
0.022
2.4
0.016
0.027
1.4
S. typhimurium
P. aeruginosa
Y.
0.0081
pseudotuberculosis
(9) Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2012, 75, 311.
(10) Ling, L. L.; Peoples, A. J.; Spoering, A. L.; Hughes, D. E.; Cohen,
D. R.; Felix, C. R.; Fetterman, K. A.; Millett, W. P.; Nitti, A. G.; Zullo, A.
M.; Schneider, T.; Engels, I.; Mueller, A.; Conlon, B. P.; Chen, C.; Lewis,
K.; Schaberle, T. F.; Epstein, S.; Jones, M.; Lazarides, L.; Steadman, V. A.
Nature 2015, 517, 455.
a
teix = teixobactin. Controls: vanc = vancomycin, line = linezolid,
cipro = ciprofloxacin. MIC against Mtb H37Rv = 1.5
Control: rifampicin MIC = 10.7 0.33 nM.
0.03 μM.
(11) Berlinck, R. G. S.; Kossuga, M. H. Nat. Prod. Rep. 2005, 22, 516.
with an MIC of 1.5 μM, S. aureus and MRSA (MIC = 1.1 μM),
and B. subtilis (MIC = 0.21 μM). Unsurprisingly, teixobactin 1
showed no activity up to the highest tested concentration against
the tested Gram-negative pathogens.
In summary, we have developed a solid-phase synthetic route
to access teixobactin (1), a potent antibacterial natural product.
The synthetic natural product was shown to possess potent
antibacterial activity against Mtb and a number of pathogenic
Gram-positive organisms. The work described here lays the
foundation for generating analogues of teixobactin with a view to
developing potential candidates for the treatment of TB and
Gram-positive infections.
(12) von Nussbaum, F.; Sussmuth, R. D. Angew. Chem., Int. Ed. 2015,
̈
54, 6684.
(13) Oppedijk, S. F.; Martin, N. I.; Breukink, E. Biochim. Biophys. Acta,
Biomembr. 2016, 1858, 947.
(14) Craig, W.; Chen, J.; Richardson, D.; Thorpe, R.; Yuan, Y. Org. Lett.
2015, 17, 4620.
(15) Jad, Y. E.; Acosta, G. A.; Naicker, T.; Ramtahal, M.; El-Faham, A.;
Govender, T.; Kruger, H. G.; de la Torre, B. G.; Albericio, F. Org. Lett.
2015, 17, 6182.
(16) Parmar, A.; Iyer, A.; Vincent, C. S.; Van Lysebetten, D.; Prior, S.
H.; Madder, A.; Taylor, E. J.; Singh, I. Chem. Commun. 2016, 52, 6060.
(17) Rudolph, J.; Hannig, F.; Theis, H.; Wischnat, R. Org. Lett. 2001, 3,
3153.
(18) Peoples, A. J.; Hughes, D.; Ling, L. L.; Millett, W.; Nitti, A.;
Spoering, A.; Steadman, V. A.; Chiva, J.-Y. C.; Lazarides, L.; Jones, M. K.;
Poullennec, K. G.; Lewis, K.; Epstein, S. WO/2013/US72838, 2013.
(19) García-Martín, F.; Quintanar-Audelo, M.; García-Ramos, Y.;
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.6b01324.
■
S
̂ ́
Cruz, L. J.; Gravel, C.; Furic, R.; Cote, S.; Tulla-Puche, J.; Albericio, F. J.
Comb. Chem. 2006, 8, 213.
(20) Goolcharran, C.; Borchardt, R. T. J. Pharm. Sci. 1998, 87, 283.
(21) Lai, M. C.; Topp, E. M. J. Pharm. Sci. 1999, 88, 489.
(22) Koide, T.; Itoh, H.; Otaka, A.; Yasui, H.; Kuroda, M.; Esaki, N.;
Soda, K.; Fujii, N. Chem. Pharm. Bull. 1993, 41, 502.
(23) Hondal, R. J.; Nilsson, B. L.; Raines, R. T. J. Am. Chem. Soc. 2001,
123, 5140.
(24) Andrushchenko, V. V.; Vogel, H. J.; Prenner, E. J. J. Pept. Sci. 2007,
13, 37.
(25) Taneja, N. K.; Tyagi, J. S. J. Antimicrob. Chemother. 2007, 60, 288.
(26) Wiegand, I.; Hilpert, K.; Hancock, R. E. W. Nat. Protoc. 2008, 3,
163.
Experimental procedures for synthesis, antibacterial
screening and characterization data, and NMR spectra of
teixobactin (1) and all novel intermediates (PDF)
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: richard.payne@sydney.edu.au.
Author Contributions
⊥A.M.G. and L.J.D. contributed equally to this work.
Notes
The authors declare no competing financial interest.
D
DOI: 10.1021/acs.orglett.6b01324
Org. Lett. XXXX, XXX, XXX−XXX