Solid-phase total synthesis of daptomycin and analogs

Article abstract of DOI:10.1021/acs.orglett.5b00043

An entirely solid-phase synthesis of daptomycin, a cyclic lipodepsipeptide antibiotic currently in clinical use, was achieved using a combination of ?±-azido and Fmoc amino acids. This methodology was applied to the synthesis of several daptomycin analogs, one of which did not contain kynurenine or the synthetically challenging amino acid (2S,3R)-methylglutamate yet exhibited an MIC approaching that of daptomycin.

Full text of DOI:10.1021/acs.orglett.5b00043

Letter
pubs.acs.org/OrgLett
Solid-Phase Total Synthesis of Daptomycin and Analogs
Chuda Raj Lohani, Robert Taylor, Michael Palmer, and Scott D. Taylor*
Department of Chemistry, University of Waterloo, 200 University Avenue West, Waterloo, Ontario, Canada, N2L 3G1
S
* Supporting Information
ABSTRACT: An entirely solid-phase synthesis of daptomycin, a
cyclic lipodepsipeptide antibiotic currently in clinical use, was
achieved using a combination of α-azido and Fmoc amino acids.
This methodology was applied to the synthesis of several
daptomycin analogs, one of which did not contain kynurenine or
the synthetically challenging amino acid (2S,3R)-methylglutamate
yet exhibited an MIC approaching that of daptomycin.
aptomycin, a natural product isolated from fermentations
of Streptomyces roseosporus, is the first approved member of
resistant strains,5 emphasizes the need for establishing detailed
structure−activity relationships (SARs). This requires a method-
ology for the ready preparation of numerous daptomycin analogs
and the interrogation of most or all residues. Although
daptomycin analogs have been prepared using chemoenzymatic6
and combinatorial biosynthetic approaches,7 these methods have
produced less than two dozen daptomycin analogs since they
were first described almost a decade ago. One approach that has
the potential to produce large numbers of daptomycin analogs is
total chemical synthesis using solid-phase methodologies;
however, the chemical synthesis of daptomycin and its analogs
has proven to be very challenging.8 Lam et al. recently reported
the total synthesis of daptomycin.9 They were unable to form the
crucial ester linkage using FmocKyn(CHO,Boc)OH as a
building block either in solution or on a solid support. Hence
they resorted to a 12-step solution-phase synthesis of a branched
tetrapeptide, during which formation of the ester linkage
between Thr4 and Kyn13 was accomplished by ozonolyis of a
suitably protected Trp residue. This tetrapeptide was used as a
building block for the preparation of a linear daptomycin
precursor using solid-phase methods. A solution-phase serine
ligation procedure was used to achieve macrocyclization. This
combination of solid- and solution-phase chemistry, along with
the need for HPLC purification immediately before and after
cyclization, makes this approach very labor-intensive. Very
recently, Martin et al. reported the synthesis of two daptomycin
analogs and their enantiomers.10 Their initial attempts at
performing the entire synthesis on the solid phase were
unsuccessful. Therefore, they developed a combined solid- and
solution-phase approach which involved the initial preparation of
linear precursors using solid-phase chemistry, followed by
cleavage of the peptides from the support, cyclization in solution,
and then side chain deprotection. To avoid the difficulties that
were encountered by Lam et al. in forming the ester linkage,
Martin opted to substitute Thr4 with L-diaminopropionic acid,
thus replacing the ester linkage with an amide linkage.
D
a novel class of antibiotics known as cyclic lipopeptides.1a−d
Daptomycin was approved in the USA in 2003, and in Europe in
2006, for the treatment of complicated skin and skin structure
infections caused by Gram-positive bacteria.1b,2 In 2006 it was
approved in the USA for the treatment of bacteremia and right-
sided endocarditis caused by Staphylococcus aureus including
methicillin resistant S. aureus (MRSA).2 It is not effective against
community-acquired pneumonia due to inhibition by pulmonary
surfactant.3
Daptomycin is a branched, cyclic lipodepsipeptide consisting
of a 10-amino acid macrolactone ring to which is attached an
exocyclic tripeptide bearing a decanoyl lipid tail (Figure 1). It
Figure 1. Structure of daptomycin. DecTrp = N-decanoyltryptophan.
contains six nonproteinogenic amino acids: D-Asn2, Orn6, D-
Ala8, D-Ser11, (2S,3R)-methylglutamate (MeGlu12), and
kynurenine (Kyn13). The depsi (ester) bond in the macro-
lactone ring is between the side chain of Thr4 and the α-COOH
of Kyn13.
Although daptomycin was first discovered in the early 1980s,
fundamental aspects of its mechanism of action (MOA) remain
unknown. Its activity is Ca2+-dependent. Some studies suggest
that daptomycin forms oligomeric pores in bacterial membranes,
which results in dissipation of membrane potential and cell
death.4a,b However, this MOA has yet to be firmly established,
and other MOAs have been proposed.1d
Daptomycin’s unresolved MOA, its lack of activity in the
presence of lung surfactant and the appearance of daptomycin-
Received: January 6, 2015
Published: January 29, 2015
© 2015 American Chemical Society
748
DOI: 10.1021/acs.orglett.5b00043
Org. Lett. 2015, 17, 748−751

Organic Letters
Letter
Unfortunately, this substitution was very detrimental to
biological activity.
Scheme 2. Attempted Esterification Reactions with Model
Peptides
The development of a solid-phase approach to the synthesis of
daptomycin and it analogs would greatly facilitate SAR studies,
which would in turn expedite MOA investigations and the search
for cyclic lipodepsipeptide antibiotics with improved activity and
decreased bacterial resistance. Here we report a solid-phase
approach to the total synthesis of daptomycin. We also apply this
methodology to the synthesis of several daptomycin analogs.
In our initial approach to daptomycin (Scheme 1), dipeptide 1,
attached to the resin by the side chain of the Asp9 residue, would
Scheme 1. First Attempted Synthesis of Daptomycin
the side chain of Thr4 was not necessary so long as DIC/HOBT
were used as coupling agents and DMF was used as the solvent.
We were unable to obtain any esterification product (peptide 10)
using 9-mer 7 as the substrate under a wide variety of conditions.
However, when the decanoyl tail in 7 was replaced with a Boc
group (peptide 8), 40% of peptide 7 was converted into branched
peptide 11 using DIC/cat. DMAP as coupling agents and
dichloromethane (DCM) as the solvent, as determined by
HPLC. When these conditions were applied to heptapeptide 9,
almost quantitative conversion into peptide 12 was achieved with
no evidence of epimerization. The use of DCM as solvent was
crucial, as the use of other solvents, such as DMF, resulted in
significantly reduced yields.11 The reason for the lack of ester
bond formation in peptide 7 is not clear. It is possible that the
decanoyl tail is interacting with the hydrophobic resin and
making the hydroxyl group on the Thr residue less accessible.
The above findings prompted us to redesign our approach to
daptomycin which involved using a combination of α-azido12
and α-Fmoc-protected amino acids (Scheme 3). Octapeptide 13
was prepared using standard SPPS techniques, employing Fmoc
amino acids for residues ten to four and N3Asp(tBu)OH for
residue three. Coupling of FmocKyn(CHO,Boc)OH to Thr4 in
peptide 13 to give peptide 14 was accomplished using DIC/cat.
DMAP in DCM. Peptide 14 was converted into peptide 4 by
installing N3Asn(Trt) and then N3Trp(Boc) and using PMe3 in
be prepared and then converted into decapeptide 2 using
standard Fmoc solid-phase peptide synthesis (SPPS). After
removal of the Thr4 side chain protecting group in peptide 2 to
give peptide 3, the ester linkage would be established with Kyn13
to give peptide 4. Residues 12 and 11 would be added, and the
allyl group in Gly10 removed to give peptide 5. An on-resin
cyclization between Ser11 and Gly10 would enable us to avoid
racemization and minimize the formation of oligomers during
this procedure. Simultaneous resin cleavage and side-chain
deprotection of the resulting peptide 6 would give daptomycin.
The synthesis of resin-bound peptide 3 proved to be
straightforward using standard SPPS methodology (diisopropyl
carbodiimide (DIC)/HOBt as coupling agents). However,
attempts to form the ester bond between FmocKyn(CHO,-
Boc)OH and Thr4 in peptide 3 using a wide variety of coupling
conditions (see the Supporting Information (SI)) were
unsuccessful. This result was consistent with the report by Lam
et al. that this ester bond is very challenging to form on a solid
support or even in solution.9
Scheme 3. Synthesis of Daptomycin
Rather than turning to a solution-phase approach, we
examined whether the decanoyl tail or peptide length had any
effect on the esterification reaction. Three peptides, 7−9, starting
from Asp9, were prepared (Scheme 2) and used as model
substrates for the on-resin esterification reaction. During the
synthesis of these model peptides, it was found that protection of
749
DOI: 10.1021/acs.orglett.5b00043
Org. Lett. 2015, 17, 748−751

Organic Letters
Letter
aq dioxane for reduction of the α-azido groups to α-amino
groups.12b The decanoyl tail was installed using decanoic acid
and DIC/HOBt. It has been shown that the ester bonds in
depsipeptides can be unstable toward piperidine, the most
common reagent used for Fmoc removal.13 We had previously
reported that S−O bond cleavage in sulfate-protected sulfotyr-
osine peptides by piperidine can be avoided by using 2-
methylpiperidine (2-MP) for Fmoc removal.14 Hence, the Fmoc
group in peptide 4 and subsequent peptides was removed using
2-MP as a precautionary measure against ester bond cleavage.
The allyl group on Gly10 was removed using cat. Pd(PPh3)4 and
dimethylbarbituric acid (DMBA). Attempts to cyclize peptide 5
using PyBOP/DIPEA in DMF or PyBOP/HOBt/DIPEA in
DMF gave only very small amounts of cyclic peptide 6. Using
PyBOP/HOAt/DIPEA in DMF, the cyclization efficiency was
about 30%. The cyclization efficiency was increased to 40% using
PyBOP/HOAt/DIPEA and DMF/DCM/NMP (3:2:2 (v/v))
containing 1% Triton X-100 (Magic mixture15) as the solvent.
Although this cyclization efficiency is modest, attempts by Lam et
al. to perform a similar macrocyclization in solution during
daptomycin synthesis using HATU/DIPEA/DMF or DEPBT/
DIPEA/DMF gave only a trace amount of cyclized product.9
Simultaneous cleavage of peptide 6 from the support and side
chain deprotection using TFA/thioanisole/triisopropylsilane
(TIS)/H2O (88:2:5:5) gave daptomycin in a 9% overall yield.
The synthetic daptomycin was identical to authentic daptomycin
Marahiel et al. reported that a daptomycin analog bearing L-
Asn at position 2, Glu at position 12, and Trp at position 13
(Dap-L-N2/E12/W13) was 33-fold less potent than daptomy-
cin.6b At the time their results were reported, it was not known
that daptomycin contains D-Asn rather than L-Asn at position 2.16
Consequently, these workers interpreted the reduced activity of
the Dap-L-N2/E12/W13 analog as being due to only the Glu12/
Trp13 double substitution when, in fact, the presence of L-Asn
may also have contributed to this loss of activity, as it has been
shown that daptomycin-L-N2 is 10-fold less active than
daptomycin.16 To determine what effect a Glu12/Trp13 double
substitution has on activity, we prepared Dap-E12/W13 (16,
3.7% yield). The MIC of peptide 16 was almost the same as the
MIC of daptomycin when assayed against B. subtilis 1046 and
only 3-fold greater when assayed against B. subtillis PY79 (Table
1). These results indicate that the significant reduction in activity
of the triply substituted analog reported by Marahiel was mainly
due to the presence of the L-Asn residue. We also prepared a
Glu12/Tyr13 analog (Dap-E12/Y13, peptide 17, 16% yield).
The MIC of peptide 17 was over 35-fold greater than the MIC of
daptomycin (5 mM CaCl2, Table 1).
The MICs of daptomycin and peptides 15 and 17 were also
determined at high CaCl2 concentrations against B. subtillis 1046.
Surprisingly, at 25 and 100 mM CaCl2 the MIC of 17 decreased
dramatically and was only 6- and 2.5-fold greater than the MIC of
daptomycin respectively (Table 1). The MIC of peptide 15 also
decreased significantly at high Ca2+ concentrations but not to the
same extent as peptide 17. In contrast to peptides 15 and 17, the
MIC of daptomycin decreased only marginally over this range of
CaCl2 (Table 1). These findings indicate that residues 12 and 13
play a role in determining the calcium affinity of daptomycin.
To summarize, we have developed a solid-phase approach to
the synthesis of daptomycin and daptomycin analogs. This opens
the door for the rapid synthesis of other daptomycin analogs to
establish SARs and for the solid-phase synthesis of other
branched, cyclic lipodepsipeptide antibiotics. We have shown
that the methyl group in the MeGlu residue in daptomycin must
be of the R-configuration for potent activity. The low potencies
exhibited by peptides 15 and 17 under standard conditions
could, to a large extent, be remedied by supplementation of Ca2+
at high concentrations suggesting that these modifications
mainly affect calcium affinity but have little effect on other
aspects of the MOA. Finally, we have shown that a Glu12/Trp13
analog (peptide 16) has an MIC approaching that of daptomycin.
This result is significant, as it reveals that highly active
daptomycin analogs that lack both Kyn and 3-MeGlu can be
obtained. The synthesis of other daptomycin analogs and the
preparation of combinatorial libraries based on Dap-E12W13 are
in progress.
1
in every respect (mass spectrum, H NMR spectrum, HPLC
retention time; see the SI) including biological activity (MIC of
0.75 μg/mL against Bacillus subtilis 1046 and B. subtilis PY79
under commonly used conditions of 5 mM CaCl2).
The approach to daptomycin outlined in Scheme 3 was
applied to the synthesis of three daptomycin analogs (15−17,
Table 1). The efficiency of the cyclization step during the
Table 1. MICs of Daptomycin and Daptomycin Analogs
15−17
MIC (μg/mL)
peptide
daptomycin
B. subtilus 1046
bc
B. subtilus PY79
a
a
,
0.75, 0.5
0.75
a
b
c
a
15 (Dap-(2S,3S)-MeGlu)
16 (Dap-E12/W13)
40, 5.0, 5.0
>40
a
a
1.0
3.0
a
b
c
17 (Dap-E12/Y13)
35, 3.0, 1.3
ND
a
b
c
5 mM CaCl2. 25 mM CaCl2. 100 mM CaCl2
synthesis of these analogs varied. Peptides 15 and 16 exhibited
cyclization efficiencies slightly lower or greater than that
experienced during the synthesis of daptomycin (35−45%),
while, for peptide 17, the cyclization efficiency was >95% even
when DIC/HOBt was used to promote the cyclization.
Since the synthesis of enantiomerically pure (2S,3R)-Fmoc-3-
MeGlu(tBu)OH is a labor-intensive process (see the SI), it
would be advantageous if this residue could be replaced with an
amino acid that is more readily accessible. However, replacing
this residue with Glu results in an 11-fold loss of activity.7a Since
the synthesis of (2S,3S)-Fmoc-3-MeGlu(tBu)OH is slightly less
labor-intensive than the synthesis of the 2S,3R isomer (see the
SI), daptomycin analog 15 was prepared in which (2S,3R)-
MeGlu was replaced with (2S,3S)-MeGlu (Dap-(2S,3S)-MeGlu)
(7.9% yield). However, analog 15 was found to be over 40-fold
less active than daptomycin (5 mM CaCl2, Table 1), revealing
that the stereochemistry of the methyl group has a significant
effect on biological activity.
ASSOCIATED CONTENT
■
S
* Supporting Information
Synthetic procedures for (2S,3R)-Fmoc-3-MeGlu(tBu)OH,
(2S,3S)-Fmoc-3-MeGlu(tBu)OH, daptomycin, and daptomycin
analogs and characterization data for these compounds. This
material is available free of charge via the Internet at http://pubs.
acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: s5taylor@uwaterloo.ca.
750
DOI: 10.1021/acs.orglett.5b00043
Org. Lett. 2015, 17, 748−751

Organic Letters
Letter
(15) Thakkar, A.; Trinh, T. B.; Pei, D. ACS Comb. Sci. 2013, 15, 120−
Notes
129.
The authors declare no competing financial interest.
(16) Miao, V.; Coeffet-LeGal, M.-F.; Brian, P.; Brost, R.; Penn, J.;
Whiting, A.; Martin, S.; Ford, R.; Parr, I.; Bouchard, M.; Silva, C. J.;
Wrigley, S. K.; Baltz, R. H. Microbiology 2005, 151, 1507−1523.
ACKNOWLEDGMENTS
■
This research was supported by a Collaborative Health Research
Projects Grant from the Natural Sciences and Engineering
Research Council (NSERC) of Canada and the Canadian
Institutes for Health Research (CHIR) to SDT and MP and a
Discovery Grant from NSERC to SDT.
REFERENCES
■
(1) For recent reviews, see: (a) Robbel, L.; Marahiel, M. A. J. Biol.
Chem. 2010, 285, 27501−27508. (b) Vilhena, C.; Bettencourt, A. Mini-
Rev. Med. Chem. 2012, 12, 202−209. (c) Bionda, N.; Pitteloud, J. P.;
Cudic, P. Future Med. Chem. 2013, 5, 1311−1330. (d) Baltz, R. H. Curr.
Opin. Chem. Biol. 2009, 13, 144−151.
(2) (a) Arbeit, R. D.; Maki, D.; Tally, F. P.; Campanaro, E.; Eisenstein,
B. I. Clin. Infect. Dis. 2004, 38, 1673−1681. (b) Fowler, V. G.; Boucher,
H. W.; Corey, G. R.; Abrutyn, E.; Karchmer, A. W.; Rupp, M. E.; Levine,
D. P.; Chambers, H. F.; Tally, F. P.; Vigliani, G. A.; Cabell, C. H.; Link, A.
S.; DeMeyer, I.; Filler, S. G.; Zervos, M.; Cook, P.; Parsonnet, J.;
Bernstein, J. M.; Price, C. S.; Forrest, G. N.; Fatkenheuer; Gareca, G. M.;
̈
Rehm, S. J.; Brodt, H. R.; Tice, A.; Cosgrove, S. E. New Engl. J. Med.
2006, 355, 653−665.
(3) Silverman, J. A.; Mortin, L. I.; VanPraagh, A. D. G.; Li, T.; Alder, J. J.
Infect. Dis. 2005, 191, 2149−2152.
(4) (a) Silverman, J. A.; Perlmutter, N. G.; Shapiro, H. M. Antimicrob.
Agents Chemother. 2003, 47, 2538−2544. (b) Zhang, T.; Muraih, J. K.;
Tishbi, N.; Herskowitz, J.; Victor, R. L.; Silverman, J.; Uwumarenogie,
S.; Taylor, S. D.; Palmer, M.; Mintzer, E. J. Biol. Chem. 2014, 289,
11584−11591.
(5) Bayer, A. S.; Schneider, T.; Sahl, H.-G. Ann. N.Y. Acad. Sci. 2013,
1277, 139−158.
(6) (a) Kopp, F.; Grunewald, J.; Mahlert, C.; Marahiel, M. A.
Biochemistry 2006, 45, 10474−10481. (b) Grunewald, J.; Sieber, S. A.;
Mahlert, C.; Linne, U.; Marahiel, M. A. J. Am. Chem. Soc. 2004, 126,
17025−17031.
(7) (a) Nguyen, K. T.; Ritz, D.; Gu, J.; Alexander, D.; Chu, M.; Miao,
V.; Brian, P. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 17462−17467.
(b) Nguyen, K. T.; He, X.; Alexander, D. C.; Li, C.; Gu, J.-Q.; Mascio;
Van Praagh, C. A.; Mortin, L.; Chu, M.; Silverman, J. A.; Brian, P.; Baltz,
R. H. Antimicrob. Agents Chemother. 2010, 54, 1404−1413.
(8) The first total chemical synthesis of daptomycin was reported in the
patent literature by Cubist Pharmaceuticals Inc. and involved a labor-
intensive segment-coupling approach. Details such as resin loadings and
% yields were not provided. See: WO Patent WO2006110185, 2007.
(9) Lam, H. Y.; Zhang, Y.; Liu, H.; Xu, J.; Wong, C. T. T.; Xu, C.; Li, X.
J. Am. Chem. Soc. 2013, 135, 6272−6279.
(10) Hart, P.; Kleijn, L. H. J.; de Bruin, G.; Oppedijk, S. F.; Kemminka,
J.; Martin, N. I. Org. Biomol. Chem. 2014, 12, 913−918.
(11) Stawikowski, M.; Cudic, P. Tetrahedron Lett. 2006, 47, 8587−
8590.
(12) α-Azido amino acids are easily and economically prepared in good
yield in a single step from the corresponding α-NH2-, α-COOH-
unprotected amino acids by diazo transfer methodology and often do
not require any chromatographic purification. The α-azido group is
atom economical and readily converted into the α-amino group using
phosphines or DTT/DIPEA. See: (a) Meldal, M.; Juliano, M. A.;
Janason, A. M. Tetrahedron Lett. 1997, 38, 2531−2534. (b) Lundquist, J.
T.; Pelletier, J. C. Org. Lett. 2001, 3, 781−783. (c) Valverde, I. E.;
Lecaille, F.; Lalmanach, G.; Aucagne, V.; Delmas, A. F. Angew. Chem., Int.
Ed. 2012, 51, 718−722. (d) Goddard-Borger, E. D.; Stick, R. V. Org. Lett.
2007, 9, 3797−3800.
(13) Spengler, J.; Koksch, B.; Albericio, F. Peptide Science 2007, 88,
823−828.
(14) Ali, A. M.; Taylor, S. D. Angew. Chem., Int. Ed. 2009, 48, 2024−
2026; et al. Angew. Chem. 2009, 121, 2058−2060.
751
DOI: 10.1021/acs.orglett.5b00043
Org. Lett. 2015, 17, 748−751