A facile Fmoc solid phase synthesis strategy to access epimerization-prone biosynthetic intermediates of glycopeptide antibiotics

Article abstract of DOI:10.1021/ol500840f

A rapid protocol based on Fmoc-chemistry for the solid phase peptide synthesis of vancomycin- and teicoplanin-type peptides is described. Epimerization of highly racemization-prone arlyglycine derivatives is suppressed through optimized Fmoc-deprotection

Full text of DOI:10.1021/ol500840f

Letter
pubs.acs.org/OrgLett
A Facile Fmoc Solid Phase Synthesis Strategy To Access
Epimerization-Prone Biosynthetic Intermediates of Glycopeptide
Antibiotics
Clara Brieke and Max J. Cryle*
Department of Biomolecular Mechanisms, Max-Planck Institute for Medical Research, Jahnstrasse 29, 69120 Heidelberg, Germany
S
* Supporting Information
ABSTRACT: A rapid protocol based on Fmoc-chemistry for
the solid phase peptide synthesis of vancomycin- and
teicoplanin-type peptides is described. Epimerization of highly
racemization-prone arlyglycine derivatives is suppressed
through optimized Fmoc-deprotection and coupling con-
ditions. Starting from easily accessible Fmoc-protected amino
acids, this strategy enables the enantioselective synthesis of
peptides corresponding to intermediates found in vancomycin
and teicoplanin biosynthesis with excellent purity and in high
yields (38%−71%).
ancomycin and teicoplanin belong to the class of
glycopeptide antibiotics produced by various Streptomyces
V
species. They act by inhibiting cell-wall biosynthesis of many
Gram-positive bacteria and remain “last-resort” antibiotics
against Enterococci or methicillin-resistant Staphylococcus aureus
(MRSA) infections.1 As bacteria have begun to evolve
resistance to these compounds, there is a great need for
novel, highly active derivatives of these antibiotics; despite
impressive advances,2 this is hampered by the highly complex
total synthesis of these natural products.3 An alternate approach
therefore is to use semisynthetic or biotechnological
approaches, which can help to develop modified derivatives.4
The unique structural framework (Figure 1) of these
glycopeptides, which is crucial for their antibiotic activity,4a is
biosynthetically derived from a nonribosomally produced
heptapeptide precursor that is oxidatively cross-linked between
the aromatic amino acids by Cytochrome P450 enzymes
(Oxys):5 the order of these cross-links and the respective Oxys
have now been assigned for vancomycin6 as well as for
teicoplanin,7 and several examples have been structurally
characterized.8 As the chemical synthesis of these cross-links
remains as one of the great challenges in the synthesis of novel
glycopeptide antibiotic derivatives, there is significant interest in
understanding these enzymatic transformations in more detail,
with an outlook to use the enzymes as biocatalysts for obtaining
access to novel aglycone derivatives with the potential for
improved antibiotic activity against resistant bacteria.
Figure 1. Core structures of different glycopeptide antibiotics.
sensitivity to epimerization of the arylglycine derivatives 4-
hydroxyphenylglycine (Hpg) and especially 3,5-dihydroxy-
phenylglycine (Dpg) in the respective aglycones:10 these
residues have previously been thought to be incompatible
with the two standard solid phase peptide synthesis (SPPS)
The group of Robinson has shown that it is possible to
perform the first oxidative cross-linking reaction with OxyB to
form the vancomycin C−O−D ring in vitro by using precursor
hexa- and heptapeptides bound to peptidyl carrier protein
domains (PCP) of the vancomycin NRPS.9 However, this
approach is currently limited by difficulties in the chemical
synthesis of the linear peptide precursors, due to the high
Received: March 20, 2014
Published: April 14, 2014
© 2014 American Chemical Society
2454
dx.doi.org/10.1021/ol500840f | Org. Lett. 2014, 16, 2454−2457

Organic Letters
Letter
Figure 2. Optimization of Fmoc cleavage and coupling conditions for arylglycine derivatives: (A) General synthesis of tripeptides 2 and 3;
nonoptimized: (i) 20% piperidine/DMF, (ii) 4 equiv of Fmoc-D-Phg-OH 4, 4 equiv of HCTU, 8 equiv of NMM; optimized: (i) 1% DBU/DMF, (ii)
4 equiv of 4 or Fmoc-D-Hpg-OH 5, 4 equiv of COMU, 4 equiv of TEA. (B−D) LC-MS traces of crude synthesis mixtures of tripeptides 2 and 3
using nonoptimized (B) and optimized conditions (C and D). Total ion counts (TIC) and single ion counts of the respective peptides are shown.
Stereochemistry of rac-3 was assigned by synthesis of all four possible diastereomers (see SI).
strategies (Boc- and Fmoc-chemistry)11,12 and required Alloc-
chemistry.11,13 Disadvantages of Alloc-chemistry are the use of
a palladium-mediated cleavage protocol and the tedious
synthesis of all protected amino acid building blocks, while,
in contrast, Fmoc chemistry has the advantages of a large range
of available building blocks, simple synthesis of such starting
materials where this is not the case, and the possibility to
monitor the cleavage step during synthesis using UV-
absorption.
phenylglycine residues in less than 30 s with very little
racemization of the Cα-carbon.
For the exploration of a coupling reagent,17 alternative
uronium-type reagents and the phosphorus reagent DEPBT
were chosen (see Table S1, Supporting Information (SI)).
DEPBT has been reported in the liquid phase coupling of
arylglycine derivatives among others by Boger and co-
workers.2,12 From these reagents HATU18 and COMU19 in
combination with TEA quickly proved to be superior in
reducing epimerization during coupling, while a coupling time
of 30 min was optimal to maximize coupling efficiency. Using
more than 1 equiv of base to activator led to significant
epimerization (entries 5−10, Table S1). HATU afforded
slightly higher coupling yields compared to COMU (entry 11
vs 15), although the ratio between coupling efficiency and
epimerization proved to be best using COMU. Also, we did not
observe advantages of performing double coupling with half the
equivalents of activator and base (entries 13/14). Thus, 4 equiv
of COMU together with 4 equiv of TEA was chosen as an
optimized Fmoc-SPPS coupling mixture (Figure 2C; Table S1,
entry 15). Interestingly, during the course of investigation we
observed a reduced tendency to racemization when substituting
phenylglycine residues for 4-hydroxyphenylglycine resulting in
tripeptide 3 (Figure 2D), most likely due to a stabilizing
electron-donating mesomeric effect from the para-hydroxyl
groups.
Therefore, we sought to develop a streamlined and efficient
SPPS protocol for vancomycin- and teicoplanin-type bio-
synthetic precursor peptides using Fmoc-chemistry. For this a
series of conditions were explored to investigate the
susceptibility of phenylglycine residues to racemization, initially
using the model tripeptide NH2-Phg-Phg-Tyr-OH 2, synthe-
sized from preloaded Fmoc-Tyr-Wang resin 1. An initial
experiment, performed using standard SPPS conditions20%
piperidine for N-deprotection and 0.4 M NMM for amino acid
activationafforded an equimolar ratio of all four possible
diastereomers, resulting from racemization of the phenylglycine
residues (Figure 2B). To overcome the effect of piperidine
different conditions for Fmoc-removal were tested, focusing on
alternative reagents reported to be compatible with the solid
phase synthesis of base-labile thioesters, such as TBAF14 or
DBU mixtures15,16 (for the full range of conditions tested; see
Table S1, SI). Experiments using DBU gave the best results.
Use of HOBt as an additive16 was equally effective as DBU
alone; however, it had the disadvantage of not allowing UV-
monitoring of the reaction due to the absorption of the DBU/
HOBt solution. The addition of a secondary amine to scavenge
the diphenylfulvene product of Fmoc deprotection was tested
and had deleterious effects on the extent of racemization. Thus,
a 1% solution of DBU in DMF was chosen for deprotection of
the N-Fmoc group, which enables the effective deprotection of
These optimized conditions (Figure 2) were then applied to
the SPPS of vancomycin-type hexapeptide 6 (Figure 3), starting
from 1 using commercially available building blocks and Fmoc-
D-Hpg-OH 5, prepared in one step from the free amino acid
(SI). Following cleavage of 6 from the resin using a TFA/TIS/
H2O (95:2.5:2.5) mixture, analytical HPLC-MS analysis
revealed one major product of correct mass (Figures 3 and
S9): thus, essentially no side reactions or epimerization had
2455
dx.doi.org/10.1021/ol500840f | Org. Lett. 2014, 16, 2454−2457

Organic Letters
Letter
Figure 3. HPLC trace of crude product from the synthesis of
vancomycin-type peptide 6.
occurred during SPPS, global deprotection, and cleavage from
the solid phase. Following HPLC-purification, hexapeptide 6
could be isolated in an overall yield of 71%, with the structure
1
confirmed by MS and by the assignment of the H and 13C
NMR spectra using 2D NMR experiments (COSY, ROESY,
HSQC, HMBC). It should be emphasized that, with this
optimized protocol for Fmoc-chemistry, not only the synthetic
effort for the synthesis of vancomycin-type peptides is reduced
but also, compared to a previously reported optimized protocol
using Alloc-chemistry, where the coupling for each amino acid
was performed overnight,11 the speed of synthesis has also been
increased tremendously.
Figure 4. HPLC traces of crude synthesis mixtures for teicoplanin-type
peptides 7 (A) and 10 (B) after cleavage from the solid phase. In the
synthesis of 10, incomplete coupling of 8 resulted in truncated peptide
11; epimerized side product of 10 is highlighted.
In the next step, we focused on teicoplanin-type peptides.
Teicoplanin is often a more potent antibiotic compared to
vancomycin,1 and it is structurally more complex. Additionally
(and to the best of our knowledge) there have been no in vitro
investigations of the respective Cytochromes P450 to date,
likely due to the difficulties in preparing the required peptide
substrates. From the synthetic point of view, the SPPS of
teicoplanin-type peptides is yet more demanding than
vancomycin-type peptides for two reasons: first, three Hpg
residues have to be incorporated into the peptide as opposed to
two Hpg residues for vancomycin-type peptides, and second, at
position 3 of the peptide the highly racemization-prone amino
acid 3,5-dihydroxyphenylglycine must be incorporated. As
enantiomerically pure Dpg is very expensive to purchase or
requires extensive synthesis, we first focused on the synthesis of
peptide 7 having substituted Dpg with Phg at position 3 of the
peptide (Figure 4). Using the same synthesis protocol as that
for peptide 6, teicoplanin-type hexapeptide 7 was obtained in
high purity (Figures 4A and S11, SI) and could be isolated in
61% yield after HPLC purification.
Table 1. Yields and Epimerization Rates for Model Peptides
2 and 9 Using Different Activating Bases
a
tri- vs dipeptide
correct diastereomer
b
b
peptide
base
TEA
equiv
[%]
[%]
9
2
2
2
2
9
4
4
4
4
2
4
75
94
92
94
64
88
57
90
90
94
96
96
TEA
TMP
2,6-DMP
DABCO
2,6-DMP
a
Fmoc cleavage: 1% DBU in DMF, 0.5 min. Coupling: COMU (4
b
equiv), 30 min, in DMF. Yields were determined by HPLC analysis.
As initial SPPS trials under the optimized conditions revealed
a low coupling efficiency and high tendency for epimerization
for Fmoc-Dpg-OH 8, we next focused on a systematic
comparison of different activating bases to alleviate this
problem (see Table 1 and Figure S7). Thus, the less basic
and sterically hindered bases 2,4,6-trimethylpyridine (TMP,
pKa = 7.4320),21 2,6-dimethylpyridine (2,6-DMP, pKa = 6.6020),
and the dibasic compound DABCO (pKa = 8.920) were chosen
for the synthesis of model tripeptides NH2-L-Phg-L-Dpg-L-Tyr-
OH 9 and 2. Best results for reducing epimerization and
enhancing coupling efficiency were obtained with the least basic
2,6-DMP, yielding 96% of the correct enantiomer of Dpg-
containing peptide 9.
Thus, for the SPPS of teicoplanin-type peptide 10 with Dpg
at position 3, the coupling of Fmoc-L-Dpg-OH 8 was
performed using the specific Dpg-conditions with 2,6-DMP
as the base. Despite a significantly reduced coupling efficiency
of 8, which resulted in a 36% yield of truncated peptide NH2-D-
Hpg-D-Tyr-D-Hpg-D-Hpg-L-Tyr-OH 11, only minimal amounts
of epimerization were detected. By performing double coupling
with 3 equiv of 8, the amount of truncated peptide 11 could be
reduced to 22% (Figure 4B and Figure S13) resulting in an
overall yield of 38% of peptide 10 following HPLC purification.
In contrast to double coupling, no reduction of truncation was
2456
dx.doi.org/10.1021/ol500840f | Org. Lett. 2014, 16, 2454−2457

Organic Letters
Letter
(5) (a) Cryle, M. J. Metallomics 2011, 3, 323. (b) Cryle, M. J.; Brieke,
C.; Haslinger, K. In Amino Acids, Peptides and Proteins; Farkas, E.,
Ryadnov, M., Eds.; Royal Society of Chemistry: Cambridge, 2014; Vol.
38, pp 1−36.
observed by increasing the excess of 8. Thus, it is likely that
truncation could be reduced yet further, if desired, by executing
multiple coupling cycles with 8 as the combination of COMU
together with 2,6-DMP, for the coupling shows almost
complete suppression of epimerization of Dpg.
The use of unprotected amino acid building blocks raised the
question of acylation of such residues on capping and/or N-
terminal modification with acetylating reagents (Figure S8, SI).
While partial acylation of Hpg residues was observed, this did
not interfere with the peptide synthesis when acylation was
used for capping and could be selectively removed postsyn-
thesis without inducing racemization of the peptide using a
NaHCO3/H2O/MeOH solution (Figure S8, SI). Thus, both
capping and N-terminal acylation can be incorporated in the
peptide synthesis route if desired.
In conclusion, we have developed a protocol enabling the
SPPS assembly of vancomycin- or teicoplanin-type peptides
possessing multiple epimerization-prone arylglycine derivatives
using Fmoc-chemistry. This approach is significantly simplified
over previously reported methods11 and shows a reduction in
time of synthesis. Thus, this protocol facilitates the preparation
of peptide substrates that we currently are using for the
investigation of the later stages of glycopeptide biosynthesis. As
arylglycine derivatives are widespread components of peptidic
natural products, we anticipate that this approach will also aid
in the exploration of other biosynthetic pathways, such as those
of β-lactam antibiotics22 or other peptide antibiotics.12,23
(6) (a) Bischoff, D.; Pelzer, S.; Holtzel, A.; Nicholson, G. J.; Stockert,
S.; Wohlleben, W.; Jung, G.; Sussmuth, R. D. Angew. Chem., Int. Ed.
̈
2001, 40, 1693. (b) Bischoff, D.; Pelzer, S.; Bister, B.; Nicholson, G. J.;
Stockert, S.; Schirle, M.; Wohlleben, W.; Jung, G.; Sussmuth, R. D.
̈
Angew. Chem., Int. Ed. 2001, 40, 4688.
(7) Hadatsch, B.; Butz, D.; Schmiederer, T.; Steudle, J.; Wohlleben,
W.; Sussmuth, R.; Stegmann, E. Chem. Biol. 2007, 14, 1078.
̈
(8) (a) Zerbe, K.; Pylypenko, O.; Vitali, F.; Zhang, W.; Rouset, S.;
Heck, M.; Vrijbloed, J. W.; Bischoff, D.; Bister, B.; Sussmuth, R. D.;
̈
Pelzer, S.; Wohlleben, W.; Robinson, J. A.; Schlichting, I. J. Biol. Chem.
2002, 277, 47476. (b) Pylypenko, O.; Vitali, F.; Zerbe, K.; Robinson, J.
A.; Schlichting, I. J. Biol. Chem. 2003, 278, 46727. (c) Cryle, M. J.;
Staaden, J.; Schlichting, I. Arch. Biochem. Biophys. 2011, 507, 163.
(9) (a) Zerbe, K.; Woithe, K.; Li, D. B.; Vitali, F.; Bigler, L.;
Robinson, J. A. Angew. Chem., Int. Ed. 2004, 43, 6709. (b) Woithe, K.;
Geib, N.; Zerbe, K.; Li, D. B.; Heck, M.; Fournier-Rousset, S.; Meyer,
O.; Vitali, F.; Matoba, N.; Abou-Hadeed, K.; Robinson, J. A. J. Am.
Chem. Soc. 2007, 129, 6887. (c) Schmartz, P. C.; Wolfel, K.; Zerbe, K.;
̈
Gad, E.; El Tamany, E. S.; Ibrahim, H. K.; Abou-Hadeed, K.;
Robinson, J. A. Angew. Chem., Int. Ed. 2012, 51, 11468.
(10) Smith, G. G.; Sivakua, T. J. Org. Chem. 1983, 48, 627.
(11) Li, D. B.; Robinson, J. A. Org. Biomol. Chem. 2005, 3, 1233.
(12) Dettner, F.; Hanchen, A.; Schols, D.; Toti, L.; Nusser, A.;
̈
Sussmuth, R. D. Angew. Chem., Int. Ed. 2009, 48, 1856.
̈
(13) Freund, E.; Robinson, J. A. Chem. Commun. 1999, 2509.
(14) Ueki, M.; Amemiya, M. Tetrahedron Lett. 1987, 28, 6617.
(15) Wade, J. D.; Bedford, J.; Sheppard, R. C.; Tregear, G. W. Pept.
Res. 1991, 4, 194.
ASSOCIATED CONTENT
* Supporting Information
■
(16) Clippingdale, A. B.; Barrow, C. J.; Wade, J. D. J. Pept. Sci. 2000,
6, 225.
(17) El-Faham, A.; Albericio, F. Chem. Rev. 2011, 111, 6557.
(18) Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397.
(19) El-Faham, A.; Funosas, R. S.; Prohens, R.; Albericio, F. Chem.
Eur. J. 2009, 15, 9404.
S
Experimental procedure, HPLC-MS chromatograms, and
characterization data. This material is available free of charge
via the Internet at http://pubs.acs.org.
(20) (a) International Union of Pure and Applied Chemistry;
Commission on Electroanalytical Chemistry; Perrin, D. D. Dissociation
constants of organic bases in aqueous solution; Butterworths: London,
AUTHOR INFORMATION
Corresponding Author
■
1965. (b) Benoit, R. L.; Lefebvre, D.; Frec
1987, 65, 996.
(21) Carpino, L. A.; Ionescu, D.; El-Faham, A. J. Org. Chem. 1996, 61,
́
hette, M. Can. J. Chem.
*E-mail: Max.Cryle@mpimf-heidelberg.mpg.de.
Notes
The authors declare no competing financial interest.
2460.
(22) Kelly, W. L.; Townsend, C. A. J. Am. Chem. Soc. 2002, 124,
8186.
(23) Liu, W. T.; Kersten, R. D.; Yang, Y. L.; Moore, B. S.; Dorrestein,
P. C. J. Am. Chem. Soc. 2011, 133, 18010.
ACKNOWLEDGMENTS
■
The authors would like to acknowledge the Deutsche
Forschungsgemeinschaft for an Emmy-Noether Fellowship
(CR 392/1-1, M.J.C.) and the Max Planck Society for financial
support. M.J.C. is also grateful for the constant support of I.
Schlichting (MPI-Hd).
REFERENCES
■
(1) Williams, D. H.; Bardsley, B. Angew. Chem., Int. Ed. 1999, 38,
1172.
(2) Xie, J.; Okano, A.; Pierce, J. G.; James, R. C.; Stamm, S.; Crane,
C. M.; Boger, D. L. J. Am. Chem. Soc. 2012, 134, 1284.
(3) Boger, D. L. Med. Res. Rev. 2001, 21, 356.
(4) (a) Nicolaou, K. C.; Boddy, C. N. C.; Brase, S.; Winssinger, N.
̈
Angew. Chem., Int. Ed. 1999, 38, 2096. (b) Wolter, F.; Schoof, S.;
Sussmuth, R. D. In Glycopeptides and Glycoproteins; Wittmann, V., Ed.;
̈
Topics in Current Chemistry; Springer: Berlin, Heidelberg, 2007; pp
143−185. (c) Crane, C. M.; Boger, D. L. J. Med. Chem. 2009, 52, 1471.
(d) Fowler, B. S.; Laemmerhold, K. M.; Miller, S. J. J. Am. Chem. Soc.
2012, 134, 9755. (e) Pathak, T. P.; Miller, S. J. J. Am. Chem. Soc. 2012,
134, 6120. (f) Pathak, T. P.; Miller, S. J. J. Am. Chem. Soc. 2013, 135,
8415.
2457
dx.doi.org/10.1021/ol500840f | Org. Lett. 2014, 16, 2454−2457