Novel 5′-deoxy nucleosyl amino acid scaffolds for the synthesis of muraymycin analogues

Article abstract of DOI:10.1039/c003092a

Naturally occurring nucleoside antibiotics such as muraymycins represent promising lead structures for the development of novel antibacterial agents. A concise synthesis of 5′-deoxy muraymycin derivatives has been developed. The key step was the highly stereoselective asymmetric hydrogenation of suitable didehydro amino acid precursors, providing unique nucleosyl amino acid structures.

Full text of DOI:10.1039/c003092a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/obc | Organic & Biomolecular Chemistry
Novel 5¢-deoxy nucleosyl amino acid scaffolds for the synthesis of muraymycin
analogues†
Anatol P. Spork and Christian Ducho*
Received 18th February 2010, Accepted 1st April 2010
First published as an Advance Article on the web 12th April 2010
DOI: 10.1039/c003092a
Naturally occurring nucleoside antibiotics such as mu-
raymycins represent promising lead structures for the de-
velopment of novel antibacterial agents. A concise synthesis
of 5¢-deoxy muraymycin derivatives has been developed. The
key step was the highly stereoselective asymmetric hydrogena-
tion of suitable didehydro amino acid precursors, providing
unique nucleosyl amino acid structures.
have been reported to act as inhibitors.7 These natural products
are distinctly characterised by the unique structures of their
nucleoside-derived moieties. Three subclasses of nucleoside an-
tibiotics, the liposidomycins, caprazamycins and muraymycins, as
well as the muraymycin-like FR-900493, have a common (5¢S,6¢S)-
uridine-derived core structure, characterised by C–C-bond linkage
of the 5¢-carbon atom of the uridine unit to a glycine moiety (C-6¢)
as well as aminoribosylation of the 5¢-hydroxy group.7
Structure–activity relationship (SAR) studies have been re-
ported for synthetic analogues of the caprazamycins,8 indicating
a crucial role of the aminoribosyl unit for the retention of
antibacterial activity towards Mycobacterium tuberculosis.8c In
case of the muraymycins, a comparison of the 19 isolated naturally
occurring derivatives provides different SAR insights: the activities
of aminoribosylated muraymycin A1 1 and non-aminoribosylated
muraymycin A5 2 against several bacteria including Staphylococ-
cus aureus do not differ dramatically (Fig. 1).9 In addition, (semi-)-
synthetic analogues of the muraymycins have been investigated.10
Remarkable results were obtained for the SAR of synthetic
truncated muraymycin derivatives (e.g. 3–7, Fig. 1): firstly, the
absolute configuration at the 5¢-position was required to be
(R) in order to obtain good activities, i.e. only 5¢-epi-analogues
with respect to the (5¢S)-configured natural products were active.
Secondly, the presence of some synthetic protecting groups (tert-
butyl ester, tert-butyldimethylsilyl (TBDMS) groups) turned out
to be a prerequisite for antibacterial potency.10b
Due to emerging resistances of bacteria towards established
antibiotics,1 the development of novel antibacterial agents has
become one of the main objectives of medicinal chemistry in
recent years. Ideally, drug candidates with antimicrobial potency
should display new or yet unexploited modes of action.2 This
is the case for the inhibition of the bacterial membrane protein
translocase I (MraY), a key enzyme in the early stages of
peptidoglycan biosynthesis,3,4 which has recently drawn attention
as a promising drug target.5
One valuable approach in drug development is to employ
natural products as lead structures.6 They can be of particular
importance whenever structural information about the target
protein is difficult to obtain, as it is often the case for membrane
proteins. For MraY, naturally occurring nucleoside antibiotics
Georg-August-University Go¨ttingen, Department of Chemistry, Institute of
Organic and Biomolecular Chemistry, Tammannstr. 2, 37077 Go¨ttingen,
Germany. E-mail: cducho@gwdg.de; Fax: +49 (0)551 39 9660; Tel: +49
(0)551 39 3285
† Electronic Supplementary Information (ESI) available: Experimental
procedures for all syntheses, 1H and 13C NMR spectra of all com-
pounds, details of the 1H-1H NOESY NMR experiment. See DOI:
10.1039/c003092a
In order to obtain novel structurally simplified nucleoside
scaffolds for the synthesis of muraymycin analogues, it was
envisaged to prepare 5¢-deoxy derivatives lacking the site for
aminoribosylation. In addition, 5¢-deoxy nucleosyl amino acid
Fig. 1 Naturally occurring nucleoside antibiotics muraymycin A1 1 and muraymycin A5 29 and synthetic truncated 5¢-epi-muraymycin analogues 3–7
displaying antibacterial activity10b (MIC = minimal inhibitory concentration, PMB = para-methoxybenzyl, TBDMS = tert-butyldimethylsilyl).
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 2323–2326 | 2323
©

View Article Online
structures should provide insights into the role of the 5¢-hydroxy
group in case of truncated muraymycin analogues such as 3–7
(Fig. 1). Our objective therefore was to synthesise target com-
pounds 8 and 9 with (6¢S)- and (6¢R)-configuration, respectively,
from a common precursor in a stereocontrolled way. Thus, the
influence of the configuration in 6¢-position can also be elucidated.
Both 8 and 9, which might also serve as suitably protected building
blocks for the synthesis of more complex muraymycin analogues,
were converted into their corresponding derivatives 10 and 11,
respectively, which represent 5¢-deoxy analogues of 3 for SAR
studies on truncated muraymycin derivatives (Fig. 2).
obtained from uridine in an overall yield of 66% over 4 steps.11
Wittig–Horner reaction of aldehyde 12 with the N-Boc-protected
phosphonate 1312 provided didehydro amino acid (Z)-14 in a
yield of 66% and, as expected,13 with high diastereoselectivity.
Thus, only 2% of the diastereomer (E)-14 was found and isolated
by column chromatography. The configuration of the didehydro
1
amino acid moiety was assigned based on established H NMR
criteria for this class of compounds.14 Only the (Z)-isomer was
required for the subsequent asymmetric hydrogenation as it is
reported that asymmetric hydrogenation using rhodium catalysts
occurs more rapidly and with significantly better stereoselectivities
for (Z)-didehydro amino acids than for the (E)-isomers.15 Conse-
quently, hydrogenation of (Z)-14 in the presence of the chiral
rhodium catalyst (+)-1,2-bis-((2S,5S)-2,5-dimethylphospholano)-
benzene-(cyclooctadiene)-rhodium(I) tetrafluoroborate ((S,S)-
Me-DUPHOS-Rh)16a provided product 15 in an excellent yield
of 92% and with high diastereoselectivity (d.r. >97 : 3 based on
1H NMR). It is well established that (S,S)-Me-DUPHOS-Rh
converts N-carbamate protected (Z)-didehydro amino acid esters
selectively into L-amino acids.12,16b As there was clear evidence
of the asymmetric homogenous hydrogenation of (Z)-14 being a
catalyst-controlled reaction (vide infra), the stereochemistry at the
C-6¢ position of 15 could therefore be assigned as (S). In contrast,
when precursor (Z)-14 was hydrogenated under heterogenous
conditions using palladium on charcoal, a surprising substrate-
Fig. 2 Target compounds 8–11.
The synthetic route reported in this work started from pro-
tected uridine-5¢-aldehyde 12 (Scheme 1), which can easily be
1
controlled selectivity (d.r. >95 : 5 based on H NMR) towards
Scheme 1 Synthesis of 5¢-deoxy nucleosyl amino acid building blocks 8 and 9.
2324 | Org. Biomol. Chem., 2010, 8, 2323–2326 This journal is The Royal Society of Chemistry 2010
©

View Article Online
the other diastereomer with (6¢R)-configuration was observed to
give the according product 16 in 95% yield. For the synthesis
of target bulding blocks 8 and 9, the Boc group then had to
be removed in the presence of both the tert-butyl ester and the
silyl protecting groups. Though similar transformations employing
hydrogen chloride in ethyl acetate have been previously reported,17
the obtained yields of 8 and 9 (34% and 28%, respectively) were not
satisfactory (Scheme 1). Most notably, prolonged reaction times
in order to obtain sufficient conversion of the starting material
led to unwanted side reactions such as cleavage of the silyl ether
moieties.
and the well-established stereoselectivity of Me-DUPHOS-Rh
catalysts,12,16b assignment of the stereochemistry at C-6¢ was
feasible.
As an efficient stereoselective synthesis of the unprecedented 5¢-
deoxy nucleosyl amino acid scaffolds 8 and 9 could be achieved, it
was desired to convert them into muraymycin derivatives 10 and
11, respectively, representing analogues of the reportedly bioactive
compound 3.10b Both amines 8 and 9 were therefore reacted
with aldehyde 2110b,20 in reductive amination transformations,
affording 22 and 23 in 75% and 84% yield, respectively. Final
Cbz deprotection under hydrogenolytic conditions then gave target
compounds 10 and 11 in nearly quantitative yields (Scheme 2).
Due to this limitation of the synthetic route utilising N-Boc
protection, it was then decided to change the amino protecting
group. Using phosphonate 1718 for the Wittig–Horner step, N-Cbz
protected didehydro amino acids (Z)-18 and (E)-18 were isolated
in 67% and 6% yield, respectively, after column chromatography
(Scheme 1). Surprisingly, application of the established 1H NMR
criteria for configurational assignment14 was inconclusive in this
case. Based on the results previously obtained with the N-Boc
strategy, it was the most likely conclusion to propose the (Z)-
configuration for the major product in this case as well. This
was experimentally supported by the results of the subsequent
1
hydrogenation reactions (vide infra) as well as a H-1H NOESY
NMR experiment.19 When diastereomerically pure (Z)-18 was
used for the asymmetric hydrogenation in the presence of (S,S)-
Me-DUPHOS-Rh, N-Cbz protected nucleosyl amino acid 19
was obtained. The N-protecting group could then be cleaved
in a simple one-pot manner by further hydrogenation after the
addition of palladium on charcoal to provide 8 directly in a
good yield of 86% and with excellent diastereoselectivity (d.r.
>97 : 3 based on 1H NMR). In order to avoid hydrogenolysis
of the Cbz group prior to reduction of the double bond, (R,R)-
Me-DUPHOS-Rh was used instead of palladium on charcoal
(vide supra) for the synthesis of 20, finally providing the (6¢R)-
configured congener 9 after subsequent hydrogenolysis of the Cbz
group. This one-pot sequence provided 9 in good yield (80%)
and excellent diastereoselectivity (d.r. >97 : 3 based on 1H NMR,
Scheme 1). The products 8 and 9 obtained via the N-Cbz route
were identical to those furnished by the N-Boc approach, which
was unambigously proven by NMR spectroscopy. Thus, nucleosyl
amino acid building blocks 8 and 9 were synthesised from uridine
in overall yields of 38% and 35%, respectively, over 6 steps via the
N-Cbz protecting group strategy.
The reaction periods of the asymmetric hydrogenation trans-
formations of (Z)-18 needed for sufficient conversions differed
significantly for the two Me-DUPHOS-Rh catalysts. This might
indicate the combination of (Z)-18 with (S,S)-Me-DUPHOS-Rh
to represent the matched case, while the reaction using (R,R)-Me-
DUPHOS-Rh might have suffered from a mismatched situation.
As diastereoselectivities were very high in both cases, catalyst (i.e.
ligand) control clearly dominated over any potential substrate
control resulting from the chiral nucleoside moiety. This was at
least the case for homogenous hydrogenation, while substrate
control could be observed for heterogenous hydrogenation of
the Boc-protected derivative (Z)-14 (vide supra). Attempts to
perform the hydrogenation of (Z)-18 in the presence of the achiral
Wilkinson catalyst (PPh3)3RhCl surprisingly gave no conversion.
However, due to the clear evidence of the asymmetric homogenous
hydrogenation reaction being a catalyst-controlled transformation
Scheme 2 Synthesis of muraymycin analogues 10 and 11.
In conclusion, we report the synthesis of novel 5¢-deoxy nucleo-
syl amino acid scaffolds derived from the nucleosidic core structure
of several natural products, including muraymycin and capraza-
mycin antibiotics. A highly efficient and stereoselective approach
using asymmetric hydrogenation of a common didehydro amino
acid precursor provided both the (6¢S)- and (6¢R)-configured
building blocks 8 and 9, which can be used for the preparation
of novel non-aminoribosylated muraymycin analogues for SAR
studies. The results obtained for this synthetic key step are of
major general importance for the synthesis of a-amino acid
derivatives with highly functionalised side chains. The obtained
excellent diastereoselectivities highlight the enormous versatility
and broad substrate scope of the Me-DUPHOS-Rh catalysts in
the hydrogenation of complex substrates. Compounds 8 and 9
were used for the concise synthesis of two analogues 10 and 11
of an established bioactive truncated muraymycin derivative. The
biological evaluation of these 5¢-deoxy muraymycins as part of a
detailed SAR study on muraymycin analogues is currently being
carried out.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 2323–2326 | 2325
©

View Article Online
9 L. A. McDonald, L. R. Barbieri, G. T. Carter, E. Lenoy, J. Lotvin, P. J.
Petersen, M. M. Siegel, G. Singh and R. T. Williamson, J. Am. Chem.
Soc., 2002, 124, 10260.
10 (a) Y.-I. Lin, Z. Li, G. D. Francisco, L. A. McDonald, R. A. Davis, G.
Singh, Y. Yang and T. S. Mansour, Bioorg. Med. Chem. Lett., 2002, 12,
2341; (b) A. Yamashita, E. Norton, P. J. Petersen, B. A. Rasmussen,
G. Singh, Y. Yang, T. S. Mansour and D. M. Ho, Bioorg. Med. Chem.
Lett., 2003, 13, 3345.
Acknowledgements
We thank the Deutsche Forschungsgemeinschaft (DFG, SFB
803 “Functionality controlled by organization in and between
membranes”) and the Fonds der Chemischen Industrie (FCI) for
financial support. Donation of laboratory equipment by the BASF
SE is gratefully acknowledged.
11 A. P. Spork, S. Koppermann and C. Ducho, Synlett, 2009, 2503.
12 C. Ducho, R. B. Hamed, E. T. Batchelar, J. L. Sorensen, B. Odell and
C. J. Schofield, Org. Biomol. Chem., 2009, 7, 2770.
13 U. Schmidt, H. Griesser, V. Leitenberger, A. Lieberknecht, R. Mangold,
R. Meyer and B. Riedl, Synthesis, 1992, 487.
14 R. Mazurkiewicz, A. Kuz´nik, M. Grymel and N. Kuz´nik, Magn. Reson.
Chem., 2005, 43, 36.
Notes and references
1 G. Taubes, Science, 2008, 321, 356.
2 C. Walsh, Nat. Rev. Microbiol., 2003, 1, 65.
3 (a) W. G. Struve and F. C. Neuhaus, Biochem. Biophys. Res. Commun.,
1965, 18, 6; (b) W. G. Struve, R. K. Sinha, F. C. Neuhaus and M. S.
Prime, Biochemistry, 1966, 5, 82; (c) M. G. Heydanek, W. G. Struve and
F. C. Neuhaus, Biochemistry, 1969, 8, 1214; (d) M. Ikeda, M. Wachi,
H. K. Jung, F. Ishino and M. Matsuhashi, J. Bacteriol., 1991, 173, 1021;
(e) D. S. Boyle and W. D. Donachie, J. Bacteriol., 1998, 180, 6429.
4 (a) A. Bouhss, D. Mengin-Lecreulx, D. Le Beller and J. van Heijenoort,
Mol. Microbiol., 1999, 34, 576; (b) A. J. Lloyd, P. E. Brandish, A. M.
Gilbey and T. D. H. Bugg, J. Bacteriol., 2004, 186, 1747; (c) A. Bouhss,
M. Crouvoisier, D. Blanot and D. Mengin-Lecreulx, J. Biol. Chem.,
2004, 279, 29974; (d) T. Stachyra, C. Dini, P. Ferrari, A. Bouhss, J. van
Heijernoort, D. Mengin-Lecreulx, D. Blanot, J. Biton and D. Le Beller,
Antimicrob. Agents Chemother., 2004, 48, 897.
5 (a) C. Dini, Curr. Top. Med. Chem., 2005, 5, 1221; (b) T. D. H. Bugg,
A. J. Lloyd and D. I. Roper, Infect. Disord.: Drug Targets, 2006, 6, 85.
6 F. J. Dekker, M. A. Koch and H. Waldmann, Curr. Opin. Chem. Biol.,
2005, 9, 232.
7 K.-I. Kimura and T. D. H. Bugg, Nat. Prod. Rep., 2003, 20, 252.
8 (a) S. Hirano, S. Ichikawa and A. Matsuda, J. Org. Chem., 2008, 73,
569; (b) S. Hirano, S. Ichikawa and A. Matsuda, Bioorg. Med. Chem.,
2008, 16, 428; (c) S. Hirano, S. Ichikawa and A. Matsuda, Bioorg. Med.
Chem., 2008, 16, 5123.
15 B. D. Vineyard, W. S. Knowles, M. J. Sabacky, G. L. Bachman and D. J.
Wienkauff, J. Am. Chem. Soc., 1977, 99, 5946.
16 (a) M. J. Burk, J. Am. Chem. Soc., 1991, 113, 8518; (b) T. Masquelin, E.
Broger, K. Mu¨ller, R. Schmid and D. Obrecht, Helv. Chim. Acta, 1994,
77, 1395.
17 (a) F. S. Gibson, S. C. Bergmeier and H. Rapoport, J. Org. Chem., 1994,
59, 3216; (b) F. Cavelier and C. Enjalbal, Tetrahedron Lett., 1996, 37,
5131.
18 (a) U. Schmidt, A. Lieberknecht, U. Schanbacher, T. Beuttler and J.
Wild, Angew. Chem., 1982, 94, 797; U. Schmidt, A. Lieberknecht, U.
Schanbacher, T. Beuttler and J. Wild, Angew. Chem., Int. Ed. Engl.,
1982, 21, 776; (b) U. Schmidt, A. Lieberknecht and J. Wild, Synthesis,
1984, 53; (c) U. Schmidt and J. Wild, Liebigs Ann. Chem., 1985, 1882;
(d) R. Hamzavi, F. Dolle, B. Tavitian, O. Dahl and P. E. Nielsen,
Bioconjugate Chem., 2003, 14, 941.
19 N-methylation of (Z)-18 at the carbamate moiety with methyl iodide
provided a derivative suitable for 1H-1H NOESY NMR investigations.
The observed cross peak of the N-methyl group to H–4¢ of the ribose
moiety represented a nuclear Overhauser effect (nOe) indicating (Z)-
configuration of the olefin motif. Details are described in the ESI.†.
20 M. Y. H. Lai, M. A. Brimble, D. J. Callis, P. W. R. Harris, M. S. Levi
and F. Sieg, Bioorg. Med. Chem., 2005, 13, 533.
2326 | Org. Biomol. Chem., 2010, 8, 2323–2326
This journal is
The Royal Society of Chemistry 2010
©