Structure-function studies on nucleoside antibiotic mureidomycin A: Synthesis of 5′-functionalised uridine models

Article abstract of DOI:10.1039/a901287g

The importance of functional groups in nucleoside antibiotic mureidomycin A (MRD A) for biological activity has been examined by derivatisation of samples of the natural product, and by synthesis of uridine-containing analogues. N-Succinyl and di- and tri-acetyl derivatives MRD A have been prepared, and were found to have reduced activity as inhibitors of E. coli translocase I. The enamide alkene of MRD A was found to be extremely resistant towards hydrogenation by a variety of reagents. Several 5′-functionalised uridine derivatives were synthesised from N³-p-methoxybenzyl-2′,3′-isopropylideneuridine. A series of 5′-aminoacyl derivatives were prepared, and the 3-aminopropionyl (IC₅₀ 260 μM) and 7-aminoheptanoyl (IC₅₀ 1.5 mM) derivatives were found to act as reversible inhibitors. An analogue mimicking the carboxy terminus of MRD A was synthesised, and also acted as an inhibitor of translocase I (IC₅₀ 1.9 mM). A phosphonate analogue designed as a possible suicide inhibitor showed modest inhibition (IC₅₀ 3.7 mM), which was shown to be irreversible.

Full text of DOI:10.1039/a901287g

View Article Online / Journal Homepage / Table of Contents for this issue
Structure–function studies on nucleoside antibiotic mureidomycin
A: synthesis of 5Ј-functionalised uridine models
Caragh A. Gentle,a Stephen A. Harrison,a Masatoshi Inukaib and Timothy D. H. Bugg*a
a Department of Chemistry, University of Southampton, Highfield, Southampton,
UK SO17 1BJ
b Sankyo Co. Ltd., 1-2-58, Hiromachi, Shinagawa-ku, Tokyo 140, Japan
Received (in Cambridge) 16th February 1999, Accepted 1st April 1999
The importance of functional groups in nucleoside antibiotic mureidomycin A (MRD A) for biological activity has
been examined by derivatisation of samples of the natural product, and by synthesis of uridine-containing analogues.
N-Succinyl and di- and tri-acetyl derivatives MRD A have been prepared, and were found to have reduced activity
as inhibitors of E. coli translocase I. The enamide alkene of MRD A was found to be extremely resistant towards
hydrogenation by a variety of reagents. Several 5Ј-functionalised uridine derivatives were synthesised from
N3-p-methoxybenzyl-2Ј,3Ј-isopropylideneuridine. A series of 5Ј-aminoacyl derivatives were prepared, and the
3-aminopropionyl (IC50 260 µM) and 7-aminoheptanoyl (IC50 1.5 mM) derivatives were found to act as reversible
inhibitors. An analogue mimicking the carboxy terminus of MRD A was synthesised, and also acted as an inhibitor
of translocase I (IC50 1.9 mM). A phosphonate analogue designed as a possible suicide inhibitor showed modest
inhibition (IC50 3.7 mM), which was shown to be irreversible.
Mrd A (1) R= H
Introduction
Mrd C R= Gly
Methylene bridge found in
Mrd E & F and in napsamycins
Pacidamycins R= H, Gly, Ala
Inhibition of bacterial cell wall peptidoglycan biosynthesis is
the site of action of many clinically important antibiotics,
including the penicillins and the glycopeptide antibiotics.1 The
emergence of widespread clinical resistance to penicillins, and
the recent emergence of bacterial resistance to vancomycin,2
threatens the efficacy of antibiotics in current use and necessi-
tates the exploration of new targets for antibacterial chemo-
therapy.
We have previously reported that the nucleoside antibiotic
mureidomycin A acts as a potent slow-binding inhibitor (Ki 36
nM, Ki* 2 nM) for phospho-MurNAc-pentapeptide† trans-
locase (translocase I) from Escherichia coli.3 This enzyme is an
integral membrane protein which catalyses the first step of the
intramembrane cycle of reactions involved in peptidoglycan
assembly.1 This enzyme is encoded by the mraY gene,4 and has
only recently been overexpressed and solubilised.3 It is therefore
of great interest to elucidate which functional groups in the
mureidomycin A skeleton are important for its biological
activity.
Mureidomycin A consists of a 3Ј-deoxyuridine nucleoside
linked via an unusual enamide linkage to a modified peptide
chain, which contains two m-tyrosine residues, one methionine
residue, and one 3-aminomethyl-3-deoxythreonine residue.5
Two closely related families of natural products, namely the
pacidamycins6 and the napsamycins,7 share the overall skeleton
but show small differences in the amino acid composition (see
Fig. 1). All contain the uridine nucleoside, the enamide linkage,
a free amino terminus, and a carboxy-terminal aromatic amino
acid.
In the previous paper we examined the role of the enamide
linkage via synthesis of enamide-containing analogues, con-
cluding that although reactive in simpler model systems the
enamide linkage in MRD A and a uridine model showed low
chemical reactivity.8 In this paper we examine the importance
of other functional groups in the mureidomycin A skeleton
by derivatisation of the natural product, and by synthesis of 5Ј-
functionalised uridine models.
NHR
MeN
O
O
HO
HN
–
CO2
O
HO
O
O
N
H
N
N
H
O
N
N
Mrd B & D
O
H
H
contain
dihydro-uracil
S
OH
Pacidamycins contain
m-Tyr, Phe or Trp at
carboxyl terminus
Found as Ala
in pacidamycins
Fig. 1 Structures of the mureidomycin, pacidamycin, and napsamycin
families of nucleoside antibiotics.
Results
Derivatisation of mureidomycin A
Mureidomycin A contains several nucleophilic groups which
can be derivatised chemically, as shown in Fig. 2. A series of
reactions were carried out on a milligram scale using samples of
mureidomycin A. The reaction products were desalted by gel
filtration, then analysed by reversed phase HPLC/electrospray
mass spectrometry. New derivatives were isolated by reversed
phase HPLC and tested as inhibitors of solubilised E. coli
translocase I. The data is summarised in Table 1.
The amino terminus was selectively modified by treatment
with succinic anhydride.9 HPLC-MS analysis showed a new
peak corresponding to the N-succinyl derivative (MHϩ 941),
which was purified by semi-prep HPLC. Reaction of this prod-
uct with Sanger’s reagent (2,4-dinitrofluorobenzene) gave no
characteristic yellow colour or UV absorbance at 340 nm,
whereas a control sample of MRD A gave a positive result,
verifying that the free amino group had been modified. Assays
with solubilised translocase I verified that the N-succinyl
derivative was still an effective inhibitor (IC50 7 µM), but with
reduced potency compared with MRD A (IC50 < 0.1 µM).
† MurNAc = N-acetylmuramyl.
J. Chem. Soc., Perkin Trans. 1, 1999, 1287–1294
1287

View Article Online
Table 1 HPLC-MS Analysis of synthetic derivatives of mureidomycin A. Reversed phase HPLC retention times (Rt), molecular masses (MHϩ)
observed by positive ion electrospray mass spectrometry, and IC50 values for inhibition of solubilised E. coli translocase I in radiochemical assay are
given
Reaction
Product
RP-HPLC Rt/min
ES-MS MHϩ
IC50 (µM)
Control
MRD A
N-succinyl
Diacetyl
Diacetyl
Triacetyl
30.0
27.0
31.5
32.5
34.0
—
841.0
940.9
925.3
925.3
967.1
865.3a
<0.1
7.0
56
Succinylation: succinic anhydride, H2O
Acetylation: Ac2O, NaOAc
132
ӷ100
Methylation: CH2N2
Reduction:
Cyclised, ϩ42 amu
NDb
NaBH3CN, pH 4
H2/cat (Pd/C; Rh(PPh3)3Cl)
MRD A only
MRD A only
HN᎐NH, pH 4
Et3SiH, 10% CF3COOH
Et3SiH, AcOH
MRD A only
Ϫuracil Ϫ H2O ϩ 2H, 6H
MRD A only
᎐
32.4, 34.2
717.3, 713.1
a Derivative analysed as mixture of desalted products by electrospray MS. b ND not determined.
A. These observations imply that the enamide alkene is highly
resistant to hydrogenation, and is only reduced after the uracil
base is cleaved. The reasons for this extremely low reactivity are
not clear, but it appears most likely that approach to the alkene
is sterically hindered.
acetylation
succinylation
NH3
O
N
O
HO
O
HN
MeN
–
HO
CO2
O
O
Synthesis of 5Ј-aminoacyluridine analogues
H
N
N
H
O
N
H
N
H
In order to assess further the role of the free amino terminus for
biological activity, a series of 5Ј-aminoacyluridine derivatives
were prepared, containing a variable length alkyl spacer. Pre-
liminary synthetic work indicated a need to protect N-3 of the
uracil base in order to achieve selective reaction at the 5Ј pos-
ition of the ribose ring. Accordingly, a number of protecting
groups were investigated using 2Ј,3Ј-isopropylideneuridine. In
our hands the protecting group of choice was p-methoxy-
benzyl (PMB),14 which was inserted in 93% yield, and could be
deprotected under mild conditions using cerium() ammonium
nitrate, which also removed the isopropylidene group.
O
S
OH
reduction
Fig. 2 Derivatisation of mureidomycin A.
Treatment of MRD A with acetic anhydride in the presence
of sodium acetate gave rise to two diacetylated derivatives
(MHϩ 925), and a triacetylated derivative (MHϩ 967). None of
the acetylated derivatives gave a positive reaction with Sanger’s
reagent, suggesting that the amino group had been modified in
each case. Assays showed that the diacetyl derivatives were
inhibitors with reduced potency (IC50 56 µM and 132 µM),
while the triacetyl derivative was a weak inhibitor (IC50 ӷ 100
µM). The data suggests that each of these groups forms a sig-
nificant interaction with the enzyme, but that none is absolutely
essential for inhibition.
Treatment with diazomethane, in an attempt to modify the
carboxy terminus, gave none of the desired methyl ester, but
gave a major product of mass MHϩ 865. This product corre-
sponds to loss of H2O, perhaps due to cyclisation of amino and
carboxy termini, followed by three successive methylations
(peaks with MHϩ 837 and 851 were also observed).
The synthetic route, shown in Scheme 1, involved initial Boc
O
N
O
N
R
N
PMBN
O
O
O
O
b
HO
BocNH
O
O
n
O
O
O
O
(4)–(8) n = 1,3,4,5,6
(2) R = H
(3) R = PMB
c
a
Attempts were made to hydrogenate the enamide alkene of
MRD A. Catalytic hydrogenation using palladium catalysts, or
using Wilkinson’s catalyst [(PPh3)3RhCl], failed to give any new
products. Reduction with borohydrides [NaBH3CN, NaBH4,
Me4N(AcO)3BH] under mildly acidic conditions, a reaction
which is precedented for reduction of enamides10 and which
worked in the case of an earlier model,8 also failed to give any
new product. Treatment with triethylsilane in 10% trifluoro-
acetic acid gave new products corresponding to MHϩ 713, 715
and 717, suggesting that the uracil base had been cleaved, fol-
lowed by elimination of water and addition of 2H, 4H, and 6H.
There is chemical precedent for the reduction of glycosidic
linkages to ethers by triethylsilane, in the presence of Lewis
acids.11 Thus it seems plausible that in this case the 3Ј-deoxy-
ribose ring is reduced eventually to a tetrahydrofuran. However,
treatment with triethylsilane under less forcing conditions
(AcOH) gave no new products. Treatment with diimide, pre-
pared either from potassium azodicarboxylate12 or from
toluene-p-sulfonyl hydrazide,13 also gave only recovered MRD
O
N
HN
O
O
+
H3N
O
O
n
AcO–
HO
OH
(9)–(13) n = 1,3,4,5,6
Scheme 1 Synthesis of ω-(aminoacyl)uridine analogues. a, p-Methoxy-
benzyl chloride, DBU, MeCN, 70 ЊC, 93%; b. BocNHCH2(CH2)nCO2H,
DCC, DMAP, CH2Cl2, 65–86%; c, (NH4)2Ce(NO3)6, MeCN–H2O,
50 ЊC.
protection of C-3, C-5, C-6, C-7 and C-8 ω-aminocarboxylic
acids. The Boc-protected acids were then coupled onto the
PMB-protected uridine 3 via DCC couplings, in 65–86% yield,
to give the 5Ј-acylated products 4–8. Deprotection of the PMB,
1288
J. Chem. Soc., Perkin Trans. 1, 1999, 1287–1294

View Article Online
Table 2 Inhibition data (IC50) for uridine-based analogues versus
solubilised E. coli translocase I in a radiochemical assay
H
N
ButO2C
Ph
NH2•HCl
ButO2C
Ph
CO2H
a
O
(14)
Compound
5Ј-Sidechain
IC50 (mM)
0.26
ӷ10
ӷ10
1.5
ӷ10
1.9
b,c
9
–CO(CH2)2NH2
–CO(CH2)4NH2
–CO(CH2)5NH2
–CO(CH2)6NH2
–CO(CH2)7NH2
–β-Ala-succ--Phe-
CO2H
10
O
11
12
PMBN
13
O
O
H
N
C-terminal analogue 16
O
O
N
ButO2C
N
H
O
Phosphonate diester 18
Phosphonate monoester 19
–(CH2)3PO(OEt)2
–(CH2)3P(OEt)O2Na
ӷ10
3.7
O
Ph
(15)
O
O
Boc and isopropylidine groups was then achieved by heating in
the presence of aqueous cerium() ammonium nitrate. Final
products 9–13 were purified by G10 gel filtration, followed by
reversed phase HPLC, and characterised by NMR spectro-
scopy and electrospray mass spectrometry.
The 5Ј-aminoacyluridine analogues were tested as inhibitors
of solublised E. coli translocase, and the data is summarised in
Table 2. The 5-aminopentanoyl 10, 6-aminohexanoyl 11 and
8-aminooctanoyl 13 analogues showed little or no inhibition
at 5 mM concentration, whereas the 3-aminopropionyl 9 (IC50
0.26 mM) and 7-aminoheptanoyl 12 (IC50 1.5 mM) analogues
showed effective enzyme inhibition. This somewhat surprising
pattern of inhibition will be discussed below.
d
O
HN
O
O
H
N
O
N
HO2C
Ph
N
H
O
O
O
(16)
HO
OH
Scheme 2 Synthesis of carboxy-terminal analogue 16 of mureido-
mycin A. a, Succinic anhydride, N-methylmorpholine, THF–DMF,
90%; b, isobutyl chloroformate, NEt3, H2N(CH2)2CO2H, THF; c, (3),
DCC, DMAP, CH2Cl2, 38% overall; d, (NH4)2Ce(NO3)6, MeCN–H2O,
50 ЊC.
Synthesis of a carboxy terminal analogue
to the phosphonate monoester 19 by partial hydrolysis using
sodium hydroxide. The phosphonates 18 and 19 were tested as
inhibitors of solubilised translocase I (see Table 2). The phos-
phonate diester 18 showed no inhibition at 13 mM concen-
tration, however the monoester 19 showed modest inhibition
(IC50 3.7 mM).
As mentioned above, all of the members of the mureidomycin,
pacidamycin and the napsamycin families contain an aromatic
amino acid at their carboxy termini. In the mureidomycins this
residue is always m-tyrosine, but in the pacidamycins it is found
as phenylalanine, m-tyrosine, or tryptophan (see Fig. 1). In
order to probe the role of this group, an analogue 16 contain-
ing -phenylalanine-COOH attached via an 8-carbon spacer
to the 5Ј-position of uridine was synthesised, as illustrated in
Scheme 2. The -isomer of phenylalanine was used, although
the configuration of the amino acid residues in mureidomycin
A has not been determined.
-Phenylalanine tert-butyl ester was reacted with succinic
anhydride to give the succinyl derivative 14 in 90% yield. This
acid was coupled to β-alanine via a mixed anhydride coupling,
and the crude product was coupled onto the PMB-protected
uridine 3 to give the fully protected derivative 15 in 38% yield
overall. Deprotection with cerium() ammonium nitrate under
acidic conditions, followed by gel filtration and HPLC purifi-
cation gave the target compound 16. Enzyme assays (see Table
2) revealed that 16 was an inhibitor of translocase I (IC50 1.9
mM).
Characterisation of enzyme inhibition
In order to examine the type of enzyme inhibition being shown
by phosphonate 19 and the 3-aminopropionyl analogue 9, the
time dependence of enzyme inhibition was examined. Incu-
bations of enzyme and inhibitor were set up, and aliquots
examined for activity at time points over 2 hours. The results,
shown in Table 3, imply that phosphonate 19 is exhibiting time-
dependent inhibition, whereas amine 9 is not. At the end of this
experiment, the incubations were dialysed overnight, and the
residual enzyme activity measured and compared with controls.
The results are shown in Table 3. The incubation containing
enzyme and phosphonate 19 showed only 18% residual activity,
and no recovery of activity, implying that this inhibitor has
resulted in an irreversible loss of enzyme activity; whereas in the
case of amine 9 the enzyme has recovered almost full activity
after dialysis. These data imply that phosphonate 21 is acting
as a time-dependent irreversible inhibitor, consistent with the
suicide inhibition mechanism, although its affinity for the
enzyme is modest. Amine 9 appears to be acting as a reversible
inhibitor, whose inhibition is reduced in the presence of the
natural substrate, consistent with binding to the enzyme active
site.
Synthesis of a phosphonate suicide inhibitor
The catalytic mechanism of the translocase I reaction is
thought to involve nucleophilic attack of an active site nucleo-
phile onto the β-phosphate of the substrate uridine-diphospho-
MurNAc-pentapeptide (see Fig. 3A,B), followed by attack of
undecaprenyl phosphate on the covalent intermediate.15 It was
therefore envisaged a phosphonate analogue 19 might act as a
suicide inhibitor of translocase I by attack of the active site
nucleophile to give a covalently bound species which could
not dissociate from the enzyme active site, as illustrated in
Fig. 4C.
The synthetic route to the desired phosphonate is shown in
Scheme 3. Diethyl [3-(p-tolylsulfonyl)oxypropyl]phosphonate16
was used to alkylate the PMB-protected uridine 3, using sodium
hydride as base, in 52% yield, to give the protected phosphonate
17. The PMB and isopropylidene groups were removed using
cerium() ammonium nitrate to give the deprotected phos-
phonate diester 18. Half of this material was then converted
Conclusions
In view of the potency of translocase I inhibition by murei-
domycin A, it is of considerable interest to elucidate the
structural features within MRD A which are responsible for
potent slow-binding inhibition. In the previous paper we have
examined the role of the unusual enamide linkage found in
mureidomycin A, and we have found that, although reactive in
simpler model systems, in mureidomycin A and a uridine model
it displays low reactivity.8 The uridine nucleoside itself appears
to be essential for activity, since all known inhibitors of this
J. Chem. Soc., Perkin Trans. 1, 1999, 1287–1294
1289

View Article Online
Table 3 A. Time-dependent inhibition of translocase I by phosphon-
ate 19 and amine 9. B. Regain of translocase I activity following
treatment with phosphonate 19 and 3-aminopropionyl analogue 9 and
overnight dialysis
uridine-P-P-MurNAc-pentapeptide
Enz-X–
A.
uridine-P
Table A
Enz-X-P-MurNAc-pentapeptide
undecaprenyl-P
% Translocase activity
Enz-X–
Time/min
ϩ1.5 mM 19
ϩ85 µM 9
undecaprenyl-P-P-MurNAc-pentapeptide
0
15
30
60
120
90
76
—
—
60
42
—
54
48
52
B.
OH
O
N
O
HO
RO
HN
O–
P
O–
P
O
NH
Ac
O
O
O
O
O
O
Table B
Enz-X–
HO
OH
% Activity
before dialysis
% Activity
after dialysis
Incubation
O
N
C.
Enzyme only (control)
Enzyme ϩ 11 µM UDPMurNAc-
pentapeptide
Enzyme ϩ 1.7 mM phosphonate 19
Enzyme ϩ 2.2 mM amine 9
Enzyme ϩ 2.2 mM amine 9 ϩ 11 µM
substrate
Enzyme ϩ 1.2 mM amine 13
Enzyme ϩ 2.4 mM C-terminal
analogue 16
100
92
100
96
HN
–O
EtO
O
O
27
37
51
18
84
121
P
O
O
HO
OH
Enz-X–
61
38
114
121
O
HN
–O
EnzX
O
O
N
P
O
showed significant enzyme inhibition, none being observed for
the 5-aminopentanoyl or 6-aminohexanoyl analogues, suggest-
ing that there are different explanations for the inhibition
observed by 9 and 12. It is known that translocase I requires
Mg2ϩ for activity, maximum activity being observed at 5–10
mM concentration.3 Analysis of an amino acid sequence
alignment has revealed a conserved Asp-Asp-Xaa-Xaa-Asp/
Asn at position 115–119 of the E. coli sequence, which is a
known sequence motif found in pyrophosphate-utilising
enzymes such as terpene cyclases.17 An X-ray crystal structure
of farnesyl pyrophosphate synthase has revealed that the
Asp–Asp pair co-ordinates a Mg2ϩ cofactor, which in turn
co-ordinates the pyrophosphate linkage.18 It is therefore a
reasonable hypothesis that such a Mg2ϩ binding site exists in the
translocase I active site, as shown in Fig. 4A.
O
HO
OH
Fig.
3
A. Two-step mechanism for translocase
I
involving an
active site nucleophile. B. Attack of enzyme active site nucleophile on
β-phosphate of UDPMurNAc-pentapeptide. C. Anticipated mechan-
ism of suicide inhibition by phosphonate analogue 19.
O
N
O
a
(EtO)2P
OTs
+
(3)
PMBN
EtO
O
O
EtO
P
O
O
O
O
(17)
The question then arises: why should amine sidechains
inhibit an enzyme which would normally bind a pyrophosphate
bridge? Since the Mg2ϩ cofactor is bound relatively weakly by
the enzyme, it is reasonable to suppose that inhibitors could
bind to the active site in the absence of Mg2ϩ, in which case
there would be a vacant cation binding site which could be
occupied by the ammonium sidechain. We therefore tentatively
suggest that the inhibition by the 3-aminopropionyl analogue 9
is due to the inhibitor binding with a syn ester conformation, as
shown in Fig. 4B, whereas the 7-aminoheptanoyl analogue 12
binds with an anti ester conformation, shown in Fig. 4C.
According to this model, the 5- and 6-carbon spacers would be
too long to bind in the syn conformation, but not long enough
for the chain to extend around when adopting the anti con-
formation. It is interesting to note that in MRD A there is a
7-atom spacer between the uridine C-5Ј position and the amino
terminus, consistent with this model (see Fig. 4D). Further
experiments are needed to confirm this hypothesis.
b
O
N
HN
RO
EtO
O
P
O
O
O
HO
OH
(18) R = Et
(19) R = Na
c
Scheme 3 Synthesis of phosphonate inhibitor (19). a, NaH, DMF,
52%; b, (NH4)2Ce(NO3)6, MeCN–H2O, 50 ЊC; c, NaOH, H2O.
enzyme, including tunicamycin and the liposidomycins, contain
a uridine nucleoside.1 The activity shown by the chemical
derivatives of MRD A reported in this work suggest that the
amino terminus and the phenolic hydroxy groups of MRD A
form significant binding interactions with the enzyme active
site, although no single functional group is essential for
inhibition.
The pattern of inhibition shown by the aminoacyluridine
analogues is intriguing, and merits discussion. Only the
3-aminopropionyl 9 and the 7-aminoheptanoyl 12 analogues
The modest inhibition shown by the carboxy terminal
analogue 16 shows that there is some binding interaction for the
C-terminal part of MRD A, but that it is not strong by itself. It
therefore seems likely that the potency of inhibition by MRD A
is due to a number of binding interactions from different parts
of the molecule, and that more complex synthetic analogues
1290
J. Chem. Soc., Perkin Trans. 1, 1999, 1287–1294

View Article Online
Asp 115
phobic and negatively charged. It is possible that the slow-
A.
116
Asp
binding inhibition is caused by a conformational change from
the EI to the EI* complex. Since the enzyme must acquire the
co-substrate undecaprenyl phosphate from the lipid bilayer, and
release the lipid-linked product into the bilayer, it is quite likely
that there is a protein conformational change which opens the
active site to the lipid bilayer. Investigation of an enzyme con-
formational change awaits the purification to homogeneity of
translocase I.
O
O
HO
O
O
O
N
2+
O
HO
RO
Mg
HN
–O
O–
NHAc
O
O
O
O
P
P
O
O
O
Experimental
HO OH
General
Asp 115
Infra-red spectra were recorded on a 1600 series Perkin-Elmer
FTIR spectrometer. 300 MHz NMR spectra were recorded on a
Bruker AC300 Fourier Transform Spectrometer. Mass spectra
were recorded on a VG platform quadruple E.S.I. mass spec-
trometer. Analytical RP-HPLC was performed on a Hewlett-
B.
116
Asp
O
O
–O
–O
O
N
H3N
Packard HP1100 ChemStation Chromatograph using
a
HN
Phenomenex Prodigy ODS3 column, 3 mm × 150 mm, 0.1%
TFA–H2O→0.04% TFA–MeCN, 20 min gradient, 0.5 mL
minϪ1, observing at 254 nm, unless stated otherwise. Semi-prep
RP-HPLC was performed on a Waters Associates Chromato-
graph using a Zorbax ODS C18 column, 10 mm × 250 mm,
0.1 M NH4OAc–H2O→0.1 M NH4OAc–1:1 H2O–MeCN, 30
min gradient, 2.5 mL minϪ1, observing at 260 nm. UV absorb-
ances were recorded on a Cary 1 UV spectrometer. Radioactiv-
ity was measured by mixing sample with 3.5 mL Optiphase Hi-
Safe 3 scintillation fluid and counting in a Beckman LS 6500
multipurpose scintillation counter.
2Ј,3Ј-O-Isopropylideneuridine 2 was prepared from uridine
using 2-methoxypropene–TsOH in 70–85% yield.19 N-Boc-ω-
aminocarboxylic acids were prepared by the method of Jorgen-
son.20 Diethyl [3-(p-tolylsulfonyl)oxypropyl]phosphonate was
prepared from 3-bromopropan-1-ol using the method of
Capson et al.16 14C-UDPMurNAc-pentapeptide (specific
activity 160 µCi µmolϪ1) was a gift from SmithKline Beecham
Pharmaceuticals.
O
O
O
O
HO OH
C.
Asp 115
116
Asp
O
O
–O
–O
O
O
N
NH3
HN
O
O
O
HO OH
D.
Succinylation of mureidomycin A
O
N
Mureidomycin A 1 (5 mg, 5.9 µmol, 1 equiv.) was dissolved in
water (1 mL) and adjusted to pH 9.0 by addition of 0.07 M
NaOH. Succinic anhyhdride (3 mg, 30 mol, 5 equiv.) was dis-
solved in water (100 µL) and added to the ice-cooled reaction
mixture in 10 µL aliquots over 30 min. After a total of 1 h the
mixture was lyophilised to yield a white solid. This was taken up
in minimum water and applied to a Sephadex G-10 gel filtration
column (12 mm × 600 mm), eluting with water. Fractions
showing a UV absorbance maximum at 260 nm were combined
and lyophilised to yield an off-white solid. The solid was taken
up in water to give A260 nm ≈ 2.0, and purified by semi-prep
RP-HPLC to yield the title compound as a white solid after
lyophilisation, 2 mg, 36%, Rt 27.0 min: LR-ESMS (ϩve ion)
MHϩ 940.9.
HN
+
NH3
Ar
O
O
O
NH
O
N
Me
OH
Fig. 4 Model for rationalisation of enzyme inhibition data for
aminoacyl-uridine analogues. A. Proposed role of Mg2ϩ cofactor and
Asp–Asp pair. B. Mode of inhibition of 3-aminopropionyl analogue
(9). C. Mode of inhibition of 7-aminoheptanoyl analogue 12. D. Struc-
ture of mureidomycin A, illustrating similarity in structure of the
amino-terminal portion.
will be needed in order to realise tight binding to the enzyme
active site. It is of interest that phosphonate 19 was shown to
act as an irreversible inhibitor for translocase I. Although not
a potent inhibitor, this result verifies that it is feasible in prin-
ciple to design an irreversible inhibitor for this enzyme. These
data may be valuable in the design of more potent translocase I
inhibitors.
Regarding the mechanism of slow-binding inhibition of
translocase I by mureidomycin A, our observations suggest that
both N-terminal and C-terminal ends of the peptide chain of
MRD A contribute towards active site binding. We tentatively
suggest that the N-terminal part of MRD A binds in the Mg2ϩ
binding site of the enzyme, which would bring the amino-
terminal m-tyrosine residue close to the MurNAc binding site
of the enzyme. The C-terminal m-tyrosine may bind in place of
the co-substrate undecaprenyl phosphate, which is also hydro-
Acetylation of mureidomycin A
Mureidomycin A 1 (5 mg, 5.9 µmol, 1 equiv.) was dissolved in
water (0.8 mL) and an equal volume of half-saturated sodium
acetate solution added. The mixture was cooled in an ice-bath
and treated with acetic anhydride (5 × 5 µL, 270 µmol, ~46
equiv.) over 3 h. Lyophilisation yielded a white solid which was
taken up in minimum water and applied to a Sephadex G-10 gel
filtration column, eluting with water. Fractions showing a UV
absorbance maximum at 260 nm were combined and lyophil-
ised to yield a white solid. The solid was taken up in water to
give A260 nm ≈ 2.0 and purified by semi-prep RP-HPLC to give
three products (after lyophilisation): (1) diacetylated MRD A,
1.8 mg, Rt 31.5 min, LR-ESMS (ϩve ion) MHϩ 925; (2)
J. Chem. Soc., Perkin Trans. 1, 1999, 1287–1294
1291

View Article Online
Table 4 LR-ESMS and HPLC retention times for (9–13)
diacetylated MRD A, 1.7 mg, Rt 32.5 min, LR-ESMS (ϩve
ion) MHϩ 925; (3) triacetylated MRD A, 1.2 mg, Rt 34.0 min,
LR-ESMS (ϩve ion) MHϩ 967.
5Ј-Sidechain of
LR-ESMS
HPLC retention
time/min
product
(ϩve ion) MHϩ
N 3-(p-Methoxybenzyl)-2Ј,3Ј-O-isopropylideneuridine 3
H2N(CH2)2CO– (9)
H2N(CH2)4CO– (10)
H2N(CH2)5CO– (11)
H2N(CH2)6CO– (12)
H2N(CH2)7CO– (13)
316.3
344.4
358.3
372.3
386.3
4.4
6.5
6.9
7.4
7.9
A solution of 2Ј,3Ј-O-isopropylideneuridine 2 (5.04 g, 17.8
mmol, 1 equiv.) in dry acetonitrile (150 mL) was treated with
DBU (5.3 mL, 35.4 mmol, 2 equiv.), resulting in a transient
colour change from colourless to green, and with p-methoxy-
benzyl chloride (5.00 g, 31.9 mmol, 1.8 equiv.) The reaction
mixture was heated to reflux under an inert atmosphere. After
24 h the solvent was removed in vacuo and the residue par-
titioned between water (50 mL) and ethyl acetate (50 mL). The
aqueous layer was acidified with 5% KHSO4 solution and re-
extracted with ethyl acetate (2 × 50 mL). The combined org-
anics were washed with saturated brine and dried over MgSO4.
Removal of the solvent in vacuo yielded a yellow oil which was
purified by column chromatography (EtOAc) to yield the title
compound as a white solid, 6.65 g, 93%, Rf 0.51 (EtOAc): mp
101–102 ЊC; LR-ESMS (ϩve ion) MHϩ 405.2, MNaϩ 427.2; IR
mmol, 1 equiv.) in acetonitrile (5 mL) was treated with a solu-
tion of ammonium cerium() nitrate (190 mg, 0.35 mmol, 2
equiv.) in water (2 mL), and the resulting yellow solution was
heated to 50 ЊC under nitrogen for 48 h. Water (10 mL) was
added, and the aqueous layer was washed twice with ethyl acet-
ate (5 mL). The aqueous phase was neutralised to pH 7.0 by
addition of aqueous sodium bicarbonate solution, filtered, and
lyophilised. The resultant brown solid was dissolved in water (2
mL) and desalted by Sephadex G10 gel filtration (12 mm × 600
mm). Fractions which showed UV absorbance at 260 nm were
combined and lyophilised. The crude product was purified by
semi-prep RP-HPLC to yield, after lyophilisation to constant
mass, the title compound as a white solid, 30 mg, 55% yield:
analytical RP-HPLC Rt 4.4 min; LR-ESMS (ϩve ion) MHϩ
316.3; δH (300 MHz, D2O) 7.71 (1H, d, J = 8 Hz, uracil 6-H),
5.88 (1H, d, J = 8 Hz, uracil 5-H), 5.81 (1H, d, J = 4 Hz, 1Ј-H),
4.47–4.25 (5H, m, 2Ј-H, 3Ј-H, 4Ј-H, 5Ј-H), 3.29 (2H, t, J = 6 Hz,
H2NCH2), 2.88 (2H, t, J = 6 Hz, NCH2CH2CO). Analogues 10,
11, 12, and 13 were prepared using the same method, on a 400
mg scale. In these cases only a portion of the final product was
purified by HPLC. Crude yields after gel filtration were in the
range 60–100% (including some inorganic contaminants). LR-
ESMS and HPLC retention times for (9–13) were as shown in
Table 4.
(nujol mull) 3400–3500 (br, OH), 1702 (C᎐O), 1650 (C᎐O) cmϪ1
;
᎐
᎐
δH (300 MHz, CDCl3) 7.44 (2H, d, J = 9 Hz), 7.33 (1H, d, J = 8
Hz, uracil 6-H), 6.83 (2H, d, J = 9 Hz), 5.76 (1H, d, J = 8 Hz,
uracil 5-H), 5.56 (1H, d, J = 3 Hz, 1Ј-H), 5.07–4.96 (4H, m),
4.31 (1H, q, J = 4 Hz, 4Ј-H), 3.92 and 3.85 (2H, 5Ј-H AB
system), 3.75 (3H, s) (OMe), 1.58 and 1.37 (2 × 3H, s, CH3);
δC (75.42 MHz, CDCl3) 162.64, 159.32, 151.19, 140.72, 131.04,
128.78, 114.46, 113.88, 102.38, 97.08, 87.07, 83.93, 80.44, 62.86,
55.40, 43.78, 27.42, 25.43.
Preparation of N-Boc-ù-aminoacyl 2Ј,3Ј-O-isopropylidene-N 3-
(p-methoxybenzyl)uridines 4–8
Procedure for 5-aminopropionyl analogue (5): A solution of
N-tert-butyloxycarbonyl-5-aminopentanoic acid18 (326 mg,
1.5 mmol, 1 equiv.) in dry dichloromethane (10 mL) was cooled
to 0 ЊC and treated with dimethylaminopyridine (20 mg) and
DCC (340 mg, 1.7 mmol, 1.1 equiv.). After 30 min N3-(p-
methoxybenzyl)-2Ј,3Ј-O-isopropylideneuridene (3) (607 mg,
1.5 mmol, 1 equiv.) was added and the reaction stirred at room
temperature. After 18 h the reaction mixture was filtered and
the filtrate washed with water (20 mL). The aqueous layer was
re-extracted with dichloromethane (20 mL) and the combined
organic extracts washed with saturated NaHCO3 (2 × 20 mL)
and saturated brine, and dried over MgSO4. Removal of the
solvent in vacuo afforded a yellow oil. Column chromatography
(2:1 EtOAc–cyclohexane) allowed the isolation of the title
compound as a viscous, colourless oil, 778 mg, 86%, Rf 0.57
(2:1 EtOAc–cyclohexane): LR-ESMS (ϩve ion) MHϩ 604.5,
(S)-4-[(1-Butoxycarbonyl-2-phenylethyl)amino]-4-oxobutanoic
acid 14
A suspension of -phenylalanine tert-butyl esterؒHCl (500 mg,
1.9 mmol, 1 equiv.) in THF (8 mL) was cooled to 0 ЊC and
N-methylmorpholine (469 µL, 432 mg, 4.3 mmol, 2.2 equiv.)
was added dropwise. A solution of succinic anhydride (213 mg,
2.1 mmol, 1.1 equiv.) in THF (2 mL) was added, followed by
DMF (40 mL) until dissolution was complete. The reaction was
allowed to warm to room temperature and stirred for 18 h. The
THF was removed in vacuo and the residue partitioned between
ethyl acetate (100 mL) and 1  HCl (200 mL). The aqueous
layer was re-extracted with ethyl acetate (40 mL) and the com-
bined organics dried over MgSO4. Removal of the solvent in
vacuo afforded the title compound as a white solid, 560 mg,
90%, Rf 0.36 (EtOAc): LR-ESMS (ϩve ion) MHϩ 322.3, MNaϩ
344.3; IR (nujol mull) 3436 (br, OH), 1695 (C᎐O), 1648 (C᎐O)
MNaϩ 626.5; IR (nujol mull) 3391 (NH), 1653 (very br, C᎐O)
᎐
cmϪ1; δH (300 MHz, CDCl3) 7.45 (2H, d, J = 9 Hz), 7.22 (1H, d,
J = 8 Hz, uracil 6-H), 6.84 (2H, d, J = 9 Hz), 5.77 (1H, d, J = 8
Hz, uracil 5-H), 5.63 (1H, d, J = 2 Hz, 1Ј-H), 5.03 (2H, ABq,
J = 14 Hz, PhCH2), 4.97 (1H, dd, J = 2, 6 Hz, 2Ј-H), 4.82
(1H, dd, J = 4, 6 Hz, 3Ј-H), 4.63 (1H, br s, NH), 4.41–4.23
(3H, m), 3.81 (3H, s, OMe), 3.11 (2H, q, J = 6 Hz, 2.32
(2H, td, J = 2, 7 Hz), 1.68–1.58 (2H, m), 1.52–1.43 (2H, m),
1.58 (3H, s, CH3), 1.44 (9H, s, C{CH3}3), 1.37 (3H, s, CH3);
δC (75.42 MHz, CDCl3) 172.97, 162.61, 159.28, 155.97,
150.77, 139.88, 130.97, 128.87, 114.63, 113.87, 102.27, 96.05,
85.63, 85.06, 81.46, 79.20, 64.32, 55.39, 43.67, 40.18, 33.65,
29.57, 28.56, 27.28, 25.48, 22.04. Other analogues were pre-
pared by the same method, using the appropriate N-Boc-
ω-aminocarboxylic acid. Yields were as follows: (4) 96%; (6)
84%; (7) 65%, (8) 78%.
᎐
᎐
cmϪ1; δH (300 MHz, CDCl3) 7.32–7.20 (3H, m), 7.15 (2H, d,
J = 7 Hz), 6.38 (1H, d, J = 8 Hz, NH), 4.76 (1H, dt, J = 6, 8 Hz,
Phe α-H), 3.10 (2H, d, J = 6 Hz, PhCH2), 2.68 (2H, ABq, J = 6
Hz, succinyl CH2), 2.52 (2H, t, J = 6 Hz, succinyl CH2), 1.43
(9H, s, C{CH3}3); δC (75.42 MHz, CDCl3) 177.02, 171.49,
170.96, 136.14, 129.67, 128.53, 127.15, 82.84, 53.86, 38.13,
30.68, 29.50, 28.08.
5Ј-O-(N-{4-[(Butoxycarbonyl-2-phenylethyl)amino]-4-oxo-
butanoyl}aminopropanoyl)-N3-(p-methoxybenzyl)-2Ј,3Ј-O-
isopropylideneuridine 15
A solution of N-succinyl derivative 14 (236 mg, 0.73 mmol, 1
equiv.) in THF (5 mL) was treated with triethylamine (206 µL,
1.5 mmol, 2 equiv.) and cooled to Ϫ20 ЊC. A solution of iso-
butyl chloroform (105 µL, 0.81 mmol, 1.1 equiv.) in dry THF
(2 mL) was then added dropwise causing the formation of a
white precipitate. The mixture was stirred at Ϫ20 ЊC for 30 min
Preparation of 5Ј-ù-aminoacyluridines (9–13)
Procedure for 3-aminopropionyl analogue 9 as follows: A solu-
tion of protected 5Ј-aminopropionyluridine 4 (100 mg, 0.17
1292
J. Chem. Soc., Perkin Trans. 1, 1999, 1287–1294

View Article Online
then a solution of β-alanine (72.0 mg, 0.81 mmol, 1.1 equiv.) in
1:1 H2O–THF (4 mL) was added dropwise. The reaction was
allowed to warm to room temperature and stirred for 18 h. The
THF was removed in vacuo and the residue partitioned between
ethyl acetate (50 mL) and 10% citric acid (50 mL). The aqueous
layer was re-extracted with ethyl acetate (50 mL) and the com-
bined organic extracts washed with 10% citric acid and dried
over MgSO4. Removal of the solvent in vacuo afforded a 14:4
hydride (60% in mineral oil, 36 mg, 0.93 mmol, 1.5 equiv.) in
dry DMF (12 mL). The mixture was stirred at room temper-
ature for 15 min, with a concurrent colour change to orange,
before the addition of diethyl {3-[(p-tolylsulfonyl)oxo]propyl}-
phosphonate16 (650 mg, 1.85 mmol, 3 equiv.). After 6 h TLC
(EtOAc) showed starting uridine (Rf 0.46) remaining so a fur-
ther 1 equiv. of tosylate was added. After a total of 24 h the
reaction mixture was partitioned between water (70 mL) and
ethyl acetate (3 × 50 mL). The combined organics were washed
with saturated brine, dried over MgSO4 and concentrated in
vacuo to yield a yellow oil. Column chromatography (EtOAc→
20% MeOH–EtOAc) allowed the isolation of the title com-
pound as a yellow oil, 186 mg, 52%, Rf 0.13 (EtOAc): LR-
ESMS (ϩve ion) MHϩ 583.2, MNaϩ 605.2; IR (liquid film)
3453, 2984, 2935, 1708, 1665, 1512 cmϪ1; δH (300 MHz, CDCl3)
7.48 (1H, d, J = 8 Hz, uracil 6-H), 7.44 (2H, d, J = 9 Hz), 6.80
(2H, d, J = 9 Hz), 5.84 (1H, d, J = 3 Hz, 1Ј-H), 5.73 (1H, d, J = 8
Hz, uracil 5-H), 5.03 (2H, ABq, J = 15 Hz, PhCH2), 4.76 (1H,
dd, J = 6, 3 Hz) and 4.73 (1H, br d, J = 5 Hz) (2Ј and 3Ј-H), 4.36
(1H, J = 4 Hz, 4Ј-H), 4.16–4.01 (4H, m, POCH2CH3), 3.77 (3H,
s, OMe), 3.69 (1H, dd, J = 12, 2 Hz) and 3.57 (1H, dd, J = 12, 2
Hz) (5Ј-H in AB system), 3.50 (2H, t, J = 7 Hz, PCH2CH2CH2),
1.97–1.65 (4H, m, PCH2CH2), 1.58 (3H, s, CH3), 1.37 (3H, s,
CH3), 1.32 (6H, t, J = 7 Hz, POCH2CH3); δC (75.42 MHz,
CDCl3) 162.74, 159.22, 150.95, 138.76, 130.94, 129.09, 114.26,
113.79, 101.73, 98.98, 85.97, 85.63, 80.93, 71.29 (d, J = 16 Hz,
POCH2CH3), 70.84, 61.76 (d, J = 6 Hz, PCH2CH2CH2), 55.36,
43.65, 27.36, 25.48, 23.42, 22.92, 22.85, 21.52 (m, PCH2CH2),
16.63 (d, J = 6 Hz, POCH2CH3).
1
mixture of the β-alanine-linked product–starting acid (by H
NMR analysis) as a colourless oil, 225 mg, 81% (Rf 0.16
(EtOAc), LR-ESMS (Ϫve ion) {M ϩ TFA Ϫ H}Ϫ 505.7),
which was used directly for the next step. A solution of this
material (194 mg) in dry dichloromethane (5 mL) was cooled to
0 ЊC. DCC (112 mg, 0.54 mmol, 1.1 equiv.) and DMAP (10 mg)
were added and the mixture stirred for 30 min. N3-(p-Methoxy-
benzyl)-2Ј,3Ј-O-isopropylideneuridine 3 (200 mg, 0.49 mmol,
1 equiv.) was added and the reaction allowed to warm to room
temperature. After 18 h the reaction was filtered and concen-
trated in vacuo. The residue was partitioned between dichloro-
methane (25 mL) and water (30 mL) and the aqueous portion
re-extracted with dichloromethane (25 mL). The combined
organics were washed with saturated NaHCO3 and saturated
brine, then dried over MgSO4 and concentrated in vacuo.
Column chromatography (2:1 EtOAc–cyclohexane) allowed
the isolation of the title compound as a yellow foam, 221 mg,
38% over two steps, Rf 0.37 (EtOAc): LR-ESMS (ϩve ion)
MHϩ 779.9, MNaϩ 801.9; IR (nujol mull) 3415 (NH), 1731
(C᎐O), 1712 (C᎐O), 1665 (C᎐O) cmϪ1; δH (300 MHz, CDCl3)
᎐
᎐
᎐
7.48 (2H, d, J = 9 Hz), 7.24–7.05 (6H, m), 6.77 (2H, d, J = 9
Hz), 6.42 (1H, t, J = 6 Hz, NHCH2), 6.35 (1H, d, J = 8 Hz, Phe
NH), 5.72 (1H, d, J = 8 Hz, uracil 5-H), 5.56 (1H, d, J = 2 Hz,
1Ј-H), 5.00–4.88 (3H, m), 4.78 (1H, dd, J = 4, 6 Hz, 3Ј-H), 4.63
(1H, dt, J = 6, 8 Hz, Phe α-H), 4.32–4.19 (3H, m, 4Ј-H and
5Ј-H), 3.70 (3H, s, OMe), 3.41 (2H, q, J = 6 Hz, NHCH2), 3.01
(2H, d, J = 6 Hz, Phe PhCH2), 2.49–2.32 (6H, m), 1.49 (3H, s),
1.32 (9H, s), 1.29 (3H, s); δC (75.42 MHz, CDCl3) 171.99,
171.70, 171.48, 170.58, 162.44, 159.18, 150.66, 140.16, 136.24,
130.70, 129.48, 128.74, 128.37, 126.92, 114.55, 113.78, 102.26,
96.02, 85.38, 84.74, 81.37, 82.27, 64.32, 55.24, 53.72, 43.56,
38.05, 34.98, 34.07, 31.40, 29.67, 27.94, 27.15, 25.34.
5Ј-O-{3-[(Diethoxy)phosphoryl]propyl}uridine 18
A solution of the protected phosphonate 17 (230 mg, 0.39
mmol, 1 equiv.) in acetonitrile (5 mL) was treated with an aque-
ous solution (2 mL) of cerium() ammonium nitrate (CAN,
102 mg, 185 µmol, 4 equiv.) and stirred at 50 ЊC. After 4 h TLC
showed starting material remaining (10:1 EtOAc–MeOH, Rf
0.60) so a further 2 equiv. of CAN were added and stirring
continued. After a further 2 h another 2 equiv. of CAN were
added and the reaction was cooled and stirred at room temper-
ature overnight. Water (20 mL) was added and the mixture
washed with ethyl acetate (2 × 20 mL). The organic fractions
were washed with water (30 mL) and the combined aqueous
portions lyophilised to yield a yellow solid. The solid was taken
up in the minimum amount of water and applied to a Sephadex
G-10 gel filtration column. Fractions showing UV absorbance
at 260 nm were analysed by LR-ESMS (ϩve ion) and frac-
tions containing a major signal of MHϩ 423.2 were combined
and lyophilised to yield an off-white solid (280 mg). Half of
this product was carried on to the monohydrolysis step; the
remainder was purified by semi-prep RP-HPLC. Lyophilisation
to constant mass yielded the title compound as a white solid,
5.1 mg, 6% yield, Rt 28.4 min; LR-ESMS (ϩve ion) MHϩ 423.3,
MNaϩ 445.3; δH (300 MHz, D2O), 7.78 (1H, d, J = 8 Hz, uracil
6-H), 5.81 (1H, d, J = 4 Hz, 1Ј-H), 5.79 (1H, d, J = 8 Hz,
uracil 5-H), 4.28–4.10 (3H, m), 4.03 (4H, quintet, J = 7 Hz,
CH3CH2OP), 3.75 (1H, dd, J = 12, 2 Hz) and 3.61 (1H, dd,
J = 12, 4 Hz) (5Ј-CH2 in AB system), 3.60–3.45 (2H, m, PCH2-
CH2CH2O), 1.90–1.70 (4H, m, PCH2CH2CH2O), 1.20 (6H, t,
J = 7 Hz, CH3CH2OP); δC (75.42 MHz, D2O), 168.78, 154.24,
144.43, 104.83, 92.36, 85.65, 76.48, 73.19 (d, J = 18 Hz, POCH2-
CH3), 72.43, 72.01, 65.92 (d, J = 6 Hz, PCH2CH2CH2), 24.36,
22.30 (m, PCH2CH2), 18.21 (d, J = 6 Hz, POCH2CH3).
5Ј-O-(N-{4-[(1-Carboxy-2-phenylethyl)amino]-4-oxobutanoyl}-
aminopropanoyl)uridine 16
A solution of the protected derivative 15 (220 mg, 0.28 mmol,
1 equiv.) in acetonitrile (4 mL) was treated with a solution of
ammonium cerium() nitrate (170 mg, 0.31 mmol, 1.1 equiv.)
in water (1 mL), and the resulting yellow solution heated to
50 ЊC under an inert atmosphere. After 18 h the solvent was
removed in vacuo and the residue neutralised with 1 M NaOH
and partitioned between water (10 mL) and ethyl acetate (10
mL). The organic layer was re-extracted with water (10 mL)
and the combined aqueous portions lyophilised. The resultant
brown solid was dissolved in water (3 mL) and desalted by
Sephadex G-10 gel filtration. Fractions which showed UV
absorbance at 260 nm were combined and lyophilised. A por-
tion of the crude product was purified by semi-prep RP-HPLC
to yield, after lyophilisation to constant mass, the title com-
pound as a white solid, 2.0 mg: analytical RP-HPLC Rt 9.3 min
(84%); LR-ESMS (ϩve ion) MHϩ 563.5, MNaϩ 585.4; HR-
FABMS (ϩve ion) 563.1996 (calc. 563.1989); δH (300 MHz,
D2O) 7.50 (1H, d, J = 8 Hz, uracil 6-H), 7.18–7.02 (5H, m), 5.66
(1H, d, J = 8 Hz, uracil 5-H), 5.64 (1H, d, J = 4 Hz, 1Ј-H), 4.32–
4.01 (6H, m), 3.22 (2H, t, J = 6 Hz, NHCH2), 2.73 (2H, dd,
J = 8, 14 Hz, PhCH2), 2.42 (2H, t, J = 7 Hz), 2.31–2.13 (4H, m).
5Ј-O-3-{3-[(Ethoxy)sodiooxyphosphoryl]propyl}uridine 19
5Ј-O-{3-[(Diethoxy)phosphoryl]propyl}uridine (18) (135 mg
crude, <0.32 mmol) was treated with 1 M NaOH (1 mL)
and stirred at 50 ЊC overnight. Analytical RP-HPLC analysis
(Phenomenex ODS3, 3 mm × 250 mm, 0.1 M NH4OAc–
H2O→0.1 M NH4OAc 1:1 H2O–MeCN, 30 min gradient, 1.0
N3-(p-Methoxybenzyl)-5Ј-O-{3-[(diethoxy)phosphoryl]propyl}-
2Ј,3Ј-O-isopropylideneuridine 17
N3-(p-Methoxybenzyl)-2Ј,3Ј-O-isopropylideneuridine (3) (250
mg, 0.62 mmol, 1 equiv.) was added to a suspension of sodium
J. Chem. Soc., Perkin Trans. 1, 1999, 1287–1294
1293

View Article Online
mL minϪ1, observing 260 nm) showed only a new product (Rt
24.0 min) and so the reaction was neutralised with 2 M HCl,
filtered to remove a yellow solid and lyophilised to yield a white
solid. The product was taken up in the minimum amount of
water and applied to a Sephadex G-10 gel filtration column and
fractions showing UV absorbance at 260 nm were combined
and lyophilised. The crude product was purified by semi-prep
RP-HPLC and lyophilisation to constant mass yielded the title
compound as a white solid, 11 mg, 14% yield, Rt 21.0 min: LR-
ESMS (ϩve ion) MHϩ 395.2; δH (300 MHz, D2O) 7.78 (1H, d,
J = 8 Hz, uracil 6-H), 5.81 (1H, d, J = 4 Hz, 1Ј-H), 5.79 (1H, d,
J = 8 Hz, uracil 5-H), 4.28–4.10 (3H, m), 3.78 (2H, quintet,
J = 7 Hz, CH3CH2OP), 3.72 (1H, dd, J = 12, 2 Hz) and 3.61
(1H, dd, J = 12, 4 Hz) (5Ј-CH2 in AB system), 3.60–3.45 (2H,
m), 1.80–1.40 (4H, m), 1.10 (3H, t, J = 7 Hz, CH3CH2OP); δC
(75.42 MHz, d6-DMSO) 168.82, 154.29, 144.42, 104.98, 92.11,
85.74, 76.43, 74.41 (d, J = 17 Hz, POCH2CH3), 72.53, 72.05,
63.11 (d, J = 6 Hz, PCH2CH2CH2), 26.53, 25.71, 24.54 (m,
PCH2CH2), 18.62 (d, J = 6 Hz, POCH2CH3).
tration 1.32 mM) ϩ 100 µL H2O; (iv) 125 µL of 16.9 mM amine
9 (final concentration 1.88 mM) ϩ 125 µL H2O; (v) 125 µL of
16.9 mM amine 9 (final concentration 1.88 mM) ϩ 125 µL
of 0.1 mM UDP-MurNAc-pentapeptide (final concentration
11 µM). After 2 h 200 µL aliquots were removed and kept on
ice. The remainder of each incubation was dialysed against 500
mL buffer (200 mM TrisؒHCl pH 7.5, 50 mM MgCl2) overnight
at 4 ЊC. 3 × 40 µL aliquots were removed from both the dialysed
and undialysed incubations and treated with 10 µL of 62.5
µM (4 nCi µLϪ1, 64 nCi nmolϪ1 14C-UDP-MurNAc-penta-
)
peptide ϩ 50 µM dodecaprenyl phosphate. This allowed radio-
chemical assays to be performed as described above. All assays
were performed in triplicate. Activities were expressed as a per-
centage of the total count observed in incubation (i).
Acknowledgements
We would like to thank EPSRC for the provision of a student-
ship (C. A. G.) and SmithKline Beecham for financial
assistance.
Characterisation of inhibitors in the radiochemical assay
References
Aqueous solutions of the inhibitors were prepared at known
concentrations, as determined by UV absorbance at 262 nm
(εuridine = 104 MϪ1 cmϪ1). Escherichia coli translocase 1 was
prepared from overexpression construct JM109/pBROC525,
as previously described,3 and was solubilised in 0.5% Triton
X-100. Radiochemical assays were performed, as previously
described,3 in the presence of 10 µL solubilised E. coli trans-
locase I (4 mg mLϪ1 protein), 12.5 µM 14C-UDP-MurNAc-
pentapeptide (64 nCi nmolϪ1) and 10 µM dodecaprenyl
phosphate. All assays were performed in triplicate. % Inhibition
was calculated as {(cpm in the absence of inhibitor Ϫ cpm in
presence of inhibitor)/cpm in the absence of inhibitor} × 100.
IC50 values were determined as the concentration of inhibitor
required to reduce the cpm of the n-butanol layer by 50% as
compared to an inhibitor-free control assay.
1 T. D. H. Bugg and C. T. Walsh, Nat. Prod. Rep., 1992, 9, 199.
2 M. Michel and L. Gutmann, Lancet, 1997, 349, 1901.
3 P. E. Brandish, M. K. Burnham, J. T. Lonsdale, R. Southgate,
M. Inukai and T. D. H. Bugg, J. Biol. Chem., 1996, 271, 7609.
4 M. Ikeda, M. Wachi, H. K. Jung, F. Ishino and M. Matsuhashi,
J. Bacteriol., 1991, 173, 1021.
5 F. Isono, M. Inukai, S. Takahashi, T. Haneishi, T. Kinoshita and
H. Kuwano, J. Antibiot., 1989, 42, 667.
6 R. H. Chen, A. M. Buko, D. N. Whittern and J. B. McAlpine,
J. Antibiot., 1989, 42, 512.
7 S. Chatterjee, S. R. Nadkarni, E. K. S. Vijayakumar, M. V. Patel,
B. N. Ganguli, H. W. Fehlhaber and L. Vertesy, J. Antibiot., 1994,
47, 595.
8 C. A. Gentle and T. D. H. Bugg, previous paper.
9 E. Eisenstein and H. K. Schachman, in Protein Function: A Prac-
tical Approach, ed. T. E. Creighton, IRL Press, Oxford, 1990,
pp. 140–142.
10 R. F. Borch, M. D. Bernstein and H. D. Hurst, J. Am. Chem. Soc.,
1971, 93, 2897.
11 S. G. Zeller, A. J. D’Ambra, M. J. Rice and G. R. Gray, Carbohydr.
Res., 1988, 182, 53.
12 C. Wanke and N. Amrhein, Eur. J. Biochem., 1993, 218, 861.
13 D. J. Hart, W. P. Hong and L. Y. Hsu, J. Org. Chem., 1987, 52,
4665.
14 S. J. Danishefsky, S. L. DeNinno, S. Chen, L. Boisvert and M.
Barbachyn, J. Am. Chem. Soc., 1989, 111, 5810.
15 M. G. Heydanek, W. G. Struve and F. C. Neuhaus, Biochemistry,
1969, 8, 1214.
Time course inhibitor assays with phosphonate 19 and amine 9
At t = 0 min 160 µL solubilised enzyme (0.5% Triton X-100), 80
µL of an aqueous solution of inhibitor (final concentrations:
1.5 mM 19; 85 µM 9) and 400 µL buffer (200 mM TrisؒHCl pH
7.5, 50 mM MgCl2) were mixed and held on ice. At time points
from 0–120 min 40 µL aliquots were removed and treated with
10 µL of 62.5 µM (4 nCi µLϪ1, 64 nCi nmolϪ1 14C-UDP-
)
MurNAc-pentapeptide ϩ 50 µm dodecaprenyl phosphate. This
allowed radiochemical assays to be performed as described
above. A control assay of solubilised enzyme incubated in the
absence of inhibitor was also performed at each time point and
activities were expressed as a percentage of this total count.
16 T. L. Capson, M. D. Thompson, V. M. Dixit, R. G. Gaughan and
C. D. Poulter, J. Org. Chem., 1988, 53, 5903.
17 P. F. Marrero, C. D. Poulter and P. A. Edwards, J. Biol. Chem., 1992,
267, 21873.
18 L. C. Tarshis, M. Yan, C. D. Poulter and J. C. Sacchettini,
Biochemistry, 1994, 33, 10871.
Recovery of activity after removal of inhibitor by dialysis
19 J. Tomasz, in Nucleic Acid Chemistry, eds. L. B. Townsend and
R. S. Tipson, Wiley, New York, 1978, p. 765.
20 E. C. Jorgenson, G. C. Windringe and T. C. Lee, J. Med. Chem.,
1970, 13, 352; 744.
Five incubations were set up and held on ice, containing 250 µL
solubilised enzyme, 625 µL buffer (200 mM TrisؒHCl pH 7.5, 50
mM MgCl2) and: (i) 250 µL H2O; (ii) 125 µL of 0.1 mM UDP-
MurNAc-pentapeptide (final concentration 11 µM) ϩ 125 µL
H2O; (iii) 150 µL of 12.4 mM phosphonate 19 (final concen-
Paper 9/01287G
1294
J. Chem. Soc., Perkin Trans. 1, 1999, 1287–1294