Studies on the interaction of the antibiotic moenomycin A with the enzyme penicillin-binding protein 1b

Article abstract of DOI:10.1016/S0968-0896(03)00187-1

The interaction of a moenomycin derivative with the enzyme penicillin binding protein 1b (PBP 1b) has been studied by means of STD NMR. The results obtained initiated the synthesis of a number of moenomycin derivatives modified in unit A including a moenomycin-ampicillin conjugate and determination of their antibiotic activities. A protocol is described that allows studying the interaction of moenomycin analogues with PBP 1b by fluorescence correlation spectroscopy.

Full text of DOI:10.1016/S0968-0896(03)00187-1

Bioorganic & Medicinal Chemistry 11 (2003) 2965–2981
Studies on the Interaction of the Antibiotic Moenomycin A
with the Enzyme Penicillin-Binding Protein 1b
Thomas Ruhl,a Mohammed Daghish,a Andrij Buchynskyy,a Karen Barche,a
Daniela Volke,a Katherina Stembera,a Uwe Kempin,a Dietmar Knoll,a Lothar Hennig,a
Matthias Findeisen,a Ramona Oehme,a Sabine Giesa,a Juan Ayalab and Peter Welzela,*
aUniversitat Leipzig, Fakultat fur Chemie und Mineralogie, Johannisallee 29, D-04103 Leipzig, Germany
¨
¨
¨
bCentro de Biologica Molecular, Universidad Autonoma, Canto Blanco, E-28049 Madrid, Spain
Received 25 October 2002; accepted 18 March 2003
Abstract—The interaction of a moenomycin derivative with the enzyme penicillin binding protein 1b (PBP 1b) has been studied by
means of STD NMR. The results obtained initiated the synthesis of a number of moenomycin derivatives modified in unit A
including a moenomycin–ampicillin conjugate and determination of their antibiotic activities. A protocol is described that allows
studying the interaction of moenomycin analogues with PBP 1b by fluorescence correlation spectroscopy.
# 2003 Elsevier Science Ltd. All rights reserved.
Introduction
C55 anchor. The transglycoslyation reaction is catalyzed
by a number of multimodular bifunctional polymerases
(that catalyze also the transpeptidation reaction) desig-
nated as class A high molecular mass penicillin-binding
proteins (PBPs). Of these PBP 1b from Escherichia coli
has been studied in great detail.2 In addition, a number of
membrane-bound monofunctional glycosyltransferases
are known. Recently, a monofunctional glycosyl-
transferase from Staphylococcus aureus that shares con-
siderable homology with the transglycosylase domain of
bifunctional (class A) high molecular mass PBPs has
been expressed as a truncated protein lacking the mem-
brane domain and purified to homogeneity.3
Peptidoglycan is an essential cell wall constituent of
almost all eubacteria. It is a heteropolymer consisting of
glycan strands crosslinked by short peptide chains. A
number of indispensible functions can be attributed to
this macromolecule that is located at the outside of the
cytoplasmic membrane. It preserves cell integrity by
withstanding the internal osmotic pressure and gives the
cell a defined shape. Furthermore, the biosynthesis of
peptidoglycan is intimately associated with the cell divi-
sion process. The final events of peptidoglycan bio-
synthesis are the formation of the macromolecular
architecture from a disaccharide oligopeptide precursor
by two major types of reaction, (i) formation of the
polysaccharide strands (transglycosylation) and (ii) for-
mation of peptide crosslinkages between the glycan
chains (transpeptidation).
The moenomycin antibiotics are known to interfere with
the transglycosylation reaction.4,5 Moenomycin A (1a)
has been demonstrated to bind reversibly to E. coli PBP
1b.1 The structural similarities between the moeno-
mycins and both glycosyl donor and glycosyl acceptor
of the transglycosylation reaction are obvious. From
structure–activity relationships6,7 it has been concluded
that the moenomycins first bind to the cytoplasmic
membrane via their lipid moiety8 and that membrane
anchoring is an essential step preceding the highly
selective binding of the sugar part to the donor binding
site of the enzyme. The structural features that are
known to be responsible for the antibiotic activity are
indicated by arrows in 2.6,7,9 The structural similarity of
units C and E of 1a and 2, respectively, with the second
A representation of the transition state of the transglyco-
sylation reaction is shown in Figure 1.1
Both the glycosyl donor (the growing peptidoglycan
chain) and the glycosyl acceptor (the disaccharide-derived
lipid II) are attached to the cytoplasmic membrane via a
*Corresponding author. Tel.: +49-341-97-36551; fax: +49-341-97-
36599; e-mail: welzel@organik.chemie.uni-leipzig.de
0968-0896/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0968-0896(03)00187-1

2966
T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
Figure 1.
and the third sugar unit of the growing peptidoglycan
chain are obvious. The different binding of the first two
sugar units (1!4 in the growing peptidoglycan strand
versus 1!2 in 1a and 2, respectively) has been suggested
to be responsible for the inhibition of the enzyme.10
through fast exchange of ligand molecules from the
bound to the free state where the saturation transfer is
detected. We repeated the published methyl b-d-galacto-
pyranoside/Ricinus communis agglutinin experiment of
Mayer and Meyer14 and found the same transfer results
that were reported but with some deviations in quantity.
The results of SAR studies are in agreement with NMR
results. For a long time it was impossible to obtain good
quality 1H NMR spectra due to the formation of
moenomycin aggregates in aqueous solution. We suc-
On the other hand, experiments with moenomycin A
(1a) and an E. coli K12 (JM109/pJP13) membrane
extract in Tris–maleate buffer (in D2O, apparent pH
6.8) which contained PBP 1b solubilized by Triton
1
ceeded to overcome this obstacle when the H NMR
spectra were measured at concentrations below the
1
cmc.11 Well-resolved H NMR spectra of the moeno-
mycins in aqueous solution (c <5 Â 10À4 mol/L) were
then obtained which could be fully assigned.12 Based on
NMR-derived distance constraints and molecular
dynamics simulations a three-dimensional structure of
moenomycin in aqueous solution was proposed by Kurz
et al.13 as illustrated in Figure 2. The arrows in this
presentation indicate groups that have been shown to be
essential for the antibiotic activity (vide supra). Since
these substituents are exposed in close proximity at the
surface of the molecule it has been speculated that they
are part of a polar binding epitope.13
It is the purpose of the present publication to extend the
knowledge on the interaction of moenomycin with the
binding site at the enzyme.
Results and Discussion
STD experiments
Recently, saturation transfer difference (STD) NMR has
been applied by the group of Meyer14 and by others15 for
mapping carbohydrate epitopes in direct contact with
lectins. STD NMR relies on the transfer of saturation
from the protein to the ligand and is strongest for those
segments of the ligand that are in closest contact with
the protein. Saturation is transferred into solution
Figure 2. NMR solution structure (D2O) of moenomycin A (from ref 13).

T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
2967
X-100 were inconclusive. The STD spectrum showed
Triton X-100 and possibly moenomycin signals which,
however, could not be separated from each other. Then
membrane preparations in phosphate buffer (in D2O,
apparent pH 7.0) were used. The concentration of the
membrane fragments was about 40 mM and moeno-
mycin A was added until a final concentration of 4 Â
10À4 mol/L (this is close to the cmc, vide supra) was
reached. No STD could be observed. We attribute this
failure to the fact that binding of moenomycin to PBP
1b is very strong7 and possibly the rate of exchange
between free and bound ligand is so slow that the con-
centration of free ligand labeled by saturation transfer is
too lowto be detected. We have shown previously that
the lipid part contributes considerably to the binding
between moenomycin A and PBP 1b.7 Therefore, for
further STD experiments the delipido derivative 3,
which was prepared as described previously,4 was used.
(22 h), the resulting STD spectrum was only of modest
quality. Only the most intensive signals of the ligand
could be observed [N-acetyl groups at d=1.91 and 1.95;
4-methyl group of unit F (d=1.10), methyl group of
unit C (d=1.27) and the signal of the chromophore A
part (at d=2.26)] demonstrating that these groups had
been in contact with the protein. In comparison to the
other signals, the intensity of the N-acetyl group signals
was amplified to a much higher degree. On the other
hand, no signal for the anomeric proton of unit F at
d=5.66 could be detected. This feature indicates site
specific differences in the saturation transfer processes.
Because of the unfavourable quality of the spectrum not
only the integrals but also the peaks heights have been
compared carefully. Both modes indicated an intensity
increase of the N-acetyl signals in comparison to the
4-methyl group signal of unit F and the four-proton
singlet of the methylene groups of unit A. This result is
in agreement with the known absolute requirement of the
N-acetyl groups for the antibiotic activity of moeno-
mycin analogues and also with the speculative model of
Kurz13 derived from their moenomycin NMR structural
work. Interestingly, the STD NMR spectrum also
shows the signal of the chromophore A part (at d=2.26)
which has been shown to be of minor importance for
the antibiotic activity. Probably, ring A has contact with
the protein at positions different from the moenomycin
binding site. The findings of Kurz et al.13 suggest, that
rings A and B display a high degree of conformational
flexibility. In any case, the STD NMR result prompted
us to look again at the role of the chromophore unit A.
1
The 600 MHz H NMR spectrum of 3 in aqueous solu-
tion is depicted in Figure 3 (top). Using APT, TOCSY,
H,H COSY, 13C HSQC/13C HMQC and HMBC the 1H
and 13C spectra could be fully assigned. The data are
collected in Table 1 (Scheme 1).
For the newSTD experiment, a PBP 1b solution
(membrane fraction without detergents, about 40 mM in
PBP 1b, phosphate buffer, D2O, apparent pH 7.0,
20 ꢀC) was used and delipidomoenomycin 3 was added
(final concentration 4 Â 10À4 mol/L). The experiment
was performed essentially as described in ref 14 (see
Experimental). In spite of the long acquisition time
Figure 3. Autocorrelation G(t) of the interaction of PBP 1b with 11. Each sample was measured in 25 mM Tris/HCl, pH 7.2, 200 mM NaCl, 1%
Triton X-100 for 20 s. Data were analysed with ConforCor 2 Version 2.5 SP2 (two-components model for 11) and three-components model for PBP
1b/11. Tetramethylrhodamine (TMR, 10 nM); 11 (M-TMR, 10 nM); 11 (10 nM)+PBP 1b (160 nM).

2968
T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
Table 1. 1H, 13C, and 31P chemical shifts of 3 (in D2O)
Table 2. MIC values of moenomycin analogues with a modified
chromophore unit (in mol LÀ1
)
Position
1H chemical
shifts
13C chemical
shifts
31P chemical
shifts
1a
1b
5a
5b
6b
6c
8a
7a
7b
8b
6a
7Â10À9
6.0Â10À8
1Â10À7
6Â10À8
5Â10À8
4Â10À8
4Â10À7
1Â10À7
2Â10À7
2Â10À7
2Â10À7
1A
—
—
2.26
2.26
—
4.50
110.08
201.45
30.56
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
À0.7
—
—
—
2A
3A
4A
30.56
5A
201.45
103.10
70.84
1B
2B
ca. 3.47
ca. 3.63
4.14
3B
4B
69.10
74.87
169.95
101.49
55.82
5B
4.18
—
4.46
3.64
CONHB
1C
2C
substituents of differing chain length in the 3-position of
the aromatic ring by the usual Japp–Klingemann route
(Scheme 2).17
CH3CONHC
—
174.79
22.58
CH3CONHC
1.91/1.95
ca. 3.55
3.42
ca. 3.52
1.27
4.38
3C
4C
83.38
71.34
16.85
5C
Furthermore amines 5a, 6a, and 6b were treated with
4-isothiocyanatobenzoic acid to provide thioureas 5b,
7a, and 8a with an extra free carboxyl group. From
amines 6a and 6b on reaction with 4-isocyanato-
benzenesulfonic acid thioureas 7b and 8b were obtained
bearing sulfonic acid groups (Schemes 2 and 3).
6C
1D
103.44
2D
3.18
3.42
3.26
3.36
3D
75.93
69.91
76.19
60.95
102.30
55.24
174.50
22.49
4D
5D
6D
ca. 3.61; 3.81
4.46
1E
In the 1H NMR spectra of 5a and of 7a some anomalies
were observed. The chemical shifts of 4C-H of 5a and of
7a in methanol-d4–DMSO-d6 solution (d=2.42 and
d=2.46, respectively) are smaller than that of 4C-H of
moenomycin A (1a, d=3.17 in methanol-d4 solution,
d=3.09 in DMSO-d6 solution, referenced to d=5.69 for
1F-H). On the other hand, the 5B-H of 5a (d=5.02) and
7a (d=4.97, both spectra in methanol-d4–DMSO-d6
solution) had a larger chemical shift than that of moe-
nomycin A (1a, d=3.85 in methanol-d4 and d=3.88 in
DMSO-d6 solution, referenced to d=5.69 for 1F-H).
The reasons for these observations are at present not
clear. All other 1H chemical shifts of 5a and of 7a are in
the same region as those of 1a. The smaller chemical
shift of 4C-H of 5a and of 7a than in moenomycin A
could be the result of a shielding effect of the aromatic
ring.18 However, in 5a also the signal of 5C-H (d=2.96
in methanol-d4–DMSO-d6) appears at higher field than
in moenomycin A (1a, d=3.39 in methanol-d4, d=3.42
in DMSO-d6). If it is assumed that this upfield shift also
originates from a shielding effect of the aromatic ring
this would indicate two conformations. We did not
observe these anomalies in the spectra with a nitro-sub-
2E
CH3CONHE
—
1.91/1.95
CH3CONHE
3E
4E
ca. 3.47
79.85
5E
6E
ca. 3.63; 3.98
68.94
94.5*1
76.4*1
75.15
158.38
73.26
14.88
1F
5.66
3.74
4.90
—
2F
3F
OCONHF2
4F
—
CHF3
5F
1.10 (s)
4.33
—
CONHF2
P
172.96
—
68.9*1
72.5*1
177.8*1
—
1H
3.90/4.09
4.12
—
2H
3H
Moenomycin derivatives with modified units A
We have previously developed a method for removing
the chromophore part selectively from moenomycine by
K3[Fe(CN)6] oxidation to give 1b.16 We have now
determined the antibiotic activity of this compound and
found a MIC of 6.0 Â 10À8 mol LÀ1. For moenomycin 7
 10À9 mol LÀ1 have been determined in a parallel
experiment (see Table 2). This result means that the
interaction of the chromophore unit A with the enzyme
contributes to some extent to the antibiotic activity of
the moenomycins. Nevertheless, 1b is still an anti-
biotically very active compound. We took this occasion
to prepare a number of newmoenomycin A derivatives
modified in ring A in order to find out if different pen-
dants at ring A would lead to changes in the antibiotic
activity in a systematic way. First, we prepared the
4-nitrophenyl- and the 4-[(2-aminoethyl)carbamoyl]-
phenyl-substituted triazoles 4 and 5a and the two
4-nitro-aryl-substituted triazoles 6b and 6c with basic
1
stituent in the 4-position. The H NMR spectra of 7a
and 5a revealed a rather fast H ! D exchange in
methanol-d4–DMSO-d6 solution at the a-position of the
keto group (2A-H, d=3.19–3.13 in 5a and d=3.14 in
7a). This caused the 3A-H signal around d=2.50 to be a
doublet after exchange of one proton and as a singlet
after exchange of both protons (see Experimental).
Antibiotic activities
Compound 1b lacking the chromophore unit A has
about 10% of the antibiotic activity of moenomycin A
(vide supra), whereas in 6a 4% of the antibiotic activity
of moenomycin A are left. As described above, we have
prepared newderivatives with neutral aryl substituents

T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
2969
Scheme 1.
The moenomycin-ampicillin conjugate 10
at the triazole ring (4), with basic substituents (5a, 6b,
6c), with weakly acidic substituents (5b, 7a, 8a) and with
strongly acidic substituents (7b, 8b). Their MIC values
against seven Staphylococcus aureus strains were deter-
mined by a serial micro dilution method on microtiter
plates as described previously.8 As Table 2 shows, for
most compounds the MIC values are in the 10À7 mol LÀ1
range. Compounds 5b, 6b, and 6c have MIC values in
the 10À8 mol LÀ1 range. Two of them have additional
basic groups and one has an additional weakly acidic
function. From these results, one may conclude that
changes in the region of unit A of moenomycin cause
some decrease in the antibiotic activity. The decrease of
activity differs somewhat for different substituents but
we could not find a systematic correlation between anti-
biotic properties and newly introduced functional
groups. All these compounds are, however, still very
active when compared to other classes of antibiotics.
In viewof the advantages which may be offered by drug
synergism (for example impeding the development of
resistance), we thought it would be interesting to develop
chemistry that could be used to fuse moenomycin with
antibiotics such as penicillin or vancomycin. All these
compounds act at the outer face of the cytoplasmic
membrane and inhibit the last two steps of the pepti-
doglycan biosynthesis. In a preliminary experiment
without suitable spacers, ampicillin and 6a were coupled
via the squaric acid moiety under mild conditions. The
lowstability of ampicillin in aqueous solution at acidic
and basic pH values demanded reaction conditions
where the b-lactam ring of ampicillin is stable. Per-
forming the reaction between 6a and the ampicillin
sodium salt in a buffered solution at pH 9.0 furnished
the coupling product but with a hydrolized b-lactam

2970
T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
Scheme 2.
ring. Then the less nucleophilic polar solvent DMF with
added Et3N was used. First the moenomycin-squaric
acid derivative 6d was treated with ampicillin in DMF/
Et3N solution at 4 ꢀC. The purification problems in this
case (very similar Rf values of the product 10 and of 6d,
see Experimental) made this coupling sequence less
advantageous than the reversed order of events. Here
ampicillin squaric acid amide ester 9 was prepared from
ampicillin sodium salt and diethyl squarate in phos-
phate buffer at pH 7.3.
likely since the ESI ICR MS spectrum gave no indica-
tion of an impurity. Then the crude 9 was attached to 6a
in DMF/Et3N solution at 4 ꢀC. The chimeric product 10
was isolated after careful purification in 19% yield (not
1
optimized). The characteristic signals in the H and 13C
NMR spectra of 10 were found with the help of 2D
NMR experiments (1H,1H COSY, HMQC, HMBC)
and by comparison with the NMR spectra of 9. The IR
spectrum of 10 showed a band at 1755 cmÀ1 which is
characteristic for the b-lactam ring (for ampicillin at
1759 cmÀ1). The high resolution ESI ICR MS of com-
pound 10 showed besides the correct ion peaks at m/z
=1131.89462 ([M+Na-3H]2À), 1120.90120 ([MÀ2H]2À),
746.93268 ([MÀ3H]3À), 559.94854 ([MÀ4H]4) also
peaks that differed from the expected ions by an addi-
tional mass unit. In conclusion, we have shown that the
squaric acid method can be applied to the synthesis of
moenomycin–ampicillin chimeric derivatives with the
1
The structure of 9 was proven by H and 13C NMR
analysis based on comparison with the described 1H and
13C NMR spectra of ampicillin in the same solvent and
by high resolution ESI ICR MS. Both in the 1H and 13
C
NMR spectra one signal was doubled (see Experi-
mental). It is unclear whether this indicates an impurity
or two conformations. The former possibility seems less

T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
2971
Scheme 3.
intact b-lactam ring. The MIC value obtained for 10
was 1.8 Â 10À7 mol/LÀ1. Attempts to couple vancomy-
cin to 6a via the squaric acid linker were not successful
until now.
action of 11 with the purified and solubilized membrane
protein PBP 1b.21 In the first experiments, the average
diffusion time of 11 was determined. The measurements
demonstrated that on addition of detergent the average
diffusion time increased from 5.6 Â 10À5 to 3.00 Â
10À4 s due to the binding of 11 to Triton X-100 micelles
(see Fig. 5). In principle in a solution of 11 and micelles,
two diffusion times could be expected: (1) the ligand in
solution and (2) the ligand incorporated into micelles.22
But in our case the diffusion time 11 in solution could
not be detected. Obviously, 11 was completely incorpo-
rated into the micelles in agreement with the previous
observation that the lipid part of moenomycin binds
very efficiently to membranes, vesicles and micelles.8
For the analysis of the moenomycin/PBP 1b interaction
constant concentrations of 11 were measured with
increasing concentrations of PBP 1b. As soon as the
protein was added to the solution a diffusion time of
Fluorescence correlation spectroscopy for studying
moenomycin binding to PBP 1b
The interaction of moenomycin with PBP 1b has been
studied previously by means of surface plasmon reso-
nance.19 Here, we report preliminary results that show
that fluorescence correlation spectroscopy (FCS) can
likewise be used for this purpose (Fig. 4). FCS measures
statistical fluctuation of the fluorescence intensity in a
small illuminated sample volume to obtain information
about the binding properties of fluorescent molecules.
Moenomycin was labeled with tetramethylrhodamine to
give 11 as described previously.20 We studied the inter-

2972
T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
Experimental
All O2- or moisture-sensitive reactions were performed
in oven-dried glassware under a positive pressure of
argon. Liquids and solutions were transferred by syr-
inge. Small-scale reactions were performed in Wheaton
serum bottles sealed with aluminum caps with open top
and Teflon-faced septum (Aldrich). Usual workup
means partitioning the reaction mixture between an
aqueous phase and CH2Cl2, drying the combined
organic solutions over Na2SO4, and removal of solvent
by distillation using a rotatory evaporator (bath tem-
perature 45 ꢀC). Solvents were purified by standard
techniques. The following materials and methods were
used for chromatographic separations: flash chromato-
graphy (FC):23 silica gel 32–63 mm (ICN Biomedicals);
medium-pressure liquid chromatography (MPLC):
silica gel 40–60 mm (Grace), Duramat pump (CfG);
analytical TLC: Merck precoated silica gel 60 F254
plates (0.2 mm), spots were identified under a UV lamp
(l=254 nm, Camag 29 200) and with a 2.22 mol/L
Figure 4. Autocorrelation G(t) of the interaction of PBP 1b with 11.
Each sample was measured in 25 mM Tris/HCl, pH 7.2, 200 mM
NaCl, 1% Triton X-100 for 20 s. Data were analysed with ConforCor
2 Version 2.5 SP2 (two-components model for 11) and three-compo-
nents model for PBP 1b/11. Tetramethylrhodamine (TMR, 10 nM); 11
(M-TMR, 10 nM); 11 (10 nM)+PBP 1b (160 nM).
4.50 Â 10À4 s appeared in the autocorrelation function.
H2SO4 solution which contained Ce(SO4)2 4H2O (10 g/
.
24
.
L) and H3[PO4(Mo3O9)4] H2O (25 g/L) and heating at
140 ꢀC, or an anisaldehyde reagent for carbohydrates
[2 mL of anisaldehyde, 8 mL of concd H2SO4 in ethanol
(190 mL)]. Medium-pressure liquid chromatography
(MPLC): LiChroprep RP-18 material 40–63 mm
(Merck) was used. The samples were applied to a pre-
column (ꢁ2 g of RP-18 material) and eluted at 1.5–2.5
bar using a Duramat (CFG) dosage pump. Analytical
HPLC was performed using a HPLC system (Jasco)
consisting of an intelligent HPLC pump PU-980, a
three-line degasser DG-980-05, a ternary gradient unit
LG-980-02 and a multi-wavelength detector MD-910.
As eluent a mixture of buffer (prepared from 0.6 g
KH2PO4, 26.2 g K2HPO4 Â 3H2O, 3.0 g 1-heptane-
sulfonic acid sodium salt monohydrate and 1000 mL of
water, pH 8.0) and acetonitrile 60:40 was used. A col-
umn 250 Â 4.6 mm, nucleosil 300, C18, 5 mm (Jasco)
with a pre-column was used. Ultrafiltration (UF) was
performed using gas-pressurized (N2, 3.5 bar) stirred
ultrafiltration cells (Amicon, model 8050, 50 mL cell
capacity and model 8400, 400 mL capacity) with mem-
brane type YM3 (Amicon, molecular weight cut-off
3000 Da). NMR and MS equipment: NMR: DRX 400
(Bruker), DRX 600 (Bruker), Gemini 200 (Varian),
Gemini 2000 (Varian); Mass spectrometry: FAB MS:
VG Autospec (Fisons, matrix: 3-nitrobenzyl–alcohol),
ESI MS: FT ICR MS APEX II (Bruker Daltonics,
water–methanol). Following the molecular formula two
masses are always communicated, the first was calcu-
lated using the International Atomic Masses, the second
is the mono-isotopic mass. IR: Genesis FTIR (ATI
Mattson). For the description of the NMR spectra the
protons and carbons are indexed according to the indi-
ces in the formulae. The MIC (minimum inhibitory
concentration) values against seven different Staphylo-
coccus aureus strains (ATCC 25923, ATCC 29213,
MRSA 1309, SG 511, PEG 18, PEG 5, KNS PEG 5)
were determined by a serial twofold micro dilution
method on microtiter plates (Iso-Sensitest medium,
Oxoid). A series of decreasing concentrations of the
compound under investigation was prepared in the
Figure 5. Increasing of the diffusion times during the interaction of PBP
1b with 11. The diffusion times of tetramethylrhodamine (TMR) and 11
(MTMR) were influenced significantly by Triton X-100. TMR1 and
MTMR1: tretramethylrhodamine or 11 in 25 mM Tris/HCl, pH 7.2,
200 mM NaCl without 1% Triton X-100. TMR2 and MTMR2: tetra-
methylrhodamine or 11 in 25 mM Tris/HCl, pH 7.2, 200 mM NaCl with
1% Triton X-100. Addition of PBP 1b in 25 mM Tris/HCl, pH 7.2,
200 mM NaCl, 1% Triton X-100 resulted in complex formation. Tetra-
methylrhodamine: 10 nM; 11: 10 nM; complex: 10 nM 11 and 160 nM
PBP 1b.
This diffusion time can be attributed to 11 bound to
the protein. Thus, three different average diffusion
times have been identified: tetramethylrhodamine, 11
incorporated into micelles and the complex of 11
with PBP 1b in Triton X-100 micelles. The auto-
correlation functions for all three species are shown
in Figure 4. A shift to higher relaxation times can be
observed which is due to the increase of the mole-
cular weight of the fluorescent species. The differences
between the molar mass of 11 bound to micelles and
the protein–ligand–micelles complex is very small.
Competition experiments with moenomycin were per-
formed to verify the experiments and showed an
increase of the 11 fraction bound into micelles but not
bound to the protein. The relative molecular mass of the
solubilized PBP 1b could be estimated to be 124 kDa
including a contribution of about 30–50 kD from
micelles. Aggregations of the protein and micelles were
observed in all experiments.

T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
2973
medium. For inoculations 1Â105 cfu mL-1 were used.
The MICs were determined (absence of visible tur-
bidity) after 24 h at 37 ꢀC. The MIC values were cal-
culated as the average values from three
measurements.
(d, 1H, 3F-H, J2F, 3F=10.2 Hz), 2.63 (d, 2H, CH2-12I,
J12, 13=7.1 Hz), 2.49 (t, 2H, CH2-3A, J2A, 3A=6.2 Hz),
1.95, 1.93 (2Âs, 6H, NHCOCHC3 ; E), 1.84–1.76 (m, 2H,
CH2-10I), 1.62 (s, 3H, CH3-25I), 1.55 (s, 3H, CH3-19I),
1.51 (s, 3H, CH3-21I), 1.49 (s, 3H, CH3-20I), 1.29–1.21 (m,
2H, CH2-9I), 1.11 (s, 3H, CH3-4F), 1.08 (d, 3H, CH3-6C,
J=6.0Hz), 0.83 (s, 6H, CH3-23I, CH3-24I).
C75H111N8O36P (1731.71, 1730.68), ESI ICR MS: m/z
(R)-3-({(5R)-5-[3-(3-Carboxypropionyl)-1-(4-nitrophenyl)-
1H-1,2,4-triazol-5-yl]-ꢀ-L-arabinopyranosyl-(1!4)-2-
acetamido-2,6-dideoxy-ꢁ-D-glucopyranosyl-(1!4)-[ꢁ-D-
glucopyranosyl-(1!6)]-2-acetamido-2-deoxy-ꢁ-D-gluco-
pyranosyl-(1!2)-3-O-carbamoyl-4-C-methyl-ꢀ-D-gluco-
pyranuronamidosyloxy}hydroxyphosphoryloxy) - 2 - ((2Z,
6E,13E)-3,8,8,14,18-pentamethyl-11-methylene-2,6,13,17-
nonadecatetraenyloxy)propionic acid (4). To a solution
of 4-nitroaniline (62.5 mg, 0.45 mmol) in 9% HCl
=864.33338 (calcd 864.33465) [MÀ2H]2À
,
m/z
=875.32560 (calcd 875.32563) [MÀ3H+Na]2À
.
(5BR)-5B-(1-{4-[(2-Aminoethyl)carbamoyl]phenyl}-3-(3-
carboxypropionyl) - 1H - 1,2,4 - triazol - 5 - yl) - 5B - de[(2-
hydroxy - 5- oxo - 1- cyclopenten - 1 - yl)carbamoyl]moeno-
mycin A (5a). To a solution of 4-(2-amino-ethylcarba-
moyl)-aniline (125 mg, 0.7 mmol) in 9% HCl (5 mL) at
0 ꢀC a precooled (0 ꢀC) solution of sodium nitrite (50 mg
0.7 mmol) in water (10 mL) was added. The mixture was
stirred at 0 ꢀC for 15 min. The solution of the diazonium
salt was then added slowly to a solution of moenomycin
complex (main component moenomycin A, 1 g) and
sodium acetate (15 g) in water (500mL). The reaction
mixture was stirred for 144 h at ambient temperature.
Progress of the reaction was monitored by TLC (n-pro-
panol–water 7:3). After addition of water the solvent
was removed by freeze-drying. Careful and repeated FC
(n-propanol–2 mol/L ammonia 10:3 and RP-18 FC (water,
acetonitrile–water 1:3 ! 1:1 furnished a pure sample of
5a (43 mg, 4%) as a pale yellowsolid. C 78H118N9O35P
(1772.81, 1771.746792), ESI MS: m/z 884.86774,
calcd 884.86612 [MÀ2H]À2. NMR spectra: Unit I
(below):
(1.2 mL)
a solution of sodium nitrite (31.3 mg,
0.45 mmol) in water (1 mL) was added slowly at 0 ꢀC.
After 20 min at 0 ꢀC, this solution was added slowly to a
solution of 1a (500 mg, 0.32 mmol) and sodium acetate
(5 g) in water (50 mL), and the mixture was stirred at
20 ꢀC for 48 h. The reaction mixture was desalted by
RP18 chromatography eluting with water (200 mL) and
then with acetonitrile–water 1:1 (UV monitoring at
l=280 nm). Solvent evaporation followed by FC
(CHCl3–methanol–water 9:6.5:1.4), Sephadex LH 20
LC (methanol–water 3:1), and lyophilization furnished
4 (416 mg, 76%) as a pale orange solid. Rt=14.8 min,
RP HPLC (buffer–acetonitrile=60:40, lmax=280 nm,
flow0.5 mL min À1). UV (MeOH): lmax (e)=272 nm
(10,123). 1H NMR (400 MHz, D2O, H,H COSY): charac-
teristic signals at d=8.52 (d, 2H, 3Ar-H, 5Ar-H), 7.85 (d,
2H, 2Ar-H, 6Ar-H, J=8.5 Hz), 5.65 (bs, 1H, 1F-H), 4.92
Assignment
1H NMR (600 MHz,
CD3OD/DMSO-d6,
H, H COSY, HMBC,
HMQC)
13C NMR
(150 MHz,
CD3OD/
Assignment
1H NMR (600 MHz,
CD3OD/DMSO-d6,
H, H COSY, HMBC,
HMQC)
13C NMR
(150 MHz,
CD3OD/
DMSO-d6)
DMSO-d6)
Unit Position
Unit Position
I
1
2
3
4
5
6
4.03, m, 3.89, m
5.26, m
—
1.93, m
1.95, m
5.10, dt, J5, 6=6 Hz,
J6, 7=16 Hz
5.19, d, J6, 7=16 Hz
—
1.19, m
1.72, m
—
2.52, d, J12, 13=7 Hz
4.97, t, J=7 Hz
—
1.87, m
1.92, m
4.94, t, J=7 Hz
—
1.51, s
1.43, s
1.44, s
67.1
124.0
140.4
33.7
32.7
127.0
24
25
0.78, s
1.57, s
—
3.80, m
4.07, m, 3.90, m
—
3.19–3.13 (exchangable)
2.50, d, J=6 Hz (probably
caused by H!D exchange
of one proton at C-2A)
—
4.33, unter H2O
3.36, m
3.46, m
4.16, broad signal
5.02, broad signal
4.24, m
3.25, m
3.23, m
24.3
?
H
A
1
2
3
1
2
3
81.6
69.0
194.0
36.1
29.9
7
8
9
141.5
36.6
42.9
32.4
151.1
36.1
123.5
137.5
41.0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
4
1
2
3
4
5
1
2
177.0
103.8
72.4
74.1
71.1
71.2
102.5
56.9
?
84.3
72.1
18.6
173.4
24.0
B
C
27.8
125.6
132.4
26.4
18.3
16.5
3
4
5
2.42, broad signal
2.96, broad signal
1.01, broad signal
—
6
(CH3CONH)
(CH3CONH) 1.85 s
4.51, s, 4.50, s
0.78, s
109.7
28.13, 28.18

2974
T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
Assignment
1H NMR (600 MHz,
CD3OD/DMSO-d6,
H, H COSY, HMBC,
HMQC)
13C NMR
(150 MHz,
CD3OD/
Assignment
1H NMR (600 MHz,
CD3OD/DMSO-d6,
H, H COSY, HMBC,
HMQC)
13C NMR
(150 MHz,
CD3OD/
DMSO-d6)
DMSO-d6)
Unit Position
Unit Position
D
1
4.29, covered by the
H2O signal
3.02, dd, J1,2=J2,3=9 Hz
3.26, m
3.07, m
3.19, m
3.70, d, JA, B=10 Hz,
3.47, m
4.46, covered by the
H2O signal
3.54, m
3.46, m
3.38, m
3.49, m
3.96, d, JA, B=10 Hz, 1H
—
104.7
F
1
2
3
4
5.69, s
3.50, m
4.88, d, J2, 3=10 Hz
—
4.26, s
—
—
1.05, s
95.9
78.6
75.8
74.3
73.7
174.2
158.8
17.0
160.0
155.3
141.7
130.1
126.8
136.2
168.4
?
2
3
4
5
6
75.1
78.2
71.9
78.3
62.8
5
6
(OCONH2)
(CH3)
3
5
1
2, 6
3, 5
4
E
1
103.2
TA
Ar
—
—
—
2
3
4
5
56.7
73.9
81.8
?
69.6
173.4
23.9
8.04, d, J=9 Hz
7.69, d, J=7 Hz
—
—
3.61, broad signal
3.52, m
6
(CH3CONH)
(CH3CONH) 1.85, s
DAE (CONH)
1
2
?
Impurity signals (from RP-18): 1H NMR: 0.74 (m); 0.80 (m); 1.13 (s); 1.14 (s); 1.24 (m) 1.48 (m); 1.74 (s); 2.75 (m); 13C NMR: 14.4; 20.8; 21.3; 30.7;
45.0; 80.0.
(5BR)-5B-{1-[4-({2-[3-(4-Carboxyphenyl)thioureido]ethyl}-
carbamoyl)phenyl]-3-(3-carboxypropionyl)-1H-1,2,4-tri-
azol-5-yl}-5B -de[(2-hydroxy-5-oxo-1-cyclopenten-1-
(methanol, 1.2 Â 10À4 mol L-1): lmax (emax)=370
.
1267 cmÀ1 1H NMR (CD3OD, 200 MHz): d=8.51
(2603), 372 (2605). IR (KBr): 3363, 1653, 1593, 1313.
yl)carbamoyl]moenomycin
A
(5b).
5a
(43 mg,
(1H, CONH, W1=2 ¼ 5:6 Hz), 7.97 (1H, d, 9.12 Hz,
Ar–H), 6.70–6.57 (1H, m, Ar–H), 6.55 (1H, s, Ar–H),
3.75–3.567 (8H, m, CH2OðCH2Þ2OCH2), 3.57–3.52
(2H, m, CONHCH2), 3.15–3.10 (2H, m, CH2NH2).
13C NMR (DMSO-d6, 50MHz): d=167.19 (CO),
0.025 mmol) was dissolved in 1:1 pyridine–water
(1.5 mL) assisted by sonication. After addition of a
solution of 4-isothiocyanatobenzoic acid (7 mg,
0.037 mmol) in 1:1 pyridine–water (1 mL) the mixture
was stirred at 50 ꢀC for 30 min and for 6 days at ambient
temperature. Then a solution of 4-isothiocyanato-
benzoic acid (3 mg, 0.017 mmol) in 1:1 pyridine–water
(1 mL) was added and stirring continued for 24 h at
20 ꢀC. Progress of the reaction was followed by TLC (n-
propanol–water 7:3). Water was added and solvents were
removed by freeze-drying furnishing 5b (23 mg, 48%) as
a pale yellowsolid. 31P NMR (81 MHz, CD3OD/DMSO-
d6): d=À1.28 (s). C86H123N10O37PS (1952.01,
1950.750892) ESI MS: m/z 634.90768, calcd 634.90775
À
Á
154.64 (CNHA2 r), 137.05 (CNOA2 r), 132.97 CCOAr
,
127.36,
ðOCH2CH2NH2Þ,
FAB MS: m/z=313.1 ([M+H]+).
112.32,
111.88
69.64
(NHCH2CH2O). C13H20N4O5 (312.33, 312.14337),
(3Â
CHAr),
ðOCH2CH2OÞ,
73.02
68.69
5-Amino-N-(13-amino-3,6,9-trioxadecane)-2-nitrobenz-
amide. The title compound was prepared as described
for 5-amino-N-(8-amino-3,6-dioxaoctane)-2-nitrobenz-
amide. Yellowoil, yield: 61%. 1H NMR (pyridine-d5,
300 MHz): d=9.31 (1H, W1=2 ¼ 5:6 Hz, CONH), 7.43
(2H, bs, Ar–NH2), 7.18–7.17 (1H, d, 8.5 Hz, Ar–H),
6.80–6.79 (1H, dd, J=9.0, 2.4 Hz, Ar–H), 6.77–6.76 (1H,
d, J=2.4 Hz, Ar–H), 3.81–3.75 (2H, q, J=6.3Hz,
CONHCH2), 3.65–3.61 (2H, t, J=6.3 Hz, OCH2), 3.60–
3.51 (10, m, 5Â OCH2), 3.14–3.09 (2H, t, J=6.6Hz,
CH2NH2), 2.10–2.02 (2H, quintet, J=6.6 Hz, CH2),
1.94–1.88 (2H, quintet, J=6.3 Hz, CH2). 13C NMR
(D2O, 50 Hz): d=172.93 (CO), 159.08 (CNHA2 r), 139.48
(CAr–NO2), 137.85 (signal that could not be assigned),
[MÀ3HÀCONH]À3
974.36913, calcd 974.36817) [MÀ2H]À2; 985.35781, calcd
985.35914) [M+NaÀ3H]À2; 996.35030, calcd 996.35011
[MÀ2H+2NaÀ2H]2.
;
649.24285,
calcd
649.24302
[MÀ3H]À3; 656.57016, calcd 656.57034 [M+NaÀ4H]À3
;
5-Amino-N-(8-amino-3,6-dioxaoctane)-2-nitrobenzamide.
To a stirred solution of 5-amino-2-nitrobenzoic acid
(700.0 g, 3.8 mmol) in dry pyridine (25 mL) CDI
(800.0 g, 4.9 mmol) was added at 20 ꢀC. After 10 min,
the mixture was added to a solution of 1,8-diamino-
3,6-dioxaoctane (2 mL) in 5 mL pyridine (5 mL).
Heating to 80 ꢀC for 24 h, solvent evaporation and
FC (methanol-chloroform 3:7) gave the title com-
pound (950.0 mg, 80%) as a yellowoil. UV/VIS
À
Á
132.75 CCOAr , 121.68, 118.75, 117.15 (3Â CHAr),
74.08, 73.94, 73.84, 73.74, 72.73 (5Â OCH2), 42.15
(CH2NH2), 41.31 (CONHCH2), 32.37, 30.93 (2Â CH2).
C17H28N4O6 (384.43, 384.201), FAB MS: m/z=385.1
([M+H]+).

T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
2975
(5BR)-5B-{1-[3-(15-Amino-6,9,12-trioxa-2-azapentadeca-
noyl)-4-nitrophenyl]-3-(3-carboxypropionyl)-1H-1,2,4-
triazol-5-yl}-5B-de(2-hydroxy-5-oxocyclopent-1-enylcarb-
amoyl)moenomycin A (6c). To a stirred solution of
5-amino-N-(13-amino-3,6,9-trioxadecane)-2-nitrobenza-
mide (31.7 mg, 77 mmol) in 6% HCl (0.3 mL) at 3 ꢀC a
precooled (3 ꢀC) solution of sodium nitrite (6.2 mg,
89.86 mmol) in water (0.25 mL) was added. After 15 min,
the diazonium salt solution was dropped within 20 min
into a solution of moenomycin A (118.0 mg, 74.6 mmol)
sodium acetate (600.0 mg) in water (50 mL, cooled to
10 ꢀC). The mixture was stirred at 20 ꢀC for 24 h. The
mixture was passed through a RP 18 column (elution
with acetonitrile–water 3:7). Product fractions were
combined and solvents were removed by evaporation.
The residue was redissolved in water and freeze-dried.
FC (H2O–n-propanol 2:7), solvent evaporation and
lyophilization provided 6c (100.0 mg, 68%) as a pale
J=7.5Hz, CHA2 ), 2.66–2.65 (2H, d, J=6.8 CH2IÀ12), 2.60
(2H, s, CH2A), 2.14–1.99 (14H, m, CH2IÀ4;5;15;16, CH2
NH2spacer, CH3CONHC;E), 1.88 (2H, m, CH2I-10), 1.75
(2H, s, CHI2-25), 1.68–1.65 (5H, m, NHCHs2pacer, CH3I-19),
1.58 und 1.57 (6H, je s, CHI3-20,21), 1.44–1.40 (2H, m,
impurity from RP-18), 1.35 (4H, m, H2NCHs2pacer, CH2I-
9), 1.25 (3H, s, CHF3 ), 0.98–0.97 (2H, t, impurity from
RP-18), 0.95 (6H, s, CHI3-23,24). 13C NMR (HMBC,
HMQC, CD3OD/DMSO-d6 4:1, 150, 100 MHz):
d=195.11 (COA), 180.45 (COOHA), 176.80 (COOHH),
À
Á
173.95, 173.19 CH3CONHC; E , 167.32 ðARCONHÞ,
159.21 (CONHF), 158.43 (CTA), 154.32 (CTA), 149.87
(CHI-11), 146.69 (CAr), 141.54 (CAr), 140.63 (CHI-7),
136.36 (CHI-3), 133.77 (CAr), 131.41 (CHI-14), 128.42
(CAr), 127.01 (CAr), 126.06 (CHI-6), 125.59 (CHI-17),
122.02 (CHI-13), 121.42 (CHI-2), 108.98 (CHI-22),
104.61, 102.73, 101.61 (CHB-1,C-1,D-1), 94.94 (CHF-1),
83.47 (CHC-4), 80.87, 80.04 (CHH-2, CHE-4), 77.30,
76.72, 76.59, 75.19, 73.82, 73.47, 73.16, 72.44, 72.25,
71.17, 69.83, 67.80 (carbohydrate signals), 70.40, 70.21
(OCHs2pacer), 66.98 (CH2I-1), 66.25 (OCHE2 ), 61.43
(HOCHD2 ), 55.53 (CHE-2,C-2), 42.00 (CH2I-9), 40.10
(CHI2-15), 39.67, 39.64 (CH2NH2, CONHCH2spacer),
35.63 (CHI2-12), 35.21 (CHI2-8), 32.52, 31.84, 31.53, 31.42
(CH2I-4;5;10, CI-8), 29.46 (CH2A), 27.34, 27.31 (CH3I-23,24),
26.86 (CHI3-19), 25.66 (CH3I-25), 23.53, 22.82
(CH3CONHC;E), 17.58 (CH3I-20), 17.15 (CH3C),
15.85 (CHI3-21), 15.48 (CH3F). C82H125N10O39P,
yellowhygroscopic podwer.
1H NMR (CD3OD/
DMSO-d6 0.5:0.4, 600 MHz): d=8.14–8.12 (1H, d,
J=8.4 Hz, Ar–H), 7.98 (1H, d, J=8.4 Hz, Ar–H), 7.91
(1H, s, Ar–H), 5.69 (1H, s, CHF-1), 4.37–3.67 (14H, m,
carbohydrate signals, 3.44–3.26 (20H, m, CHI2, CH2spacer),
2.89–2.77 (2H, m, CHD-2,3), 2.52–2.51 (2H, d, J=7.08
Hz, CHI2-12), 2.44–2.42 (2H, m, CHA2 ), 2.00–1.82 (10H,
m, CH2I-4,5,15,16 CH3CONHC, E), 1.78–1.71 (4H, m, CHI2-10
,
CHs2pacer), 1.57 (3H, s, CH3I-25), 1.49 (3H, s, CH3I-19),
1.43 (3H, s, CH3I-20), 1.42 (3H, s, CHI3-21), 1.21–1.11
(7H; CH2spacer, CH2I-9, CH3CÀ6), 1.04 (3H, s, CHF3 ), 0.79
(6H, s, CH3I-23, 24).- 13C NMR (CD3OD/DMSO-d6
0.5:0.4, 100, 150 MHz): d=195.97 (COA), 181.53
1905.91,
1904.78),
ESI
MS:
m/z=951.38600
([MÀH]2À), calcd 951.38487, 633.91943 ([MÀ3]3À),
calcd 633.92082.
À
Á
(COOHA), 174.85, 173.19 CH3CONHC; E , 167.65
ðARÀCONHÞ, 159.65 (CONHF2 ), 159.02 (COTA),
154.74 (CTA), 150.10 (CHI-11), 147.15 (CAr–NO2),
142.00, 141.50, 141.10 (CHI-7, CI-3, CI-14), 136.78 (CI-
18), 134.37 (CAr), 131.68 (CHI-6), 126.69 (CHAr), 125.23
(CHI-17), 123.10 (CHI-2), 109.66 (CHI2-22), 104.29, 103.56
(CHB-1, C-1, D-1, E-1), 77.12, 74.03 (Cpentasaccharide), 70.60,
69.35 (bs, 8Â CHs2pacer), 61.92 (HOCH2D), 56.03 (CHE-2,C-
2), 42.56 (CHI2-9), 40.56 (CHI2-15), 38.68, 38.10
(CH2NH2, CONHCH2spacer), 36.19, 35.71 (CH2I-12, CI-8),
33.11, 32.32, 31.92 (CHI2-4, 5, 10), 29.28 (CHA2 ), 28.01,
27.67, 27.51 (CH3I-23;24, CH2I-16), 26.40 (CH3I-19), 24.24
(5BR)-5B-[1-(3-{[2-(2-{2-[3-(4-Carboxyphenyl)thioureido]-
ethoxy}ethoxy)ethyl]-carbamoyl}-4-nitrophenyl)-3-(3-car-
boxypropionyl)-1H-1,2,4-triazol-5-yl]-5B-de[(2-hydroxy-
5-oxo-1-cyclopenten-1-yl)carbamoyl]moenomycin A (8a).
To a solution of 6b (60 mg (0.031 mmol) in 1:1 pyri-
dine–water (1 mL, sonication) a solution of 4-iso-
thiocyanatobenzoic acid (14 mg 0.081 mmol) in 1:1
pyridine–water (2 mL, sonication) was added in two
portions. The mixture was stirred at 50 ꢀC for 1 h and
at 20 ꢀC for 12 h. Progress of the reaction was mon-
itored by TLC (n-propanol–water 7:3). Addition of
water, lyophilization and RP-18 chromatography
(water, acetonitril–water 1:4) provided 8a (28 mg, 43%)
as a pale yellowsolid. 31P NMR (122 MHz, CD3OD/
DMSO-d6): d=0.04 (s). C90H130N11O41PS (2085.11,
2083.788400) ESI MS: m/z 1040.88804, calcd
À
C; DÁ
(CHI3-25), 23.44 CH3CONH
, 18.33 (CHI3-21), 17.68
(CHC3 ), 16.62 (CH3I-20), 16.00 (CHF3 ). 31P NMR
(CD3OD, 80 MHz): d=À1.18. C86H133N10O40P
(1978.01, 1976.842), ESI MS: m/z=987.41356
([MÀ2H]2À), calcd 987.41363, 657.94185 ([MÀ3H]3À),
calcd 657.9400.
1040.88692 [MÀ2H]À2
.
(5BR)-5B-{1-[3-(10-Amino-5,8-dioxa-2-azadecanoyl)-4-
nitrophenyl]-3-(3-carboxypropionyl)-1H-1,2,4-triazol-5-
yl}-5B-de(2-hydroxy-5-oxocyclopent-1-enylcarbamoyl)-
moenomycin A (6b). 6b was prepared from moenomycin
A and 5-amino-N-(8-amino-3,6-dioxaoctane)-2-nitro-
benzamide as described for 6c. Yield: 55%. Pale yellow
hygroscopic powder. 1H NMR (CD3OD, 600 MHz):
d=8.47–8.45 (1H, d, J=8.4 Hz, Ar–H), 8.10 (1H, d,
J=8.4Hz, Ar–H), 8.05 (1H, s, Ar–H), 5.85 (1H, s, CHF-1),
(5BR)-5B-{1-[3-({2-[3-(4-Carboxyphenyl)thioureido]ethyl}-
carbamoyl)-4-nitrophenyl]-3-(3-carboxypropionyl)-1H-
1,2,4-triazol-5-yl}-5B-de[(2-hydroxy-5-oxo-1-cyclopen-
ten-1-yl)carbamoyl]moenomycin
A
(7a). 6a (45 mg
(0.025 mmol) was converted to 7a by reaction with
4-isothiocyanatobenzoic acid (13.7 mg, 0.074 mmol) as
described for 8a. RP18 chromatography (acetonitrile–
water 3:7 gave 7a (32 mg, 65%) as a colourless solid.
31P NMR (81 MHz, CD3OD/DMSO-d6): d=À0.06
(s). C86H122N11O39PS (1997.00, 1995.73597), ESI MS:
5.44 (1H, bs, CHI-7), 5.36–5.06 (8H, m, CHI-2,6,17,13, CHB-5
,
CHF-3), 4.67–4.66 (2H, d, J=7.8 Hz, CHI2-22), 4.61–3.32
(38H, m, carbohydrate signals), OCHs2pacer, CH2I-signals),
3.27–3.20 (3H, m, CHD-2, CH2A), 3.00–2.98 (2H, t,
m/z
664.23662,
calcd
664.23805
[MÀ3H]À3
,
996.86145, calcd 996.86071 [MÀ2H]À2. NMR spectra
(below):

2976
T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
Assignment
1H NMR (600 MHz,
CD3OD/DMSO-d6,
H,H COSY)
13C (150 MHz,
CD3OD/
DMSO-d6,
HMBC/
HMQC)
Assignment
1H NMR (600 MHz,
CD3OD/DMSO-d6,
H,H COSY)
13C (150 MHz,
CD3OD/
DMSO-d6,
HMBC/
HMQC)
Unit Position
Unit Position
I
1
2
3
4
5
6
3.99, m, 3.87, m
5.21, m
—
1.90, m
1.87, m
5.08, dt, J5, 6=6 Hz,
J6, 7=16 Hz
5.16, d, J6, 7=16 Hz
—
1.16, m
1.69, m
—
2.48, d, J12, 13=7 Hz
4.93, t, J=8 Hz
—
1.81, m
1.89, m
4.89 t, J=7 Hz
—
1.46, s
1.39, s
1.40, s
4.46, covered by the H2O
signal
0.75, s
0.75, s
1.53, s
—
3.76, broad signal
4.03, m, 3.89, m
—
67.1
123.6
140.8
33.6
32.7
127.1
4
5
6
2.71, m
3.30, m
1.16, m (not separated
from CH2-9I)
85.2
72.4
17.9
(CH3CONH)—
(CH3CONH)1.81 s
1
2
3
4
173.6
23.7
104.8
75.1
78.0
71.8
D
E
4.24, m
3.00, m
3.25, m
3.07, m, covered by the
methanol signal
3.15, m
3.66, m, 3.46, m
4.22, m
3.64, m
3.38, m
3.17, m
7
8
9
141.5
36.5
42.9
32.4
151.1
36.0
123.6
137.4
41.0
10
11
12
13
14
15
16
17
18
19
20
21
22
5
6
1
2
3
4
5
6
78.2
62.7
104.3
55.8
74.4
82.6
74.9
70.0
174.1
23.7
95.9
79.1
76.2
74.3
73.9
174.1
159.1
16.6
27.8
125.5
132.3
26.2
18.0
16.4
3.26, m
3.85, m 1H
(CH3CONH)—
(CH3CONH)1.80, s
1
2
3
F
5.69, s
3.41, m
4.83, d, J=10 Hz
—
4.25, s
—
—
1.01, s
—
—
—
7.95, s
—
—
109.4
23
24
25
1
2
3
1
2
3
28.1
28.1
24.2
?
80.9
69.0
193.9
4
5
6
(OCONH2)
(CH3)
H
A
TA
Ar
3
5
1
2
3
4
5
6
?
155.7
147.8
126.7
135.3
142.5
127.3
128.8
167.8
? (40.7)
? (41.0)
?
3.14, m (mainly exchanged) 35.6
2.46, s, (through H!D
exchange of both protons
at C-2A)
29.9
4
1
2
3
4
5
1
2
3
—
177.4
105.4
72.1
73.9
71.3
71.1
102.9
56.7
73.5
8.14, d, J5, 6=9 Hz
7.90, d
—
3.51, m
3.44, m
B
C
4.35, d, J=7 Hz
3.42, m
3.47, m
4.05, broad signal
4.97, broad signal
4.22, m
DAE(CONH)
1
2
(NHCSNH) —
—
BL
1
143.6
131.6
123.4
3.38, m
3.15, m
2, 6
3, 5
7.71, d, J=8 Hz
7.34, d, J=8 Hz
Impurity signals (from RP-18): 1H NMR: 0.74 (m); 0.80 (m); 1.13 (s); 1.14 (s); 1.24 (m) 1.48 (m); 1.74 (s); 2.75 (m); 13C NMR: 14.4; 20.8; 21.3; 30.7;
45.0; 80.0.
(5BR)-5B-{3-(3-Carboxypropionyl)-1-[4-nitro-3-({2-[3-(4-
sulfophenyl)thioureido]ethyl}carbamoyl)phenyl]-1H-1,2,4-
triazol-5-yl}-5B-de[(2-hydroxy-5-oxo-1-cyclopenten-1-yl)-
carbamoyl]moenomycin A (7b). 6a (48 mg, 0.026 mmol)
was dissolved in 1:1 pyridine–water (1 mL) assisted by
19 h. Progress of the reaction was followed by TLC
(n-propanol–water 7:3). Water addition, lyophilization
and RP 18 chromatography (water, acetonitrile–water
1:4) gave 7b (40 mg, 75% as a pale yellowsolid. 31P
NMR (243 MHz, CD3OD/DMSO-d6): d=À1.45 (s).
C85H122N11O40PS2 (2033.05, 2031.702957) ESI MS:
sonication.
4-Isothiocyanatobenzenesulfonic
acid
(15 mg, 65 mmol) dissolved (sonication) in 1:1 pyri-
dine–water (2 mL) was added in two portions. The
mixture was stirred at 50 ꢀC for 1 h and at 20 ꢀC for
m/z
676.22747,
calcd
676.22704
[MÀ3H]À3
;
1014.84557, calcd 1014.84420 [MÀ2H]À2. NMR spectra
(below):

T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
2977
1H NMR (600 MHz,
CD3OD/DMSO-d6,
H, H COSY, HMBC,
HMQC)
13C NMR
(150 MHz,
CD3OD/
Assignment
1H NMR (600 MHz,
CD3OD/DMSO-d6,
H, H COSY, HMBC,
HMQC)
13C NMR
(150 MHz,
CD3OD/
Assignment
DMSO-d6)
DMSO-d6)
Unit Position
Unit Position
I
1
4.12, broad signal,
3.98, broad signal
5.32, m
—
2.02, m
2.01, m
5.20, dt, J5, 6=6 Hz,
J6, 7=16 Hz
5.31, d, J6, 7=16 Hz
—
1.27, m (not separated
from CH3-6C)
1.81, m
66.9
3
4
5
6
3.24, m
2.83, m
3.39, m
1.26, partial signal
overlap
—
73.3
85.5
72.3
18.0
2
3
4
5
6
123.9
140.5
33.7
32.7
127.0
(CH3CONH)
(CH3CONH) 1.909, s
1
2
3
4
173.2
23.9
104.6
D
E
F
4.30, m
7
8
9
141.1
36.6
42.9
3.12, dd, J1, 2=J2, 3=8 Hz 75.1
3.31, m
78.2
71.7
3.15, m, covered by the
methanol signal
3.23, m
3.75, d, JA, B=11 Hz,
3.55, m
4.25, d, J1, 2=8 Hz
3.66, m
3.45, m
3.32, m
3.52, m
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
1
32.4
151.1
36.0
123.5
137.4
40.9
5
6
78.2
62.7
—
2.60, d, J12, 13=8 Hz
5.05, t, J=7 Hz
—
1.93, t, J=8 Hz
2.00, m
5.01, t, J=7 Hz
—
1.57, s
1.51, s
1.52, s
4.59, s, 4.58, s
0.87, s
0.87, s
1.65, s
—
3.84, broad signal
4.12, broad signal,
3.98, broad signal
—
3.21–3.25, (exchangeable)
2.52, covered by the
DMSO signal
—
1
2
3
4
104.3
55.9
?
82.4
?
27.8
125.5
132.3
26.3
18.2
16.5
109.6
28.1
28.1
5
6
(CH3CONH)
(CH3CONH) 1.906, s
1
2
3
3.94, d, JA, B=10 Hz, 1H
—
69.4
173.4
23.8
95.9
78.8
75.9
74.2
73.8
174.2
158.8
16.9
160.3
155.7
147.1
126.8
135.4
142.6
127.2
128.5
5.69, s
3.51, m
4.90, d, J2, 3=10 Hz
—
4.29, broad signal
—
—
1.13, s
—
—
—
8.05, s
—
—
8.27, d, J5, 6=9 Hz
8.04, d (overlapping
with 2-HAr
—
3.63, m, 1H
3.54, m
—
—
7.66, d, J=9 Hz
4
5
24.3
H
A
177.0
81.2
69.1
6
(OCONH2)
(CH3)
2
3
TA
Ar
3
5
1
2
3
4
5
6
1
2
3
194.1
36.0
30.8
4
1
177.9
105.4
B
C
4.48, covered by the
H2O signal
3.49, m
3.55, m
4.16, s
5.12, s
4.37, covered by the
H2O signal
3.48, m
2
3
4
5
1
72.1
74.0
71.3
71.3
)
DAE (CONH)
167.5
1
2
? (40.5)
? (40.5)
182.4
141.6
128.0
102.9
(NHCSNH)
1
2, 6
BL
2
56.7
Impurity signals (from RP-18 material) as described above.
(5BR)-5B-{3-(3-Carboxypropionyl)-1-[4-nitro-3-({2-[3-(4-
sulfophenyl)thioureido]ethyl}carbamoyl)phenyl]-1H-1,2,4-
triazol-5-yl}-5B-de[(2-hydroxy-5-oxo-1-cyclopenten-1-yl)-
carbamoyl]moenomycin A (8b). 6b (58 mg, 0.030 mmol)
was converted into 8b as described for 7b. RP-18 chro-
matography (water, acetonitrile–water 1:4) gave 8b
34 mg, (53%) as a pale yellowsolid.
(243 MHz, CD3OD/DMSO-d6):
C89H130N11O42PS2 (2121.16, 2119.755386) ESI MS: m/z
528.93100, calcd 528.93157 [MÀ4H]À4; 705.57685, calcd
705.57785 [MÀ3H]À3; 1058.87021, calcd 1058.87042
[MÀ2H]À2. NMR spectra: (below).
31P NMR
d=À0.97
(s).

2978
T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
Assignment
1H NMR (600 MHz,
CD3OD/DMSO-d6,
H, H-COSY, HMBC
HMQC)
13C NMR
(150 MHz,
CD3OD/
Assignment
1H NMR (600 MHz,
CD3OD/DMSO-d6,
H, H-COSY, HMBC
HMQC)
13C NMR
(150 MHz,
CD3OD/
DMSO-d6)
DMSO-d6)
Unit Position
Unit Position
I
1
4.07, broad signal,
3.92, broad signal
5.29, broad signal
—
1.99, m
1.98, m
5.17, dt, J6, 7=15 Hz
5.26, d, J6, 7=16 Hz
—
1.27, m (overlapping with
other signals)
1.78, m
—
2.57, d, J12, 13=7 Hz
5.02, broad signal
—
1.90, t, J=7 Hz
1.97, m
4.98, t, J=6 Hz
—
1.55, s
1.48, s
1.49, s
4.56, s, 4.55, s
0.84, s
0.84, s
1.62, s
—
3.77, broad signal
4.07, broad signal,
3.92, broad signal
—
3.19–3.16 (covered by
the methanol signal)
2.45, covered by the DMSO
signal
—
4.40, m
3.44, m
3.47, m
3.99, broad signal
5.02, broad signal
4.40, m
3.45, m
3.27, m
2.88, dd, J3, 4=J4, 5=8 Hz
?
1.21, d (partial signal overlap) 18.3
66.9
(CH3CONH) —
(CH3CONH) 1.85, s
172.8
24.1
104.8
2
3
4
5
6
7
8
9
124.2
140.2
33.7
D
1
4.30, covered by the H2O
signal
2
3
4
3.04, dd, J1, 2=J2, 3=7 Hz
3.28, m
3.10, covered by the methanol 71.9
signal
75.2
78.2
32.8
127.1
141.6
36.7
5
6
1
3.19, broad signal
3.70, d, JA, B=11 Hz, 3.50, m 62.9
4.33, covered by the H2O
signal
78.4
42.9
E
104.2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
1
32.5
151.2
36.1
123.6
137.6
41.0
27.9
125.7
132.4
26.6
18.5
16.7
109.8
28.3
28.3
24.5
176.9
81.6
?
2
3
4
5
6
3.59, m
3.40, m
3.33, m
3.38, m
3.95, d, JA, B=10 Hz, 1H
56.1
74.3
82.4
75.2
69.8
173.1
24.1
95.9
78.8
76.0
74.3
73.8
(CH3CONH) —
(CH3CONH) 1.85, s
F
1
2
3
4
5
5.69, s
3.49, m
4.87, d, J2, 3=10 Hz
—
4.25, covered by the H2O
signal
6
(OCONH2)
(CH3)
—
—
1.08, s
—
—
—
7.89, s
—
—
174.1
158.7
17.2
H
A
2
3
TA
Ar
3
5
1
2
3
4
5
6
160.5
156.0
148.1
127.2
135.4
142.4
127.2
129.1
167.1
41.2
1
2
194.2
?
3
31.3
8.18, d, J5, 6=9 Hz
8.01, d
—
4
1
2
3
4
5
1
2
3
4
5
6
178.0
105.5
72.0
74.1
71.5
71.4
103.0
56.9
73.4
86.0
72.4
DAO (CONH)
B
C
1
2
3
4
5
6
3.42, m
3.54, broad signal
3.54, broad signal
3.54, broad signal
3.54, broad signal
3.42, m
70.4
71.5
71.5
71.5
41.2
(NHCSNH)
1
—
—
182.3
142.1
127.8
123.5
143.5
BL
2, 6
3, 5
4
7.57, d, J=8 Hz
7.41, d, J=8 Hz
—
Impurity signals (from RP-18) as described above.
6ꢁ-[(R)-2-(2-Ethoxy-3,4-dioxo-1-cyclobuten-1-ylamino)-
2-phenylacetamido]-penicillanic acid (9). The mixture of
ampicillin sodium salt (Sigma, 0.2 g, 0.53 mmol) and
diethyl squarate (0.105 g, 0.62 mmol) in phosphate buf-
fer (10 mL, pH 7.28) was stirred at rt. After 6 h, when
ampicillin sodium salt could be not observed anymore
by TLC (MeCN/H2O=4:1); the solvent was removed
by lyophilization. From the crude product (1.032 g,
mixture of compound 9 and inorganic salts) the salts
could not be removed by UF. The sample for NMR
analysis was prepared as follows: the mixture was
poured in DMSO-d6, filtered, and the filtrate was used
1
for NMR measurements. H NMR (200 MHz, DMSO-
1
1
d6, H H COSY): d=9.44, 9.10 (s, broad s, 1H, 1H,
2-NH, 6Pen-NH), 7.49–7.25 (m, 5H, 2,3,4,5,6Ph-H), 6.01,
5.55 (2 broad s, 0.4H and 0.3H), 5.41–5.25 (m, 2H, 5Pen-H,
6
Pen-H), 4.63 (m, 2H, CH3CH2O), 3.83 (s, 1H, 2Pen-H),
1.47, 1.39 (s, s, 3H, 3H, CH3-9Pen, CH3-10Pen), 1.32 (m,
3H, CH3CH2O). 13C NMR (50 MHz, DMSO-d6):
d=177.49, 171.61, 169.45 (C-1,2,3,4SA), 172.34 (C-1),
169.07 (C-7Pen), 128.32 (C-2,6Ph), 128.01 (C-4Ph), 127.43
(C-3,5Ph), 74.07 (C-2Pen), 68.80 ðCH3CH2OÞ, 66.65

T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
2979
À
Á
(C-5Pen), 64.28 (C-3Pen), 60.30, 57.60 (C-6Pen, C-2),
CONHAr , 166.71, 164.65 (C-1,2SA), 158.63 (C-3TA),
À
Á
31:72=31:28
(C-9Pen),
27.43
(C-10Pen),
15.63
156.60 OCONHF2 , 154.51 (C-5TA), 149.25 (C-11I),
146.70 (C-1Ar), 139.87 (C-7I), 137.40 (C-3Ar), 135.87 (C-
14I), 130.77 (C-18I), 128.42, 127.46, 126.72 (broad signals,
C-2,6Ar, C-2,3,4,5,6Ph), 125.47 (C-6I), 124.09 (C-17I),
123.22 (C-2I), 121.79 (C-13I), 108.77 (C-22I), 102.93,
102.28, 101.15 (broad s, C-1B,C,D,E), 93.77 (C-1F), 81.00–
67.00 (many broad overlapping signals, mainly of sugar
carbons), 64.87 (C-3Pen), 61.37 (C-6D), 59.30, 59.06 (C-
ðCH3CH2OÞ, the signal of C-8Pen was not observed.
C22H23N3O7S (473.60, 473.12567), ESI ICR MS (neg.
mode, MeCN/H2O): m/z=963.25480 (calcd 963.25404)
[2M+H2OÀH]À,
945.24287
(calcd
945.24407)
[2MÀH]À, 472.11825 (calcd 472.11839) [MÀH]À.
(R)-3-({(5R)-5-[1-(3-{[2-(2-{[(R)-((2S,5R,6R)-2-Carboxy-
3,3-dimethyl-7-oxo-4-thia-1-azabicyclo[3.2.0]hept-6-yl-
carbamoyl)phenylmethyl]amino}-3,4-dioxo-1-cyclobuten-
1-ylamino)ethyl]carbamoyl}-4-nitrophenyl)-3-(3-carboxy-
propionyl)-1H-1,2,4-triazol-5-yl]-ꢀ-L-arabinopyranosyl-
(1!4)-2-acetamido-2,6-dideoxy-ꢁ-D -glucopyranosyl-
(1!4)-[ꢁ-D-glucopyranosyl-(1!6)]-2-acetamido-2-deoxy-
ꢁ-D-glucopyranosyl-(1!2)-3-O-carbamoyl-4-C-methyl-
ꢀ-D - glucopyranuronamid - osyloxy}hydroxyphosphory-
loxy)-2-((2Z,6E,13E)-3,8,8,14,18-pentamethyl-11-methyl-
ene-2,6,13,17-nonadecatetraenyloxy)propionic acid (10).
(a) 6a (53 mg, 0.029 mmol) and the crude compound 9
(0.52 g, ca. 90 mg of 10, 0.182 mmol) in a mixture of
DMF (10 mL) and Et3N (1 mL) were stirred at 4 ꢀC for
24 h. Progress of the reaction was monitored by TLC
(MeCN/H2O=4:1, Rf (33)=0.27, Rf (5)=0.1). Then the
solvents were removed under vacuum (10À2 bar) at rt.
To the residue water was added and lyophilized. Inor-
ganic salts and an excess of 32 were removed by UF
(YM3, 6 Â 40 mL) in the refrigerator. Water was lyo-
philized. The crude product was dissolved in a small
amount of water and kieselguhr was added. After lyo-
philization the kieselguhr with the adsorbed compounds
was transferred onto the top of a FC column. A water-
cooled column (4 ꢀC, water-ice) was used. After solvent
evaporation and lyophilization 12.6 mg (19%) of pure
product 10 were obtained. (b) The mixture of com-
pound 6d (20 mg, 0.010 mmol) and ampicillin sodium
salt 31 (4.2 mg, 0.011 mmol) in DMF (3 mL) and Et3N
(0.5 mL) was stirred at 4 ꢀC for 26 h. Progress of the
reaction could not be efficiently monitored by TLC
(MeCN/H2O=4:1) because of the very close Rf values
of 9 (Rf=0.26) and 6d (Rf=0.28). In another solvent
system (n-propanol/water=7:2) the separation was not
better. After workup procedure as described above and
lyophilization the product was analyzed by ESI ICR
MS. IR (KBr): 3345, 1755, 1674, 1602, 1508, 1408, 1327,
759, 698 cmÀ1. IR (KBr) ampicillin: 3355, 1759, 1675,
6
Pen, C-2), 55.33, 54.50 (broad s, C-2E,C), 42.86 (C-2A),
40.99 (C-9I), 39.50 (C-15I, hidden by solvent signals,
HMQC), 36.91 (C-2AE), 35.80 (C-1AE), 35.25 (C-8I), 35.11
(C-3A), 34.53 (C-22I), 31.96 (C-4I), 30.98 (C-5I), 30.75 (C-
10I), 29.04 (C-9Pen), 28.29, 27.92 (C-10Pen), 27.16 (C-
23,24I), 26.14 (C-16I), 25.55 (C-19I), 23.37 (C-25I), 23.09,
À
Á
22.99 CH3CONHC;E , 17.59 (C-20I), 17.24 (CHC3 ), 16.16
(CHF3 ), 15.76 (C-21I). C98H134N13O43PS 2245.23,
2243.81568, ESI ICR MS (neg. mode, MeCN/H2O): m/z
1154.42339 (calcd 1153.87345) [M+3NaÀ5H]2À
,
,
,
1143.42962 (calcd 1142.88194) [M+2NaÀ4H]2À
1131.89462
(calcd
1131.89099)
[M+NaÀ3H]2À
1120.90120 (calcd 1120.90004) [MÀ2H]2À, 769.28044
(calcd 768.91266) [M+3NaÀ6H]3À, 761.95228 (calcd
761.59154)
[M+2NaÀ5H]3À
,
, 571.21341 (calcd 570.93707)
746.93268
(calcd
746.93076) [MÀ3H]3À
[M+2NaÀ6H]4À
,
559.94854
(calcd
559.94612)
[MÀ4H]4À
.
Preparation of PBP 1b-containing membrane fractions
Cells of Escherichia coli strain JM109 [K-12 recA1,
ꢀ(lac-proAB), endA1, gyrA96, thi1, hsdR17, supE44,
relA1, F0(tra-D36, proAB+, lacI9, lacZꢀM15)] carrying
the plasmid pJP13 (this strain expresses the structural
gene of PBP 1b) were grown in modified Lennox broth
(10 g/L Bacto-Trypton, 5 g/L yeast extract, 5 g/L NaCl,
100 mg/mL ampicillin) to an OD550 nm of 0.25 and the
temperature was increased from 30 to 42 ꢀC to induce
transcription and overexpression of PBP 1b.21 After 4 h,
cells were harvested by centrifugation (5000g) and
washed. The cells were resuspended in 0.01 M Tris–
maleate buffer (pH 6.8) containing 0.1 mM MgCl2 and
150 mM NaCl and sonicated (18 mm, 3 Â 30 s). The
lysed cell suspension was centrifuged at 100,000g for
90 min to obtain the membrane fraction. The super-
natant was discarded. The pellet was resuspended in the
same buffer and mixed with an equal volume of buffer
containing 2% (w/v) detergent (Triton X-100) and
phenylmethanesulfonyl fluoride (final concentration
2 mM) at 30 ꢀC for 30 min to extract PBP 1b. Insoluble
residues were removed by centrifugation. Protein con-
centrations were quantified by the bicinchoninic acid
(BCA) method. The membrane extracts were stored at
À70 ꢀC.
1604, 1508, 1404, 1325 cmÀ1
.
1H NMR (600 MHz,
DMSO-d6, H,1H COSY, HMBC, HMQC): Only char-
acteristic signals could be assigned, since most of the
signals were very broad. d=7.50–7.15 (broad m, 5H,
2,3,4,5,6Ph-H), 5.34 (d, 7I-H, J=15.1 Hz), 5.25 (broad s,
6I-H), 5.10 (broad s, 1H, 13I-H), 5.06 (broad s, 1H,
17I-H), 4.89 (broad s, 1H), 4.65 (s, 2H, CH2-22I), 2.65
(broad d, 2H, CH2-12I, J=5.8 Hz), 2.10–1.95 (broad s,
8H, CH2-4,5,15,16I), 1.84 (broad s, 8H, CH2-10I,
CH3CONHC,E), 1.67 (s, 3H, CH3-25I), 1.63 (s, 3H,
CH3-19I), 1.56, 1.55 (s, s, 6H, CH3-20,21I), 1.47 (broad
s, CH3-10Pen), 1.32 (broad s, CH2-9I), 1.23 (broad s,
3H), 1.07 (broad s, 3H, CHF3 ), 0.93 (s, 6H, CH3-23,24I).
13C NMR (150 MHz, DMSO-d6, HMBC, HMQC):
d=192.71 (C-1A), 183.07, 181.99 (C-3,4SA, COOHA),
173.94 (COOHH), 172.23 (C-1), 171.70 (CONHF2 ),
1
For STD NMR measurements, the buffer of the mem-
brane extract was exchanged against the same buffer [in
D2O, 0.01 M Tris–maleate (apparent pH 6.8), 0.1 mM
MgCl2, 150 mM NaCl]. For the exchange we used 30 K
filtron units (PALL, 3 Â 12,000g, 1 h, 4 ꢀC). The NMR
results were unconclusive as discussed in the general
part. Further experiments were carried out with mem-
brane particles. Thus, the harvested E. coli cells (see
À
Á
169.86, 169.54 CH3CONHC;E , 168.84 (C-7Pen), 168.30

2980
T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
Table 3. STD experiments that were performed
Entry
1
Protein preparation
Membrane extract
Ligand
Buffer in D2O
Detergent
Temperature
rt
Result
Moenomycin A
Tris–maleate, pH 6.8
1% Triton X-100
STD signals could
not be assigned
No STD signals
No STD signals
See Figure 3
2
3
4
Membrane fraction
Membrane fraction
Membrane fraction
Moenomycin A
Delipidomoenomycin A
Delipidomoenomycin A
Sodium phosphate, pH 7.0
Sodium phosphate, pH 7.0
Sodium phosphate pH 7.0
—
—
—
rt
rt
20 ꢀC
below) were washed with 0.2 M sodium phosphate buf-
fer (in D2O, apparent pH 7.0) and were resuspended in
the same buffer. After sonication (18 mm, 3Â30 s) the
lysed cell suspension was centrifuged at 15,000g for
20 min and 4 ꢀC to get the membrane particles. The
membrane particles were stored at À20 ꢀC, but not
longer than 24 h to avoid degradation.
Acknowledgements
We thank Prof. B. Meyer for the STD NMR pulse
sequence and Prof. S. Berger for the implementation of
the pulse sequence and verification of the experimental
conditions. Financial support by the Deutsche For-
schungsgemeinschaft, BC Biochemie GmbH (Frank-
furt), and the Fonds der Chemischen Industrie is
gratefully acknowledged.
STD experiments
1D STD NMR experiments were performed according
to ref 14 on a Bruker Avance DRX 600-MHz spectro-
meter equipped with a 5-mm inverse triple-resonance
probe head. A Gauss shape pulse of 50 ms at À0.6 ppm
(on-resonance irradiation) and +40 ppm (off-resonance
irradiation) with a repetition duration of 2 s was used. A
trim pulse of 20 ms was applied to eliminate the protein
signals. The total experimental time was about 22 h
(16 k scans) (Table 3).
References and Notes
1. Review: van Heijenoort, J. Glycobiology 2001, 13, 25 R-
36R.
2. Terrak, M.; Ghosh, T. K.; van Heijenoort, J.; van Beemen,
J.; Lampilas, M.; Aszodi, J.; Ayala, J. A.; Ghuysen, J.-M.;
`
Nguyen-Desteche, M. Mol. Microbiol. 1999, 34, 350, and
references therein.
3. Wang, Q. M.; Peery, R. B.; Johnson, R. B.; Alborn, W. E.;
Yeh, W.-K.; Skatrud, P. L. J. Bacteriol. 2001, 183, 4779.
4. Vogel, S.; Buchynskyy, A.; Stembera, K.; Richter, K.;
Hennig, L.; Muller, D.; Welzel, P.; Maquin, F.; Bonhomme,
C.; Lampilas, M. Bioorg. Med. Chem. Lett. 2000, 10, 1963,
and references therein.
5. For a recent reviewon transgylcosylase inhibition, see:
Goldman, R. C.; Gange, D. Curr. Med. Chem. 2000, 7, 801.
6. El-Abadla, N.; Lampilas, M.; Hennig, L.; Findeisen, M.;
Welzel, P.; Muller, D.; Markus, A.; van Heijenoort, J. Tetra-
hedron 1999, 55, 699, and references therein.
FCS measurements
FCS measurements were performed with the Con-
foCor1 2 (Evotec Biosystems, Hamburg and Carl Zeiss
Jena, Jena). Excitation of tetramethylrhodamine (TMR)
was performed at a wavelength of 543 nm (HeNe-laser)
with an OD-filter of 1.0 mm. For calibration 5 nM rho-
damine 6G was used. For the measurements a droplet
of 50 mL of the sample solution was placed on a glass
cover slide (Nunc1 chamber) on the microscope table.
Each sample was measured for 30 s (bleach time 2 s) and
repeated 10 times. For every experiment, three indepen-
dent measurements were performed. The solubilized
PBP 1b was prepared as described previously.21 A pur-
ified and characterized solution of PBP 1b was carefully
defrosted in ice, centrifuged to remove undissolved par-
ticles. For the binding studies a series of different protein
concentrations (20 nM bis 2 mM) in 10 mM Tris–maleate
(pH 6.8) was preincubated with 11 (5–15 nM) in the
dark. The sample solutions were prepared 20 min before
the measurements to ensure that equilibrium had been
reached. Stock solutions of all further additives (Triton
X-100 and NaCl) were diluted in binding buffer. Con-
trol experiments were done with binding buffer, tetra-
methylrhodamine in binding buffer, PBP 1b in binding
buffer and PBP 1b with tetramethylrhodamine. For the
competition experiments preincubated samples of PBP
1b/11 were mixed with moenomycin A in different con-
centrations (100–1000 mM). Data analysis was per-
formed with the software FCS Access1 2.5 SP2 (Evotec
Biosystems, Hamburg, Germany). Depending on the
number of fluorescent labelled species in the sample an
one- or two-component model was used.
7. Stembera, K.; Vogel, S.; Buchynskyy, A.; Ayala, J. A.;
Welzel, P. Chembiochemistry 2002, 3, 559.
8. For a discussion, see: Anikin, A.; Buchynskyy, A.; Kempin,
U.; Stembera, K.; Welzel, P.; Lantzsch, G. Angew. Chem.
1999, 111, 3931. Anikin, A.; Buchynskyy, A.; Kempin, U.;
Stembera, K.; Welzel, P.; Lantzsch, G. Angew. Chem., Int. Ed.
1999, 38, 3703.
9. Donnerstag, A.; Marzian, S.; Muller, D.; Welzel, P.; Bott-
ger, D.; Stark, A.; Fehlhaber, H.-W.; Markus, A. Tetrahedron
1995, 51, 1931.
10. (a) For a detailed discussion, see: Ritzeler, O.; Hennig, L.;
Findeisen, M.; Welzel, P.; Muller, D.; Markus, A.; Lemoine,
G.; Lampilas, M.; van Heijenoort, J. Tetrahedron 1997, 53,
1675. (b) Kosmol, R.; Hennig, L.; Welzel, P.; Findeisen, M.;
Muller, D.; Markus, A.; van Heijenoort, J. J. Prakt. Chem.
1997, 339, 340.
11. For a discussion of the aggregation behaviour of moeno-
mycin A, see: Lantzsch, G.; Binder, H.; Heerklotz, H.; Welzel,
P.; Klose, G. Langmuir 1998, 14, 4095.
12. Hennig, A.; Findeisen, M.; Weelzel, P.; Haessner, R.
Magn. Reson. Chem. 1998, 36, 615, and references therein.
13. Kurz, M.; Guba, W.; Vertesy, L. Eur. J. Biochem. 1998,
252, 500.
14. Mayer, M.; Meyer, B. J. Am. Chem. Soc. 2001, 123, 6108
and references therein.
15. Vogtherr, M.; Peters, T. J. Am. Chem. Soc. 2000, 122,
6093.

T. Ruhl et al. / Bioorg. Med. Chem. 11 (2003) 2965–2981
¨
2981
16. Marzian, S.; Happel, M.; Wagner, U.; Muller, D.; Welzel,
P. Tetrahedron 1994, 50, 5299, and references therein.
17. Kempin, U.; Hennig, L.; Knoll, D.; Welzel, P.; Muller,
D.; Markus, A.; van Heijenoort, J. Tetrahedron 1997, 53,
17669.
deisen, M.; Muller, D.; Giesa, S.; Knoll, H.; Welzel, P. Eur. J.
Org. Chem. 2002, 1149.
21. Stembera, K.; Buchynskyy, A.; Vogel, S.; Knoll, D.;
Osman, A. A.; Ayala, J. A.; Welzel, P. Chembiochemistry
2002, 3, 332.
18. For leading references, see: Martin, N. H.; Allen, N. W.;
Moore, J. C. J. Mol. Graph. Model. 2000, 18, 242.
19. Stembera, K.; Vogel, S.; Buchynskyy, A.; Ayala, J.; Wel-
zel, P. Chem. Biochem. 2002, 3, 559.
22. Wohland, T.; Friedrich, K.; Hovius, R.; Vogel, H. Bio-
chemistry 1999, 8671.
23. Still,W. C.;Kahn, M.;Mitra, A.J. Org. Chem. 1978, 43, 2923.
24. Kritchevsky, D.; Kirk, M. R. Arch. Biochem. Biophys.
1952, 35, 346.
20. Buchynskyy, A.; Kempin, U.; Vogel, S.; Hennig, L.; Fin-