Solid phase synthesis of vancomycin mimics

Article abstract of DOI:10.1002/ejoc.200300044

A solid phase synthesis of a model system of the DE ring system of vancomycin is described. The synthesis involved biocatalytic resolutions of unnatural amino acids, is compatible with conventional solid phase peptide synthesis and contains as the key step: an on-be.ad S_NAr cyclization. Binding of a cyclic peptide to the carboxylate of N-Ac-D-Ala was demonstrated. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejoc.200300044

FULL PAPER
Solid Phase Synthesis of Vancomycin Mimics
Christopher J. Arnusch[a] and Roland J. Pieters*[a]
Keywords: Antibiotics / Bioorganic chemistry / Peptides / Cyclization / Receptors
A solid phase synthesis of a model system of the DE ring
system of vancomycin is described. The synthesis involved
biocatalytic resolutions of unnatural amino acids, is compat-
ible with conventional solid phase peptide synthesis and con-
of a cyclic peptide to the carboxylate of N-Ac-
demonstrated.
D-Ala was
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
tains as the key step: an on-bead SNAr cyclization. Binding Germany, 2003)
Introduction
ried amino acids at the C-terminal end and concluded from
ITC binding studies in unbuffered water that binding to
Vancomycin is an important antibiotic due to its efficacy relevant carboxylates took place with selected receptors.
against Gram-positive bacteria, including many methicillin Our interest is to vary the carboxylate-binding pocket itself,
resistant strains.[1,2] Unfortunately, also vancomycin-resist- i.e. amino acid 3[14] in order to see what modifications are
ant bacteria have emerged. Their dominant resistance tolerated and what the effects of alternative residues would
mechanism is based on the replacement of the natural acyl- be. In the complex of vancomycin with its ligands the pri-
ated -Ala--Ala sequence of the growing cell wall by acyl- mary amide group of amino acid 3, an asparagine in vanco-
ated -Ala--Lac.[3] Vancomycin binds the latter sequence mycin, is in close proximity to the ligand’s carboxylate oxy-
1000-fold less strong due to electrostatic repulsion in its gen atoms, and in fact has been implicated in hydrogen
complex.[4] For this reason it is an important goal to find bonding to it.[15] Despite the fact that X-ray structures did
high affinity compounds for the altered sequence. Besides not confirm the hydrogen bond, proximity to the side chain
its clinical importance, vancomycin is also a remarkable of amino acid 3 is certain[16,17] and its replacement with
compound from a molecular recognition point of view. It is e.g. a positively charged residue may have a sizeable effect.
a relatively small molecule that binds its small, polar and Varying amino acid 1 (Figure 1), which is -leucine in van-
flexible target in water by a combination of hydrogen bonds
and hydrophobic effects. Many carboxylate receptors have
been synthesized[5] and many efforts to design and synthe-
size vancomycin mimics de novo, have led to a great deal
of innovative and creative chemistry but not thus far to very
effective systems.[6Ϫ10]
The structure of vancomycin clearly carries subtle fea-
tures that make it effective. One way to gain insights into
these features is to simplify the structure and see what bind-
ing power remains. We[11] and others[12,13] have previously
made simplified versions of the carboxylate-binding pocket.
In these simplifications two rings, the AB and the CD rings,
of the vancomycin skeleton were deleted, leaving only the
DE ring framework, i.e. the carboxylate binding pocket, in-
tact. With our system we observed tight binding of relevant
carboxylates in organic solvents.[11] Ellman and co-work-
ers[13] have taken the intact native binding pocket and va-
[a]
Department of Medicinal Chemistry, Utrecht Institute of
Pharmaceutical Sciences, Utrecht University
P.O. Box 80082, 3508 TB Utrecht, The Netherlands
Fax: (internat.) ϩ31-30-2536655
E-mail: r.j.pieters@pharm.uu.nl
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Figure 1. Structure of vancomycin and a generic structure of a
simplified mimic
Eur. J. Org. Chem. 2003, 3131Ϫ3138
DOI: 10.1002/ejoc.200300044
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3131

C. J. Arnusch, R. J. Pieters
FULL PAPER
comycin, is also of interest for the proximity of its N-ter- synthesis, as described here, represents a powerful tool to
minus and its side chain to the bound carboxylate. These the medicinal chemist.
issues led to the following conditions of our synthetic work:
Results and Discussion
1) The synthesis would have to allow variation of both am-
ino acids 1 and 3; and 2) be compatible with common solid
phase peptide chemistry. Thus vancomycin mimics of gen-
eral structure 1 were targeted by application of the SNAr
Preparation of Unnatural Amino Acid Starting Materials
The synthesis of the vancomycin mimic started with the
cyclization reaction.[18Ϫ23] This reaction was used success- preparation of the two unnatural amino acids 7 and 17,
fully by Zhu and co-workers in the synthesis of vancomycin a substituted phenylalanine derivative and a phenylglycine
model compounds and other biaryl ether containing com- derivative, respectively. Fluoro(nitro)phenylalanine deriva-
pounds of medicinal interest.[18] By using this reaction am- tive 7 (Scheme 1) was prepared from 4-fluorobenzaldehyde
ino acids 3 (and, 1) can be chosen at will and by performing (2) following known,[35] but slightly modified procedures.
it on-bead it proved compatible with solid phase Fmoc-pep- Starting with a nitration reaction, and subsequent NaBH4
tide chemistry and should be suitable for the future genera- reduction of the aldehyde group the resulting benzylic al-
tion of libraries.
cohol was converted into 4 with PBr3. According to work
Medicinal relevance of compounds like 1 is also present of Zhu et al.[36] 4 was used to alkylate acetamidomalonate;
in arenas other than the glycopeptide antibiotics as is exem- the formed adduct was decarboxylated in concd. HCl and
plified by the compounds shown in Figure 2. K-13 is an (؉/-)-6 was obtained after esterification of the carboxyl
inhibitor of angiotensin I converting enzyme;[24] the com- group with MeOH and trifluoroacetylation of the amino
pounds OF4949 I-IV show anti-tumor activity,[25Ϫ29] and group. Compound (؉/-)-6 was resolved in a kinetic resolu-
(ϩ)-piperazinomycin is an antibiotic.[30] In fact, a few mem- tion using the protease type VIII-A from Bacillus Licheni-
bers in the vancomycin-class of antibiotics, namely chloro- formus which selectively hydrolyzed the methyl ester of the
peptin and complestatin have shown activity against HIV-1 -enantiomer and the required -enantiomer remained fully
integrase (IC50 ϭ 0.3Ϫ0.5 µ).[31] The structural similarity protected. After separation by extraction, it was found to
of all these molecules is noticeable despite the distinct med- be enantiomerically pure as determined by optical rotation.
icinal targets of these compounds. In general, cyclic pep- After acidic hydrolysis and N-terminus protection with a
tides are currently popular targets for medicinal purposes Boc group, 7 was obtained in 40% yield from the protected
due to their increased in vivo stability, reduced confor- racemic amino acid (؉/-)-6, (maximum yield 50%) and
mational mobility and better-defined geometry.[32] Natural could be run on a multigram scale.
cyclic peptides are often good starting points in structure-
Phenylglycine derivative 17 was also prepared from a ra-
based drug design.[33] Furthermore cyclic peptides are often cemic amino acid involving an alternative enzymatic kinetic
used to mimic the important recognition units: the turns resolution (Scheme 2). To this end 3-hydroxybenzaldehyde
in proteins.[23,34] To explore these and other advantages, a (8) was converted into the corresponding hydantoin 9 using
cyclization method that is compatible with common peptide the BuchererϪBergs reaction.[37] The hydantoin 9 was sub-
Figure 2. Biologically active compounds with structural features related to vancomycin (mimics)
3132
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 3131Ϫ3138

Solid Phase Synthesis of Vancomycin Mimics
FULL PAPER
Scheme 1. Reagents and conditions: a) HNO3, H2SO4, Ϫ5 °C, 60%; b) NaBH4, room temp. then c) PBr3; d) CH3CONCH(CO2Et)2,
NaH; e) 6  HCl, 65 °C; f) SOCl2, MeOH; g) (CF3CO)2O, TEA; h) protease type VIII-A from Bacillus licheniformus, pH 7.5, 37 °C; i)
6  HCl, AcOH, 65 °C; j) Boc2O, NaHCO3 Steps dϪj according to ref.[36]
.
sequently hydrolyzed under acidic conditions to obtain the of TBAF in DMF. It proved necessary for the mixture to
free amino acid hydrochloride salt. Its carboxylic acid react for 40 h to complete the reaction. In our previous
group was converted into a methyl ester using thionyl chlo- paper,[11] CsF in DMF was used as the fluoride source for
ride in methanol and the amino function was subsequently deprotection and cyclization. However, when used for re-
outfitted with a trifluoroacetate group to yield (؉/-)-12. ceptors on the solid phase, mass spectrometry results
This protected (,)-phenylglycine derivative was first sub- showed incomplete cyclization. This is likely be due to the
jected to protease VIII-A from Bacillus licheniformus as was poor solubility of CsF in DMF. The fluoride ion in this
done with (؉/-)-6, under several conditions in order to ob- reaction has a dual purpose since it first has to act as a
tain the -enantiomer but with limited success. The solvent nucleophile to deprotect the silyl phenyl ether and then acts
mixtures, temperature, pH, and phenolic protecting groups as a base to facilitate the SNAr cyclization. After treatment
were all factors that were varied. The methyl ester was with TFA in the presence of triisopropylsilane and water
therefore hydrolyzed with an aqueous LiOH/dioxane solu- amide 18 was obtained. HPLC analysis of the crude reac-
tion to obtain the free racemic carboxylic acid and was sub- tion mixture showed the result of an efficient synthesis[40]
sequently treated with the acylase AMANO 30000.[21,38,39] with a major peak flanked by a minor peak in a 9:1 ratio.
The non-deacylated and resolved compound 14 was ob- Since the enantiomeric purity of the substituted phenylgly-
tained in 45% yield with an ee of 82%. Its trifluoroacetate cine used was 82% these two peaks were likely due to the
group was removed under acidic conditions in nearly quan- formation of diastereoisomers (Figure 3, a). A sample was
titative yield to give the unprotected amino acid 15. The purified and the major peak was separated using prepara-
amino group of compound 15 was first protected with an tive HPLC and was isolated in 30% yield (Figure 3, b).
1
Fmoc group using Fmoc-OSu at pH 8.5 in a 1:1 mixture of Based on H NMR spectra obtained from this peak, it was
acetonitrile and water. The phenolic hydroxy group of the concluded that it consisted of a mixture of two compounds.
resulting compound 16 was then silylated with TBDMS-Cl The ratio was based on the characteristic signals derived
to yield the amino acid derivative 17, ready for solid phase from the proton on the phenylglycine benzene ring, located
synthesis, in 76% yield after lyophilization.
between the ether linkage and the link to the α-carbon
atom. These signals were located at around δ ϭ 6 ppm and
had moved upfield upon ring closing due to the shielding
effect of the phenylalanine ring. Since epimerization was
not expected to occur under the conditions used,[41,42] the
two compounds (ratio 4:1) were presumed to be two atropi-
somers. The atropisomeric ratio is not expected to change
based on the work of Boger and co-workers where they
showed that under ambient conditions, no rotation is seen
about the biaryl ether linkage once ring closing is
achieved.[43] Additional experiments confirmed our in-
terpretation about compound 18. The single peak could in-
deed be resolved into two peaks (atropisomers) using a dif-
ferent HPLC column. A resynthesis of compound 18 using
the racemic building block 17, yielded material which
Solid-Phase Synthesis
The solid-phase synthesis was performed as shown in
Scheme 3. Argogel OH resin outfitted with a Rink linker
was chosen to test the validity of the on-bead SNAr cycliza-
tion to be performed. Fmoc chemistry therefore was used
for amino acid chain elongation with BOP as the amino
acid coupling reagent. Normally, 4 equivalents of the amino
acid are used relative to the loading of the precursor on the
solid phase. However, with the more precious amino acids
7 and 17, 2.5 equivalents were used combined with longer
reaction times. The tetra peptide 18 was synthesized in this
fashion and cyclization was realized by treatment of the un-
cyclized receptor on the solid phase with a 0.1  solution
Eur. J. Org. Chem. 2003, 3131Ϫ3138
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3133

C. J. Arnusch, R. J. Pieters
FULL PAPER
Scheme 3. Synthesis of receptor on solid phase. Reaction con-
ditions: a) Coupling Ϫ 2.5 equiv. 17, 2.5 equiv. BOP, 5 equiv. DI-
PEA, then deprotection Ϫ 20% piperidine in NMP, (3 ϫ 8 min),
and washing Ϫ NMP, CH2Cl2, (3 ϫ 2 min); b) Coupling Ϫ 4 equiv.
Fmoc-Leu-OH, 4 equiv. BOP, 8 equiv. DIPEA, deprotection, wash-
ing; c) Coupling Ϫ 2.5 equiv. 7, 2.5 equiv. BOP, 5 equiv. DIPEA;
d) 7 equiv. (0.1 ) TBAF in DMF, 40 h; e) TFA/TIS/H2O,
95:2.5:2.5 v:v:v
resolution mass spectra was therefore obtained and upon
analysis of the isotope patterns, as well as the technique of
tandem mass spectrometry (MS/MS), unambiguously con-
firmed structure of monomeric 18.
Scheme 2. Reaction conditions: a) (NH4)2CO3, KCN, then HCl,
40%; b) 6  HCl, 120 °C, quant.; c) SOCl2, MeOH, 80%; d)
(CF3CO)2O, NEt3, 82%; e) LiOH(aq.)/dioxane, quant.; f) AMANO
30000, pH 7.5, 5 d, 25 °C, 45%; g) 6  HCl/AcOH (1:1), 65 °C,
quant.; h) Fmoc-OSu, NEt3, pH 8.5, 96%; i) TBDMS-Cl, pyridine,
65 °C, 76%.
As in our previous paper,[11] 1H NMR titration studies
were performed on compound 18 as an indication of bind-
ing of carboxylate ions. Addition of the tetra-n-butyl am-
monium salt of NAc--Ala (0.75Ϫ42 m) to a solution of
receptor 18 in CD3CN (1 m) resulted in downfield shifts
showed four peaks under these conditions (see Supporting of at least two of the NH resonances to δ ϭ 9.02 and
Information, for Supporting Information see also the foot- 10.52 ppm. These shifts are even further downfield than
note on the first page), strongly indicating that the dia- that of our previously reported receptor molecule[11] where
stereomeric integrity can be maintained during the syn- the maximum shift was to δ ϭ 9.00 ppm (Figure 4). This
thesis. To assess the versatility of applying this methodology suggests that the hydrogen bonding in compound 18 to
for a future parallel library synthesis, 2 additional receptors NAc--Ala is taking place in a similar fashion but differ-
were synthesized with a Lys and Tyr incorporated at posi- ences are also apparent.
tion 3. As before, mass spectrometric analysis indicated a
clean and efficient synthesis of these compounds (see Sup- measure binding affinities and directly observe interaction
porting Information). of the receptor-ligand complex. Titrations with vancomycin
UV spectroscopy was previously employed in our lab to
Another consideration that should be addressed is the were performed as well in aqueous solution and consistently
possibility of covalent dimerization upon ring closing. gave binding constants that agreed with previously reported
Analysis of the electrospray mass spectra obtained for com- values.[44Ϫ46] A UV titration is compatible with the biaryl
pound 18 indicated a small peak at m/z ϭ 2M (singly ether-type receptors since the aryl systems have an ab-
charged particle) that could indicate a dimer of 18. High sorbance in the UV region where the ligands do not. In the
3134
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 3131Ϫ3138

Solid Phase Synthesis of Vancomycin Mimics
FULL PAPER
impact on binding and varying this position may further
alter the delicate hydrogen bonding system of the vancomy-
cin(mimic)-carboxylate complexes.
Figure 5. Structure of receptor 19
Conclusion
Methodology to build vancomycin-like mimics on the
solid phase was developed. Using biocatalysis it was pos-
sible to prepare the components needed for the solid phase
peptide chemistry that was used to assemble the system. An
on-bead SNAr reaction was successfully applied to effect the
necessary cyclization of two peptide side chains. Variation
of the amino acid in the middle of the carboxylate binding
pocket seems to have an effect on binding to small car-
boxylates and can possibly disturb or enhance the delicate
hydrogen bonding system. Parallel libraries will be prepared
to determine this. The synthetic strategy used here in the
preparation of cyclic peptides may also be useful in prepar-
ing other new compounds with e.g. anti-viral properties or
enzyme-inhibitory properties.
Figure 3. HPLC traces of a) crude, and b) purified 18
Experimental Section
General Remarks: Chemicals and resins were obtained from com-
mercial sources and used without further purification unless stated
otherwise. The solvents DMF, NMP, CH2Cl2, MeOH and dioxane
were purchased from Biosolve, the Netherlands. The solvents
CH2Cl2, DMF, NMP, and Et2O were kept dry with molecular si-
˚
˚
eves (4 A); for MeOH 3-A molecular sieves was used. The base
iPr2NEt was distilled from ninhydrin and KOH. Column chroma-
tography was performed on ICN Silica 32Ϫ63, 60 A. Thin layer
˚
Figure 4. NMR spectra of a 1 m solution of 18 in CD3CN with
the addition of a) a saturating amount of N-Ac--Ala-OϪϩNBu4
(42 m) and b) with no ligand added
chromatography (TLC) was performed on Merck precoated Silica
60 plates. Visualization was accomplished with UV light and stain-
ing with ninhydrin. Protease type VIII-A from Bacillus licheni-
formus was obtained from Sigma. Acylase ‘‘Amano 30000’’ was pur-
chased from the Amano Pharmaceutical Co., Ltd. Nagoya, Japan.
same way[4], titration of a dilute solution of 18 (1 ϫ 10Ϫ4 
in CH3CN) with a solution of the tetra-n-butyl ammonium
salt of N-Ac--Ala (while maintaining a constant receptor Amino acids were purchased from Multisyntech. Rotations were
1
measured with a Jasco P-1010 polarimeter. H NMR spectra were
recorded with a Varian G-300 spectrometer (300 MHz); NMR ti-
trations and spectra for 18 were obtained with a Varian INOVA
spectrometer. (500 MHz). 13C NMR spectra were recorded with a
Varian G-300 spectrometer at 75.4 MHz. All spectra were refer-
enced to the solvent signal: CD3CN (1H: δ ϭ 1.93 ppm, 13C: δ ϭ
1.3 ppm), CD3OD(H) (1H: δ ϭ 3.3 ppm, 13C: δ ϭ 49.0 ppm),
CDCl3 (1H: δ ϭ 7.24 ppm, 13C: δ ϭ 77.0 ppm) and [D6]DMSO
(δ ϭ 2.49, 39.5 ppm). 13C NMR spectra were recorded using the
attached proton test (APT) sequence where indicated. Electrospray
concentration), resulted in an increased absorption around
270Ϫ280 nm. This increase was fitted to a 1:1 binding iso-
therm and indicated an association constant of 7.3 ϫ 103
Ϫ1 or 5.2 kcal molϪ1 of binding free energy. This value is
actually about 4Ϫ5 times weaker than our related receptor
19, where alanine is present in the middle of the carboxylate
binding pocket.[11] The lower association constant and also
the NMR chemical shift positions mentioned above, might
be an indication that the small variation of the amino acid
in the center of the carboxylate binding pocket can have an ionization (ESI) mass spectrometry was carried out using a Shim-
Eur. J. Org. Chem. 2003, 3131Ϫ3138
www.eurjoc.org  2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 3135

C. J. Arnusch, R. J. Pieters
FULL PAPER
adzu LCMS QP-8000 single quadrupole bench top mass spec-
(RS)-4-Fluoro-3-nitrophenylalanine [(؉/-)-5], (RS)-4-Fluoro-3-nitro-
trometer (m/z range Ͻ 2000), coupled with a QP-8000 data system, N-(trifluoroacetyl)phenylalanine Methyl Ester [(؉/-)-6] and (R)-N-
except for compounds 13, 14 and 18 which were carried out using
electrospray MS (nano ESI-TOF-MS) run on a Micromass LC-
TOF mass spectrometer by spraying a CH3CN/H2O solution from
a gold-coated glass capillary in a Z-spray nanospray ionization
source (cone voltage 30Ϫ40). Elemental Analysis was measured by
the Firma H. Kolbe, Mikroanalytisches Laboratorium, Mülheim
a.d. Ruhr, Germany. UV measurements were carried out with a
Helios-β Spectrophotometer. Analytical HPLC was performed with
a Shimadzu Class-VP automated HPLC using an analytical re-
Boc-4-Fluoro-3-nitro-phenylalanine (7): These compounds were pre-
pared from 4-fluoro-3-nitrobenzyl bromide (4) according to ref.[36]
.
(3-Hydroxy)phenylhydantoin (9): 3-Hydroxybenzaldehyde (8, 6.1 g,
0.05 mol) was dissolved in ethanol (80 mL) and added to a solution
of ammonium carbonate (24 g, 0.25 mol) and KCN (3.9 g, 0.06
mol) in water (80 mL). The flask was fitted with a condenser and
heated to 60 °C for 150 min. After cooling in ice, the solution was
acidified in a well-ventilated fumehood to pH 2 with 6  HCl while
bubbling N2 through the mixture. After acidification, the mixture
was degassed for an additional 30 min to ensure complete removal
of HCN. The solution was poured into a beaker and dried by evap-
oration. The solid was dissolved in dry MeOH and solution de-
canted from the salts (4 ϫ). The MeOH fractions were combined
and the solvents evaporated in vacuo and the residue dissolved in
100 mL MeOH/DCM (5:1) and filtered to remove remaining salt
content to yield a crude solid (6 g). Column chromatography was
used to purify the hydantoin using a gradient of solvents (DCM/
MeOH, 8:1 to 1:3) to yield 3.87 g (40%) as a light beige powder,
m.p. 204Ϫ205 °C, (ref.[47] 212 °C). Rf ϭ 0.5 (DCM/MeOH, 6:1).
1H NMR (CD3OD): δ ϭ 5.05 (s, 1 H, Hα), 6.79 (m, 3 H, ArH),
˚
versed-phase column (Alltech Adsorbosphere C18, 300 A, 5µm,
250 ϫ 4.6 mm) and a UV detector operating at 220 nm and
254 nm. Preparative HPLC was performed with a Gilson auto-
mated HPLC using a preparative reversed-phase column (Alltech
Adsorbosphere C18, 10 µm, 250 ϫ 22 mm) and a UV detector
operating at 220 nm and 254 nm. Elution was effected using an
appropriate gradient from 0.1% TFA in water to 0.085% TFA in
acetonitrile/water (95:5, v/v) using a flow rate of 1 mL·minϪ1 (ana-
lytical) or 11.5 mL·minϪ1 (preparative). Separation of the enanti-
omers of 17 was performed by HPLC (Gilson) using a Chirobiotic
T column from Advanced Separation Technologies Inc. Elution
was effected using 50% MeOH and 50% (1% TEAA in water) using
1
7.20 (t, 1 H, ArH) ppm. H NMR ([D6]DMSO): δ ϭ 5.04 (s, 1 H,
a flow rate of 1 mL·minϪ1
.
Hα), 6.73 (m, 3 H, ArH), 7.17 (t, 1 H, ArH), 8.36 (s, 1 H, NH),
9.54 (s, 1 H, NH), 10.75 (s, 1 H, ΟΗ) ppm. 13C NMR (CD3OD):
δ ϭ 63.5, 114.5, 116.7, 118.8, 131.0, 138.0, 159.1, 160.1, 176.4 ppm.
4-Fluoro-3-nitrobenzaldehyde (3): 4-Fluorobenzaldehyde (2,
15.6 mL, 18 g, 0.145 mol) was slowly added dropwise to a solution
of H2SO4 (72 mL), and HNO3 (10 mL) at Ϫ5 °C. The temperature
was kept below 5 °C. The solution was warmed to room temp. (1
h), and after pouring into ice, (400 g) the precipitate was filtered
and washed with H2O. The aqueous phase was extracted with
CH2Cl2 and combined with the precipitate. The solution was
washed with NaHCO3, brine, and dried with Na2SO4. The solvent
was removed in vacuo and the residue was recrystallized from di-
ethyl ether to yield 14.5 g (60%) of large clear white crystals. Rf ϭ
0.48 (CH2Cl2/hexane, 2:1). 1H NMR (CDCl3): δ ϭ 7.50 (t, J ϭ
9 Hz, 1 H, ArH), 8.20 (oct, J ϭ 2 Hz, 1 H, ArH), 8.60 (dd, J ϭ 7,
J ϭ 2 Hz, 1 H, ArH), 10.04 (s, 1 H, CHO) ppm. C7H4FNO3
(169.11): calcd. C 49.72, H 2.38, N 8.28; found C 49.68, H 2.36,
N 8.22.
(RS)-(3-Hydroxy)phenylglycine (10): In a round-bottom flask fitted
with a condenser (3-hydroxy)phenylhydantoin (5 g, 0.026 mol) was
heated to 120 °C in 6  HCl for 5 d. This solution was evaporated
1
to dryness and used without further purification. H NMR (D2O):
δ ϭ 4.81 (s, 1 H, Hα), 6.76Ϫ6.84 (m, 3 H, ArH), 7.19 (t, J ϭ 8 Hz,
1 H, ArH) ppm. 13C NMR ([D6]DMSO): δ ϭ 55.7, 115.3, 116.5,
118.7, 130.1, 134.5, 158.0, 169.9 ppm.
(RS)-(3-Hydroxy)phenylglycine Methyl Ester (11): This compound
was prepared according to ref.[48] Spectroscopic data: 1H NMR
(CD3OD): δ ϭ 3.66 (s, 3 H, OCH3), 4.47 (s, 1 H, Hα), 6.70 (dd,
J ϭ 8, J ϭ 1.5 Hz, 1 H, ArH), 6.80 (m, 2 H, ArH), 7.15 (t, J ϭ
8 Hz, 1 H, ArΗ) ppm. 13C NMR ([D6]DMSO): δ ϭ 52.3, 58.6,
114.3, 114.9, 118.1, 129.9, 142.9, 157.9, 175.1 ppm.
4-Fluoro-3-nitrobenzyl Bromide (4): In
a round-bottom flask
equipped with an addition funnel 4-fluoro-3-nitrobenzaldehyde (3,
14.5 g, 0.085 mol) was dissolved in methanol, (90 mL). NaBH4
(12.3 g, 0.325 mol) in water (22 mL) was added dropwise, while
keeping the reaction mixture at room temp. After the addition was
complete, a small test portion of the reaction mixture was tested in
dilute H2SO4. H2(g) was formed and indicated a finished reaction.
The methanol was removed in vacuo and water was added
(200 mL). The aqueous phase was extracted with diethyl ether (5
ϫ 150 mL) and the diethyl ether layer was washed with water, brine,
and dried with MgSO4. The solvent was removed in vacuo to give
an oil (14.6 g) and without further purification redissolved in di-
ethyl ether (275 mL). To the stirred solution was added dropwise
PBr3 (8.2 mL, 23.37 g, 0.086 mol). After 1 hour water (400 mL)
was carefully added. The aqueous phase was extracted with diethyl
ether and the combined organic fractions were washed with
NaHCO3, brine, and dried with Na2SO4. The solvents were re-
moved in vacuo and the product purified by recrystallization in
(RS)-(3-Hydroxy)phenyl-N-trifluoroacetylglycine Methyl Ester
[(؉/؊)-12]: (ϩ/Ϫ)-(3-hydroxy)phenylglycine methyl ester (11, 3.3 g,
15 mmol) was dissolved in DCM (100 mL) and TEA (3 mL) and
trifluoroacetic anhydride (3 mL, 21.3 mmol) were added dropwise.
The solution was stirred for 4 h and kept at a basic pH with TEA.
The solvent was removed in vacuo and DCM (100 mL) was added.
The mixture was washed with NaHCO3 (2 ϫ 75 mL), dilute aque-
ous HCl (2 ϫ 75 mL), brine (2 ϫ 75 mL) and dried with Na2SO4.
The solvent was removed in vacuo and (؉/-)-12 was purified by
recrystallization (EtOAc/hexane) to yield 2.91 g (70%). Rf ϭ 0.5
1
(CH2Cl2/MeOH, 7:1). H NMR (CDCl3): δ ϭ 3.73 (s, 3 H, CH3),
5.49 (d, J ϭ 7 Hz, 1 H, Hα), 6.74 (br. s, 1 H, OH), 6.79Ϫ6.88 (3
H, ArH), 7.20 (t, J ϭ 8 Hz, 1 H, ArH), 7.70 (d, J ϭ 7 Hz, 1 H,
NH) ppm. 13C NMR APT (CDCl3): δ ϭ 53.4, 56.6, 114.3, 116.5,
119.1, 130.5, 115.4 (q, JC,F ϭ 287 Hz), 135.5, 156.5, 156.9 (q,
JCCF ϭ 39 Hz), 170.0 ppm. C11H10F3NO4 (277.20): calcd. C 47.66,
H 3.64, N 5.05; found C 47.73, H 3.67, N 4.97.
hexane for a yield of 10 g over 2 steps. (50% sparkling yellow solid).
1
Rf ϭ 0.33 (hexane/EtOAc, 5:1). H NMR (CDCl3): δ ϭ 4.49 (s, 2 (RS)-(3-Hydroxy)phenyl-N-trifluoroacetylglycine [(؉/-)-13]: (؉/-)-
H, CH2Br), 7.29 (t, J ϭ 7 Hz, 1 H, ArH), 7.67 (oct, J ϭ 2 Hz, 1
12 (1.84 g, 6.6 mmol) was dissolved in tBuOH (100 mL) and water
H, ArH), 8.10 (dd, J ϭ 7, J ϭ 2 Hz, 1 H, ArH) ppm. C7H5BrFNO2 (50 mL) and LiOH·H2O (0.545 g, 13 mmol) was added . The mix-
(234.02): calcd. C 35.93, H 2.15, N 5.99; found C 35.81, H 2.17,
N 5.86.
ture was stirred at room temp. for 100 min and then diluted with
water (100 mL). The mixture was acidified with 1  KHSO4, the
3136
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 3131Ϫ3138

Solid Phase Synthesis of Vancomycin Mimics
FULL PAPER
volitile components were removed in vacuo and the remaining
aqueous phase was extracted with EtOAc (3 ϫ 80 mL). The organic
fractions were combined and extracted with 5% NaHCO3 (3 ϫ
100 mL), acidified with concd. HCl (15 mL), and extracted with
EtOAc (3 ϫ 100 mL). The organic fractions were washed with
brine, dried with Na2SO4, and the solvents were removed in vacuo
to yield pure (؉/-)-13 in quantitative yield. (1.74 g). All data
matched 14 (below).
then washed with water, brine, and dried with Na2SO4. The solvent
was removed in vacuo and the residue was purified with column
chromatography (eluent: CH2Cl2/MeOH/AcOH, 20:0:0.1 gradient
to 20:1:0.1). Lyophilization gave 382 mg (76%) of compound 17 as
a white powder, m.p. 85 °C (dec.). Rf ϭ 0.6 (CH2Cl2/MeOH, 3:1).
[α]D ϭ Ϫ48.3 (c ϭ 0.18, MeOH). Chiral HPLC analysis: 80% ee
1
Rt: 5.4 min (S), Rt: 18.6 min (R). H NMR (CD3OD): δ ϭ 0.19 [s,
6 H, Si(CH3)2], 0.98 [s, 9 H, SiC(CH3)3], 4.22 (t, J ϭ 6 Hz, 1 H,
CH-Fmoc), 4.30, 4.32 (2 H, CH2-Fmoc), 5.17 (s, 1 H, Hα), 6.78
(dd, J ϭ 8.5, J ϭ 1.5 Hz, 1 H, ArH), 6.94 (s, 1 H, ArH), 7.02 (d,
J ϭ 8 Hz, 1 H, ArH), 7.21 (t, J ϭ 8 Hz, 1 H, ArH), 7.29 (t, J ϭ
6.6 Hz, 2 H, ArH-Fmoc), 7.38 (t, J ϭ 6.6 Hz, 2 H, ArH-Fmoc),
7.65 (d, J ϭ 7.5 Hz, 1 H, ArH-Fmoc), 7.78 (d, J ϭ 7.5 Hz, 2 H,
ArH-Fmoc) ppm. 13C NMR (CD3OD): δ ϭ Ϫ4.5, 18.1, 25.6, 46.8,
47.0, 57.7, 58.1, 67.3, 68.3, 99.9, 118.8, 119.1, 119.9, 120.4, 124.9,
125.0, 127.1, 127.7, 129.7, 130.0, 137.1, 138.7, 141.1, 141.2, 143.3,
143.5, 143.6, 143.7, 155.5, 156.0, 157.2, 173.2, 175.1 ppm. ES-MS:
m/z ϭ 504.4 [M ϩ H]ϩ, 526.3 [M ϩ Na]ϩ. C29H33NO5Si (503.21):
calcd. C 69.16, H 6.60, N 2.78; found C 69.28, H 6.53, N 2.85.
Enzymatic Resolution of (14): (؉/؊)-13 (0.41 g, 1.56 mmol) was dis-
solved in phosphate buffer (40 mL, pH 7) and CoCl2·6H2O (0.5 mg
2.1 µmol) and NaN3 (0.8 mg, 13 µmol) were added to it. After
addition of ‘‘Amano 30000’’ (0.144 g) the reaction was stirred for
5 d at 25 °C. The mixture was then acidified to pH 2Ϫ3 with a
citric acid solution and extracted with EtOAc (3 ϫ 100 mL). The
organic layer was filtered through celite and washed with brine (2
ϫ 150 mL). The organic solvents were removed in vacuo and co-
evaperated with CHCl3 (2 ϫ) to yield a white solid. The solid was
then dissolved in H2O (4 mL) and lyophilized giving 14 in 45%
yield (185 mg) with an ee of 82% (based on rotation of 15, see
below) [α]D ϭ Ϫ161 (c ϭ 0.39, MeOH). Rf ϭ 0.31 (hexane/EtOAc/
Synthesis of Receptor 18: Synthesis was performed on Fmoc-Rink-
Argogel with a loading of 0.335 mmol/g. 0.5 g Resin was depro-
tected (see Fmoc Deprotection) and then coupled (see Amino Acid
Coupling) with sequentially: Fmoc-Ala-OH, 17, Fmoc-Leu-OH,
and 7. After cyclization, deprotection, and cleavage (see below), the
crude material was purified using preparative HPLC to obtain a
white solid in 30% yield (27 mg). Rt: 18.0 min. 1H NMR (CD3OH):
δ ϭ 0.94 [d, J ϭ 5.5 Hz, 6 H, Leu(CH3)2), 1.30 (d3, J ϭ 7.0 Hz, 3
H, AlaCH), 1.60 (m, 2 H, LeuβCH2), 1.69 (m, 1 H, LeuγCH), 2.14
(s, 2 H, NH2), 3.10 (dd, J ϭ 14.5, J ϭ 5.5 Hz, 1 H, PheβCH), 3.53
(dd, J ϭ 14.5, J ϭ 4.0 Hz, 1 H, PheβCH), 4.24 (m, 3 H, 3NH),
5.10 (d, J ϭ 5.5 Hz, 1 H, PhgαH), 6.06 (s, 1 H, PhgC2H], 7.02,
7.56 (s, 2 H, NH2), 7.07 (d, J ϭ 7.0 Hz, 1 H, ArH), 7.13 (d, J ϭ
8.5 Hz, 1 H, ArH), 7.19 (d, J ϭ 7.5 Hz, 1 H, ArH), 7.37 (t, J ϭ
7.5 Hz, 1 H, ArH), 7.57 (s, 1 H, ArH), 7.99 (s, 1 H, ArH), 8.24 (d,
J ϭ 6.5 Hz, 1 H, NH), 8.31 (d, J ϭ 4.5 Hz, 1 H, NH) ppm. ES-
MS: m/z ϭ 541.3 [M ϩ H]ϩ, 563.3 [M ϩ Na]ϩ.
1
AcOH, 1:1:0.05). H NMR (CD3OD): δ ϭ 5.43 (s, 1 H, Hα), 6.78
(d, J ϭ 8 Hz, 1 H, ArH), 6.90 (m, 2 H, ArH), 7.19 (t, J ϭ 8 Hz, 1
H, ArH) ppm. 13C NMR (CD3OD): δ ϭ 58.3, 116.0, 116.7, 120.1,
130.9, 117.4 (q, JC,F ϭ 286 Hz), 137.9, 159.0, 158.6 (q, JCCF
ϭ
45 Hz), 172.3 ppm. C10H8F3NO4 (263.17): calcd. C 45.64, H 3.06,
N 5.32; found C 45.38, H 3.11, N 5.26.
(3-Hydroxy)phenylglycine·HCl (15): To 14 (0.1 g, 0.4 mmol) was ad-
ded aqueous HCl (6 , 4 mL) and acetic acid (4 mL) in a round
bottom flask fitted with a condenser. The mixture was heated to
65 °C overnight and then evaporated to dryness to quantitatively
yield the free amino acid hydrochloride salt (0.08 g) after lyophiliz-
ation. [α]D ϭ Ϫ128 (c ϭ 0.13, 6  HCl), 82% ee, (ref.[49] [α]D
Ϫ151 (97% ee, c ϭ 0.13, 6  HCl).
ϭ
Fmoc-(3-Hydroxy)phenylglycine
(16):
(3-Hydroxy)phenyl-
glycine·HCl (15, 0.245 g, 1.2 mmol) was dissolved in water (4 mL)
and the pH was adjusted to 9 with NEt3 (20 drops). Fmoc-OSu
(0.410 g, 1.2 mmol) was dissolved in acetonitrile (4 mL) and added
all at once to the stirred solution. The reaction was stirred at room
temperature for 30 min and the pH was kept at 8.5 with NEt3. The
clear light-beige reaction mixture was neutralized with HCl and the
acetonitrile was removed in vacuo. 1  KHSO4 (50 mL) was added
and then extracted with EtOAc (3 ϫ 100 mL). The organic solvents
were then combined, washed with brine, and dried with Na2SO4.
The solvent was removed in vacuo and purified using column chro-
matography (eluent: CH2Cl2/MeOH/AcOH, 10:1:0.1) yielding
0.448 g (96%) of 16 as a white powder, m.p. 173Ϫ174 °C. Rf ϭ 0.49
(CH2Cl2/MeOH/AcOH, 5:1:0.1). 1H NMR (CD3CN): δ ϭ 4.22 (m,
1 H, CH-Fmoc), 4.32 (d, 2 H, CH2-Fmoc), 5.17 (d, 1 H, NH), 6.48
(d, 1 H, αCH), 6.79 (dd, 1 H, ArH), 6.75 (m, 2 H, ArH), 7.21 (t,
1 H, ArH), 7.28Ϫ7.43 (m, 4 H, ArH-Fmoc), 7.66 (d, 2 H, ArH-
Fmoc), 7.81 (d, 2 H, ArH-Fmoc) ppm. 13C NMR (CD3CN): δ ϭ
47.9, 58.8, 67.5, 115.4, 116.3, 119.9, 121.0, 126.2, 128.4, 131.0,
142.1, 144.9, 158.1, 172.3 ppm. MS: m/z ϭ 390.1 [M ϩ H]ϩ, 412.1
[M ϩ Na]ϩ. C23H19NO5 (389.13): calcd. C 70.94, H 4.92, N 3.60;
found C 71.04, H 4.86, N 3.51.
Fmoc Deprotection: To the resin was added 20% piperidine in NMP
(3 ϫ 8 min) and agitated by bubbling nitrogen through the solution.
The resin was washed with NMP (3 ϫ 2 min) and CH2Cl2 (3 ϫ 2
min). The ‘‘Kaiser test’’[50] was performed on a small portion of
beads for elucidation of the free amine terminus.
Amino Acid Coupling: To the reaction vessel containing the depro-
tected resin (500 mg) was added sequentially the appropriate amino
acid (4 equivalents, 0.67 mmol) in NMP (3 mL), BOP (296 mg,
0.67 mmol) in NMP (3 mL) and DIPEA (233 µL, 1.34 mmol). The
reaction mixture was agitated by bubbling nitrogen through the
solution for 75 min. For amino acids 7 and 17, 2.5 equivalents were
used with reaction times of 150 min. The beads were then washed
with NMP (3 ϫ 2 min) and CH2Cl2 (3 ϫ 2 min) and the ‘‘Kaiser
test’’ was performed on a small portion of beads. The negative test
provided evidence for successful coupling.
On-bead Cyclization: TBAF·3H2O (1.29 g, 4.1 mmol) was dissolved
˚
in DMF (45 mL) and dried with molecular sieves (4 A, 3 h). The
TBAF solution in DMF (11 mL) was then added to the resin and
shaken (40 h). The resin was washed exhaustively with NMP,
CH2Cl2, NMP, CH2Cl2, 3% AcOH in dioxane, Et2O, NMP, and
CH2Cl2. (each 2 ϫ 2 min).
Fmoc-(3-TBDMSO)phenylglycine (17): 16 (390 mg, 1.0 mmol) was
dissolved in pyridine (1.5 mL) and then a solution of TBDMS-Cl
(380 mg, 2.5 mmol) in pyridine (2 mL) was added. The solution
was fitted with a condenser, and stirred overnight at 70 °C. The
solvent was removed in vacuo, then 1  KHSO4 (30 mL) and
EtOAc (80 mL) were added to the residue. The organic phase was
Deprotection and Cleavage from the Resin: The resin was treated
with (TFA/water/TIS, 95:2.5:2.5) for 3 h. The resin was washed
with TFA (1 mL, 2 ϫ 2 min) and the organic solvents were removed
Eur. J. Org. Chem. 2003, 3131Ϫ3138
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3137

C. J. Arnusch, R. J. Pieters
FULL PAPER
in vacuo. After coevaporation with ethanol (2 ϫ 10 mL), the solu-
[22]
W. D. Kohn, L. Zhang, J. A. Weigel, Org. Lett. 2001, 3,
971Ϫ974.
Y. Feng, K. Burgess, Chem. Eur. J. 1999, 5, 3261Ϫ3272.
H. Kase, M. Kaneko, K. Yamado, J. Antibiot. 1987, 40,
450Ϫ454.
S. Sano, K. Ikai, H. Kuroda, T. Nakamura, A. Obayashi, Y.
Ezure, H. Enomoto, J. Antibiot. 1986, 39, 1674Ϫ1684.
S. Sano, K. Ikai, K. Katayama, K. Takesako, T. Nakamura, A.
Obayashi, Y. Ezure, H. Enomoto, J. Antibiot. 1986, 39,
1685Ϫ1696.
S. Sano, M. Ueno, K. Katayama, T. Nakamura, A. Obayashi,
J. Antibiot. 1986, 39, 1697Ϫ1703.
S. Sano, K. Ikai, Y. Yoshikawa, T. Nakamura, A. Obayashi, J.
Antibiot. 1987, 40, 512Ϫ518.
S. Sano, H. Kuroda, M. Ueno, Y. Yoshikawa, T. Nakamura,
A. Obayashi, J. Antibiot. 1987, 40, 519Ϫ525.
S. Tamai, M. Kaneda, S. Nakamura, J. Antibiot. 1982, 35,
1130Ϫ1136.
S. B. Singh, H. Jayasuriya, D. L. Hazudda, P. Felock, C. F.
Homnick, M. Sardana, M. A. Patane, Tetrahedron Lett. 1998,
39, 8769Ϫ8770.
For a review on cylic peptides see: J. N. Lambert, J. P. Mitchell,
K. D. Roberts, J. Chem. Soc., Perkin Trans. 1 2001, 471Ϫ484.
As exemplified by the conotoxins, see: J. M. McIntosh, A. D.
Santos, B. M. Olivera, Ann. Rev. Biochem. 1999, 68, 59Ϫ88.
A. J. Souers, J. A. Ellman, Tetrahedron 2001, 57, 7431Ϫ7448.
F. Micheel, D. Noffz, Chem. Ber. 1957, 90, 1586Ϫ1589.
C. Vergne, M. Bois-Choussy, J. Ouazzani, R. Beugelmans, J.
Zhu, Tetrahedron: Asymmetry 1997, 8, 391Ϫ398.
H. T. Bucherer, W. Steiner, J. Prakt. Chem. 1934, 140, 291Ϫ316.
M. Bois-Choussy, J. Zhu, J. Org. Chem. 1998, 63, 5662Ϫ5665.
Acylase AMANO 30000 was obtained from AMANO Phar-
maceutical Co. LTD. Nagoya Japan.
Yields based on the HPLC spectra of the crude product are
typically between 50 and 70% over 11 steps.
D. L. Boger, S. L. Castle, S. Miyazaki, J. H. Wu, R. T. Beresis,
O. Loiseleur, J. Org. Chem. 1999, 64, 70Ϫ80.
A. J. Pearson, M. V. Chelliah, J. Org. Chem. 1998, 63,
3087Ϫ3098.
D. L. Boger, O. Loiseleur, S. L. Castle, R. T. Beresis, J. H. Wu,
Bioorg. Med. Chem. Lett. 1997, 7, 3199Ϫ3202.
J. P. Mackay, U. Gerhard, D. A. Beauregard, R. A. Maple-
stone, D. H. Williams, J. Am. Chem. Soc. 1994, 116,
4573Ϫ4580.
J. P. Mackay, U. Gerhard, D. A. Beauregard, M. S. Westwell,
M. S. Searle, D. H. Williams, J. Am. Chem. Soc. 1994, 116,
4581Ϫ4590.
J. Rao, L. Yan, J. Lahiri, G. M. Whitesides, R. M. Weis, H. S.
Warren, Chem. Biol. 1999, 6, 353Ϫ359.
H. R. Henze, R. J. Speer, J. Am. Chem. Soc. 1941, 64, 522Ϫ523.
M. J. Stone, M. S. van Dijk, P. M. Booth, D. H. Williams, J.
Chem. Soc., Perkin Trans. 1 1991, 1629Ϫ1635.
M. J. Garcia, R. Azerad, Tetrahedron: Asymmetry 1997, 8,
85Ϫ92.
tion was lyophilized from a 1:1 mixture of acetonitrile and water.
[23]
[24]
[25]
[26]
Acknowledgments
We like to thank Professor R. M. J. Liskamp for helpful support
and the Royal Netherlands Academy of Arts and Sciences
(KNAW) is gratefully acknowledged for a fellowship to R. J. P.
[27]
[28]
[29]
[30]
[31]
[1]
D. H. Williams, B. Bardsley, Angew. Chem. Int. Ed. 1999, 38,
1172Ϫ1193.
[2]
K. C. Nicolaou, C. N. C. Boddy, S. Bräse, N. Winssinger, An-
gew. Chem. Int. Ed. 1999, 38, 2096Ϫ2152.
T. D. H. Bugg, S. Dutka-Malen, M. Arthur, P. Courvalin, C.
[3]
T. Walsh, Biochemistry 1991, 30, 2017Ϫ2021.
[4]
I. A. D. Lessard, C. T. Walsh, Proc. Natl. Acad. Sci. USA 1999,
96, 11028Ϫ11032.
R. J. Fitzmaurice, G. M. Kyne, D. Douheret, J. D. Kilburn, J.
[5]
[32]
[33]
Chem. Soc., Perkin Trans. 1 2002, 841Ϫ864.
M. C. F. Monnee, A. J. Brouwer, L. M. Verbeek, A. M. A. van
[6]
Wageningen, R. M. J. Liskamp, Bioorg. Med. Chem. Lett. 2001,
11, 1521Ϫ1525.
U. Botana, S. Ongeri, R. Arienzo, M. Demarcus, J. G. Frey, U.
[7]
[34]
[35]
[36]
Piarulli, D. Potenza, C. Gennari, J. K. Kilburn, Chem. Com-
mun. 2001, 1358Ϫ1359.
A. Casnati, M. Fabbi, N. Pelizzi, A. Pochini, F. Sansone, R.
[8]
[37]
[38]
[39]
Ungaro, Bioorg. Med. Chem. Lett. 1996, 6, 2699Ϫ2704.
L. Frish, F. Sansone, A. Casnati, R. Ungaro, Y. Cohen, J. Org.
[9]
Chem. 2000, 65, 5026Ϫ5030.
[10]
B. Hinzen, P. Seiler, F. Diederich, Helv. Chim. Acta 1996, 79,
[40]
[41]
[42]
[43]
[44]
942Ϫ960.
[11]
R. J. Pieters, Tetrahedron Lett. 2000, 41, 7541Ϫ7145.
[12]
N. Pant, A. D. Hamilton, J. Am. Chem. Soc. 1988, 110,
2002Ϫ2003.
[13]
R. Xu, G. Greiveldinger, L. E. Marenus, A. Cooper, J. A. Ell-
man, J. Am. Chem. Soc. 1999, 121, 4898Ϫ4899.
Vancomycin numbering, see: N. J. Skelton, D. H. Williams, M.
[14]
J. Rance, J. C. Ruddock, J. Chem. Soc., Perkin Trans. 1 1990,
77Ϫ81.
[15]
This study involved A82846B, a close relative of vancomycin,
see: W. G. Prowse, A. D. Kline, M. A. Skelton, Biochemistry
1995, 34, 9632Ϫ9644.
P. J. Loll, A. E. Bevivino, B. D. Korty, P. H. Axelsen, J. Am.
[45]
[46]
[16]
Chem. Soc. 1997, 119, 1516Ϫ1522.
P. J. Loll, R. Miller, C. M. Weeks, P. H. Axelsen, Chem. & Biol.
[17]
1998, 5, 293Ϫ298.
[18]
[47]
[48]
For a review on SNAr cyclizations see: J. Zhu, SynLett 1997,
133Ϫ144 (and references cited therein).
K. Burgess, D. Lim, M. Bois-Choussy, J. Zhu, Tetrahedron
[19]
[49]
[50]
Lett. 1997, 38, 3345Ϫ3348.
[20]
A. S. Kiselyov, L. Smith, II, P. Tempest, Tetrahedron 1999,
55, 14813Ϫ14822.
M. Goldberg, L. Smith, II, N. Tamayo, A. S. Kiselyov, Tetra-
E. Kaiser, R. L. Colescott, C. D. Bossinger, P. I. Cook, Anal.
Biochem. 1970, 34, 595Ϫ598.
[21]
hedron 1999, 55, 13887Ϫ13898.
Received January 27, 2003
3138
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 3131Ϫ3138