Tight binding of a dimeric derivative of vancomycin with dimeric L-Lys-D-Ala-D-Ala

Article abstract of DOI:10.1021/ja971225l

The ligand/receptor pair consisting of a synthetic dimeric derivative of vancomycin (V), linked at the C terminus by p-xylylenediamine (V-CONHCH₂C₆H₄CH₂NHCO-V), and a dimeric derivative of L-Lys-D-Ala-D-Ala, [CH

Full text of DOI:10.1021/ja971225l

10286
J. Am. Chem. Soc. 1997, 119, 10286-10290
Tight Binding of a Dimeric Derivative of Vancomycin with
Dimeric L-Lys-D-Ala-D-Ala
Jianghong Rao and George M. Whitesides*
Contribution from the Department of Chemistry and Chemical Biology, HarVard UniVersity,
12 Oxford Street, Cambridge, Massachusetts 02138
ReceiVed April 17, 1997X
Abstract: The ligand/receptor pair consisting of a synthetic dimeric derivative of vancomycin (V), linked at the C
terminus by p-xylylenediamine (V-CONHCH2C6H4CH2NHCO-V), and a dimeric derivative of L-Lys-D-Ala-D-Ala,
[CH2CONꢀH(NR-Ac)-L-Lys-D-Ala-D-Ala-CO2-]2, provides a new system with which to study the influence of divalency
on the strength of binding. A competitive assay using affinity capillary electrophoresis (ACE) has been developed
and used to estimate the dissociation constant of the divalent complex (Kdd ≈ 1.1 nM) and the enhancement in
binding (∼103) relative to the corresponding monomeric interaction between unmodified monomeric vancomycin
and diacetyl-L-Lys-D-Ala-D-Ala.
Introduction
dimerization of vancomycin and structurally related glycopep-
tides is responsible for their tight binding to DADA groups in
the bacterial cell wall,22 and thus that simultaneous interaction
of the dimer with the two ligands contributes to their antibiotic
action.23-26 We now describe a system comprising synthetic
dimeric derivatives of vancomycin and of DADA that demon-
strates an increase in the strength of binding of these components
of approximately 103 relative to the analogous binding of
monovalent vancomycin and DADA. These complexes do not
have the geometry required to model the head-to-tail dimeric
structure hypothesized for the action of the unmodified antibi-
otic, but they suggest the entropic advantage that can be obtained
through divalency.
Divalent interactions in molecular recognition involving two
linked ligands interacting with two linked receptors have been
studied in many systems, including dimers of cyclodextrin with
divalent ligands,1-3 dimers of sialyl Lewis x with E-selectin,4
dimers of immunophilin ligands such as FK506 and cyclosporin
A with receptors active in controlling cell signal transduction,5
and divalent antibodies with a surface antigen.6,7 We have
begun to develop a new model systemsbased on polyvalent
derivatives of vancomycin (V) and of its ligand, D-Ala-D-Ala
(DADA)sto study the physical organic chemistry and physical
biochemistry of polyvalency. The binding of V with DADA
has been extensively characterized thermodynamically and
structurally,8-14 especially by Williams15-18 and Griffin.19-21
There is inferential evidence that spontaneous noncovalent
Results and Discussion
We chose p-xylylenediamine as the linker Rd in the dimeric
derivative of vancomycin V-Rd-V because it is structurally rigid
and easily modified synthetically. Previous studies have showed
that modification at the C-terminus of V has little effect on its
* To whom correspondence should be addressed.
X Abstract published in AdVance ACS Abstracts, October 1, 1997.
(1) Many dimers of cyclodextrin have been reported as divalent receptors,
and a wide range of enhancement of affinity of dimers has been observed.
Recently, Zhang and Breslow described a careful study of cyclodextrin
dimers (J. Am. Chem. Soc. 1996, 118, 8495-8496 and references therein)
and concluded that the binding constant of â-cyclodextrin (â-CD) with
cholesterol was 1.7 × 104 M-1, while that of a dimer of â-CD with
cholesterol was 5.5 × 106 M-1; the enhancement was therefore a factor of
300.
(12) Popieniek, P. H.; Pratt, R. F. Anal. Biochem. 1987, 165, 108-113.
(13) Loll, P. J.; Bevivino, A. E.; Korty, B. D.; Axelsen, P. H. J. Am.
Chem. Soc. 1997, 119, 1516-1522.
(14) Chu, Y.-H.; Whitesides, G. M. J. Org. Chem. 1992, 57, 3524-
3525.
(15) Pierce, C. M.; Williams, D. H. J. Chem. Soc., Perkin Trans. 2 1995,
2, 153-157.
(16) Williams, D. H.; Williamson, M. P.; Butcher, D. W.; Hammond,
S. J. J. Am. Chem. Soc. 1983, 105, 1332-1339.
(17) Kannan, R.; Harris, C. M.; Harris, T. M.; Waltho, G. P.; Skelton,
N. J.; Williams, D. H. J. Am. Chem. Soc. 1988, 110, 2946-2953.
(18) Bongini, A.; Feeney, J.; Williamson, M. P.; Williams, D. H. J. Chem.
Soc., Perkin Trans. 2 1981, 201-206.
(19) Shi, Z.; Griffin, J. H. J. Am. Chem. Soc. 1993, 115, 6482-6486.
(20) Sundram, U. N.; Griffin, J. H. J. Org. Chem. 1995, 60, 1102-1103.
(21) Sundram, U. N.; Griffin, J. H.; Nicas, T. I. J. Am. Chem. Soc. 1996,
118, 13107-13108.
(22) Beauregard, D. A.; Williams, D. H.; Gwynn, M. N.; Knowles, D.
J. C. Antimicrob. Agents Chemother. 1995, 39, 781-785.
(23) Gerhard, U.; Mackay, J. P.; Maplestone, R. A.; Williams, D. H. J.
Am. Chem. Soc. 1993, 115, 232-237.
(24) Groves, P.; Searle, M. S.; Waltho, J. P.; Williams, D. H. J. Am.
Chem. Soc. 1995, 117, 7958-7964.
(2) Beeson, J. C.; Czarnik, A. W. Bioorg. Med. Chem. 1994, 2, 297-
303.
(3) Venema, F.; Rowan, A. E.; Nolte, R. J. M. J. Am. Chem. Soc. 1996,
118, 257-258.
(4) Defrees, S. A.; Kosch, W.; Way, W.; Paulson, J. C.; Sabesan, S.;
Halcomb, R. L.; Huang, D.-H.; Ichikawa, Y.; Wong, C.-H. J. Am. Chem.
Soc. 1995, 117, 66-79. The concentration of the 3,6-linked SLex dimer
required to block adhesion of 50% of the cells was 0.2 mM, and the value
for the tetrasaccharide monomer was 1.2 mM.
(5) Spencer, D. M.; Wandless, T. J.; Schreiber, S. L.; Crabtree, G. R.
Science 1993, 262, 1019-1024.
(6) Hornick, C. L.; Karush, F. Isr. J. Med. Sci. 1969, 5, 163-170. The
binding constant of rabbit γG anti-2,4-dinitrophenyl (DNP) antibody with
the monovalent hapten Nꢀ-DNP-lysine was 6 × 106 M-1, while that of the
antibody with conjugated bateriophage φX174 with DNP groups was 3.5
× 1011 M-1; the enhancement was a factor of 6 × 104.
(7) Hornick, C. L.; Karush, F. Immunochemistry 1972, 9, 325-340.
(8) Nieto, M.; Perkins, H. R. Biochem. J. 1971, 123, 773-787.
(9) Nieto, M.; Perkins, H. R. Biochem. J. 1971, 123, 789-803.
(10) Sheldrich, G. M.; Jones, P. G.; Kennard, O.; Williams, D. H.; Smith,
G. A. Nature 1978, 271, 223-224.
(25) Mackay, J. P.; Gerhard, U.; Beaurgard, D. A.; Westwell, M. S.;
Searle, M. S.; Williams, D. H. J. Am. Chem. Soc. 1994, 116, 4581-4890.
(26) Westwell, M. S.; Bardsley, B.; Dancer, R. J.; Try, A. C.; Williams,
D. H. J. Chem. Soc., Chem. Commun. 1996, 589-590.
(11) Sheldrick, G. M.; Paulus, E.; Vertesy, L.; Hahn, F. Acta Crystallogr.
1995, B51, 89-98.
S0002-7863(97)01225-0 CCC: $14.00 © 1997 American Chemical Society

Binding of Dimeric Vancomycin with Dimeric L-Lys-D-Ala-D-Ala
J. Am. Chem. Soc., Vol. 119, No. 43, 1997 10287
Scheme 1. The Dimeric Derivative of Vancomycin V-Rd-V and Ligands L-R'd-L and P. HBTU:
2-(1H-Benzotriazol-1-yl)-1,1,3,3-tetramethyluronium Hexafluorophosphate
affinity for DADA.19 We synthesized V-Rd-V by coupling two
molecules of V with one of p-xylylenediamine (Scheme 1). The
1H-NMR spectroscopy and ESI-MS of the product confirmed
its structure. We formed the dimeric derivative of D-Ala-D-
Ala, L-R'd-L, by coupling the ꢀ-amino groups of the lysyl
moieties of the tripeptide, NR-Ac-L-Lys-D-Ala-D-Ala, with the
two carboxylic acid groups of a succinic acid (Scheme 1).
Examination of CPK molecular models of the complex sug-
gested that L-R'd-L fitted well into both pockets of V-Rd-V.
We first performed a fluorescence titration to compare the
binding of V-Rd-V to a monomeric ligand and to L-R'd-L. On
complexation with vancomycin, there is an increase in the
fluorescence intensity of the ligand NR-dansyl-Nꢀ-Ac-L-Lys-D-
Ala-D-Ala (fL).12 A solution containing V-Rd-V (∼1 µM) and
fL (∼2 µM) had stronger fluorescence than did a solution of
fL (∼2 µM) alone; this difference in the fluorescence intensity
Figure 1. The competitive fluorescence titration of a solution
between the two solutions reflected the binding of fL to V-Rd-
V. We titrated the solution containing V-Rd-V (∼1 µM) and
fL (∼2 µM) with L-R'd-L (Figure 1). Upon addition of L-R'd-
L, the fluorescence intensity of the solution decreased; this
decrease reflected the dissociation of the complex of V-Rd-V
with fL, as a result of competitive binding of L-R'd-L to V-Rd-
V. We defined ∆F as the difference in the fluorescence intensity
between the solution being titrated and a reference solution
containing only fL (∼2 µM). Before the addition of L-R'd-L,
∆F was at its initial value, ∆Fi. At the end of the titration,
there was no or little complexation between V-Rd-V and fL,
and ∆F was at its final value, ∆Ff. A similar titration was
performed using the monomeric ligand diacetyl-L-Lys-D-Ala-
D-Ala (L) for comparison. The observation that a plot of the
data from the titration of V-Rd-V and fL with L-R'd-L was
nearly linear indicated that the binding of L-R'd-L to V-Rd-V
was tight and stoichiometric at that concentration (Figure 1).
containing V-Rd-V (∼1.0 µM) and fL (∼2.0 µM) with L-R'd-L and
with L; the expanded inset shows the region from [L-R'd-L] ) 0 to
1.0 µM. The excitation wavelength was 330 nm and the fluorescence
intensity was monitored at 550 nm at 24 °C. ∆F, ∆Fi, and ∆Ff are
defined as in the text.
In order to establish the influence of divalency accurately,
we have developed a new assay using affinity capillary
electrophoresis (ACE).27 In general, ACE measures the change
of mobility of a receptor as a function of the concentration of
a ligand in the CE buffer; Scatchard analysis yields the affinity
constant. This assay is based on competition of species for the
active sites (Scheme 2); the reagents were chosen so that the
influence of their charges on the electrophoretic mobility of the
aggregate indicated the composition of the aggregate. Ligand
(27) For review of general principles of ACE, see: Chu, Y.-H.; Avila,
L.; Gao, J.; Whitesides, G. M. Acc. Chem. Res. 1995, 28, 461-468 and
references therein.

10288 J. Am. Chem. Soc., Vol. 119, No. 43, 1997
Rao and Whitesides
Scheme 2. Schematic Representation of Equilibrium among All Possible Species in the Competitive Assaya
a The Kmd1 and Kdm2 are the two dissociation constants for the complex P•V-Rd-V•P; Kdd1 and Kdd2 are the two dissociation constants for the complex
•
V-Rd-V•L-R'd-L (the divalent species). The equilibrium constant K1 defines the exchange reaction between P•V-Rd-V•P and P•V-Rd-V•L-R'd-L;
•
K2 is the equilibrium constant for the exchange between P•V-Rd-V•L-R'd-L and V-Rd-V• L-R'd-L.
P was designed for this purpose; it was a monovalent ligand
but had the same number of carboxylate groups as ligand L-R'd-
L.
tion of L-R'd-L added to this buffer. These species will be
present at equilibrium: P•V-Rd-V•P, P•V-Rd-V•L-R'd-L, and
•
V-Rd-V•L-R'd-L (the divalent species).29 The values of ∆Z
(defined by analogy with eq 1) for the first two of these species
is -4; that for the divalent species is -2. The mass of these
three species is similar: in particular, that of the first and second
differs by only 3.5%. We therefore expect the influence of
differences in mass (or, more properly, hydrodynamic drag) on
electrophoretic mobilities, µ, to be small among these three
aggregates. Thus only the fraction of V-Rd-V present in divalent
•
species V-Rd-V•L-R'd-L will change the electrophoretic mobility
relative to that of P•V-Rd-V•P. Analysis of µ for species
containing the V-Rd-V group as a function of [L-R'd-L] thus
indicates the relative concentration of the divalent complex
•
There are four amino groups on the V-Rd-V, and at pH 7.0
this compound is expected to have a net charge between +3
and +4; P has a net charge of -2 at pH 7.0. The complex of
V-Rd-V with P has a net negative charge at pH 7.0 (the complex
emerged after the neutral marker when [L-R'd-L] ) 0; Figure
2). This observation indicated that V-Rd-V bound 2 equiv of
P and formed a 1:2 complex.
We began the development of the assay by establishing the
average dissociation constant, Kmd (“m” defines monovalent
binding), of the aggregate P•V-Rd-V•P. We assumed the two
active sites of V-Rd-V to be independent and identical, and we
estimated a value of Kdm ≈ 4.8 µM;28 this value is comparable
to those observed by others for similar compounds.21 On the
basis of this value, we selected a concentration of P for the
assay system that was sufficiently high to saturate both active
V-Rd-V•L-R'd-L and allows the value of Kd of that species to
be determined relative to the known value for the species P•V-
Rd-V•P (eqs 2, 3). We define the dissociation constant of the
•
sites of V-Rd-V ([P] ) 1.0 mM ≈ 200Km). At this concentra-
d
tion, the only vancomycin-derived species in solution at
significant concentration is P•V-Rd-V•P (Scheme 2). Since P
has a net charge of -2 at pH 7.0, the difference in charge, ∆Z,
between this complex and V-Rd-V is -4 (eq 1).
divalent complex V-Rd-V•L-R'd-L as Kd (the superscript “d”
d
signifies divalent binding), R as the ratio of the change in the
electrophoretic mobility of V-Rd-V on adding L-R'd-L ∆µ, to
its maximal value ∆µmax; Scatchard analysis uses eq 2 to
•
∆Z ) Z(P•V-Rd-V•P) - Z(V-Rd-V) ) -4
(1)
estimate K, the dissociation constant of V-Rd-V•L-R'd-L relative
to the two dissociation constants of P•V-Rd-V•P, Km and
We then examined the mobility of the V-Rd-V group in a
buffer containing [P] ) 1.0 mM as a function of the concentra-
d1
(29) Our analysis does not take into account another possible 1:2 complex
of V-Rd-V, L-Rd-L•V-Rd-V•L-R'd-L. The value of ∆Z for this complex is
-4; that is the same as that for P•V-Rd-V•P. The presence of a small amount
of L-R'd-L•V-Rd-V•L-R'd-L should, therefore, not change the electrophoretic
mobility µ.
(28) At 280 nm, P has little UV absorbance. This value was an average
of three independent runs, determined by UV difference spectroscopy
according to the literature reported method (ref 8).

Binding of Dimeric Vancomycin with Dimeric L-Lys-D-Ala-D-Ala
J. Am. Chem. Soc., Vol. 119, No. 43, 1997 10289
Figure 3. Scatchard plot of data from the ACE assays; R ) ∆µ/∆µmax
.
The intercept along y-axis of the straight line indicates that K/[P]2 has
a value of 20 nM-1 (eq 2): this value establishes K ) 20 nM-1 × [P]2
) 20 mM. The inset plots the electrophoretic mobilities of V-Rd-V
and of V as a function of concentrations of L-R'd-L in the running
buffer.
8 kcal/mol.8, 9 The second binding site in the divalent ligand
therefore contributes an additional 3.9 kcal/mol to the binding
energy. If we make the plausible assumption that Kmd1 and
Kdd1 are indistinguishable, the ratio of Kmd2 to Kdd2 which, in fact,
equals K, gives the effective molarity characterizing the second
binding event in the divalent system.31 We estimated this
effective molarity to be 20 mM. For concentrations of P lower
than 20 mM, intramolecular binding will, therefore, be more
favorable than intermolecular binding in a system containing P
and L-R'd-L.
Figure 2. Representative electropherograms of a mixture of V-Rd-V
and V in the running buffer (20 mM phosphate; pH 7.0) containing
L-R'd-L and 1.0 mM P; the concentrations of L-R'd-L were as indicated
in the plot. Mesityl oxide (MO) was used as a neutral marker (b); the
V-Rd-V group is represented by * and V by 4. The inverted peaks are
due to the loss of ligands in the buffer as a result of initial binding.
The time marker (1) corresponds to 100 s after injection for all samples.
Kmd2 (eq 3).30
In a study of the binding of AcDADA and N-decanoyl-D-Ala-
D-Ala (N-dec-DADA) with ristocetin A in the presence of sodium
dodecylsulfate, Williams et al. reported that there is a favorable
intramolecular binding between a noncovalent ristocetin A dimer
and N-dec-DADA anchored in micelles, and that intramolecu-
larity gives an overall contribution of free energy to the binding
of 3.5 kcal/mol.25
K1
R
K
K
[P]2
) -
+
R +
(2)
(3)
2
(
)
[L-R'd-L]
[P]
[P]
Kmd1 Kmd2 (Kmd )2
Kdd1 Kdd2
K ) K1K2 )
)
Kdd
The electrophoretic mobility of V-Rd-V changed systemati-
cally with the concentration of L-R'd-L. Figure 2 shows a
representative series of electropherograms of V-Rd-V in 20 mM
phosphate buffer at pH 7.0 containing 1.0 mM P and various
concentrations of L-R'd-L. We used unmodified vancomycin
as the control to estimate the effect of the hydrodynamic drag
on the complex. Figure 3 indicates that the change in the
electrophoretic mobility of vancomycin was insignificant on
addition of L-R'd-L; thus the effect of the small differences in
the hydrodynamic drag of the aggregates of V-Rd-V with P and
L-R'd-L on their electrophoretic mobility was negligible. Hence
we can interpret the data in terms of the effect of divalency on
binding. Scatchard analysis yielded the value of K ≈ 20 mM
(Figure 3). We therefore estimated Kdd for the divalent species
•
Conclusions
By tethering two molecules of vancomycin and two of DADA
we have assembled and demonstrated a new model system with
which to study divalency. This model system has several
advantages. (i) It is biologically relevant. The study helps to
understand the antibacterial action of vancomycin-type glyco-
peptides. A tight binder might be a potential antibiotic
candidate.21 (ii) The structures are relatively rigid, and the effect
of entropy on binding is small. (iii) The well-defined specificity
of the interaction between vancomycin and DADA minimizes
nonspecific interaction that might complicate the analysis of
the origin of increased binding energy.
The ACE competition assay we developed here in principle
can be applied to other systems. It has all the advantages of
the ACE method,27 including the ability to measure the influence
of divalency on binding. It is easy to apply; for analysis using
ACE, this method requires only a competing monomeric ligand
V-Rd-V•L-R'd-L to be 1.1 nM (eq 3).
In terms of the binding free energy, ∆G°, the value of Kdd ≈
1.1 nM corresponds to 11.9 kcal/mol; the corresponding value
for vancomycin with diacetyl-L-Lys-D-Ala-D-Ala (L) is about
(31) Page, M. I.; Jencks, W. P. Proc. Natl. Acad. Sci. U.S.A. 1971, 68,
1678-1683.
(30) See Supporting Information for detailed derivations.

10290 J. Am. Chem. Soc., Vol. 119, No. 43, 1997
Rao and Whitesides
temperature for 3 h; a white cloudy precipitate formed. The precipiate
was collected after filtration and washed with saturated NaCl and ethyl
acetate. The hexapeptide ester (56 mg) was obtained as white solid
(0.066 mmol, 52%). The crude ester was deprotected with TFA and
then fractionated by reverse phase HPLC. 1H-NMR (DMSO-d6) δ 8.14
(d, 2H), 8.08 (d, 2H), 8.00 (d, 2H), 7.77 (t, 2H), 4.32-4.24 (m, 2H),
4.20-4.12 (m, 4H), 2.97 (q, 4H), 2.25 (s, 4H), 1.82 (s, 6H), 1.60-1.40
(m, 4H), 1.36-1.27 (m, 4H), 1.27-1.17 (m, 4H), 1.27 (d, 6H), 1.17
(d, 6H); FAB-MS, for C32H55N8O12 (M + H+), calcd m/z 743.3939,
found m/z 743.3913.
Diacetyl-L-lysyl-D-alanyl-D-alanine (L) was prepared using acetic
anhydride in acetone to acetylate the ꢀ-amine of NR-Ac-L-Lys-D-Ala-
D-Ala-O-tBu. The tert-butyl ester was cleaved by TFA, and the peptide
was purified by reverse phase HPLC. 1H-NMR (DMSO-d6) δ 8.14
(d, 1H), 8.08 (d, 1H), 8.00 (d, 1H), 7.77 (t, 1H), 4.32-4.24 (m, 1H),
4.20-4.12 (m, 2H), 2.96 (q, 2H), 1.82 (s, 3H), 1.76 (s, 3H), 1.58-1.40
(m, 2H), 1.36-1.17 (m, 2H), 1.27 (d, 3H), 1.17 (d, 3H); FAB-MS, for
C16H29N4O6 (M + H+), calcd m/z 373.2087, found m/z 373.2096.
Nr-Dansyl-NE-acetyl-L-lysyl-D-alanyl-D-alanine (fL) was synthe-
sized by coupling of NR-dansyl-Nꢀ-Ac-L-Lys with D-Ala-D-Ala-O-tBu.
TFA cleavage and reverse phase HPLC purification afforded the pale
yellow fluorescent dansyl ligand with excitation maxima 330 nm and
emission maxima 550 nm in water.12 1H-NMR (D2O) δ 8.65 (d, 1H),
8.34 (d, 1H), 8.27 (d, 1H), 7.94 (d, 1H), 7.77 (m, 2H), 4.14 (q, 1H),
3.91 (q, 1H), 3.58 (t, 1H), 3.34 (s, 6H), 2.56 (t, 2H), 1.78 (s, 3H),
1.44-1.40 (m, 2H), 1.27 (d, 3H), 1.02 (d,3H), 1.05-0.96 (m, 1H), 0.90-
0.83 (m, 2H), 0.70-0.65 (m, 1H); FAB-MS, for C26H37N5O7SNa (M
+ Na+), calcd m/z 586.2311, found m/z 586.2322.
ACE Binding Study. The capillary tubing (Polymicro Technolo-
gies, Phoenix) was of uncoated fused silica with an internal diameter
of 50 mm, a total length of 74 cm, and a length from the inlet to the
detector of 44 cm. A representative sample for injection into the
electrophoresis capillary consisted of V-Rd-V (0.7 mg/mL), 20 mM
mesityl oxide (MO), and vancomycin (0.3 mg/mL). The sample
solution (∼8 nL) was introduced into the capillary by vacuum injection.
The general conditions used during each CE experiment were as
follows: voltage, 30 kV; current uncontrolled, but generally 30 mA;
buffer, 20 mM phosphate (pH 7.0), 1.0 mM ligand P, 0-320 µM ligand
L-R'd-L; detection, 214 nm; temperature, 30 ( 2 °C.
Fluorescence Titration. The fluorescent-labeled ligand (fL, 2.0 µM)
in phosphate buffer (20 mM and at pH 7.0) was used in the titration.
The change in fluorescence intensity at 550 nm (excitation at 330 nm)
was monitored with a Perkin-Elmer MPF-4 spectrofluorometer at 24
°C. The fluorescence intensity of a cell containing 3.0 mL of 2.0 µM
fL and V-Rd-V (∼1.0 µM) and a reference cell containing 3.0 mL of
2.0 µM fL was measured, respectively. Small aliquots of a 5.4 mM
of L-R'd-L solution in the same buffer was added to a cell containing
3.0 mL of 2.0 µM fL and V-Rd-V (∼1.0 µM), and the fluorescence
intensity decreased (reflecting the binding of L-R'd-L to V-Rd-V). The
titration was stopped until no more decrease of fluorescence intensity
was observed. During the titration experiment, the same aliquots of
buffer as that of L-R'd-L solution were added to the reference cell as
calibration. The same procedure was applied to the titration using
ligand L.
having one binding site with the same charge as that of the
dimeric ligand of interest.
Experimental Section
General Procedure. Chemical solvents were obtained from Aldrich.
The 1H-NMR spectra were recorded at 400 MHz on a Bruker
spectrometer. Chemical shifts are reported in parts per million
downfield of tetramethylsilane. Reverse-phase HPLC was carried out
with a Waters Model 600E chromatography system and Vydac C18
columns. A 4.6 mm i.d. column was used for analytical purpose and
a 21.4 mm i.d. column was used for preparative separations. Linear
gradients of 0.2% trifluoroacetic acid (TFA) in acetonitrile and 0.2%
TFA in water were used in HPLC elution. ACE binding study was
performed on an ISCO Model 3140 or a Beckman P/ACE 5010 system.
The fluorescence binding study was performed using a Perkin-Elmer
fluorescence spectrophotometer, Model MPF-4. Vancomycin hydro-
chloride was purchased from Sigma and used without further purifica-
tion. Amino acids were purchased from Sigma except D-Ala-O-tBu
from BACHEM Bioscience, and peptide coupling reagent 2-(1H-
benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate
(HBTU) was purchased from Applied Biosystem.
Synthesis of Vancomycin Derivative V-Rd-V. The dimeric van-
comycin V-Rd-V was synthesized according to a reported coupling
protocol.20 To the solution of 100 mg of vancomycin hydrochloride
(67 µmol) in 0.5 mL of dry dimethyl sulfoxide (DMSO) were added
0.5 mL of dry dimethylformamide (DMF) and 4.5 mg (0.5 equiv) of
p-xylylenediamine. The mixture was cooled to 0 °C, and 38 mg (1.5
equiv) of HBTU was added, followed by 20 mg (2.3 equiv) of
diisopropylethylamine (DIEA). The solution was allowed to warm to
room temperature and stirred overnight. Analytical reverse-phase
HPLC showed nearly complete loss of vancomycin and appearance of
a much less polar major product. Removal of the solvent afforded
100 mg of crude product, and 20 mg of this crude product was purified
by preparative reverse-phase HPLC and lyophilized to afford 8.6 mg
(2.5 µmol, 38%) of the dimeric derivative of vancomycin, V-Rd-V, as
its trifluoroacetate salt. The 1H-NMR spectroscopy showed resonances
attributable to vancomycin as well as the linker: a new triplet at δ
7.18 (amide NH), a single at δ 7.14 (phenyl CHs), and a singlet at δ
4.45 (CH2PhCH2); ESI-MS exhibited an ion at m/z 2998.6, consistent
with the molecular weight calculated for the parent ion (M + H+),
C140H159N20O46Cl4, 2998.
Syntheses of Peptide Ligands. Peptide ligands L, P, fL, and
L-R'd-L were synthesized by solution method32 and final products were
purified by reverse phase HPLC. HBTU was used to activate all the
peptide coupling in our syntheses. Benzyloxycarbonyl protection was
used to protect amines and cleaved by palladium-catalyzed hydrogena-
tion under a hydrogen balloon. The tert-butyl ester was used to protect
the carboxylic acid and cleaved by 1:1 TFA/methylene chloride.
Ligands L, P, and L-R'd-L were synthesized from the same tripeptide
precursor, NR-Ac-L-Lys-D-Ala-D-Ala-O-tBu.
Nr-Carbobenzoxy-L-glutamyl-(Nr-acetyl)-L-lysyl-D-alanyl-D-ala-
nine (P). Coupling of NR-Ac-L-Lys-D-Ala-D-Ala-O-tBu with NR-Z-L-
Glu-γ-tert-butyl ester afforded the tetrapeptide ester. After cleavage
of the tert-butyl ester, the free peptide was purified by HPLC. 1H-
NMR (DMSO-d6) δ 8.13 (d, 1H), 8.07 (d, 1H), 7.99 (d, 1H), 7.85 (t,
1H), 7.37-7.29 (m, 5H), 5.0 (q, 2H), 4.30-4.20 (m, 1H), 4.17-4.13
(m, 2H), 3.93-3.88 (m, 1H), 3.03-2.96 (m, 2H), 2.20 (t, 2H), 1.90-
1.80 (m, 1H), 1.81 (s, 3H), 1.75-1.65 (m, 1H), 1.65-1.40 (m, 2H),
1.40-1.36 (m, 2H), 1.36-1.16 (m, 2H), 1.26 (d, 3H), 1.17 (d, 3H);
FAB-MS, for C27H39N5O10Na (M + Na+), calcd m/z 616.2595, found
m/z 616.2595.
NE,N'E-[1,4-(1,4-Dioxobutanediyl)]bis(NR-acetyl-L-lysyl-D-alanyl-
D-alanine) (L-R'd-L). To 98 mg (0.254 mmol) of NR-Ac-L-Lys-D-
Ala-D-Ala-O-tBu and 15 mg (0.127 mmol) of succinic acid in 10 mL
of acetonitrile solution were added 120 mg (0.317 mmol) of HBTU
and 33 mg of (0.256 mmol) DIEA. The reaction was stirred at room
Acknowledgment. J. Rao gratefully acknowledges G. Sigal
for helpful discussions. This work was supported by NIH grants
GM 30367 and GM 51559. J. Rao was supported by an Eli
Lilly predoctoral fellowship in 1996. The ESI-MS spectrum was
acquired by the Harvard Microchemistry Facility. The NMR
facilities at Harvard were supported by NIH grant 1-S10-
RR04870-01 and NSF grant CHE88-14019. The Harvard
University Mass Spectrometry Facility was supported by grants
from NSF (CHE-9020043) and NIH (SIO-RR06716).
Supporting Information Available: A detailed derivation
of eq 2 (5 pages). See any current masthead page for ordering
and Internet access instructions.
(32) Dourtolou, V.; Gross, B.; Lambropoulou, V.; Zioudrou, C. J. Chem.
Soc., Chem. Commun. 1984, 572-274.
JA971225L