Synthesis and binding studies of two new macrocyclic receptors for the stereoselective recognition of dipeptides

Article abstract of DOI:10.3762/bjoc.6.5

We present here the design, synthesis, and analysis of a series of receptors for peptide ligands inspired by the hydrogen-bonding pattern of protein β-sheets. The receptors themselves can be regarded as strands 1 and 3 of a three-stranded β-sheet, with crosslinking between the chains through the 4-position of adjacent phenylalanine residues. We also report on the conformational equilibria of these receptors in solution as well as on their tendency to dimerize. ¹H NMR titration experiments are used to quantify the dimerization constants, as well as the association constant values of the 1:1 complexes formed between the receptors and a series of diamides and dipeptides. The receptors show moderate levels of selectivity in the molecular recognition of the hydrogen-bonding pattern present in the diamide series, selecting the α-amino acid-related hydrogen-bonding functionality. Only one of the two cyclic receptors shows modest signs of enantioselectivity and moderate diastereoselectivity in the recognition of the enantiomers and diastereoisomers of the Ala-Ala dipeptide (DDG⁰ ₁ (DD-DL) = -1.08 kcal/mol and DDG⁰ ₁(DD-LD) = -0.89 kcal/mol). Surprisingly, the linear synthetic precursors show higher levels of stereoselectivity than their cyclic counterparts.

Full text of DOI:10.3762/bjoc.6.5

Synthesis and binding studies of two new
macrocyclic receptors for the stereoselective
recognition of dipeptides
Ana Maria Castilla1, M. Morgan Conn2,3 and Pablo Ballester*1,4
Full Research Paper
Open Access
Address:
Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
1Institute of Chemical Research of Catalonia (ICIQ), Avgda. Països
Catalans 16, 43007 Tarragona, Spain, 2Amherst College, Amherst,
MA 01002, USA, 3now at PTC Therapeutics, Inc., 100 Corporate
Court, South Plainfield, NJ 07080, USA and 4Catalan Institution for
Research and Advanced Studies (ICREA), Passeig Lluís Companys,
23, 08018 Barcelona, Spain
doi:10.3762/bjoc.6.5
Received: 22 September 2009
Accepted: 06 January 2010
Published: 19 January 2010
Guest Editor: C. A. Schalley
Email:
Pablo Ballester* - pballester@iciq.es
© 2010 Castilla et al; licensee Beilstein-Institut.
License and terms: see end of document.
* Corresponding author
Keywords:
dipeptides; host–guest; macrocyclic; molecular recognition; receptors;
stereoselective
Abstract
We present here the design, synthesis, and analysis of a series of receptors for peptide ligands inspired by the hydrogen-bonding
pattern of protein β-sheets. The receptors themselves can be regarded as strands 1 and 3 of a three-stranded β-sheet, with cross-
linking between the chains through the 4-position of adjacent phenylalanine residues. We also report on the conformational equi-
libria of these receptors in solution as well as on their tendency to dimerize. 1H NMR titration experiments are used to quantify the
dimerization constants, as well as the association constant values of the 1:1 complexes formed between the receptors and a series of
diamides and dipeptides. The receptors show moderate levels of selectivity in the molecular recognition of the hydrogen-bonding
pattern present in the diamide series, selecting the α-amino acid-related hydrogen-bonding functionality. Only one of the two cyclic
receptors shows modest signs of enantioselectivity and moderate diastereoselectivity in the recognition of the enantiomers and
diastereoisomers of the Ala-Ala dipeptide (ΔΔG01 (DD-DL) = −1.08 kcal/mol and ΔΔG01 (DD-LD) = −0.89 kcal/mol). Surpris-
ingly, the linear synthetic precursors show higher levels of stereoselectivity than their cyclic counterparts.
Introduction
Manipulation of protein–protein interactions is gaining interest tion, and apoptosis. Rational protein surface recognition poses a
as they are known to play a critical role in important biological challenging test to our actual knowledge of molecular design.
processes such as the normal function of cellular/organelle Nevertheless, its practice and developments will provide a
structure, immune response, enzyme inhibitors, signal transduc- better understanding of protein–protein interactions. Interest-
Page 1 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
ingly, in molecule-based disease therapy, the disruption of efficient receptors prepared might be useful as scaffolds for
protein–protein interactions by small molecules constitutes an future antibiotics.
alternative approach to the classical active-site enzyme inhibi-
tion design. One of the strategies employed for binding protein Herein, we report the design and synthesis of two new macro-
surfaces relies on the use of arrays of synthetic receptors origin- cyclic receptors, 1 and 2, conceived for the binding of
ally designed for the recognition of oligopeptides. dipeptides, in particular for the selective recognition of D-Ala-
Consequently, the selective recognition of oligopeptides repres- D-Ala. We also report on the studies performed using these two
ents an intermediate step toward the recognition of protein macrocyclic receptors, as well as their linear precursors, in the
surfaces [1]. The studies of host–guest complexes as model molecular recognition of a series of dipeptides and diamides
systems of peptide–peptide interactions are of particular interest with diverse hydrogen-bonding patterns. We rationalize the
because they may provide insight into the structural basis of the observed modulation of their binding affinity as a function of
high size/shape specificities and enantioselectivities exhibited the hydrogen-bonding pattern exhibited by the target molecule.
by the complex protein–protein recognition processes that occur We also describe the levels of stereoselectivity displayed by
in biology. Moreover, short oligopeptides are themselves worth- these receptors in the recognition of the diastereoisomers and
while targets for recognition and their conformational flexib- enantioisomers of Ala-Ala dipeptide. We explain the differ-
ility represents an added challenge to achieve selective binding. ences observed in their binding abilities as a function of
The preparation of synthetic receptors for the selective binding conformational rigidity (macrocyclic vs linear receptors).
of short oligopeptides has potential applications in the develop-
Results and Discussion
ment of diagnostic sensors, separation techniques, and thera-
Design of the synthetic receptors: general
peutic agents.
considerations
With respect to this latter point, there is a significant interest in
the advance of receptors that selectively bind the D-Ala-D-Ala
dipeptide, the common target for the vancomycin antibiotics.
This group of antibiotics is active against certain aerobic and
anaerobic Gram-positive bacteria, and has been used for many
years as treatment of last resort in clinical wards [2-4].
However, vancomycin resistance has recently been identified
among clinical isolates of several Gram-positive species [5-8].
Therefore, although many examples already exist in the litera-
ture [9,10], the design and synthesis of new synthetic receptors
for this dipeptide is still a relevant endeavor not only in terms of
understanding the interactions that take place during vanco-
mycin action, but also because the structures of the most
The design of the receptors described in this article is based on
the interactions that occur in the β-sheets commonly found in
the secondary structure of many biologically relevant proteins.
We start from a schematic termolecular complex mimicking a
three-stranded β-sheet in which the central strand corresponds
to the target guest peptide and the two outer strands constitute
the structure of the host (Figure 1). In this design, we employ
some of the properties of the β-sheet structure – the conver-
gence of hydrogen-bonding patterns and the presence of
exposed side chains. It is worth mentioning that the β-sheet
structure has already been used as the inspirational binding
motif for the preparation of other synthetic receptors for
peptides [11-16]. However, we believe that our design includes
Figure 1: Schematic representation of the design of a host–guest complex based on antiparallel β-sheet geometry. *The presence of a stereogenic
center.
Page 2 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
some novelties. To reduce the conformational flexibility and arrangement, the successive β-strands alternate directions so
confer a certain degree of preorganization to this type of that the N-termini of two adjacent strands are at opposite ends.
receptor, the use of one or two linkers connecting the two In a parallel arrangement, all of the N-termini of successive
peptide strands is mandatory. The principal difference with strands are oriented in the same direction [22]. In contrast,
respect to previous designs of β-sheet-based synthetic receptors successive strands in a mixed-mode arrangement may be
is that the connection between the two peptide strands, used as parallel or antiparallel to each other. To examine the influence
the receptor’s binding sites, emerges from their side chains and of the relative orientation of the two receptor strands on their
not from their C- or N-terminus. In our final design, we propose binding abilities, we conceived and synthesized two analogous
the introduction of two linkers connecting the two peptide receptors mimicking these two types of arrangements present in
strands affording, a macrocyclic structure. In doing so, we the β-sheet structure (Figure 2). The outer peptide strands of
expect that the molecular recognition properties of the designed receptor 1 are arranged antiparallel to each other (“antiparallel
receptor will also benefit from the macrocyclic effect [17-20]. receptor”). That is, the stereogenic center of the C-terminus of
Simple molecular modeling studies [21] revealed that a benzo- one strand is covalently connected to the stereogenic center of
phenone unit would be ideally suited to span the gap between the N-terminus of the other strand. Receptor 1 is anticipated to
two methyl side chains emerging from alanyl residues of the form a mixed-mode sheet structure with an included peptide
outer strands in the three-stranded β-sheet complex (Figure 1). ligand. Conversely, the two outer strands of receptor 2 are
oriented parallel to each other (“parallel receptor”), such that
In proteins, adjacent β-strands can form hydrogen bonds in anti- the covalent connections between strands join similar stereo-
parallel, parallel, or mixed arrangements. In an antiparallel genic centers, C-terminus with C-terminus and N-terminus with
Figure 2: Molecular structures of the two designed receptors 1 and 2 having different relative orientations of the peptide strands.
Page 3 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
N-terminus. Receptor 2 is anticipated to form an antiparallel incurring any substantial steric clashes. We also observed
sheet structure with the included peptide ligand. This change in appropriate complementarity between the hydrogen-bonding
connectivity does not involve any inversion of the stereogenic groups of substrate and receptor (Figure 3). The analysis of the
centers but only a modification in the sequence of peptide-coup- structures of the minimized complexes revealed that they are
ling reactions that yield the cyclic structure, and will be stabilized by the formation of the same number of hydrogen
explained below.
bonds, that is, five. The hypothesized “endo” structure for the
complexes of 1 and 2 with n-C6H13CO-D-Ala-D-Ala-NH2, in
The exploration of the conformational space of both macro- which the ligands thread through the receptor’s macrocycle,
cycles, using molecular modeling, indicated the existence of a also allows for the possibility of binding short amino acid
built-in cavity. These studies also suggested that the reduced sequences not necessarily located on the edges of larger
conformational flexibility of the receptors avoids the complete peptides.
collapse of the cavity through the formation of intramolecular
hydrogen bonds. Moreover, we were able to minimize struc- Other considerations, apart from preventing intramolecular
tures for the complexes formed between both receptors and hydrogen bond formation, related to the use of
n-C6H13CO-D-Ala-D-Ala-NH2 in which the dipeptide is bis(alanyl)benzophenone rigid linkers include: a) to avoid steric
threaded through the macrocycle (Figure 3). In these minim- clashes between the methyl groups of the target peptide and the
ized structures, the hydrogen-bonding groups of the receptor benzophenone linking chains, the stereogenic centers in the
converge toward the center of the macrocycle. The macrocycle linkers must have the (S) configuration, opposite to that of the
is large enough to accommodate the threading dipeptide without bound peptide (R); b) the benzophenone aromatic ring will also
provide a hydrophobic pocket for the neighboring side chains of
the target peptide, and may promote the formation of additional
CH–π and π–π intermolecular interactions. The pleat of the
bound D-Ala-D-Ala peptide is inverted because of the unnat-
ural stereochemistry and the resulting complex is not exactly a
β-sheet. Attaching the linking groups in the side chains leaves
the ends of the receptor strands open, allowing the introduction
of additional interactions between the receptor and the chain of
a larger peptide.
As schematically depicted in Figure 2, we planned that both
receptors could be obtained through cyclic dimerization,
through the formation of two peptide bonds, of two S,S-
bis(alanyl)benzophenone units 3. In turn, the synthesis of the
protected S,S-bis(alanyl)benzophenone units 3 could be easily
achieved by means of Stille carbonylative cross-coupling reac-
tions of two adequately bisprotected S-phenylalanine deriva-
tives, iodo-aryl 4 and trimethylstannyl-aryl 5, following experi-
mental procedures described in recent literature reports [23].
Synthesis
The synthetic strategy designed for the construction of receptors
1 and 2 involves the use of a carbonylative cross-coupling reac-
tion between two aryl derivatives (iodo-aryl 4 and trimethyl-
stannyl-aryl 5) to prepare 4,4’-bis(alanyl)benzophenones 3,
followed by macrocyclization of two molecular units of 3. The
macrocyclization reaction of two 4,4’-bis(alanyl)benzophen-
ones 3 will be promoted by the sequential and regioselective
Figure 3: CAChe minimized structures for the “endo” complexes
formed between receptors 1 (a) and 2 (b) and the n-C6H13CO-D-Ala-
D-Ala-NH2 dipeptide. The absolute configuration of the stereogenic
formation of two peptide bonds between them, the first one
through an intermolecular reaction and the second one
centers are indicated with a capital letter. Five intermolecular hydrogen
bonds are also shown as black dashed lines.
intramolecularly, affording the desired macrocyclic structures 1
and 2. The main dissimilarity between the two synthetic
Page 4 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
strategies resides in the type of functional groups that each 4,4’- peptide synthesis [24]. This characteristic of the carbonylative
bis(alanyl)benzophenone 3 supplies to the macrocyclization cross-coupling reaction allowed us to achieve the differential
reaction. Thus, for the synthesis of antiparallel receptor 1 each protection of the two amino acid moieties present in the 4,4’-
benzophenone unit will provide, in an alternative way, one bis(alanyl)benzophenones 3 by protecting separately the func-
carboxylic and one amino function to the final macrocyclic tional groups in the reaction partners, 4 and 5, before attempting
skeleton. Conversely, for the synthesis of antiparallel receptor the cross-coupling. We prepared a single orthogonally protected
2, one benzophenone unit will donate its two carboxylic acid benzophenone 3a for the synthesis of the antiparallel macro-
functions while the other will participate with its two amino cycle 1. In contrast, the synthesis of parallel receptor 2 called
groups. To achieve the regioselective control demanded in the for the preparation of two differently protected
macrocyclization reactions, a precise selection of the ortho- bis(alanyl)benzophenone units, 3b and 3c. All iodo-phenyl de-
gonal protecting groups to be included in the bis-amino acid rivatives 4 were prepared in high yields using standard proced-
functionalities of the benzophenone derivatives 3 is needed. The ures (Scheme 1). Thus, 4-iodo-L-phenylalanine 6 (I-Phe) was
starting material for both synthetic routes is 4-iodo-L- converted into the methyl ester hydrochloride by treatment with
phenylalanine (6). We prepared 6 in multigram scale starting thionyl chloride in methanol followed by acylation of the amino
from commercial L-phenylalanine (7) by following a described group with tert-butyl dicarbonate to yield Boc-I-Phe-OMe, 4a.
procedure [24] consisting in the iodination of 7 in acetic acid Iodo-L-phenylalanine 6 was acylated under Schotten–Baumann
solution in the presence of I2, NaIO3, and sulfuric acid conditions with benzyl chloroformate to obtain the N-protected
(Scheme 1). We obtained (S)-6 in enantiomerically pure form in amino acid Cbz-I-Phe. This compound was esterified with
50% yield. Since we plan to assemble the 4,4’-bis(alanyl)benzo- 2-(trimethylsilyl)ethanol using DCC as coupling agent,
phenones 3 by a Stille carbonylative cross-coupling reaction, providing Cbz-I-Phe-TMSE, 4c. In a different reaction, Cbz-I-
the required trimethylstannyl derivatives 5 should be easily Phe was treated with 4-nitrobenzyl bromide and triethylamine
prepared from adequately diprotected phenylalanine iodides 4 to afford Cbz-I-Phe-PNB, 4b [25]. Finally, 6 was treated with
(Scheme 1).
Fmoc hydroxysuccinimide (Fmoc-OSu) [26-29] to obtain the
Fmoc N-protected amino acid that was subsequently esterified
The mild reaction conditions used in the carbonylative cross- with diazomethane affording Fmoc-I-Phe-OMe, 4d.
coupling permit the use of common protecting groups of
Scheme 1: Synthesis of tetraprotected bis(alanyl)benzophenones 3 from L-phenylalanine 7.
Page 5 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
The diprotected aryl iodides, 4a and 4b, were converted
uneventfully to the corresponding diprotected aryl trimethyl-
stannane derivatives, 5a and 5b, by reaction with hexamethyl-
ditin catalyzed by Pd(0) under inert atmosphere. The organo-
stannanes 5 showed signs of decomposition with time, and they
were freshly prepared just before being used in the cross-coup-
ling reaction.
The 4,4’-bis(alanyl)benzophenones 3 were prepared in an ortho-
gonally protected form by carbonylative coupling between
diprotected iodo-aryl derivatives 4 and diprotected aryl
trimethylstannanes 5, using the experimental conditions
described by Morera and Ortar for similar substrates [23]. The
reactions were performed at 90 °C under atmospheric CO pres-
sure in the presence of PdCl2/PPh3, proceeding smoothly to
give derivatives 3 with isolated yields, after column chromato-
graphy, in the range of 53–61%. This complete synthetic
sequence is reminiscent of the work of Lei et al. for the prepara-
tion of phosphinate bis-amino acids [24]. This convergent route
allows the installation of diverse and differentiable function-
ality in a small molecule like 3.
The sequential peptide coupling of two units of 4,4’-
bis(alanyl)benzophenones 3a should lead to the construction of
the designed macrocyclic bis-dipeptide receptor 1. Benzo-
Scheme 2: Deprotection reactions of bis(alanyl)benzophenone units 3.
phenone 3a was converted into the carboxylic acid 8 by treat- However, the subsequent sequence of reactions directed toward
ment with tetrabutylammonium fluoride [25] (Scheme 2). In a the macrocyclic receptors utilized, as starting material, the
separate reaction, 3a was treated with trifluoroacetic acid to diastereomeric mixture of 12 or 13 obtained by flash chromato-
remove the Boc group and produce the trifluoroacetic salt of graphy purification of the reaction crude. The deprotections of
amine 9 [25]. Both deprotection reactions proceeded unevent- the diastereoisomeric mixtures were carried out using standard
fully in almost quantitative yields. The PNB group in 3b was methods. First, we used fluoride to cleave the TMSE group, and
removed using a mixture of SnCl2 and phenol in acid media and subsequently, we removed the Boc group by the action of TFA.
the Fmoc [30] in 3c using piperidine to obtain 10 and 11, We obtained the bis-deprotected tetrapeptides 15 and 17 in high
respectively (Scheme 2).
yield (70–80%).
Next, we carried out intermolecular peptide-coupling reactions The macrocyclization reactions of the linear tetrapeptides, 15
between 8 and 9, as well as between 10 and 11 to obtain the and 17, were carried out under high-dilution conditions. Using a
linear tetrapeptides 12 and 13, direct precursors of receptors 1 syringe pump and under inert atmosphere, a DMF solution of
and 2, respectively. The best results for the coupling reactions the corresponding linear tetrapeptide was added dropwise, over
were obtained when using a combination of HATU/NMM [31, a period of 12 h, to a stirred DMF solution containing the coup-
32] in DMF at room temperature (Scheme 3). The analysis of ling agent and the base. The purification of the crude macrocyc-
the crude reaction mixtures by HPLC and 1H NMR spectro- lization reactions using flash chromatography afforded the
scopy revealed that both tetrapeptides, 12 and 13, were obtained expected macrocyclic products in acceptable yields but as
as mixtures of two diastereoisomers. Most likely, the stereo- complex mixtures of diastereoisomers. HPLC–MS analysis of
genic center in the α-position with respect to the carboxylic the isolated fraction showed the presence of four different peaks
group undergoing activation during peptide coupling was in the chromatogram producing ions with molecular mass value
partially epimerized. The all-S diastereoisomers, (S)-12 and (S)- corresponding to the expected cyclic structure. We tentatively
13, were the major products detected in the crude reaction mix- assigned the two major peaks to cyclic diastereoisomers formed
ture. They were isolated as pure compounds using preparative during the intramolecular peptide-coupling reaction of the all-S
reverse-phase HPLC and fully characterized by a complete set linear tetrapeptide. As discussed above, one of the two diaste-
of high-resolution spectra.
reoisomers is probably the outcome of the epimerization reac-
Page 6 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
Scheme 3: Synthesis of the linear tetrapeptides 15 and 17 as mixtures of diastereoisomers.
tion experienced by the stereogenic center in the α-position with Initially, we used HBTU/NMM [25,33,34] for activation of the
respect to the carboxylic group undergoing activation. Like- intramolecular peptide bond formation. We observed consider-
wise, the two minor peaks should correspond to cyclic diaste- able epimerization at the stereogenic α-carbon. We assessed the
reoisomers formed from macrocyclization and concomitant coupling reaction using different coupling methods, HATU/
epimerization reactions experienced by the minor linear NMM [35] and PyAOP/DIEA [36], and found that although the
tetrapeptide S,S,S,R also incorporated into the starting material. overall reaction yields were independent of the coupling
Figure 4b depicts the HPLC chromatogram obtained from the method, the epimerization diminished substantially when the
analysis of the purified fraction containing the mixture of PyAOP/DIEA [35,36] combination was used (Figure 5c).
diastereoisomers of receptor 1. Using normal-phase preparative
HPLC, we isolated the two major products of the macrocycliza- Conformational studies
tion reaction of 15 as pure compounds. The structures of the The 1H NMR spectra of chloroform-d solutions of the diaste-
isolated products were assigned by means of standard spectro- reomerically pure all-S cyclic receptors 1 and 2, as well as those
scopic techniques and symmetry considerations to cyclic diaste- of their linear tetraprotected precursors, 15 and 17, were tem-
reoisomers of receptor 1. Furthermore, the structure of the perature-dependent (Figure 6). We attribute this temperature de-
major product of the cyclization of 15 was also characterized in pendence to the existence of conformational equilibria that are
the solid state by X-ray diffraction and proved to be the desired in a slow chemical exchange regime with respect to the NMR
all-S antiparallel cyclic receptor 1. The results obtained in the time scale, i.e., the rotation of the C–N single bond in the
macrocyclization of tetrapeptide 17 were completely analogous. carbamate protecting groups [37,38].
The all-S diastereoisomer corresponds to macrocyclic receptor
2, and was the major product isolated from the purification of Upon increasing the temperature of chloroform-d solutions of 1,
the reaction mixture using normal-phase preparative HPLC. 2, 15, and 17, the proton signals became sharper and well
Receptor 2 was fully characterized by means of standard spec- defined, which is indicative that the chemical exchange due to
troscopic techniques.
the conformational equilibria has been accelerated. Conversely,
cooling the samples slows down the rate of the chemical
Page 7 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
Figure 4: a) Molecular structures of the two major diastereoisomers of the cyclic receptors obtained from the intramolecular coupling reactions of 15
and 17. b) HPLC chromatogram of the purified fraction containing the mixture of diastereoisomers of 1. c) X-ray structure of the receptor 1.
exchange. Thus, at low temperature, we observed the appear- tion constants in chloroform. The calculation of the dimeriza-
ance of new proton signals that were assigned to different tion constants relies on the chemical shift changes observed for
conformations. We observed another general trend in the vari- certain proton signals of the receptors when their 1H NMR
able-temperature 1H NMR spectra, that is, as the temperature spectra are acquired at different concentrations. In particular,
was lowered, the NH signals shifted downfield. This behavior the receptors’ NH signals shift downfield when the concentra-
suggested that the cyclic and acyclic peptides may dimerize or tion of the solution is increased, indicating the formation of
oligomerize in chloroform solution through the formation of aggregates in the solution that are stabilized through hydrogen
intermolecular NH···O hydrogen bonds. We have already bonding. The observed chemical shifts for the NH signals were
observed the formation of intermolecular hydrogen bonds in the analyzed mathematically using the HypNMR software and a
solid-state structure of receptor 1 (Figure 7).
simple theoretical dimerization binding model [39,40]. We
obtained a good fit between the experimental and theoretical
Before undertaking the study of the binding and molecular data. Additional conclusions can be drawn from the data
recognition properties of the receptor series, and due to their presented in Table 1. Macrocyclic receptors 1 and 2 show
tendency to aggregate in solution, we quantified their dimeriza- greater tendency to dimerize than their linear precursors.
Page 8 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
Figure 6: Variable-temperature 1H NMR experiments of 1 in chloro-
form-d solution. The proton signals that appeared at low temperature
are marked with an asterisk.
Figure 5: Reverse-phase HPLC chromatograms of the purified frac-
tion obtained from macrocyclization reactions yielding 1 using different
coupling agents. The all-S cyclic receptor 1 has a retention time of tr =
26.8 min, and is the major component in the three analyzed mixtures.
The peak with retention time of tr = 27.8 min corresponds to the
R,R,R,S-1 diastereoisomer.
binding affinity. We also investigated the molecular recogni-
tion properties of the receptor series with a set of dipeptides.
The effect of the size of the amino acid substituents in the
dipeptide series (Ala-Ala vs L-Phe-L-Phe) was investigated to
shed some light on the geometry of the complexes formed with
the cyclic receptors. Finally, the stereoselective recognition
properties of the receptors were derived from their binding
interactions with the four diastereoisomers of Ala-Ala.
A stronger dimerization tendency for the antiparallel cyclic
receptor 1 also becomes apparent.
We used a wide range of guest molecular structures to examine
the molecular recognition properties of receptors 1, 2, 15, and The molecular structures of all selected guests have several
17 (Figure 8). We selected a series of diamides to evaluate the hydrogen-bonding groups, making them natural candidates to
effect that the hydrogen-bonding pattern produces in the dimerize in solution. Therefore, before studying the interac-
tions of these guests with the receptors, we studied their dimer-
ization behavior in chloroform solution. Using the same meth-
odology described above for the receptors, we calculated the
dimerization constants of all guest molecules. The values
obtained are summarized in Table 2. With an additional amide
group with respect to diamides, the dipeptide dimers can be
stabilized by a higher number of hydrogen bonds. The value of
Table 1: Calculated dimerization constant values for the receptor
series.
Receptors
Kd (M−1)a
1
112
60
2
15
17
27
Figure 7: Small fraction of the columnar arrangement observed in
solid-state packing of receptor 1. Two adjacent molecules of 1 interact
through the establishment of four hydrogen bonds (yellow dashed
lines). For clarity nonpolar hydrogen atoms are omitted.
48
aValues determined in chloroform-d solution at 298 K using 1H NMR
dilution experiments. All values are associated with at least a 10% error.
Page 9 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
the dimerization constant of the diamide of fumaric acid stands D-Ala-D-Ala-NH2. We monitored the chemical shift changes
out from the rest, likely due to the higher conformational experienced by the NH proton signals of the receptor and of the
rigidity of this compound (preorganization). Figure 9 depicts guest when a 1 mM chloroform-d solution of the receptor is
the 1H NMR spectra acquired in the variable-concentration treated with incremental amounts of the guest. The titration data
experiments used for the calculation of the dimerization were fitted to a theoretical binding model considering the
constant of fumaramide. The NH proton signals experience a exclusive formation of a 1:1 complex, and the existence of
significant downfield shift on increasing the concentration dimeric aggregates of both the receptor and the guest.
of fumaramide.
Figure 10b depicts the experimental data of the titration fitted to
the theoretical binding isotherm derived from the above-
Having determined the dimerization tendency of host and guest mentioned theoretical model. The values of the calculated
molecules, we initiated the study of the molecular recognition stability constants for the 1:1 complexes are summarized in
properties of the receptors toward the different guests. All Table 3 and Table 4.
binding constants were determined using 1H NMR titration
experiments. As an example, Figure 10a shows a series of The analysis of the tabulated data allowed us to draw several
spectra acquired during the titration of receptor 2 with n-C6H13- conclusions (Table 3 and Table 4). The macrocyclic receptors
Figure 8: Molecular structures of the guests used in the binding experiments.
Page 10 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
Table 2: Dimerization constant values calculated for the guests used
in this study.
Diamides of
Kd (M−1)a
Dipeptides
Kd (M−1)a
Succinic acid
Malonic acid
66
11
D-Ala-D-Ala
L-Ala-L-Ala
D-Ala-L-Ala
L-Ala-D-Ala
370
346
331
316
Ethylenediamine 44
Propane-1,3-dia
mine
44
Glutaric acid
Maleic acid
Fumaric acid
Gly
<10
31
L-Phe-L-Phe 346
478
22
aValues determined in chloroform-d solution at 298 K using 1H NMR
dilution experiments. All values are associated with at least a 10% error.
do not show any affinity for the complexation of diamides in
which the two amide groups are spanned by three methylene
groups (glutaramide and propane-1,3-diamine). However, these
receptors do form complexes with the rest of diamides showing
certain degree of selectivity in response to the hydrogen-
bonding pattern (Table 3). The antiparallel macrocycle 1
exhibits a moderate preference for the hydrogen-bonding
pattern DAAD (D = hydrogen bond donor, A = hydrogen bond
acceptor) instead of ADAD when just one methylene group
spans the two amide groups (ΔΔG0 (malonamide-Gly) = −0.96
kcal/mol). In contrast, parallel receptor 2 effectively discrimin-
ates in favor of the hydrogen-bonding pattern ADDA when two
methylene groups span the amide groups (ΔΔG0 (ethylene-
Figure 10: a) Selected region of a series 1H NMR spectra acquired
during titration of receptor 2 with n-C6H13-D-Ala-D-Ala-NH2; b) fit of the
experimental data of the titration to the theoretical binding isotherm of
the formation of a complex with 1:1 stoichiometry.
diamine-succinamide) = −2.35 kcal/mol). The calculated in comparison with the rest of diamides resides in the reduced
stability constants are, in general, lower than the values conformational flexibility of the substrate (ΔΔG01 (fumara-
expected for a complex that can be stabilized by an array of mide-succinamide) = −2.46 kcal/mol and ΔΔG02 (fumaramide-
not adjacent four hydrogen bonds in chloroform solution succinamide) = −2.79 kcal/mol). When the association constant
(K ≈ 104 M−1). The stability constant values determined for the values obtained for the DAAD hydrogen-bonding pattern are
complexes formed by both cyclic receptors and fumaramide are compared, it becomes evident that both cyclic receptors exhib-
more consistent with our estimate. Most likely, the high thermo- ited a marked preference for the diamides in which the NH–CO
dynamic stability calculated for the complexes of fumaramide groups are separated by just one methylene group (ΔΔG01
(malonamide-succinamide) = −1.56 kcal/mol and ΔΔG02 (malo-
namide-succinamide) = −1.01 kcal/mol).
Not surprisingly, the binding affinities calculated for the cyclic
and acyclic receptors toward the dipeptide series were higher
than those for the diamides. Dipeptides have an additional
amide hydrogen-bonding group. The degree of stereoselectivity
displayed by the cyclic and acyclic receptors was low (two
possible binding geometries for the complexes formed between
the macrocyclic receptors and n-C6H13-L-Phe-L-Phe-NH2 are
shown in Figure 11).
The cyclic antiparallel receptor 1 showed reduced signs of
Figure 9: Selected region of the variable-concentration 1H NMR
spectra acquired using chloroform-d solutions of fumaramide. The
signal of the NH proton is marked with an asterisk.
enantioselectivity and moderate diastereoselectivity in the
recognition of the enantiomers and diastereoisomers of the Ala-
Page 11 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
Table 3: Binding constants (Kass) and free energies of complexation (−ΔG0 at 298 K) of the 1:1 complexes formed between the cyclic receptors 1 and
2 and the different guests used in this study.
Cyclic receptors
“Antiparallel”
Receptor 1
“Parallel”
Receptor 2
Kassa (M−1)
−ΔG0 (kcal/mol)
Kassa (M−1)
−ΔG0 (kcal/mol)
Guests (DAAD)b
Malonamide (1CH2)
Succinamide (2CH2)
Glutaramide (3CH2)
Maleamide (2CH)
Fumaramide (2CH)
Guests (ADDA)b
Ethylenediamine (2CH2)
Propane-1,3-diamine (3CH2)
Guest (ADAD)b
Gly (1CH2)
1380
100
<5
4.2
2.7
<1
398
72
3.5
2.5
<1
–c
<5
692
6309
3.8
5.1
–c
7940
5.3
123
<5
2.8
<1
3800
<5
4.8
<1
275
3.3
457
3.6
Guests (ADADAD)b
D-Ala-D-Ala
6606
3311
1047
1445
5012
5.2
4.8
4.1
4.3
5.0
1047
912
4.1
4.0
4.2
3.9
4.5
L-Ala-L-Ala
D-Ala-L-Ala
1148
759
L-Ala-D-Ala
L-Phe-L-Phe
2089
aAll values are associated with at least a 10% error. bHydrogen-bonding pattern; D = donor, A = acceptor. cNot calculated.
Ala dipeptide (ΔΔG01 (DD-DL) = −1.08 kcal/mol and ΔΔG01 same substrates (Table 3). The difference in free energy meas-
(DD-LD) = −0.89 kcal/mol). The parallel receptor 2 showed ured for the complexes of 2 with the four diastereoisomers of
neither enantio- nor diastereoselectivity in the recognition of the Ala-Ala was in the order of 0.3 kcal/mol.
Table 4: Binding constants (Kass) and free energies of complexation (−ΔG0 at 298 K) of the 1:1 complexes formed between the acyclic receptors 15
and 17 and the different guests used in this study.
Acyclic receptors
“Antiparallel”
15
“Parallel”
17
Kassa (M−1)
−ΔG0 (kcal/mol)
Kassa (M−1)
−ΔG0 (kcal/mol)
Guests (DAAD)b
Malonamide (1CH2)
Succinamide (2CH2)
Glutaramide (3CH2)
Maleamide (2CH)
Fumaramide (2CH)
Guests (ADDA)b
Ethylenediamine (2CH2)
Propane-1,3-diamine (3CH2)
Guest (ADAD)b
Gly (1CH2)
104
<5
2.7
<1.0
–c
91
2.6
2.9
2.8
–c
158
126
–c
–c
95
2.6
4.0
973
7413
5.2
446
78
3.6
2.5
33
–c
2.0
–c
158
2.9
417
3.5
Guests (ADADAD)b
D-Ala-D-Ala
6309
1202
436
794
–c
5.2
4.2
3.6
3.9
–c
8318
4365
190
436
–c
5.3
4.9
3.1
3.6
–c
L-Ala-L-Ala
D-Ala-L-Ala
L-Ala-D-Ala
L-Phe-L-Phe
aAll values are associated with at least a 10% error. bHydrogen-bonding pattern; D = donor, A = acceptor. cNot calculated.
Page 12 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
We also investigated the complexation affinity of the cyclic The low stereoselectivity exhibited by the cyclic receptors,
receptors toward n-C6H13-L-Phe-L-Phe-NH2, with the aim of together with the lack of selectivity for the size of the amino
gaining some information about the geometry of the complex. acid side chain, encourages us to propose that the geometry of
Molecular modeling suggested that although the formation of an the 1:1 complex is, most likely, exo-cyclic. In other words, the
endo-complex in which n-C6H13-L-Phe-L-Phe-NH2 is threaded dipeptide is not threaded through the cyclophane skeleton of the
through the macrocyclic ring of the receptor is plausible, the receptor but bound externally. This hypothesis is also supported
steric clashes detected between the dipeptide side chains and the by the fact that we were unable to observe upfield shifts in any
benzophenone linking units should significantly reduce the of the protons of the dipeptide during the binding experiments.
binding affinity of the cyclic receptors for this substrate or even The inclusion of the dipeptide in the aromatic cavity of the
favor the formation of an alternative complex with exo- receptor should produce the shielding of some of its protons
geometry. Unexpectedly, the stability constant values that we due to the anisotropic magnetic current produced by the
calculated for the 1:1 complexes of the cyclic receptors and aromatic rings.
n-C6H13-L-Phe-L-Phe-NH2 were higher than those for any of
the complexes with Ala-Ala (Table 3). Probably, additional The linear receptors 15 and 17 seem to be more promiscuous in
intermolecular interactions between the receptors and the the interaction with the diamides (Table 4). In general the
phenyl side chains are responsible for the increase in affinity. binding affinities are low, except for the fumaramide. The linear
receptor 17 shows moderate selectivity for the hydrogen-
bonding pattern DAAD instead of ADAD when n = 2 (ΔΔG0
(succinamide-ethylenediamine) = −0.92 kcal/mol) but selects
the hydrogen-bonding pattern ADAD when n = 1 (ΔΔG0 (Gly-
malonamide) = −0.90 kcal/mol).
Surprisingly, linear receptors 15 and 17 exhibited higher levels
of stereoselectivity than their cyclic counterparts (Table 4).
Receptor 15 displayed the highest enantioselectivity we have
measured in the molecular recognition of the D-Ala-D-Ala
dipeptide (ΔΔG015 (DD-LL) = −1 kcal/mol) and an acceptable
level of diastereoselectivity (ΔΔG015 (DD-DL) = −1.60 kcal/
mol). Even higher values of diastereoselectivity were obtained
when studying the interaction between the linear receptor 17
and Ala-Ala diastereomers (ΔΔG017 (DD-DL) = −2.18 kcal/mol
and ΔΔG017 (DD-LD) = −1.70 kcal/mol). We attribute the
surprising and superior stereoselectivity measured for the linear
receptors to their higher conformational flexibility compared
with the cyclic analogs. This enhanced conformational flexib-
ility allows them to adopt a more effective binding conforma-
tion for the sensing of the substrate’s chirality.
Conclusion
We have designed two macrocyclic receptors for the stereose-
lective recognition of dipeptides on the basis of the interactions
that occur in the β-sheets commonly found in the secondary
structure of many biologically relevant proteins. The geometry
of the putative complex used in the design of the receptors
implies the threading of the dipeptide guest through the macro-
cyclic skeleton of the receptor. The two designed macrocycles,
1 and 2, have been synthesized and fully characterized. One of
the key synthetic steps, which is common to both synthetic
Figure 11: CAChe minimized structures for two possible binding
geometries, a) exo and b) endo complexes formed between receptor 1
routes, consists in the use of a Stille carbonylative cross-coup-
and n-C6H13-L-Phe-L-Phe-NH2. The macrocyclic receptor is shown as
CPK model and the dipeptide in yellow stick representation.
ling reaction that affords orthogonally tetraprotected 4,4’-
bis(alanyl)benzophenone units in good to acceptable yields.
Page 13 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
Sequential deprotection reactions combined with the formation We conclude with the caveat that the analysis here pre-supposes
of two consecutive amide bonds between two units of 4,4’- that the receptors respond to different ligands with similar
bis(alanyl)benzophenone produced the macrocyclic receptors in binding modes. Due to the complexity of the system, we have
low yield. Notwithstanding the epimerization reactions not attempted to analyze the possibility that multiple binding
observed in the formation of the peptide bonds of the macro- modes – exo-binding, endo-binding – all operate simultan-
cyclic structures, both receptors have been isolated as single eously and to varying degrees depending on the ligand.
diastereoisomers. The molecular structure of receptor 1 has
been confirmed by single-crystal X-ray diffraction analysis.
Supporting Information
Although molecular modeling suggested that the cyclic
receptors can adopt a conformation with a cavity size large
Supporting information features experimental procedures,
enough to include a peptidic substrate, the X-ray structure
characterization data, NMR spectra of selected compounds
obtained for antiparallel receptor 1 shows the collapse of the
and crystallographic data for the solid-state structure of 1.
designed cavity. Although crystal packing may contribute to
Supporting Information File 1
this conformational change to some degree, the solid-state
structure of 1 suggests that the optimal conformation for
binding is probably not the lowest-energy conformation. The
prepared macrocyclic receptors 1 and 2 as well as their acyclic
tetraprotected precursors 15 and 17 show a moderate tendency
to aggregate in chloroform solution. Dilution studies carried out
at room temperature show that the variation in chemical shift
fits a simple theoretical dimerization model, although higher
order aggregation cannot be ruled out. Using 1H NMR titration
experiments we have determined the association constant values
of the 1:1 complexes formed between receptors 1, 2, 15, and 17
and a series of diamides and dipeptides. We have observed that
Synthesis and binding studies of two new macrocyclic
receptors for the stereoselective recognition of dipeptides
[http://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-6-5-S1.doc]
Supporting Information File 2
CIF for the solid-state structure of 1
[http://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-6-5-S2.cif]
each receptor shows different selectivities in the recognition of Acknowledgements
the hydrogen-bonding patterns present in the diamide series as We thank DGPYTC, Ministerio de Ciencia e Innovación
well as of the number of methylene groups used to separate the (CTQ2008-00222/BQU, Consolider Ingenio 2010 Grant
two amide functions. However, when the association constant CSD2006-0003), ICIQ Foundation, and Generalitat de
values obtained for the DAAD hydrogen-bonding pattern are Catalunya (Grant 2009SGR686) for generous financial support.
compared, it becomes clear that both cyclic receptors exhibited
References
1. Peczuh, M. W.; Hamilton, A. D. Chem. Rev. 2000, 100, 2479–2494.
a marked preference for the diamides in which the NH–CO
groups are separated by just one methylene group. It is worth
doi:10.1021/cr9900026
noting that a single methylene unit was used as the spacer for
2. Neu, H. C. Science 1992, 257, 1064–1073.
the diamide guest used in the receptors’ design. We also investi-
doi:10.1126/science.257.5073.1064
gated the stereoselective recognition properties of the synthe-
3. Gold, H. S. Clin. Infect. Dis. 2001, 33, 210–219. doi:10.1086/321815
sized receptors using the four diastereoisomers of the Ala-Ala
4. Williams, D. H.; Bardsley, B. Angew. Chem., Int. Ed. 1999, 38,
dipeptide as guests. The low stereoselectivity displayed by the
cyclic receptors, together with their insensitivity to the size of
the amino acid chain of the dipeptide guest, allows us to
propose that the topology of the 1:1 complexes is not a pseu-
dorotaxane as initially proposed in our design. Most likely, the
guests, dipeptides and diamides, bind to the hydrogen-bonding
groups that are directed toward the exterior of the aromatic
cavity. If macrocyclization results in the receptor adopting a
low-energy conformation different from that envisioned in the
modeled structures, then preorganization will have created an
additional energetic barrier to endo-complexation. Finally, the
affinity and surprising stereoselectivity exhibited by the linear
receptors 15 and 17 are very difficult to rationalize with an
endo-complex geometry.
1172–1193.
doi:10.1002/(SICI)1521-3773(19990503)38:9<1172::AID-ANIE1172>3.
0.CO;2-C
5. Uttley, A. H. C.; Collins, C. H.; Naidoo, J.; George, R. C. Lancet 1988,
331, 57–58. doi:10.1016/S0140-6736(88)91037-9
6. Woodford, N.; Johnson, A. P.; Morrison, D.; Speller, D. C.
Clin. Microbiol. Rev. 1995, 8, 585–615.
7. Hiramatsu, K.; Hanaki, H.; Ino, T.; Yabuta, K.; Oguri, T.; Tenover, F. C.
J. Antimicrob. Chemother. 1997, 40, 135–136.
doi:10.1093/jac/40.1.135
8. Roderich, D. S. ChemBioChem 2002, 3, 295–298.
doi:10.1002/1439-7633(20020402)3:4<295::AID-CBIC295>3.0.CO;2-G
9. Chamorro, C.; Liskamp, R. M. J. Tetrahedron 2004, 60, 11145–11157.
doi:10.1016/j.tet.2004.08.074
10.Xu, R.; Greiveldinger, G.; Marenus, L. E.; Cooper, A.; Ellman, J. A.
J. Am. Chem. Soc. 1999, 121, 4898–4899. doi:10.1021/ja990240i
Page 14 of 15
(page number not for citation purposes)

Beilstein Journal of Organic Chemistry 2010, 6, No. 5.
11.Davies, M.; Bonnat, M.; Guillier, F.; Kilburn, J. D.; Bradley, M.
J. Org. Chem. 1998, 63, 8696–8703. doi:10.1021/jo981741+
12.Jeong, K. S.; Muehldorf, A. V.; Rebek, J., Jr. J. Am. Chem. Soc. 1990,
112, 6144–6145. doi:10.1021/ja00172a049
38.Marcovici-Mizrahi, D.; Gottlieb, H. E.; Marks, V.; Nudelman, A.
J. Org. Chem. 1996, 61, 8402–8406. doi:10.1021/jo961446u
39.Frassineti, C.; Ghelli, S.; Gans, P.; Sabatini, A.; Moruzzi, M. S.;
Vacca, A. Anal. Biochem. 1995, 231, 374–382.
13.LaBrenz, S. R.; Kelly, J. W. J. Am. Chem. Soc. 1995, 117, 1655–1656.
doi:10.1021/ja00110a029
doi:10.1006/abio.1995.9984
40.Reniero, F.; Guillou, C.; Frassineti, C.; Ghelli, S. Anal. Biochem. 2003,
319, 179–194. doi:10.1016/S0003-2697(03)00300-2
14.Schmuck, C.; Geiger, L. J. Am. Chem. Soc. 2004, 126, 8898–8899.
doi:10.1021/ja048587v
15.Yoon, S. S.; Georgiadis, T. M.; Still, W. C. Tetrahedron Lett. 1993, 34,
6697–6700. doi:10.1016/S0040-4039(00)61678-7
16.Yoon, S. S.; Still, W. C. J. Am. Chem. Soc. 1993, 115, 823–824.
doi:10.1021/ja00055a083
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
17.Cabbiness, D. K.; Margerum, D. W. J. Am. Chem. Soc. 1969, 91,
6540–6541. doi:10.1021/ja01051a091
(http://creativecommons.org/licenses/by/2.0), which
permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
18.Hinz, F. P.; Margerum, D. W. J. Am. Chem. Soc. 1974, 96, 4993–4994.
doi:10.1021/ja00822a050
19.Hinz, F. P.; Margerum, D. W. Inorg. Chem. 1974, 13, 2941–2949.
doi:10.1021/ic50142a032
20.Izatt, R. M.; Bradshaw, J. S.; Nielsen, S. A.; Lamb, J. D.;
Christensen, J. J.; Sen, D. Chem. Rev. 1985, 85, 271–339.
doi:10.1021/cr00068a003
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
(http://www.beilstein-journals.org/bjoc)
21.CAChe WorkSytem Pro, Version 6.1.12.33; Fujitsu Limited..
22.Levin, S.; Nowick, J. S. J. Am. Chem. Soc. 2007, 129, 13043–13048.
doi:10.1021/ja073391r
The definitive version of this article is the electronic one
which can be found at:
23.Morera, E.; Ortar, G. Bioorg. Med. Chem. Lett. 2000, 10, 1815–1818.
doi:10.1016/S0960-894X(00)00344-9
doi:10.3762/bjoc.6.5
24.Lei, H.; Stoakes, M. S.; Herath, K. P. B.; Lee, J.; Schwabacher, A. W.
J. Org. Chem. 1994, 59, 4206–4210. doi:10.1021/jo00094a036
25.Schwarz, H.; Arakawa, K. J. Am. Chem. Soc. 1959, 81, 5691–5695.
doi:10.1021/ja01530a041
26.Paquet, A. Can. J. Chem. 1982, 60, 976–980. doi:10.1139/v82-146
27.Sigler, G. F.; Fuller, W. D.; Chaturvedi, N. C.; Goodman, M.;
Verlander, M. Biopolymers 1983, 22, 2157–2162.
doi:10.1002/bip.360221002
28.Milton, R. C. d. L.; Becker, E.; Milton, S. C. F.; Baxter, J. E. H.;
Elsworth, J. F. Int. J. Pept. Protein Res. 1987, 30, 431–432.
29.Lapatsanis, L.; Milias, G.; Froussios, K.; Kolovos, M. Synthesis 1983,
671–673. doi:10.1055/s-1983-30468
30.Zajdel, P.; Subra, G.; Bojarski, A. J.; Duszyńska, B.; Tatarczyńska, E.;
Nikiforuk, A.; Chojnacka-Wójcik, E.; Pawłowski, M.; Martinez, J.
Bioorg. Med. Chem. 2007, 15, 2907–2919.
doi:10.1016/j.bmc.2007.02.018
31.Carpino, L. A.; Chao, H. G.; Ghassemi, S.; Mansour, E. M. E.;
Riemer, C.; Warrass, R.; Sadat-Aalaee, D.; Truran, G. A.; Imazumi, H.;
El-Faham, A.; Ionescu, D.; Ismail, M.; Kowaleski, T. L.; Han, C. H.;
Wenschuh, H.; Beyermann, M.; Bienert, M.; Shroff, H.; Albericio, F.;
Triolo, S. A.; Sole, N. A.; Kates, S. A. J. Org. Chem. 1995, 60,
7718–7719. doi:10.1021/jo00129a005
32.Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397–4398.
doi:10.1021/ja00063a082
33.Dourtoglou, V.; Ziegler, J.-C.; Gross, B. Tetrahedron Lett. 1978, 19,
1269–1272. doi:10.1016/0040-4039(78)80103-8
34.Dourtoglou, V.; Gross, B.; Lambropoulou, V.; Zioudrou, C. Synthesis
1984, 572–574. doi:10.1055/s-1984-30895
35.Carpino, L. A.; El-Faham, A. Tetrahedron 1999, 55, 6813–6830.
doi:10.1016/S0040-4020(99)00344-0
36.Carpino, L. A.; El-Faham, A.; Albericio, F. Tetrahedron Lett. 1994, 35,
2279–2282. doi:10.1016/0040-4039(94)85198-0
37.Lauvergnat, D.; Hiberty, P. C. J. Am. Chem. Soc. 1997, 119,
9478–9482. doi:10.1021/ja9639426
Page 15 of 15
(page number not for citation purposes)