Chemical modulation of peptoids: Synthesis and conformational studies on partially constrained derivatives

Article abstract of DOI:10.1002/chem.201100216

The high conformational flexibility of peptoids can generate problems in biomolecular selectivity as a result of undesired off-target interactions. This drawback can be counterbalanced by restricting the original flexibility to a certain extent, thus leading to new peptidomimetics. By starting from the structure of an active peptoid as an apoptosis inhibitor, we designed two families of peptidomimetics that bear either 7-substituted perhydro-1,4- diazepine-2,5-dione 2 or 3-substituted 1,4-piperazine-2,5-dione 3 moieties. We report an efficient, solid-phase-based synthesis for both peptidomimetic families 2 and 3 from a common intermediate. An NMR spectroscopic study of 2 a,b and 3 a,b showed two species in solution in different solvents that interconvert slowly on the NMR timescale. The cis/trans isomerization around the exocyclic tertiary amide bond is responsible for this conformational behavior. The cis isomers are more favored in nonpolar environments, and this preference is higher for the six-membered-ring derivative 3 a,b. We propose that the hydrogen-bonding pattern could play an important role in the cis/trans equilibrium process. These hydrogen bonds were characterized in solution, in the solid state (i.e., by using X-ray studies), and by molecular modeling of simplified systems. A comparative study of a model peptoid 10 containing the isolated tertiary amide bond under study outlined the importance of the heterocyclic moiety for the prevalence of the cis configuration in 2 a and 3 a. The kinetics of the cis/trans interconversion in 2 a, 3 a, and 10 was also studied by variable-temperature NMR spectroscopic analysis. The full line-shape analysis of the NMR spectra of 10 revealed negligible entropic contribution to the energetic barrier in this conformational process. A theoretical analysis of 10 supported the results observed by NMR spectroscopic analysis. Overall, these results are relevant for the study of the peptidomimetic/biological-target interactions. Holding in place: Conformationally constrained peptidomimetics derived from peptoids have been identified as active compounds against biological targets. A structural study of some selected peptidomimetics by using NMR spectroscopic analysis in different solvents determined the relative stability (thermodynamic) and interconversion barriers (kinetic) of the cis/trans conformers present around the exocyclic tertiary amide bond (see scheme). Copyright

Full text of DOI:10.1002/chem.201100216

FULL PAPER
DOI: 10.1002/chem.201100216
Chemical Modulation of Peptoids: Synthesis and Conformational Studies on
Partially Constrained Derivatives
Alejandra Moure,[a] Glꢀria Sanclimens,[a] Jordi Bujons,[b] Isabel Masip,[a]
Angel Alvarez-Larena,[c] Enrique Pꢁrez-Payꢂ,[d] Ignacio Alfonso,[b] and
Angel Messeguer*[a]
Dedicated to Professor Josep Font on occasion of his 70th birthday
Abstract: The high conformational
flexibility of peptoids can generate
problems in biomolecular selectivity as
a result of undesired off-target interac-
tions. This drawback can be counter-
balanced by restricting the original
flexibility to a certain extent, thus lead-
ing to new peptidomimetics. By start-
ing from the structure of an active pep-
toid as an apoptosis inhibitor, we de-
signed two families of peptidomimetics
that bear either 7-substituted perhydro-
1,4-diazepine-2,5-dione 2 or 3-substitut-
ed 1,4-piperazine-2,5-dione 3 moieties.
We report an efficient, solid-phase-
based synthesis for both peptidomimet-
ic families 2 and 3 from a common in-
termediate. An NMR spectroscopic
study of 2a,b and 3a,b showed two spe-
cies in solution in different solvents
that interconvert slowly on the NMR
timescale. The cis/trans isomerization
around the exocyclic tertiary amide
bond is responsible for this conforma-
tional behavior. The cis isomers are
more favored in nonpolar environ-
ments, and this preference is higher for
the six-membered-ring derivative 3a,b.
We propose that the hydrogen-bonding
pattern could play an important role in
the cis/trans equilibrium process. These
hydrogen bonds were characterized in
solution, in the solid state (i.e., by
using X-ray studies), and by molecular
modeling of simplified systems. A com-
parative study of a model peptoid 10
containing the isolated tertiary amide
bond under study outlined the impor-
tance of the heterocyclic moiety for the
prevalence of the cis configuration in
2a and 3a. The kinetics of the cis/trans
interconversion in 2a, 3a, and 10 was
also studied by variable-temperature
NMR spectroscopic analysis. The full
line-shape analysis of the NMR spectra
of 10 revealed negligible entropic con-
tribution to the energetic barrier in this
conformational process. A theoretical
analysis of 10 supported the results ob-
served by NMR spectroscopic analysis.
Overall, these results are relevant for
the study of the peptidomimetic/biolog-
ical-target interactions.
Keywords: conformation analysis ·
heterocycles · NMR spectroscopy ·
peptidomimetics
·
solid-phase synthesis
Introduction
[a] Dr. A. Moure, Dr. G. Sanclimens, Dr. I. Masip,
Prof. Dr. A. Messeguer
Oligomers of N-substituted glycines (peptoids) were intro-
duced by Bartlett and co-workers as a distinctive family of
peptidomimetics.[1] The availability of peptoids as individual
compounds rationally designed to address a defined target
or as libraries of controlled mixtures (constructed under a
split-pool or positional-scanning format) made them attrac-
tive for high-throughput screening programs.[2] In this con-
text, the development of a convenient modular synthesis for
their generation (i.e., the submonomer approach) makes
possible the introduction of a wide chemical diversity at the
nitrogen atom of the glycine moiety.[3] In comparison with
peptides, it was anticipated that peptoids might exhibit
higher resistance to the action of proteases as well as im-
proved cellular-uptake features.[4] As a result, interesting
bioactive compounds as ligands for pharmaceutically targets,
such as G-protein-coupled receptors,[5] urokinase receptor,[6]
SH3 domains of signaling proteins,[7] HDM2 protein,[8] neu-
tralizers of bacterial endotoxins,[9] antimicrobial agents,[10]
antagonists of TRPV1 channel,[11] lung surfactants,[12] and
Department of Chemical and Biomolecular Nanotechnology
Instituto de Quꢀmica Avanzada de CataluÇa
Consejo Superior de Investigaciones Cientꢀficas
J. Girona, 18, 08034 Barcelona (Spain)
Fax : (+34)932045904
E-mail: angel.messeguer@cid.csic.es
[b] Dr. J. Bujons, Dr. I. Alfonso
Department of Biological Chemistry and Molecular Modeling
Instituto de Quꢀmica Avanzada de CataluÇa
Consejo Superior de Investigaciones Cientꢀficas
J. Girona, 18, 08034 Barcelona (Spain)
[c] Dr. A. Alvarez-Larena
Servicio de Difracciꢁn de Rayos X
Universidad Autꢁnoma de Barcelona
08193 Cerdanyola (Spain)
[d] Prof. Dr. E. Pꢂrez-Payꢃ
Laboratory of Peptide and Protein Chemistry
Centro de Investigaciꢁn Prꢀncipe Felipe
46012 Valencia (Spain)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201100216.
Chem. Eur. J. 2011, 17, 7927 – 7939
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7927

drug and gene delivery agents,[13] among others, have been
identified.
From a structural point of view, although peptoids can
adopt secondary structures depending on their size and the
residues linked to the nitrogen atoms on the backbone,[14] in
general they are highly flexible molecules. In peptoids, the
cis conformation of the amide bonds that connect the mono-
mer residues can be more highly populated than in pep-
tides.[15] Nevertheless, not all the cis/trans conformational
isomers of the different amide bonds present in a defined
peptoid will be biologically active in front of a specific
target. By taking advantage of the intensive research into
short peptoids as a result of their potential application in
drug discovery, Borchardt and Rabenstein recently reported
the use of NMR spectroscopic techniques to study cis/trans
isomerization by rotation around the amide bonds of small
peptoid models.[16] These authors observed a slow cis/trans
exchange rate for even the most labile of the amide bonds
between the two C-terminal residues. This feature can be of
importance with respect to the on/off rates for the ligand/re-
ceptor binding or when facing undesired interactions with
other targets. In parallel studies related to the structural re-
quirements of peptoids for biological function, Blackwell
and co-workers explored interactions between the amide
bonds and different aromatic side chains, thus concluding
that modulation of the n!p* effects can alter the ratio of
cis/trans amide bond conformers.[17]
Scheme 1. Peptoid 1a has been identified as an inhibitor of the formation
of the apoptosome and the formal chemical modulation of 1 to generate
constrained heterocyclic analogues 2 and 3.
In connection with our interest on the biological activity
of peptoids, it is now well established that multiprotein com-
plexes are important points of regulation in cellular-signal-
ing pathways, thus raising attention as targets for the devel-
opment of chemical modulators. Peptoids features are at-
tractive for application to the modulation of proteins[8,10,18]
as well as protein/protein interactions. In this context, pro-
grammed cell-death mechanisms (apoptosis) constitute an
example of the pivotal interest in controlling different path-
ologies.[19] We previously reported that a medium-through-
put screening of a positional-scanning combinatorial library
of N-alkylglycine trimers[20] allowed the identification of
peptoid 1a (Scheme 1) as an inhibitor of the apoptosome-
dependent activation of procaspase-9.[21] In this context, we
also identified peptoids as pharmacological inhibitors of
noncanonical polyubiquitylation; these compounds compete
with ubiquitin E2 variant (UEV) for its interaction with
UBC13 and inhibit its enzymatic activity.[22]
However, in contrast with the wide possibilities of identi-
fying hits against pharmaceutical targets, the high conforma-
tional flexibility of peptoids can generate selectivity prob-
lems because of undesired off-target interactions. This draw-
back can be counterbalanced by a second generation of pep-
tidomimetics in which the original conformational flexibility
could be restricted to a certain extent. The structural sim-
plicity of peptoids makes them amenable to chemical modu-
lation, thus facilitating the optimization of hit molecules for
druglike properties.[23] We deemed that the formal cycliza-
tion through selected points of the peptoid molecule would
lead to constrained analogues that could improve their po-
tency and selectivity. Moreover, if our approach successfully
identified more potent modulators of apoptotic protease ac-
tivating factor 1 (Apaf-1), these molecules, being conforma-
tionally more restricted, could be useful for defining more
precisely the interaction that takes place with the protein
target.
Covalent constraints have been explored in peptoid re-
search mainly by employing macrocyclization between side
chains,[24] head-to-tail strategies,[25] or Ugi four- and three-
component reactions.[26] In our case, from the different pos-
sibilities contemplated to obtain constrained peptidomimet-
ics derived from the identified peptoid hits, those ap-
proaches that generated the novel 7-substituted perhydro-
1,4-diazepine-2,5-dione 2 and 3-substituted 1,4-piperazine-
2,5-dione 3 derivatives were selected (Scheme 1). In these
systems, two of the three tertiary amide bonds present in
the molecules are forced to adopt the cis configuration by
the heterocycle formation. Both 2a and 3a were shown to
be potent inhibitors of the apoptosome formation and capa-
ble of decreasing cell death in different cellular models of
apoptosis.[27] Furthermore, 2b and 3b are potent inhibitors
of noncanonical polyubiquitylation and produce significant
biological effects with potential therapeutic applications in
cancer and inflammation areas.[22,28]
We report herein on an efficient solid-phase-based syn-
thetic approach to the peptidomimetic families 2 and 3. In
addition, an NMR-based structural study of 2a and 3a in
different solvents was carried out to determine the thermo-
7928
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 7927 – 7939

Conformational Flexibility of Peptoids
FULL PAPER
dynamic and kinetic parameters
that define the relative stability
of the cis/trans conformers po-
tentially present around the
exocyclic amide bond. Such an
amide bond is the only one that
still shows conformational flexi-
bility relative to the original
peptoid hit. To the best of our
knowledge, there are no prece-
dents on structural studies on
these families of peptidomimet-
ics. In addition to our general
interest in studying the influ-
ence of the frozen cis configura-
tion around the two amide
bonds on the exocyclic bond,
we anticipated that this struc-
tural study would be highly
useful for a better understand-
ing of the interaction between
the peptidomimetics with the
biological target, thus giving
more precise information for
the design of improved com-
pounds.
Results and Discussion
Synthesis of compounds 2 and
3: We developed a route for the
synthesis of 7-substituted perhy-
dro-1,4-diazepine-2,5-dione
and 3-substituted 1,4-pipera-
zine-2,5-dione derivatives
2
3
(Scheme 1) from the common
intermediate 6 (Scheme 2). The
compounds were synthesized by
using solid-phase chemistry. A
solid support shows several in-
disputable advantages over so-
lution chemistry: purification is
facilitated by simple filtration,
thus avoiding time-consuming
separation techniques; subse-
quent building blocks and re-
agents can be added in excess
to drive reactions to comple-
tion; and the “pseudo-dilution
effect”, which is the result of
using the polymeric solid sup-
port, makes intramolecular cyc-
lization a suitable reaction that
could be carried out efficiently
on solid phase rather than in
solution. Additionally, the pro-
Scheme 2. Solid-phase synthesis of the conformationally restricted peptidomimetics 2 and 3. a) Piperidine
20%, DMF, MW 2 min at 608C; b) BrCH2COOH, DIC, DMF/CH2Cl2, MW 1 min at 358C; c) RNH2, Et3N,
DMF, MW 2 min at 908C; d) allyl maleate, HOBt, DIC, DMF/CH2Cl2, MW 1 min at 458C; e) RNH2, Et3N,
DMF, 16 h, RT; f) 0.5n KOH, dioxane 1:3, MW 4 min at 1108C; g) PyBOP, HOBt, DIPEA, DMF, 20 min at
608C; h) TFA/CH2Cl2/H2O, 30 min, RT; i) dioxane, MW 4 min at 1208C; j) 4n NaOH, allyl alcohol, dioxane,
MW 4 min at 908C; k) peptoidyl-resin 4, HOBt, DIC, DMF/CH2Cl2, 3 h at RT. DIC=N,N’-diisopropylcarbo-
diimide, DIPEA=diisopropylethylamine, Fmoc=9-fluorenylmethoxycarbonyl, HOBt=1-hydroxybenzotria-
zole, MW=microwave, PyBOP=benzotriazole-1-yloxy-tris-pyrrolidinophosphonium hexafluorophosphate,
TFA=trifluoroacetic acid.
Chem. Eur. J. 2011, 17, 7927 – 7939
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7929

A. Messeguer et al.
cedures were optimized by microwave activation to decrease
the reaction times and improve conversion yields.
the hydrolysis of the ester was carried out in the solid phase
followed by cleavage and cyclization as already described
(see the Supporting Information for details of 3 f).
Preparation of key intermediate 6: Upon deprotection of
the Fmoc group of the polystyrene AM RAM resin, the free
amine group was acylated with bromoacetic acid in the pres-
ence of DIC (Scheme 2). A primary amine (first diversity
source) was coupled to the bromo derivative to yield 4, fol-
lowed by treatment with allyl maleate in the presence of
DIC and HOBt to obtain ester intermediate 5. However,
when the acylation reactions were assayed on substrates
bearing an additional tertiary amino group, the conversion
yields were much lower and complex mixtures profiles were
obtained. In our experience, the use of this class of primary
amine in a solid-phase synthesis of peptoids has resulted in
the occurrence of problematic side reactions.[29] In these
cases, the best results were obtained by using O-(benzotri-
The crude reaction mixture containing free carboxylic
acid 9 was used for coupling with the corresponding N-alkyl-
glycinamide bearing the third diversity source. Initially, the
amide formation was assayed in solution. To this aim, the
previous synthesis of the N-alkylglycinamide by reaction of
2-chloroacetamide with the corresponding primary amine
was attempted. Unfortunately, the results obtained were not
satisfactory. The use of different solvents, reaction condi-
tions, or reagent, that is, 2-bromoacetamide, resulted in in-
complete conversion of the haloacetamide and, more impor-
tantly for the further purification process, the concomitant
generation of dialkylation products. Therefore, we turned to
a solid-phase preparation of the desired N-alkylglycinamides
following the procedure developed for intermediate 4. The
reaction of crude compound 9 with selected N-alkylglycina-
mides was assayed. First, the N-alkylglycinamides were re-
leased and the amide formation was performed in solution;
however, the results obtained were not satisfactory in terms
of conversion yield. Therefore, the amide formation was car-
ried out by linking the N-alkylglycinamide to the resin with
DIC and HOBt as coupling agents. By using this procedure,
title compounds 3a–e were isolated in 30–40% overall yield
after release from the resin. The only exception was the
case of the 1,4-piperazine-2,5-dione 3 f in which TBTU and
DIPEA should be employed as coupling agents for the final
amide formation (see the Supporting Information for syn-
thetic details and characterization for compounds 3a–f).
Constriction by the formation of a heterocycle in 2 and 3
also produces a stereogenic center. The enantiopure com-
pounds 2a and 3a were synthesized and characterized; how-
ever, we believe that the description of the enantiopure de-
rivatives is beyond the scope of this study and will be re-
ported elsewhere because the chirality of these compounds
does not affect the present conformational study.
azol-1-yl)-N,N,N,N-tetramethyluronium
tetrafluoroborate
(TBTU) and DIPEA to promote the reaction. Next, the
second primary amine group was introduced by an aza-Mi-
chael reaction to render the intermediate ester 6, which
bears two of the three desired diversity sources. This reac-
tion was the only one of the whole sequence that should be
carried out at room temperature because nonreproducible
results were obtained under microwave-assisted conditions.
The NMR spectroscopic analysis of aliquot samples after
their release from the resin showed that the addition of the
amine group was regioselective and took place at the a-
carbon atom to the amide group.
Preparation of 7-substituted perhydro-1,4-diazepine-2,5-
dione derivatives 2: The secondary amine group of the inter-
mediate ester 6 was acylated with bromoacetic acid and
treated with the third primary amine to obtain 7. The allyl
ester group of 7 was removed. This deprotection could be
conducted by treatment with [PdACHTNUTRGNEUNG(PPh3)4] in the presence of
PhSiH3 in CH2Cl2. Alternatively, we observed that saponifi-
cation could be performed by using 0.5m KOH in dioxane
(1:3) at room temperature or under microwave-assisted acti-
vation conditions. Finally, the solid-phase cyclization was
promoted with PyBOP and HOBt in the presence of
DIPEA to render final compound 2 after release from the
resin using a TFA/CH2Cl2/H2O mixture (see the Supporting
Information for the synthetic details and characterization of
2a–f). As anticipated, the use of a polymeric solid support
provided a “pseudo-dilution effect” that favored the desired
intramolecular cyclization.
Structure of the conformationally constrained peptidomi-
metics 2 and 3 in solution: The knowledge of the conforma-
tion of 2 and 3 in solution is extremely important in under-
standing their biological activity and for the future redesign
and optimization of a next generation of inhibitors. There-
fore, we performed a conformational study by using NMR
spectroscopic analysis and molecular modeling to under-
stand both the thermodynamics and the kinetics of struc-
tures of 2 and 3 in solution. We selected 2a and 3a
(Scheme 3) as initial compounds to study for several rea-
sons. First, they displayed the less overlapping in the
1H NMR spectra among the two families of peptidomimet-
ics. Second, both molecules have shown interesting biologi-
cal properties as inhibitors of the formation of the apoptoso-
me.[21,27a]
Preparation of 3-substituted 1,4-piperazine-2,5-dione 3: In
this case, the release of intermediate 6 from the resin afford-
ed a crude reaction mixture, which under microwave condi-
tions in dioxane, rendered the cyclized product 8 with quan-
titative conversion yields. Removal of the allyl moiety af-
forded the expected free carboxylic acid 9. When an amine
with an additional tertiary amino group was used in the first
or second diversity position, the extraction of acid 9 into the
organic phase was troublesome (20–30% yield). In this case,
1
The main feature in the H and 13C NMR spectra of 2a
and 3a was the presence of two species in solution (in
CDCl3, CD3CN, or deuterated dimethyl sulfoxide
([D6]DMSO)), which interconvert slowly on the NMR time-
7930
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 7927 – 7939

Conformational Flexibility of Peptoids
FULL PAPER
Thermodynamic
parameters:
The unambiguous assignation
of the NMR signals that corre-
spond to each rotamer of 2a
and 3a was accomplished by
conclusive evidence (i.e., NOE
interactions and chemical-shift
differences). For example, in
the case of 2a, strong ROESY
cross peaks were observed be-
tween H6eq and H2’ for the
major isomer in CDCl3, thus
suggesting spatial proximity be-
tween the seven-membered ring
and the methylene group of the
Scheme 3. Dynamic process responsible for the conformational behavior of 2a and 3a (an arbitrary numbering
of the atoms has been adopted).
terminal
glycine
unit
(Scheme 4). This outcome im-
plies that the major species in
CDCl3 must be the cis rotamer.
On the other hand, the ROESY
spectrum showed different cor-
relations for the major isomer
in [D6]DMSO. The methylene
group in the 2,4-dichlorophe-
nethyl moiety directly attached
to the nitrogen atom of the exo-
cyclic amide group showed
cross peaks with both the
seven-membered-ring
moiety
and the methylene group of the
terminal glycine unit, thus im-
plying a preferred trans config-
uration
in
this
solvent
(Scheme 4).
These assignations are also
consistent with the observed
chemical-shift differences be-
tween both rotamers. Thus, the
trans species (minor in CDCl3,
but major in [D6]DMSO) would
set the aromatic ring of the 2,4-
Figure 1. 1H NMR spectra of 2a in [D6]DMSO, CD3CN, and CDCl3 (500 MHz, 303 K; see Scheme 3 for num-
bering).
scale (as observed at 300, 400, or 500 MHz). This fact was
evident by the splitting of most of the 1H and 13C NMR
peaks (see Figure 1 and the Supporting Information) for the
corresponding proton signals interconnected by exchange
processes (EXSY). For example, 2a showed two different
signals for the H7 proton of the chiral center of the mole-
cule, which yielded negative cross peaks between them in
the 2D ROESY spectra both in CDCl3 and [D6]DMSO (see
the Supporting Information). Additionally, the dynamic
nature of 2a,b and 3a,b has been further demonstrated by
variable-temperature (VT) NMR experiments (see below).
By considering previous studies on peptoids and the identity
dichlorophenethyl group attached to the exocyclic amide
unit close to the H7 proton, thus explaining the observed
lower chemical shift of this signal in this conformer
(Figure 1). Although with a slightly larger overlapping of
1
the H NMR signals, a closely similar behavior was found
for the six-membered-ring derivative 3a. Likewise, the cis
and trans isomers of 2b and 3b were assigned by chemical
correlation and careful comparison of the corresponding
NMR spectra. Integration of several proton bands rendered
the corresponding population of the species, which were
used to calculate the cis/trans equilibrium constants in differ-
ent solvents (Table 1).
Comparison of the data in Table 1 shows that cis isomers
are more favored in nonpolar environments, and that this
preference is higher for the six-membered ring derivatives
3a,b. When the polarity of the medium was increased, the
1
of the H NMR signals that were split more significantly,[16]
we proposed the cis/trans isomerization around the exocyclic
tertiary amide bond to be the dynamic process responsible
for this conformational behavior (Scheme 3).
Chem. Eur. J. 2011, 17, 7927 – 7939
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7931

A. Messeguer et al.
terns could be established (Figure 2a). For 2a,b and 3a,b,
the trans isomers can establish intramolecular hydrogen
bonds between the terminal amide NH group and one of
the heterocyclic carbonyl groups, thus leading to the forma-
tion of either seven- or ten-membered rings, respectively.
On the contrary, the cis isomers of the same molecules can
only form the corresponding ten-membered rings because
the seven-membered ring is geometrically impossible.[30] Fi-
nally, the simpler compound 10, in which the ten-membered
ring possibility was ruled out, can establish a seven-mem-
bered-ring hydrogen-bonding interaction in the trans isomer
exclusively.
By taking into account these facts and the data shown in
Table 1, we concluded that the seven-membered-ring hydro-
gen-bonding interaction stabilizes the trans configuration in
10, although the trans isomer is also slightly preferred in the
absence of this intramolecular hydrogen-bonding interaction
(as it is expected to be the case in pure DMSO). The scenar-
io is clearly more complicated with 2 and 3 because different
hydrogen-bonding patterns can coexist. Theoretical calcula-
tions on simplified analogues of
Scheme 4. Observed NOE interactions for the assignation of the cis/trans
isomers for 2a and 10.
Table 1. Populations of the cis/trans isomers (P), estimated cis/trans equilibrium constants (Keq), and Gibbs
free energies (DG8) for 2a,b and 3a,b at 303 K in different solvents.
2 and 3 suggested that the hy-
drogen bonding of the ten-
membered ring is more favored
than the seven-membered ring
for the trans isomer, but there
is practically no energetic dif-
ference between the cis and
trans amide isomers in the ten-
Compound
Solvent
Pcis [%]
Ptrans [%]
Keq (cis!trans)
DGo (cis!trans)
[kcalmolꢀ1
]
2a
CDCl3
CD3CN
[D6]DMSO
CDCl3
CD3CN
[D6]DMSO
CDCl3
CD3CN
[D6]DMSO
CDCl3
CD3CN
77.0
50.0
32.6
77.0
46.1
30.4
83.2
58.9
50.0
82.3
54.1
47.9
17.2
38.4
46.3
23.0
50.0
67.4
23.0
53.9
69.6
16.8
41.1
50.0
17.7
45.9
52.1
82.8
61.6
53.7
0.30
1.00
2.07
0.30
1.17
2.29
0.20
0.70
1.00
0.21
0.85
1.09
4.81
1.60
1.16
0.72
0.00
ꢀ0.44
2b
3a
3b
10
0.72
ꢀ0.09
ꢀ0.50
0.97
0.21
0.00
0.92
0.10
ꢀ0.05
ꢀ0.92
ꢀ0.28
ꢀ0.09
membered-ring
hydrogen-
bonded geometries (Figure 2b).
Compound 2a yielded crys-
tals suitable for X-ray diffrac-
tion analysis by very slow evap-
oration of a solution of the pep-
tidomimetic in acetonitrile.
Contrary to observations in so-
lution at room temperature,
only the trans isomer is present
in the solid state. The intramo-
[D6]DMSO
CDCl3
CD3CN
[D6]DMSO
amount of trans isomer also increased. These data suggested
that some intramolecular hydrogen-bonding interactions
could play an important role in the cis/trans equilibrium pro-
cess. To obtain deeper insight into this topic, model peptoid
10 was prepared and analyzed because this molecule repre-
sents an isolated version of the tertiary amide bond under
study (Scheme 4). The assignations of the corresponding sig-
nals of the cis/trans isomers of 10 were done by 1D NOESY
experiments (see the double-headed arrows in Scheme 4).
Interestingly, for this compound, the trans configuration was
favored in [D6]DMSO, CD3CN, and CDCl3 (54, 62, and
83%, respectively). This observation suggested that the pop-
ulation increase of the cis isomer in the full-length mole-
cules (i.e., 2a,b and 3a,b) was induced by the presence of
the heterocyclic moiety.
lecular hydrogen bond HN-H···O=C(diazepine) partially
fixes the conformation, thus closing a ten-membered ring
(Figure 3, Table 2). Moreover, in the solid state, a centro-
symmetric heterochiral dimer is formed as a result of the
presence of two intermolecular HN-H···O=C-NH2 hydrogen
bonds (Figure 3, Table 2). These observations support the
preference for the ten-membered ring in the trans isomer.
We believe that the intermolecular hydrogen bonding in the
crystal structure could be comparable to the effect of
DMSO molecules in solution, thus providing a reasonable
explanation for the preference of this rotamer both in the
solid state and in polar environments. Taken all together, we
found a good agreement between the behavior in solution,
crystal structure, and theoretical results.
To shed additional light onto the behavior in solution,
1H NMR titrations of solutions of 10, 3a, and 2a in CDCl3
with increasing amounts of [D6]DMSO were performed
By considering the molecular structure of the studied
compounds, different intramolecular hydrogen-bonding pat-
7932
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 7927 – 7939

Conformational Flexibility of Peptoids
FULL PAPER
for every isomer, whereas the population of the cis and
trans isomers at every titration point were evaluated by the
integration of the corresponding 1H NMR signals. In this
way, it was possible to monitor how the intramolecular hy-
drogen-bonding pattern was gradually broken by the action
of DMSO and to evaluate the influence of this solvent on
the cis/trans conformational equilibrium. An interesting be-
havior was observed in every case. For the simpler com-
pound 10, the cis isomer shows a large dependence of the
chemical shifts for the NH signal on added DMSO, as ex-
pected for a non-hydrogen-bonded amide proton. On the
contrary, the NH protons in the trans isomer are less accessi-
ble to the solvent (i.e., DMSO) as a result of the occurrence
of an intramolecular hydrogen bond, which is reflected in a
lower dependence of the chemical shifts for the NH signal
on the addition of DMSO. Concomitantly, when the propor-
tion of DMSO increases, the percentage of the cis isomer
also increases, although the trans isomer is always more
stable in this model system. These experiments are in ac-
cordance with our initial proposal on the effect of the hy-
drogen-bonding interactions on the cis/trans equilibrium of
10. On the other hand, the trans isomer for 3a showed a
larger dependence of the chemical shifts for the NH signal
on the added DMSO, thus suggesting that the possible hy-
drogen-bonding pattern is less favored by this rotamer. In
this case, the NH protons in the cis isomer are less accessi-
ble to the solvent (i.e., DMSO) as shown by the lower de-
pendence of the chemical shifts for the NH signal. When the
proportion of DMSO increases, the amount of trans isomer
also increases, thus suggesting that the intermolecular hy-
drogen bonding stabilizes the trans isomer, whereas the in-
tramolecular hydrogen bonding stabilizes the cis isomer.
Only one NH proton could be monitored for every rota-
mer of 2a, but the trends were similar to those observed for
3a. However, other structurally interesting parameters were
observable as a result of decreased signal overlapping in the
aliphatic proton region. The coupling constants of H7/H6 in
3
the cis rotamer (3JH7,H6 =3.7 and JH7,H6 =6.3 Hz) and the
eq
ax
large anisochrony of the methylene protons at C6, C3, and
C2’ (Dd=0.234, 1.159, and 0.199 ppm, respectively) in pure
CDCl3 support the presence of a conformation rigidified by
an intramolecular hydrogen bond. As previously mentioned,
this behavior has to be through a ten-membered ring. When
adding DMSO, the amount of trans isomer increased and,
for this rotamer, the anisochrony of the diastereotopic pro-
tons H3 and H6 is lower, in the same way as those for the
terminal glycine H2’. In addition, the corresponding cou-
pling constants between H7 and H6ax/H6eq are similar, which
are the consequence of breaking the intramolecular hydro-
gen-bonded ten-membered ring to give a more flexible
structure in solution. All these data support the fundamental
roles that the polarity of the medium and the intramolecular
hydrogen-bonding pattern play in the cis/trans conforma-
tional equilibrium of these peptidomimetics. The intramolec-
ularly hydrogen-bonded ten-membered ring (only possible
with 2a and 3a) favors the cis isomer in nonpolar environ-
ments. As this hydrogen-bonding interaction is gradually
Figure 2. a) Schematic representation of the possible hydrogen-bonding
patterns for 2a, 3a, and 10. b) Optimized geometries and gas-phase DG
values (kcalmolꢀ1) calculated at the B3LYP/6-31G** level for the hydro-
gen-bonded 10- and 7-membered rings of the cis and trans rotamers of
simplified analogues of 2a and 3a in which the substituents R1–R3 have
been replaced by methyl groups.
(Figure 4). Variations in the chemical shifts for the amide
NH signals versus the amount of added DMSO were plotted
Chem. Eur. J. 2011, 17, 7927 – 7939
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7933

A. Messeguer et al.
Table 2. Geometries of the main hydrogen bonds in the crystal structure of 2a (see Figure 3).[a]
temperatures to be reached and
also minimized the intramolec-
ular hydrogen-bonding effects.
Hydrogen-bond lengths [ꢅ]
Hydrogen-bond angles [8]
Hydrogen-bond type
ꢀ
N22···O2
2.945(4)
N22···N21
2.835(4)
H22B···O2
2.18
H22B···N21
2.46
N22 H22B···O2
149
intra
intra
A severe broadening of the
1H NMR spectra was obtained
upon heating (see the Support-
ing Information), thus leading
to the coalescence of several
signals and allowing the ener-
getic barrier for the amide rota-
tion to be estimated (Table 3).
As 2a exists in unequal popula-
ꢀ
N22 H22B···N21
107
O2···H22B···N21
102
N22···O22
3.040(4)
H22A···O22
2.21
N22-H22A···O22
161
inter
ꢀ
ꢀ
[a] Distances N22 H22A and N22 H22B are fixed at 0.86 ꢅ.
tions of cis and trans rotamers in DMSO, the corresponding
energetic barriers were calculated as described by Shanan-
Atidi and Bar-Eli.[31] The simplified formula for isoenergetic
species was applied for 3a.[32] For both compounds, several
proton signals could be used as probes and showed a good
internal consistency within the estimated error for these ap-
proximated formulae (ꢁ0.4 kcalmolꢀ1). The energetic barri-
ers are in the range of 19–20 kcalmolꢀ1, which is in good
agreement with those previously reported for similar sys-
tems.[16] A slightly lower interconversion barrier was ob-
tained for the six-membered-ring derivative, although the
difference is close to the experimental error by using the
corresponding approximated formulae at the coalescence
temperature. As the coalescence of different signals of a
given compound occurred at different temperatures, we
roughly estimated a small effect of temperature on the ener-
getic barrier. Therefore, there must be a small entropic con-
tribution to DG°, as expected for a dynamic process based
ꢀ
on the rotation of a single C C bond.
Unfortunately, attempts to perform the full line-shape
analysis of the VT-NMR experiments for 2a,b and 3a,b
were unsatisfactory as a result of the complex spin systems
implicated and the severe signal overlapping. Therefore, the
entropic and enthalpic contributions to the activation pa-
rameters could not be obtained accurately. As the rotational
barrier did not seem to be significantly affected by the sub-
stitution on the tertiary amide bond (R in Scheme 3), we de-
cided to perform VT-NMR spectroscopic experiments with
the simpler peptidomimetic model 10 bearing a methyl
group instead of the heterocyclic scaffold. In this case, the
dynamic process was isolated and could be studied more
precisely. VT-NMR spectroscopic experiments in DMSO
produced the coalescence of four signals, which are the asso-
ciated energy barriers summarized in Table 3. The results
obtained are in reasonable agreement with those previously
found (DG° ꢂ19 kcalmolꢀ1). In this case, we could carry out
the full line-shape analysis of the methyl signal within the
temperature range 333–393 K (Figure 5). The accurate
Eyring and Arrhenius plots yielded the corresponding acti-
vation parameters for the cis/trans isomerization in both di-
rections of the equilibrium. The results demonstrate the neg-
ligible entropic contribution to the energetic barrier for this
conformational process (Table 4). These data support our in-
itial estimation for 2a and 3a, which are also in good agree-
ment with results reported on related systems.[16]
Figure 3. Molecular structure of 2a in the solid state (up, ellipsoid plot
with 50% probability) and dimer of 2a in the crystal structure (down,
dashed lines represent hydrogen bonds).
broken by increasing the polarity of the medium, the corre-
sponding proportion of the trans rotamers increases. Fur-
thermore, the proposed ten-membered rings would be less
strained in 3a,b than in 2a,b, thus explaining the slightly
higher population of the cis isomers in the 1,4-piperazine-
2,5-dione family.
Kinetic parameters: Once the thermodynamic parameters of
the cis/trans equilibria for the representative peptidomimet-
ics 2a,b and 3a,b were characterized, we also attempted to
study the kinetics of the interconversion. With this aim, VT-
NMR spectroscopic experiments were performed in
[D6]DMSO for both derivatives. This solvent allowed higher
7934
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 7927 – 7939

Conformational Flexibility of Peptoids
FULL PAPER
Hence, the lowest free-
energy conformer of 10 (trans-
1) shows the expected hydrogen
bond between the terminal
amide NH2 group and the car-
bonyl oxygen atom of the acet-
amide group, in accord with the
NMR results. The hydrogen
bond is maintained in the
lowest energy transition state
detected for the interconversion
between cis/trans amide con-
formers (TS-1). The calculated
free energy of TS-1 is 19.3 kcal
molꢀ1, in agreement with the
experimental results obtained
by NMR spectroscopic analysis.
Transition state TS-1 is connect-
ed to conformers trans-3 and
cis-3 by rotation of dihedral
angle w with concomitant rota-
tion of dihedral angle c1 and
pyramidalization of the planar
amide nitrogen atom. The mini-
mum energy conformer trans-3
is 1.17 kcalmolꢀ1 above the
lowest-energy trans-1 confor-
mer, through which it is related
by a low barrier rotation of the
dihedral angles c1 and c3.
Whereas the minimum cis-3 is
1.35 kcalmolꢀ1
above
the
lowest-energy cis conformer
(cis-1), with which it is connect-
ed again through a low barrier
rotation around the dihedral
angle c3.
Figure 4. Plots of the chemical shifts for the amide NH signal (left) and proportions of the cis/trans isomers
(right) of 10, 3a, and 2a (CDCl3, 303 K, 500 MHz) versus the percentage of [D6]DMSO (cis/trans isomers are
given as hollow and solid symbols, respectively).
Conclusion
The partial restriction of the
The computational conformational analysis of 10 support-
ed the above observations. The conformational space of 10
was systematically explored by molecular mechanics meth-
ods to find relevant energy minima. In this way, 68 different
conformers with energies within 10 kcalmolꢀ1 from the
lowest-energy minimum were located and reoptimized by
DFT methods at the B3LYP/6-31G** level (see the Support-
ing Information). The interconversion among different pairs
of low-energy cis/trans amide conformers was studied and
the corresponding transition states were determined (see the
Supporting Information). Figure 6 shows the structures of
the most-relevant energy minima and of the lowest-energy
transition state detected in the gas phase, and Table 5 re-
ports the geometric and energetic parameters for the same
structures.[34]
conformational flexibility of peptoids by means of designing
two families of peptidomimetics based on the perhydro-1,4-
diazepine-2,5-dione 2 and the 1,4-piperazine-2,5-dione 3
scaffolds has been studied. To this end, we have developed a
versatile (for diversity-oriented strategies), solid-phase syn-
thesis of both families from a common intermediate 6 in
good overall yield. The synthetic pathway involves the gen-
eration of 6 by a regioselective aza-Michael addition and
the extensive use of microwave activation to decrease reac-
tion times while maintaining purity and conversion yield of
the different intermediates and final products.
From the conformational behavior of these peptidomimet-
ic families, some conclusions can be drawn that are of im-
portance for the study of the peptidomimetic/biological-
target interactions and for further peptidomimetic structural
Chem. Eur. J. 2011, 17, 7927 – 7939
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7935

A. Messeguer et al.
Table 3. Frequency difference (Du), coalescence temperature (Tc), and energetic barrier at Tc (DG‡) for
the conformational process in peptidomimetics 2a,b, 3a,b, and 10.
the trans configurations prevail
in solution, although important
proportions of the cis rotamers
are also present.[16] In addition,
the trans (cis) configuration is
favored even more when the
other amide bonds have trans
(cis) disposition. In our peptido-
mimetic systems, two amide
bonds are frozen in a cis config-
uration as a result of the con-
straints of the corresponding
cyclic structure. Accordingly, an
increased proportion of the cis
rotamer should be expected for
the free exocyclic amide bond,
which is exactly what we ob-
served. More interestingly, we
have demonstrated that the pro-
Compound
Signal
Du [Hz]
Tc [K]
DG° (cis!trans)[a]
DG° (trans!cis)[a]
[kcalmolꢀ1
]
[kcalmolꢀ1
]
2a
H7
H3
H3
31.5
7.5
8.1
388
363
363
19.6
19.3
19.2
19.4
19.4
19.1
19.3
19.1
19.2
19.2
19.2
19.1
18.7
18.8
18.7
19.2
18.9
20.1
19.8
19.8
19.9
20.0
19.6
19.8
–
–
–
–
–
18.9
19.0
19.0
19.5
19.1
average[b]
H7
2b
3a
36.6
5.2
388
355
H3
average[b]
H3
10.5
6.2
18.2
363
341
373
H6
H6
average[b]
H6
3b
10
18.0
13.5
16.8
32.4
8.1
371
360
365
373
363
H2’
NCH2
CH2Ar
Me
average[b]
[a] Estimated error: ꢁ0.4 kcalmolꢀ1. [b] Average value using different proton signals.
Figure 6. Relevant low-energy cis and trans amide conformers and the
lowest-energy transition state determined at the B3LYP/6-31G** level
for 10. The numbering of the cis/trans conformers and transition states
corresponds to increasing free energy (see the Supporting Information).
portion of cis/trans isomers is highly affected by the polarity
of the solvent as a result of the possibility of establishing an
intramolecular hydrogen-bonding pattern. Therefore, the
population of the cis/trans isomers of the compounds studied
herein under thermodynamic control could serve as a molec-
ular probe for the polarity of the microenvironment that sur-
rounds these molecules. This parameter could be used in
ideal systems to map the polarity of binding pockets on tar-
geted biomolecules. In addition, the kinetic parameters of
the binding must be also carefully considered. Usually, the
kinetics of the receptor/ligand interaction is fast on the
NMR timescale.[35] On the other hand, our studies showed
that the cis/trans isomerization of the amide bond is slow on
the NMR timescale. Therefore, the binding phenomena
must be faster than the amide-bond rotation and accordingly
both rotamers would look like different molecules to the
biomolecular receptor. Moreover, we should keep in mind
that binding rates are concentration dependent but amide
rotation is a unimolecular process. Therefore, knowledge of
Figure 5. Experimental (left) and simulated (right) VT-NMR spectra of
the methyl signal of 10 (300 MHz). Temperatures and computed ex-
change rates k are also given for every trace. The simulated spectra were
obtained with the gNMR program.[33]
Table 4. Activation parameters for the methyl signal of 10.
Activation parameters
cis!trans
19.28
0.1
19.28
19.96
trans!cis
19.39
0.2
19.32
20.07
[a]
DH° [kcalmolꢀ1
]
DS° [cal(Kmol)ꢀ1
ACHTUNGTRENNUNG ]
[b]
DG°
[kcalmolꢀ1
]
[a]
298K
[a]
Ea [kcalmolꢀ1
]
[a] Estimated error: ꢁ0.15 kcalmolꢀ1
.
[b] Estimated error: ꢁ0.5 cal
ACHTUNGTRENNUNG .
(Kmol)ꢀ1
optimization toward more active compounds. First, the large
population of cis amide isomers in these systems is rather
remarkable. It has been recently reported that in peptoids
7936
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 7927 – 7939

Conformational Flexibility of Peptoids
FULL PAPER
Table 5. Geometries ({w, f, y, c1, c2, c3} and dNH···O), DE, and DG values determined in the gas phase for the
conformations shown in Figure 6.
detector diffractometer at room tem-
perature with graphite-monochomat-
ized MoKa radiation. Lorentz polari-
zation and absorption corrections were
applied using Bruker SAINT[36] and
SADABS[37] software. The structure
was solved by direct methods and re-
fined by full-matrix least squares on
F2 for all unique measured data using
SHELXTL.[38] Non-hydrogen atoms
were refined anisotropically. Hydrogen
atoms were included with riding-
model constraints and isotropic dis-
placement parameters equal to 1.2
times the Ueq values of the corre-
sponding carbon or nitrogen atoms.
The following crystal structure has
been deposited at the Cambridge
Crystallographic Data Centre and allo-
cated the deposition number CCDC
808273.
Conformer w [8]
f [8]
ꢀ83.5 79.3
ꢀ86.6 80.4 ꢀ103.8 ꢀ175.6 ꢀ85.4 2.09
ꢀ98.9 5.6 ꢀ81.3 177.4 85.2 3.62
ꢀ83.8 ꢀ173.5 ꢀ82.9 3.61
ꢀ91.0 55.3 ꢀ163.5 ꢀ175.2 ꢀ84.4 1.99
y [8] c1 [8]
c2 [8]
c3 [8] dNH···O [ꢅ] DE [kcalmolꢀ1
]
DG [kcalmolꢀ1
]
trans-1
trans-3
cis-1
ꢀ176.8
96.0
173.6
81.7 2.03
0.79
2.00
4.36
5.70
0.00
1.17
1.80
3.15
19.3
ꢀ178.5
7.7
cis-3
TS-1
6.7 ꢀ102.1 11.5
ꢀ109.2
20.7
Computational methods: Molecular
the strength (and kinetics) of the interaction should be cru-
cial for the selection of a suitable experimental technique
and the optimal concentration range in carrying out biomo-
lecular binding studies with these peptidomimetic systems.
mechanics (MM) calculations were performed with the program MOE
2007.09 (Chemical Computing Group, Montreal, Canada) and quantum-
mechanics calculations with the program Jaguar 7.5.[39] The MMFF94x
force field, a modified version of the MMFF94s force field[40] implement-
ed in the MOE program, was used for all MM energy calculations by
using distance-dependent electrostatics (e=1) without cut-off points to
model the nonbonded interactions. The systematic conformational search
method implemented in the same program was used to determine the
minimum energy conformations on the potential-energy hypersurface of
10. For this purpose, the dihedral angles w, f, y, c1, c2, and c3 were varied
systematically with angle steps of 180, 120, 60, 120, 120, and 608, respec-
tively. The generated conformations were then minimized and, after re-
moving duplicate and isoenergetic symmetry-related conformations
(those with dihedral angles {w, f, y, c1, c2, c3} and {ꢀw, ꢀf, ꢀy, ꢀc1, ꢀc2,
ꢀc3}), all the resulting minima within 10 kcal from the lowest-energy
minimum were used as starting points for subsequent geometry optimiza-
tions by using quantum mechanical methods. All the minima were opti-
mized at the B3LYP/6–31G** level of theory, that is, by using the hybrid
of the Becke three-parameter exchange functional (B3)[41] combined with
the Lee, Yang, and Parr (LYP) exchange-correlation functional[42] and the
6–31G** basis set. Transition states for the interconversion between se-
lected low-energy cis/trans pairs of conformers were determined at the
same level of theory. For this purpose, four possible transition-state struc-
tures for each cis/trans pair were considered. These four structures arise
from two possible directions, namely, clockwise and counterclockwise, for
the twist of the amide bond and concomitant pyramidalization of the
amide nitrogen atom above or below the plane of the amide group.[43]
Frequency analyses were carried out to characterize the nature of the de-
termined stationary points as either minima (no imaginary frequency) or
transition states (single imaginary frequency) and to calculate the zero-
point vibrational energies and the thermal and entropic corrections from
which the conformational free energies in the gas phase were deter-
mined.
Experimental Section
General: All the reagents were obtained from commercial sources and
used without further purification. Polystyrene AM RAM resin
(0.75 mmolgꢀ1) was purchased from Rapp Polymere GmbH (Germany).
Polypropylene syringes fitted with a polyethylene disc were used for the
reactions carried out in a HS501 Digital IKA Labortechnik stirrer. Mi-
crowave-assisted reactions were performed in a CEM Discover micro-
wave reactor using a 10 mL glass reaction vessel. Coupling efficiencies
were monitored using the 2,4,6-trinitrobenzenesulfonic acid (TNBS) test
for primary amines, the chloranil test for secondary amines, and the mal-
achite green test for carboxylic acids. Analytical reverse-phase (RP)
HPLC was performed with a Hewlett Packard Series 1100 (UV detector
1315A) modular system on a reverse-phase Kromasil 100 C8 column
(15ꢆ0.46 cm, 5 mm) and a X-Terra C18 column (15ꢆ0.46 cm, 5 mm). Sol-
vent mixtures of CH3CN/H2O containing 0.1% TFA at a flow rate of
1 mLminꢀ1 were used as the mobile phase and the monitoring wave-
length was set at l=220 nm. Semipreparative RP-HPLC was performed
on a Waters system (Milford, MA, USA) on a X-Terra C18 column (19ꢆ
250 mm, 5 mm). High-resolution mass spectra (HRMS-FAB) were carried
out at the Mass Spectrometry Service of the University of Santiago de
Compostela (Spain) and at the IQAC Mass Spectrometry Service.
NMR spectroscopic analysis: The NMR samples were prepared at con-
centrations of 10 mm. The NMR spectroscopic experiments were carried
out either on a Varian INOVA 500 spectrometer (500 and 125 MHz for
1H and 13C, respectively) or a Varian MERCURY 400 spectrometer (400
and 100 MHz for 1H and 13C, respectively). To take advantage of the
magnetic field value, measurements that required temperatures higher
than room temperature for observing coalescence were performed in an
Synthesis of perhydro-1,4-diazepine-2,5-dione derivatives 2 (Scheme 2):
The synthesis of these compounds was carried out by using a general pro-
cedure on a 1% cross-linked polysterene resin bearing the Fmoc-protect-
ed Rink amide linker AM RAM (0.75 mmolgꢀ1). The resin was filtered
and washed with DMF (3ꢆ4 mL), iPrOH (3ꢆ4 mL), and CH2Cl2 (3ꢆ
4 mL) after each reaction. The ten-step synthesis proceeded as follows:
apparatus with
a proton resonance frequency of 300 MHz (Varian
UNITY 300 spectrometer). Bidimensional NMR spectroscopic experi-
ments (gDQCOSY, gHSQC, gHMBC, ROESY, and 1D NOESY) were
acquired by using the standard pulse sequences and parameter sets as
found in VNMRJ. Chemical shifts d are given in ppm relative to the in-
ternal reference trimethylsilane (TMS), and the coupling constants J are
reported in Hertz (Hz).
1) Removal of the Fmoc group: The Rink amide resin was treated with
a solution of piperidine in DMF (20%) and the reaction mixture was
stirred under microwave activation for 2 min at 608C.
2) Acylation with bromoacetic acid: The resin was treated with a solu-
tion of bromoacetic acid (5 equiv) and DIC (5 equiv) in DMF. The reac-
tion mixture was stirred under microwave activation for 1 min at 358C.
The resin was drained and washed.
X-ray crystallographic studies: Crystals of 2a were grown from a concen-
trated solution in acetonitrile. One crystal was mounted on a glass fiber,
and measurements were made with a Bruker SMART-APEX CCD area-
Chem. Eur. J. 2011, 17, 7927 – 7939
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7937

A. Messeguer et al.
3) Amine coupling: A solution of appropriate primary amine (5 equiv)
and triethylamine (5 equiv) in DMF was added to the resin and the sus-
pension was stirred for 2 min at 908C under microwave activation. The
supernatant was removed, and the residue was drained and washed.
agent. The reaction mixtures were stirred for 3 h at room temperature
and filtered. The resin was drained and washed.
10) Cleavage of the resin: Treatment of the resin as described above af-
forded a crude reaction mixture containing the title compound 3, which
was purified by semipreparative RP-HPLC with the appropriate CH3CN/
H2O gradient.
4) Acylation with allyl maleate: The resin was treated with a solution of
allyl maleate (5 equiv), HOBt (5 equiv), and DIC (5 equiv) in CH2Cl2/
DMF (2:1). The reaction mixture was stirred under microwave activation
for 1 min at 458C and filtered.
5) Michael addition: A solution of the corresponding primary amine
(5 equiv) and triethylamine (5 equiv) in DMF (4 mL) was added to the
resin and the suspension was stirred for 16 h at 208C. The supernatant
was removed, and the residue was drained and washed.
Acknowledgements
Support from MICINN (Grants CTQ2005-00995, SAF2008-00048,
BIO2007 60066) is acknowledged. The authors thank F. J. Morales and
M. Camargo for technical assistance. A JAE-CSIC fellowship to A.M. is
also acknowledged.
6) Acylation with bromoacetic acid: The resin was treated as described
in step 2.
7) Amine coupling: The resin was treated as described in step 3.
8) Removal of the allyl ester: The resin was treated with 0.5m KOH
(5 equiv) in dioxane for 4 min at 1108C under microwave activation. The
supernatant was removed and the residue was drained and washed.
[1] R. J. Simon, R. S. Kania, R. N. Zuckermann, V. D. Huebner, D. A.
Jewell, S. Banville, S. Ng, L. Wang, S. Rosenberg, C. K. Marlowe,
D. C. Spellmeyer, R. Tan, A. D. Frankel, D. V. Santi, F. E. Cohen,
P. A. Bartlett, Proc. Natl. Acad. Sci. USA 1992, 89, 9367–9371.
[2] a) G. M. Figliozzi, R. Goldsmith, S. C. Ng, S. C. Banville, R. N.
Zuckermann, N. A. John, Methods Enzymol. 1996, 267, 437–447;
b) H. J. Olivos, P. G. Alluri, M. M. Reddy, D. Salony, T. Kodadek,
Org. Lett. 2002, 4, 4057; c) B. C. Gorske, S. A. Jewell, E. J. Guerard,
H. E. Blackwell, Org. Lett. 2005, 7, 1521–1524.
9) Cyclization on a solid phase: The intramolecular cyclization was pro-
moted by treatment with PyBOP (1.5 equiv), HOBt (1.5 equiv), and
DIPEA (3 equiv) in DMF (4 mL). The mixture was stirred for 20 min at
608C under microwave activation and filtered. The solvent was removed
by filtration and the resin was washed. The cyclization was checked by
testing with chloranil, and if necessary the process was repeated under
the same conditions.
10) Cleavage of the resin: The resin was treated with a mixture of TFA/
CH2Cl2/H2O (60:40:2, v/v/v) for 30 min at room temperature. The cleav-
age mixture was filtered and the solvent was removed under reduced
pressure. The residue obtained was purified by semipreparative RP-
HPLC with the appropriate CH3CN/H2O gradient.
[3] R. N. Zuckermann, J. M. Kerr, S. B. H. Kent, W. H. Moos, J. Am.
Chem. Soc. 1992, 114, 10646–10647.
[4] S. M. Miller, R. J. Simon, S. Ng, R. N. Zuckermann, J. M. Kerr,
W. H. Moos, Bioorg. Med. Chem. Lett. 1994, 4, 2657–2662.
[5] R. N. Zuckermann, E. J. Martin, D. C. Spellmeyer, G. B. Stauber,
K. R. Shoemaker, J. M. Kerr, G. M. Figliozzi, D. A. Goff, M. A.
Siani, R. J. Simon, S. C. Banville, E. G. Brown, L. Wang, L. S. Richt-
er, W. H. Moos, J. Med. Chem. 1994, 37, 2678–2685.
[6] T. M. Ross, R. N. Zuckermann, C. Reinhard, W. H. Frey II, Neuro-
sci. Lett. 2008, 439, 30–33.
[7] J. T. Nguyen, M. Porter, M. Amoui, W. T. Miller, R. N. Zuckermann,
W. A. Lim, Chem. Biol. 2000, 7, 463–473.
[8] T. Hara, S. R. Durell, M. C. Myers, D. H. Appella, J. Am. Chem.
Soc. 2006, 128, 1995–2004.
[9] P. Mora, I. Masip, N. Cortꢂs, R. Marquina, R. Merino, J. Merino, T.
Carbonell, I. Mingarro, A. Messeguer, E. Pꢂrez-Payꢃ, J. Med. Chem.
2005, 48, 1265–1268.
Synthesis of 3-substituted 1,4-piperazine-2,5-dione derivatives
3
(Scheme 2): The synthesis of these compounds was carried out by using a
general procedure on the resin described above. Unless stated otherwise,
the reaction time and further treatment of the crude reaction mixtures
were analogous to those described above for the synthesis of 2. The resin
was filtered and washed with DMF (3ꢆ4 mL), iPrOH (3ꢆ4 mL), and
CH2Cl2 (3ꢆ4 mL) after each reaction. First, the Rink amide resin was
treated as described above in steps 1–5.
6) Cleavage of the resin: The resin was treated with a mixture of TFA/
CH2Cl2/H2O (60:40:2, v/v/v) for 30 min at room temperature. The cleav-
age mixture was filtered and the solvent was removed by evaporation
under reduced pressure. Traces of water and acid were removed by wash-
ing with CH3CN (3ꢆ5 mL), which was also removed under reduced pres-
sure.
[10] J. A. Patch, A. E. Barron, J. Am. Chem. Soc. 2003, 125, 12092–
12093.
[11] C. Garcia-Martinez, M. Humet, R. Planells-Cases, A. Gomis, M.
Caprini, F. Viana, E. De La Pena, F. Sanchez-Baeza, T. Carbonell,
C. De Felipe, E. Perez-Paya, C. Belmonte, A. Messeguer, A. Ferrer-
Montiel, Proc. Natl. Acad. Sci. USA 2002, 99, 2374–2379.
[12] C. W. Wu, S. L. Seurynck, K. Y. C. Lee, A. E. Barron, Chem. Biol.
2003, 10, 1057–1063.
[13] P. A. Wender, D. J. Mitchell, K. Pattabiraman, E. T. Pelkey, L. Stein-
man, J. B. Rothbard, Proc. Natl. Acad. Sci. USA 2000, 97, 13003–
13008.
7) Cyclization in solution to obtain intermediate 8: Cyclization was pro-
moted by treating the cleavage mixture with dioxane for 4 min at 1208C
under microwave conditions (HPLC monitoring).
8) Removal of the allyl ester in solution: A solution of 4m NaOH and
allyl alcohol (1:2 v/v) was added to intermediate 8 in dioxane (1:4) and
the mixture was stirred for 4 min at 908C under microwave activation
(monitoring by HPLC). The crude reaction mixture was acidified with
1m HCl and the organic solvent was evaporated. The resulting residue
was extracted with EtOAc, dried over anhydrous MgSO4, and concentrat-
ed in vacuo.
[14] K. Kirshenbaum, A. E. Barron, R. A. Goldsmith, P. Armand, E. K.
Bradley, K. T. V. Truong, K. A. Dill, F. E. Cohen, R. N. Zuckermann,
Proc. Natl. Acad. Sci. USA 1998, 95, 4303–4308.
The removal of the allyl ester to obtain 9e was carried out on a solid
phase under the conditions described previously to obtain 2. After cleav-
age of the resin with a solution of TFA/CH2Cl2/H2O (60:40:2 v/v/v) for
30 min at room temperature, the residue obtained from the evaporation
of the crude reaction mixture was treated with dioxane for 30 min under
reflux.
[15] a) M. Kuemin, Y. A. Nagel, S. Schweizer, F. W. Monnard, C. Och-
senfeld, H. Wennemers, Angew. Chem. 2010, 122, 6468–6471;
Angew. Chem. Int. Ed. 2010, 49, 6324–6327; b) W. J. Wedemeyer, E.
Welker, H. A. Scheraga, Biochemistry 2002, 41, 14637–14644.
[16] The cis/trans conformers of the tertiary amide bond in peptoids
were named following the nomenclature in this field; for example,
see: Q. Sui, D. Borchardt, D. L. Rabenstein, J. Am. Chem. Soc.
2007, 129, 12042–12048.
9) Amide formation: The different peptoidyl resins 4 were synthesized
as described above in steps 1–3. The corresponding intermediate
9
(0.25 mmol) was coupled onto the resin containing the N-alkylglycine
residue (1 equiv) in the presence of HOBt (0.4 mmol) and DIC
(0.4 mmol) in CH2Cl2/DMF (1:1). In the case of 3e, the mixture PyBOP
(0.4 mmol) and DIPEA (0.8 mmol) in DMF was used as the coupling
[17] a) B. C. Gorske, B. L. Bastian, G. D. Geske, H. E. Blackwell, J. Am.
Chem. Soc. 2007, 129, 8928–8929; b) S. A. Fowler, H. E. Blackwell,
Org. Biomol. Chem. 2009, 7, 1508–1524.
7938
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 7927 – 7939

Conformational Flexibility of Peptoids
FULL PAPER
[18] a) J. A. Patch, A. E. Barron, Curr. Opin. Chem. Biol. 2002, 6, 872–
877; b) R. C. Elgersma, G. E. Mulder, J. A. W. Kruijtzer, G. Posthu-
ma, D. T. S. Rijkers, R. M. J. Liskamp, Bioorg. Med. Chem. Lett.
2007, 17, 1837–1842.
[19] E. Pꢂrez-Payꢃ, M. Orzꢃez, L. Mondragꢁn, D. Wolan, J. A. Wells, A.
Messeguer, M. J. Vicent, Med. Res. Rev. 2010, DOI: 10.1002/
med.20198.
[20] I. Masip, N. Cortes, M. J. Abad, M. Guardiola, E. Perez-Paya, J. Fer-
ragut, A. Ferrer-Montiel, A. Messeguer, Bioorg. Med. Chem. 2005,
13, 1923–1929.
[29] J. Messeguer, I. Masip, M. Montolio, J. A. del Rio, E. Soriano, A.
Messeguer, Tetrahedron 2010, 66, 2444–2454.
[30] The possibility of a hydrogen bond with the second heterocyclic car-
bonyl group was discarded on the basis of the conformational analy-
sis of 2a,b.
[31] H. Shanan-Atidi, K. H. Bar-Eli, J. Phys. Chem. 1970, 74, 961–963.
[32] J. Sandstrom in Dynamic NMR Spectroscopy, Academic Press, New
York, 1982.
[33] The line-shape simulation was performed by using the gNMR v4.1.0
program.
[34] Free-energy calculations in solution were also performed by running
single-point energy determinations by using the standard Poisson–
Boltzmann solver included in Jaguar (not reported), although the re-
sults did not correlate with the experimental data; clearly, a more
accurate solvation treatment, or even calculations with an explicit
solvent, would be required to obtain more reliable results, but this
was beyond the scope of this study.
[35] a) L. Fielding, Prog. Nucl. Magn. Reson. Spectrosc. 2007, 51, 219–
242; b) B. Meyer, T. Peters, Angew. Chem. 2003, 115, 890–918;
Angew. Chem. Int. Ed. 2003, 42, 864–890.
[36] SAINT, version 6.22, Bruker AXS Inc., Madison, WI, 2001.
[37] SADABS, version 2.03, G. M. Sheldrick, University of Gçttingen,
Gçttingen (Germany), 2002.
[38] SHELXTL, version 6.10, G. M. Sheldrick, Bruker AXS Inc., Madi-
son, WI, 2000.
[21] G. Malet, A. G. Martin, M. Orzꢃez, M. J. Vicent, I. Masip, G. Sancli-
mens, A. Ferrer-Montiel, I. Mingarro, A. Messeguer, H. O. Fearn-
head, E. Pꢂrez-Payꢃ, Cell Death Differ. 2006, 13, 1523–1532.
[22] J. Scheper, M. Guerra-Rebollo, G. Sanclimens, A. Moure, I. Masip,
D. Gonzꢃlez-Ruiz, N. Rubio, B. Crosas, ꢇ. Meca-Cortꢂs, N. Loukili,
V. Plans, A. Morreale, J. Blanco, A. R. Ortiz, ꢈ. Messeguer, T. M.
Thomson, PLoS ONE 2010, 5, e11403.
[23] M. M. Hann, A. R. Leach, G. Harper, J. Chem. Inf. Comput. Sci.
2001, 41, 856–864.
[24] J. M. Holub, H. Jang, K. Kirshenbaum, Org. Lett. 2007, 9, 3275–
3278.
[25] S. B. Y. Shin, B. Yoo, L. J. Todaro, K. Kirshenbaum, J. Am. Chem.
Soc. 2007, 129, 3218–3225.
[26] O. E. Vercillo, C. K. Z. Andrade, L. A. Wessjohann, Org. Lett. 2008,
10, 205–208.
[39] Schrçdinger, LLC, New York, NY, 2008.
[27] a) L. Mondragꢁn, M. Orzꢃez, G. Sanclimens, A. Moure, A.
ArmiÇꢃn, P. Sepulveda, A. Messeguer, M. J. Vicent, E. Pꢂrez-Payꢃ,
J. Med. Chem. 2008, 51, 521–529; b) L. Mondragꢁn, L. Galluzzi, S.
Mouhamad, M. Orzꢃez, J. Vicencio, I. Vitale, A. Moure, A. Messe-
guer, E. Pꢂrez-Payꢃ, G. Kroemer, Apoptosis 2009, 14, 182–190.
[28] T. Thomson, J. Scheper, A. Messeguer, ꢉ. Ramirez Ortiz, G. Sancli-
mens, I. Masip, A. Moure, D. Gonzꢃlez, A. Morreale, WO/2008/
009758, 2008.
[40] a) T. A. Halgren, J. Comput. Chem. 1999, 20, 720–729; b) T. A.
Halgren, J. Comput. Chem. 1999, 20, 730–748.
[41] A. D. Becke, J. Chem. Phys. 1993, 98, 1372–1377.
[42] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
[43] S. Fischer, R. L. Dunbrack, M. Karplus, J. Am. Chem. Soc. 1994,
116, 11931–11937.
Received: January 20, 2011
Published online: May 24, 2011
Chem. Eur. J. 2011, 17, 7927 – 7939
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7939