Ψ[CH(CF3)NH]Gly-peptides: Synthesis and conformation analysis

Article abstract of DOI:10.1039/b901718f

Ψ[CH(CF₃)NH]Gly peptides, a conceptually new class of peptidomimetics having a stereogenic trifluoroethylamine group as a natural peptide-bond surrogate, have been synthesized by stereoselective addition of α-amino acid esters to trans-3,3,3-tr

Full text of DOI:10.1039/b901718f

View Article Online / Journal Homepage / Table of Contents for this issue
www.rsc.org/obc
Volume 7 | Number 11 | 7 June 2009 | Pages 2225–2468
ISSN 1477-0520
FULL PAPER
Alessandro Volonterio et al.
y
[CH(CF3)NH]Gly-peptides: synthesis
and conformation analysis
1477-0520(2009)7:11;1-G

PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
W[CH(CF3)NH]Gly-peptides: synthesis and conformation analysis†
Marco Molteni,a Maria Cristina Bellucci,b Serena Bigotti,a Stefania Mazzini,b Alessandro Volonterio*a and
Matteo Zanda*c,d
Received 27th January 2009, Accepted 10th March 2009
First published as an Advance Article on the web 7th April 2009
DOI: 10.1039/b901718f
W[CH(CF3)NH]Gly peptides, a conceptually new class of peptidomimetics having a stereogenic
trifluoroethylamine group as a natural peptide-bond surrogate, have been synthesized by stereoselective
addition of a-amino acid esters to trans-3,3,3-trifluoro-1-nitropropene. Long range nuclear Overhauser
effects, detected via ROESY experiments, provided evidence that model W[CH(CF3)NH]Gly-
tetrapeptides incorporating a trifluoroethylamine mimic of the dipeptide loop Pro-Gly can be
represented by an ensemble of unfolded and folded conformations. The latter are driven by the
formation of a hydrogen bond between a carbonyl group and the aminic proton of the
trifluoroethylamine unit. MD calculations indicate a population of conformers which adopt folded b
turn structures for all the L-peptides; on the other hand, a D-stereochemistry at the Pro residue induces
a natural folding towards a b hairpin conformation driven by the formation of a second hydrogen bond,
regardless of the stereochemistry of the stereogenic peptide bond surrogate.
The stereogenic trifluoroethylamine6 function is a conceptually
Introduction
new peptide bond surrogate that has recently found the first vali-
Peptides occur throughout nature in a wide range of roles
dation in drug discovery thanks to the highly potent and metaboli-
essential to virtually every biochemical process. However, the
cally stable Cathepsin K inhibitor Odanacatib, that is now in Phase
III clinical trials for the therapy of postmenopausal osteoporosis.7
This peptide-bond replacement has peculiar properties (Fig. 1).
First of all, the sp3 N atom of the trifluoroethylamine function
has little Lewis basicity and is a bad hydrogen bond acceptor.
In fact, the electron-withdrawing CF3 group is instrumental in
neutralizing the amine function generating a poorly-basic NH
moiety. The latter is essentially non-protonated at physiological
pH, in close similarity with the NH of an amide/peptide group.
In fact, the pKa of protonated trifluoroethylamine mimics was
reported to be lower than 1.5.7c Moreover, the trifluoroethy-
lamine NH moiety is a good hydrogen-bond donor, due to
the increased acidity arising from the presence of the a-CF3
group. On the contrary, the CF3 group is a weak hydrogen-
bond acceptor,8 thus rendering the trifluoroethylamine function
an effective peptide bond replacement only when the carbonyl
group of the original ligand’s amide/peptide-bond is not involved
in essential hydrogen-bonding with the receptor. One should also
notice that the trifluoroethylamine unit has an sp3 tetrahedral
configuration that can contribute to the optimization of the
geometry and spatial orientations of the interactions (including
pharmacological properties of most peptides preclude their use as
drugs mainly because of their rapid in vivo degradation by action
of proteolytic enzymes, and therefore by their low bioavailability.1
The “druggability” of peptides could be achieved by the synthesis
of suitable peptidomimetics able to retain both the activity and
potency of the parent peptides, while at the same time be more
metabolically stable, orally active and selective. In this context, the
isosteric replacement of a scissile peptide bond represents a viable
and popular approach in the rational design of peptidomimetics2
because: 1) nearly all the peptide bond surrogates, with the
notable exception of the ester and thioester functions, are more
stable to enzymatic hydrolysis than the natural peptide bond;3
2) in the area of protease inhibitors, a nonscissile peptide bond
replacement at the substrate cleavage site, particularly one that
can mimic the transition state of amide bond hydrolysis, is key to
achieving bioactivity;2b,4 3) most peptide bond surrogates influence
the conformational preference of the residues in close proximity
to them, and sometimes impart a conformational preference to a
peptide region; 4) alterations of electronic properties introduced
by peptide bond surrogates can significantly affect the transport
properties of their parent peptides.5
aDipartimento di Chimica, Materiali, e Ingegneria Chimica “Giulio Natta”,
Politecnico di Milano, via Mancinelli 7, 20131, Milano, Italy. E-mail:
alessandro.volonterio@polimi.it; Fax: +39-02-23993080; Tel: +39-02-
23993139
bDipartimento di Scienze Molecolari Agroalimentari, Universita` degli Studi
di Milano, via Celoria 2, 20133, Milano, Italy
cC.N.R.-Istituto di Chimica del Riconoscimento Molecolare, Sezione “Adolfo
Quilico”, via Mancinelli 7, 20131, Milano, Italy. E-mail: matteo.zanda@
polimi.it; Fax: +39-02-23993080; Tel: +39-02-23993084
dInstitute of Medical Sciences, University of Aberdeen, Foresterhill, Ab-
erdeen, AB25 2ZD, Scotland, (UK)
† Electronic supplementary information (ESI) available: 1H, 13C, 2D
ROESY and 1H-13C 2D HMBC spectra. See DOI: 10.1039/b901718f
Fig. 1 Comparison between the peptide bond and the trifluoroethylamine
function.
2286 | Org. Biomol. Chem., 2009, 7, 2286–2296
This journal is
The Royal Society of Chemistry 2009
©

hydrogen-bonds) between the original planar amide/peptide-
moiety and the receptor. Last but not least, there is substantial
evidence that the trifluoroethylamine unit has high metabolic
stability.7
In our first accomplishment in this area,9 we introduced the
retro-trifluoroethylamine unit that was incorporated in partially
modified retropeptides (PMR) A (Fig. 2) having a [CH(CF3)NH]
unit instead of the retro [CONH] peptide bond.
blocks for further functionalization because they are much more
reactive than the unfluorinated counterparts.12 While for the
stereoselective synthesis of PMR-W[NHCH(CF3)]Gly-peptides
we exploited the reactivity of enantiomerically and geometrically
pure fluorinated enoyl oxazolidin-2-ones, that reacted even with
poorly nucleophilic a-amino acid esters, in this case the key step
for the assembling of the amide bond surrogate was achieved
using highly reactive trans-2-trifluoromethyl-1-nitroethene 2 that
was prepared by Henry reaction13 of aqueous fluoral with an
excess of nitrometane, followed by dehydration of the intermediate
nitroaldol on P2O5 and distillation (Scheme 1).14
Fig. 2 Structure of natural peptides, PMR-W[NHCH(CF3)]Gly-peptides
and W[CH(CF3)NH]Gly-peptides 1.
Scheme 1 Synthesis of 2-trifluoromethyl-1-nitroethene 2.
Nucleophilic addition of a-amino acid esters,15 generated in situ
from the corresponding hydrochloride salts 3 in the presence of
a base (Scheme 2), to 2 gave rise to the formation of a¢-Tfm-b¢-
nitro a-amino esters diastereoisomers 4 (major) and 5 (minor).
The reaction is operatively very simple, took place almost instan-
taneously at room temperature and the final diastereoisomers were
in all cases easily separable by flash chromatography.
We showed9e that this unit is a sort of hybrid between a
peptide bond mimic and a proteolytic transition state analogue,
as it combines some of the properties of a peptidyl–CONH-
group (low NH basicity, a CH(CF3)-NH-CH backbone angle
close to 120◦, a C-CF3 bond substantially isopolar with the
=
C O) with properties of the tetrahedral intermediate B involved
in the protease-mediated hydrolysis reaction of a peptide bond
(high electron density on the trifluoromethyl group, tetrahedral
backbone carbon). Moreover, the presence of the bulky CF3
group, rather than the presence of an intramolecular hydrogen
bond, is probably the driving force for the high stability of the
turn-like conformations displayed by appropriately configured
retropeptides A both in low polarity organic solvent solutions
and in the solid state.
Next, we communicated the stereocontrolled synthesis of a re-
lated class of peptidomimetics, the W[CH(CF3)NH]Gly-peptides
1 (Fig. 2, R = H).10 These peptidomimetics are much closer to
the natural peptides, as they feature a trifluoroethylamine unit re-
placing the native peptidic amide bond. More recently we also de-
scribed the synthesis of differently fluorinated W[CH(RF)NH]Gly-
peptides (RF = CF2H, CF2Cl, CF2CH3, etc.).11 In this paper we
report a full account on the stereoselective synthesis of these
peptidomimetics with a special emphasis on the synthesis of ratio-
nally designed W[CH(CF3)NH]Gly-tetrapeptides and the studies
on their propensity to assume turn-like conformations in organic
solvents, as demonstrated by means of NMR and molecular
modeling analysis. Thus, for the first time, the trifluoroethylamine
NH moiety is shown to be a good hydrogen bond donor.
Attempts to synthesize more complex W[CH(CF3)NH]-peptides
1 (Fig. 2, R = alkyl) through the same chemistry will be also
discussed.
Scheme 2 The key aza-Michael step.
The diastereoselection of the reaction was dependent on differ-
ent factors such as the solvent, the structure and the amount of
the base, the side chain R of the a-amino acid esters 3 and the
temperature.16 To fine tune such parameters in order to optimize
diastereoselectivities we carried out a series of experiments at rt
using L-Val benzyl and tert-butyl ester hydrochlorides 3a,b as
model nucleophiles (Table 1). First we investigated the influence
of the base (1.1 equiv) using DCM as a model solvent (entries 1–4).
The best results were achieved with organic bases and in particular
with the more sterically hindered DIPEA (entry 4) as compared
to TMP (entry 2) and DABCO (entry 3). Unsatisfactory results
were obtained using an inorganic base such as NaHCO3 (entry 1).
Next, using DIPEA as a model base we focussed on the role of the
solvent (entries 4–8). As expected,16 the best results were achieved
using low-polarity solvents such as CCl4 (entry 6) and in particular
toluene (entries 7, 8) while with more polar DCM (entry 4) and
THF (entry 5) less stereoselective reactions were observed. To our
surprise, carrying out the reaction with the free a-amino acid ester
derived from 3b, therefore without using any base,17 a remarkable
Results and discussion
Chemistry
Electron poor carbon–carbon double bonds having a fluoroalkyl
group in b position constitute interesting and versatile building
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2286–2296 | 2287
©

Table 1 Aza-Michael reaction with a-amino acid esters nucleophiles
Entrya
a-amino acid ester
Major product
R
X
Base (equiv)
Solvent
drc
1.7:1.0
Yield (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
21
L-3a
L-3a
L-3a
L-3a
L-3b
L-3b
L-3a
L-3b
L-3b
L-3b
L-3b
L-3b
L-3b
L-3b
L-3b
L-3c
L-3c
L-3d
L-3d
L-3e
L-3f
D-3g
4a
4a
4a
4a
4b
4b
4a
4b
4b
4b
4b
4b
4b
4b
4b
4c
4c
4d
4d
4e
4f
iPr
iPr
iPr
iPr
iPr
iPr
iPr
iPr
iPr
iPr
iPr
iPr
iPr
iPr
iPr
Me
Me
sBu
sBu
Bn
iBu
Bn
Bn
Bn
Bn
Bn
t-Bu
t-Bu
Bn
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
Me
NaHCO3 (1.1)
TMP (1.1)
DABCO (1.1)
DIPEA (1.1)
DIPEA (1.1)
DIPEA (1.1)
DIPEA (1.1)
DIPEA (1.1)
noneb
DIPEA (1.0)
DIPEA (1.3)
DIPEA (1.5)
DIPEA (1.7)
DIPEA (2.0)
DIPEA (5.0)
DIPEA (1.1)
noneb
DCM
DCM
DCM
DCM
THF
CCl4
>98c
>98c
79d
2.5:1.0
3.0:1.0
4.4:1.0
3.7:1.0
7.9:1.0
11.2:1.0
11.7:1.0
3.8:1.0
4.0:1.0
8.5:1.0
5.1:1.0
3.6:1.0
3.7:1.0
3.2:1.0
5.8:1.0
1.6:1.0
7.5:1.0
4.4:1.0
8.5:1.0
6.0:1.0
1.8:1.0
70d
>98c
>98c
79d
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
65c
>98c
93c
88c
77c
70c
75c
74c
60d
87d
DIPEA (1.1)
72d
Me
noneb
88d
t-Bu
t-Bu
Me
DIPEA (1.1)
DIPEA (1.1)
noneb
60d
75d
4g
98d
b
1
a All the reactions were carried out at rt. a-Amino acid ester was preliminarily treated with NaHCO3 c Determined by 19F and H NMR. d Overall
isolated yield.
drop of stereoselectivity was observed (from a dr of about 12:1 to
less than 4:1, entries 8 and 9 respectively). We next investigated the
influence of the base stoichiometry on the stereoselection of the
reaction (entries 10–15), using the optimized conditions (DIPEA
as base and toluene as solvent). The best result was obtained with
a catalytic amount of the base (1.1 equiv., entries 7 and 8) while an
increasing amount of the base resulted in a progressive decrease of
diastereoselectivity (entries 11–15). Very low diasteroselection was
achieved also when the added base was completely neutralized by
the hydrochloride 3 (entry 10). As expected, the X group on the
ester function of the nucleophile has a negligible effect (compare
entries 7 and 8). Finally, the effect of the bulkiness of the R side
chains of 3 was investigated performing the reaction with different
a-amino acid esters (entries 16–21). Accordingly, the bulkier is
the R group the higher resulted to be the diastereoisomeric ratio
of the process, as exemplified by R = iso-propyl (entries 7 and 8)
as compared to benzyl (entry 20), iso-butyl (entry 21) and methyl
(entry 16). An important effect of the stereogenic sec-butyl side-
chain is more than likely in the case of L-Ile-OMe (entries 18,
19), so these results should be weighed in a different manner. In an
attempt to improve further the diastereoselectivities, we undertook
a series of experiments lowering the temperature of the reaction.
Rather surprisingly, room temperature seems to be essential in
order to obtain the desired products because carrying out the
reaction at -40 ◦C and -70 ◦C in different solvents invariably
resulted in the formation of complex mixtures of unidentified
compounds.
The stereochemical outcome may be rationalized by using the
predictive model depicted in Fig. 3, which resembles those of simi-
lar aza-Michael reactions involving fluorinated acceptors.19 The a-
amino ester nucleophile reacts from the less hindered diastereoface
of the five-membered ring resulting from an intramolecular
H-bond involving a carboxyl oxygen and an amine hydrogen
Ha. The trans-nitroalkene acceptor places the sterically demand-
ing CF3 group in the less crowded position, namely pseudo-
equatorially. One of the nitro O-atoms could be involved in H-
bonding with the amino-ester’s Hb.20 A further stabilizing polar
interaction between one of the F-atoms and the amino-ester’s Ha
is not unlikely.21 In this picture, the Hunig’s base would play a key
role by promoting the transfer of the Hb proton to the carbon in
a-position to the nitro group.
The stereochemistry of the minor diastereoisomer 5a was as-
sessed by X-ray diffraction (see ESI†),18 whereas the configurations
of the other adducts 4 and 5 were confidentially assigned on the
basis of their spectroscopic features in comparison with those of 4a
and 5a. In fact, the 19F NMR signals of the fluorinated group of the
major diasteroisomers 4 were always observed at higher fields than
those of the minor diastereoisomer 5 (see Experimental Section).
Fig. 3 Predictive model for the formation of the major diastereomers
4a–g.
The synthesis of more complex W[CH(CF3)NH]Ala-peptides
1 (R = Me, Fig. 2) was attempted by means of an aza-
Michael addition of L-Ile-OMe 3d to 3,3,3-trifluoro-1-methyl-
1-nitropropene 6 under optimized conditions (Scheme 3). The
2288 | Org. Biomol. Chem., 2009, 7, 2286–2296
This journal is
The Royal Society of Chemistry 2009
©

assembling peptide strands into b-sheets. The simplest way to bring
two antiparallel strands together is to connect them by a short
dipeptide segment (loop) between the C-terminus of one strand
and the N-terminus of the other able to generate a ten-membered
=
intramolecularly hydrogen-bonded (C O ◊ ◊ ◊ H-N) ring involving
four amino-acid residues, namely a b-turn motif.25 This strand-
loop-strand arrangement is referred to as “b-hairpin” (Fig. 4).
Scheme 3 Synthesis of nitro-adducts 7.
latter was obtained in geometrically pure form following the same
strategy outlined for compound 2 (see Scheme 1).22 The reaction,
although moderately stereoselective, gave rise to the formation of
all of the four diastereoisomers 7 which could not be isolated in
analytically pure form. Owing to the difficulties connected with
the isolation and purification of compounds like 7, the synthesis
of W[CH(CF3)NH]Ala-peptides was not investigated further.23
With a number of a¢-fluoroalkyl-b¢-nitro a-amino esters
4,5 in hand we addressed their elaboration into the target
W[CH(CF3)NH]Gly peptides 8 (Scheme 4).
Fig. 4 The b-turn and b-hairpin motif.
A number of recent studies have reported the incorporation
of peptidomimetics into b-sheet or b-hairpin structures such as
1,6-dehydro-3(2H) pyridinone, referred to as @-tide residue,26
templated b-sheet,27 alkene isosteres28 and cis-azobenzene turn
mimic.29 It is known that in small tetrapeptides, in which in-
tramolecular hydrogen bonding provides a major driving force
for secondary structure formation, the presence of proline at
the second of the four residues promotes formation of turns.30
Specifically, we focussed our attention on two tetrapeptides,
previously studied by Gellman et al., namely LVal-LPro-Gly-LLeu
9 and LVal-DPro-Gly-LLeu 10 that differ from each other only
for the configuration of the proline moiety on the dipeptide loop
Pro-Gly.31 By means of NMR studies in a nonpolar solvent (i.e.
DCM) Gellman et al. demonstrated that the presence of L-proline
in tetrapeptide 9 promotes formation of type I and type II turns
while the “mirror image” b-turns (D-Pro-Gly in tetrapeptide 10)
promotes b-hairpin formation (Fig. 5).
Recently we have demonstrated that the trifluoroethylamine
peptide-bond replacement, when incorporated into PMR-
peptides, is responsible for an increased tendency of these peptides
to assume turn-like conformations in organic solvents even if
they are not directly involved in the formation of intramolecular
hydrogen bonds.9e In order to improve our knowledge on the
influence of this stereogenic surrogate on the conformation of
peptides we decided to assess whether some analogies exist
between the Gellman’s tetrapeptides 9 and 10 and their fluorinated
parent peptides. For this purpose we synthesized four diastereoiso-
meric W[CH(CF3)NH]Gly tetrapeptides 11–14 that differ from
each other for the configurations of (1) the key proline amino
acid and (2) the CF3-substitued stereogenic center (Fig. 6). It
should be stressed that the key amide bond between the Pro-
Gly loop and H-LLeu-NMe2, in which the NH proton is involved
in an intramolecular hydrogen bond stabilizing the secondary
structures, is replaced by the trifluoroethylamine function in which
the above mentioned hydrogen is an a-CF3-amine proton.
The key diastereomeric intermediates 17a,b were obtained in
good overall yields by performing an aza-Michael addition to 2
with H-LLeu-NMe2 hydrochloride 16, using non-stereoselective
conditions (excess of NaHCO3, DCM, 1:1 dr) (Scheme 5). The
amide 16 was obtained by condensation of Cbz-LLeu-OH 15 with
Me2NH followed by N-deprotection.32
Scheme 4 Elaboration of aza-Michael adducts 4 into tripeptide mimetics
8.
Reduction of the nitro group was accomplished by using
Pd(OH)2 catalyst in the presence of a 1 N solution of HCl to
trap the free amino function as hydrochloride salt. Coupling with
Cbz-LPhe-OH using HOBt/EDC or HATU/HOAt led to the
formation of tripeptide mimics 8b-d,g in good overall yields.
Secondary structure of PMR-W[CH(CF3)NH]Gly-peptides
Three types of regular secondary structures are recognized in
proteins: helix, sheet and turn. While many model studies of
a-helices and b-turn stability have been reported, sheets have
received less attention probably because b-sheets in proteins
often form from discontinuous polypeptide segments.24 For this
reason it is important to identify the most effective strategies for
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2286–2296 | 2289
©

Scheme 5 Synthesis of intermediates 17a,b.
Fig. 5 Major folding patterns for tetrapeptides 9 and 10 in DCM solution.
The nitro group of diastereomerically pure 17a was hydro-
genated to an amino group and the latter was coupled with
Boc-LPro-OH using HOBt/EDC, affording the tripeptide 18
(Scheme 6). Standard Boc deprotection using TFA/DCM fol-
lowed by neutralization of the resulting trifluoroacetate salt and
coupling with Boc-LVal-OH afforded the tetrapeptide 19. The tar-
get Ac-LVal-LPro-(S)W[CH(CF3)NH]Gly-LLeu-NMe2 tetrapep-
tide 11 was obtained by deprotection of the amino moiety of
19 and subsequent acetylation with Ac2O in DCM. Using the
same synthetic pathway and Boc-DPro-OH in the second step
of the synthesis we prepared also the epimeric tetrapeptide
12. Analogously, starting from diastereomerically pure 17b we
Scheme 6 Synthesis of Ac-LVal-LPro-(S)W[NHCH(CF3)]Gly-LLeu-NMe2
11.
L
obtained the tetrapeptides 13 and 14 containing Pro and DPro,
respectively, at the second of the four residues.
Fig. 6 W[CH(CF3)NH]Gly tetrapeptides.
This journal is
2290 | Org. Biomol. Chem., 2009, 7, 2286–2296
The Royal Society of Chemistry 2009
©

NOE experiments
is an effective mimic for the peptide bond.37 We also considered
the distances between C-terminal and N-terminal methyl groups.
The lack of cross peaks between the terminal CH3 groups of the
diethylamide function and the CH3 of the acetyl group and/or
NHa allows us to exclude the presence of a hairpin conformation.
It is noteworthy that the different configuration of the stereogenic
trifluoromethylamino group for peptides 11 and 13 doesn’t affect
their conformation.
The CDCl3 solution structure of the peptides 11, 12 and 13
were investigated using 2D ROESY experiments (Rotating frame
Overhauser Enhancement SpectroscopY) at 298 K and 273 K.
The peptides’ sequential assignment was performed starting from
the NHa and NHb amidic signals by means of an integrated series
of 1H-1H (COSY, TOCSY and ROESY)33–35 and 1H-13C (one-bond
and long range)36 2D experiments.
The main problem is the high flexibility of small peptides, which
assume different conformations in solution. Although ROEs
cannot be interpreted in terms of a unique structure due to the
conformational averaging, we obtained a model structure using
MD simulations with a minimum number of distance constraints
converting the intensities of the cross peaks in the ROESY spectra
in distances.
Information on the conformations adopted by these peptides
can be obtained examining the results reported in Table 2, which
lists the intermolecular NOE interactions and interproton distance
˚
values (A) obtained from MD calculations. It should be pointed
out that, in comparison with the Gellman’s tetrapeptides, our
peptides have two amidic protons, NHa and NHb, and an aminic
proton NHc vicinal to the trifluoromethyl group. The experimental
problem associated to the aminic nature of the NHc was mainly
that this proton resonates in the same region of the methylene
group of Pro (see ESI†). In order to detect the important ROEs
contacts defining a folded conformation, we had therefore to
assign, case by case, the NHc. The COSY experiments showed a
cross peak connecting by scalar coupling the Ha Leu with the NHc.
For compounds 11 and 13 we detected ROEs contacts between
NHb and the aminic proton NHc, and NHa and NHc with Ha of
the L-Pro that give a clear indication of the presence of a folded
structure (Fig. 7).
Fig. 8 Conformation of 13 obtained from molecular modeling.
We performed also experiments to determine the temperature
coefficients for the exchangeable amidic protons of the peptides,
in CDCl3 solution, over a temperature range of 233 to 298K.
Literature data38 report that values of -Dd/DT ¥ 103 ppb/K below
4 for amidic NH are consistent with intra-molecular hydrogen-
bonding, while values ranging between 4 and 6 or greater, are
typical of an exposition to the solvent. The coefficient values
obtained for the NH amidic protons (NHa showed values of 10
and 5 ppb/K and NHb showed values of 5 and 6 ppb/K for 13
and 11 respectively) suggest the lack of hydrogen bonds.
Basically for the compound 12 we detected the same ROEs
interactions: NHb with the aminic proton NHc, NHa and NHc with
the Ha of D-Pro where the last interaction (i, i+2) is diagnostic
for a folded structure. The temperature coefficient obtained for
the amidic NHb showed a value of 7 ppb/K suggesting a solvent
exposition of this proton, while NHa showed a “borderline” value
of 4.0 ppb/K. In addition to these interactions we found two other
weak long range cross peaks between NHa and NHc, and between
NHa and the N-terminal methyl groups, which were not observed
for the L-peptides (Fig. 9).
Fig. 7 Schematic representation of the structure of 13 in CDCl3. The
arrows indicate the diagnostic NOE interactions. No NOEs were found
between terminal CH3 and NHa with the N-methyl terminal groups.
The NHb and NHc as well as NHa and Ha of the L-Pro
protons belong to adjacent residues, so that the cross-peaks can
be consistent with a folded conformation but can also derive from
other conformations available to flexible peptides. On the other
hand, the cross-peak between aminic NHc and Ha of the Pro (i,
i+2) is more diagnostic for a folded structure because a comparison
of the value of this distance in an “all-extended” conformation
˚
results in a distance longer than 5 A. Consequently, compounds 11
and 13 can be realistically represented by an ensemble of unfolded
and folded conformations. These findings are in accordance with
the MD calculations which indicate a population of conformers
which adopt a folded b turn structure forming a 10 membered
ring featuring a hydrogen bond between the aminic proton NHc
Fig. 9 Schematic representation of the structure of 12 in CDCl3. The
arrows indicate the diagnostic NOE interactions. NOEs were found
between NHa an NHc and between NHa and the N-methyl terminal group.
˚
of L-Leu and the CO of L-Val (2.3 A, Fig. 8). This observation
validates further the concept that the trifluoroethylamino function
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2286–2296 | 2291
©

˚
Table 2 Intramolecular NOE interactions and interproton distances values (A) for 11, 12 and 13
Model distances
Experimental NOE
Intensitiesa
11
12
13
NHb ◊ ◊ ◊ NHa
ww
w
s
w
s
4.47
4.83
2.49
5.47
3.07
2.51
4.04
4.67
4.44
3.63
4.50
5.01
3.06
5.67
4.40
5.50
2.43
4.55
2.50
2.29
2.47
2.64
3.52
4.19
3.78
4.52
4.64
2.34
2.73
3.89
4.20
4.76
2.94–4.20
2.37–3.88
2.94
2.56, 4.20
3.07
2.47
3.07
2.52
2.45
2.32
4.27
4.22
2.74
4.70, 5.10
4.00
2.75
3.03
5.04
5.33
2.29
5.28
4.56
3.50
5.00
3.09
2.79
3.93
2.84
4.01
2.73
4.61
3.63
3.06
5.37 (zero NOE)
4.76
5.33 (no NOE)
2.44
4.63
2.61
2.24
3.06
2.79
2.79
3.43
3.92
4.53
5.26
3.07
2.42
2.78
3.86
4.23
2.90–4.20
2.40–3.90
2.43
2.54, 4.18
3.08
3.12
2.52
3.07
2.47
2.37
4.27
3.77
2.88
4.65, 5.00
3.24
3.00
2.64
4.70
4.84
2.70
4.31
4.87
2.44
5.62
2.44
2.85
3.86
4.57
4.42
2.99
4.65
5.01
3.08
5.53
4.33
5.05
2.43
4.58
2.52
2.33
3.06
2.81
2.79
4.19
3.76
4.61
4.71
2.38
2.73
3.89
4.16
4.11
2.97–4.19
2.40–3.00
2.85
2.70, 4.22
3.02
3.11
3.05
2.50
3.06
2.43
4.49
4.50
2.78
5.26, 4.82
4.50
2.44
2.43
4.98
5.02
2.42
NHb ◊ ◊ ◊ Ha Val
NHb ◊ ◊ ◊ Ha Pro
NHb ◊ ◊ ◊ Ha Leu
NHb ◊ ◊ ◊ Ha Gly (pro S)
NHb ◊ ◊ ◊ Ha Gly (pro R)
NHb ◊ ◊ ◊ CH(CF3)
s
NHb ◊ ◊ ◊ Hb Pro (pro S)
NHb ◊ ◊ ◊ Hb Pro(pro R)
NHb ◊ ◊ ◊ NHc
m
m
m
w
NHb ◊ ◊ ◊ Hd (pro S)
NHb ◊ ◊ ◊ Hd (pro R)
NHa ◊ ◊ ◊ Ha Val
m
NHa ◊ ◊ ◊ Ha Pro
wb
wb
NHa ◊ ◊ ◊ Hd (pro S)
NHa ◊ ◊ ◊ Hd (pro R)
NHa ◊ ◊ ◊ terminal CH3
Ha Val ◊ ◊ ◊ . Ha Pro
s
w
s
Ha Val ◊ ◊ ◊ Hd Pro(pro S)
Ha Val ◊ ◊ ◊ Hd Pro (pro R)
Ha Val ◊ ◊ ◊ Hb Val
s
s
Ha Val ◊ ◊ ◊ gCH3 Val(pro S)
Ha Val ◊ ◊ ◊ gCH3 Val (pro R)
Ha Pro ◊ ◊ ◊ Hd Pro (pro S)
Ha Pro ◊ ◊ ◊ Hd Pro (pro R)
Ha Pro ◊ ◊ ◊ Ha Gly (pro S)
Ha Pro ◊ ◊ ◊ Ha Gly (pro R)
Ha Pro ◊ ◊ ◊ Hb Pro (pro S)
Ha Pro ◊ ◊ ◊ Hb Pro (pro R)
Ha Pro ◊ ◊ ◊ Hg (pro S)
Ha Pro ◊ ◊ ◊ Hg (pro R)
Ha Pro ◊ ◊ ◊ NHc
m
w
s
s
w
s
s
s
m
s
s
s
Hd,d’ ◊ ◊ ◊ Hb, b’ Pro
Hd,d’ ◊ ◊ ◊ Hg, g’ Pro
Ha Leu ◊ ◊ ◊ CH(CF3)
Ha Leu ◊ ◊ ◊ NCH3 terminals
Ha Leu ◊ ◊ ◊ NHc
Ha Leu ◊ ◊ ◊ Hg Leu
Ha Leu ◊ ◊ ◊ Hb Leu (pro S)
Ha Leu ◊ ◊ ◊ Hb Leu(pro R)
CH(CF3) ◊ ◊ ◊ Ha Gly (pro S)
CH(CF3) ◊ ◊ ◊ Ha Gly (pro R)
CH(CF3) ◊ ◊ ◊ Hb Leu (pro S)
CH(CF3) ◊ ◊ ◊ Hb Leu (pro R)
CH(CF3) ◊ ◊ ◊ NHc
s
m
w
s
m
s
s
NCH3 terminals ◊ ◊ ◊ NHc
NHc ◊ ◊ ◊ Hg Leu
NHc ◊ ◊ ◊ Hb Leu (pro S)
NHc ◊ ◊ ◊ Hb Leu (pro R)
NHc ◊ ◊ ◊ dCH3 Leu (pro S)
NHc ◊ ◊ ◊ dCH3 Leu (pro R)
NHc ◊ ◊ ◊ Ha Gly (pro S)
NHc ◊ ◊ ◊ Ha Gly (pro R)
Hb Val ◊ ◊ ◊ g, g’ NCH3
NHa ◊ ◊ ◊ NCH3 terminal
NHa ◊ ◊ ◊ NHc
m
m
s
m
s
3.39
2.13, 2.40
5.88
3.54
2.16, 2.16
4.28
3.42
2.16, 2.16
5.58
wc
wc
6.50
4.06
7.05
a Cross peaks intensities referred to the compounds 11, 12 and 13. b Cross peaks intensities referred to 11 and 13. c Cross peaks intensities referred to 12.
On the basis of these interactions the MD calculations showed
structures forming a 10 membered-ring with a hydrogen bond
Conclusions
The design and synthesis of metabolically stable peptide analogs
that can either mimic or block the bioactivity of natural peptides or
enzymes is an important constituent of bioorganic and medicinal
chemistry. Isosteric replacement of a scissile peptide bond with
surrogates able to maintain the “favorable” features of the latter
between the aminic proton NHc of L-Leu and the CO of L-Val
˚
(1.94 A) and a second hydrogen bond between NHa and the CO of
˚
L-Leu (2.55 A, Fig. 10). This evidence confirmed that the change
of stereochemistry at the Pro stereocenter induces a natural folding
towards a b hairpin conformation.
2292 | Org. Biomol. Chem., 2009, 7, 2286–2296
This journal is
The Royal Society of Chemistry 2009
©

1
was performed with silica gel 60 (60–200 mm, Merck). H-, 13C-,
and 19F-NMR spectra were run at 250, 400 or 500 MHz. Chemical
shifts are expressed in ppm (d), using tetramethylsilane (TMS) as
internal standard for 1H and 13C nuclei (dH and dC = 0.00), while
C6F6 was used as external standard (dF -162.90) for 19F.
Experimental for the synthesis of nitropropene 2
See ref. 14.
Experimental for the aza-Michael reaction
Synthesis of 4,5d is described as an example. To a stirred solution
of 2 (0.76 mmol, 107 mg) and 3d (0.51 mmol, 92 mg) in toluene
(7 ml) at rt, DIPEA was added (0.56 mmol, 73 ml). After half an
hour at rt, the solvent was removed in vacuo, the crude material
was dissolved in EtOAc and washed once with 1 N HCl. The
organic layer was dried over anhydrous Na2SO4. The solvent was
removed in vacuo, and the crude product was purified by FC
(hexane/diisopropyl ether 9:1) affording 106 mg (72%) of the two
pure diastereoisomers 4d and 5d in a 7.5:1 ratio.
Fig. 10 Conformation of 12 obtained from molecular modeling.
represents a viable and popular approach in the rational design
of peptidomimetics. In this sense, we introduced recently the
trifluoroethylamino function that proved to be an excellent peptide
bond surrogate because: 1) it has a better metabolic stability; 2) it
can behave as a sort of hybrid between a peptide bond mimic and
a proteolytic transition state analogue; 3) the stereo-electronic fea-
tures of the trifluoromethyl group are responsible for an increased
tendency of PMR-W[NHCH(CF3)]Gly-peptides to assume turn-
like conformations in organic solvents. To investigate further the
latter issue, several W[CH(CF3)NH]Gly tetrapeptides containing
the trifluoroethylamino function in a Pro-Gly loop have been
synthesized and their conformation has been investigated by 1D
and 2D NMR experiments and MD calculations. Independently
from the stereochemistry of the sterogenic trifluoroethylamino
function these tetrapeptides were found to exist in well defined
secondary structures stabilized by an intramolecular hydrogen
bond involving the aminic proton of the surrogate. These findings
are important because one of the most significant differences
between the peptide bond and the trifluoroethylamine surrogate
is the nature of the hydrogen atom which is amidic in the former,
and aminic in the latter. However, we have shown herein that
due to the presence of the strongly electron-withdrawing a-CF3
group, the NH moiety of the trifluoroethylamine function is
functionally very close to the amidic NH. Indeed, it can act
intramolecularly as a hydrogen bond donor like an amidic NH,
thus inducing the formation of stable secondary structures in small
peptidomimetics. All the experimental evidence described here and
in previous reports confirm that the trifluoroathylamino function
is a promising peptide- and amide-bond replacement that will
likely find further application in medicinal chemistry and drug
discovery.
4d. Rf = 0.31, (n-Hex/iPr2O 7:3); [a]23 = -16.2◦ (c = 1.1,
D
CHCl3); FT IR (film): nmax = 3365, 2967, 2930, 1732, 1567,
1379 cm-1; 1H NMR (500 MHz, CDCl3): d = 4.67 (dd, J = 13.1,
4.7 Hz, 1H), 4.52 (dd, J = 13.1, 7.8 Hz, 1H), 3.86 (m, 1H), 3.68
(s, 3H), 3.32 (d, J = 5.13 Hz, 1H), 1.91 (br s, 1H), 1.71 (m, 1H),
1.41(m, 1H), 1.14 (m, 1H), 0.91 (d, J = 6.6 Hz, 3H), 0.86 (t, J = 7.5,
3H); 19F NMR (470.6 MHz, CDCl3): d = -75.9 (d, J = 7.0 Hz);
13C NMR (62.9 MHz, CDCl3): d = 174.5, 124.5 (q, J = 282.9 Hz),
74.4, 65.8, 58.2 (q, J = 29.0 Hz), 52.0, 38.6, 24.6, 15.7, 11.4; MS
(70 eV): e/z (%): 287 [M+ + 1] (3), 227 (100).
5d. Rf = 0.41 (n-Hex/iPr2O 7:3); 19F NMR (470.6 MHz,
CDCl3): d = -75.9 (d, J = 7.0 Hz); [a]23
= +34.1◦ (c =
D
1.8, CHCl3); FT IR (film): nmax = 3349, 2967, 1735, 1561, 1382,
1156 cm-1; 1H NMR (500 MHz, CDCl3): d = 4.60 (dd, J = 12.8,
3.7 Hz, 1H), 4.40 (dd, J = 12.8, 10.3 Hz, 1H), 3.96 (m, 1H), 3.72
(s, 3H), 3.22 (d, 5.1 Hz, 1H), 2.19 (br s, 1H), 1.60 (m, 1H), 1.39
(m, 1H), 1.11 (m, 1H), 0.86 (d, J = 6.9 Hz, 3H), 0.83 (t, J = 7.5,
3H); 19F NMR (470.6 MHz, CDCl3): d = -75.1 (d, J = 7.0 Hz);
13C NMR (62.9 MHz, CDCl3): d = 174.2, 124.7 (q, J = 283.2 Hz),
74.8, 65.3, 59.2 (q, J = 29.7 Hz), 52.0, 39.5, 24.5, 15.4, 11.4.
Typical procedure for the synthesis of
W[CH(CF3)NH]Gly-tetrapetides
Synthesis of 8c is described as an example. A solution of 4c
(0.69 mmol, 197 mg) and 1 N HCl (0.69 mmol, 690 ml), in
MeOH (7 ml), in the presence of a catalytic amount of palladium
hydroxide on carbon, was stirred at rt for five hours, under a
hydrogen atmosphere. Then, the mixture was filtered on a Celite
pad, and the solvent removed in vacuo. The crude material was
dissolved in dry DMF (5 ml) and Cbz-L-Phe-OH, (0.90 mmol,
258 mg), sym collidine (2.07 mmol, 274 ml), followed by solid HOAt
(0.69 mmol, 94 mg) and solid HATU (0.69 mmol, 262 mg) were
added at rt. The mixture was stirred overnight, quenched with 1 N
HCl, and extracted with Et2O. The organic layer was washed once
with water and then dried on Na2SO4. The solvent was removed
in vacuo, and the crude purified by FC (hexane/AcOEt 70:30)
affording 270 mg of 8c (73%).
Experimental
General details
Commercially available reagent-grade solvents were employed
without purification. All reactions where an organic solvent was
employed were performed under a nitrogen atmosphere, after
flame-drying of the glass apparatus. Melting points (m.p.) are
uncorrected and were obtained on a capillary apparatus. TLC
was run on silica gel 60 F254 Merck. Flash chromatography (FC)
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2286–2296 | 2293
©

8c. [a]23 = -12.4◦ (c = 1.9, CH3COCH3); FT IR (micro-
Typical procedure for the coupling with Boc-L,DPro-OH and
Boc-Val-OH
D
scope): nmax = 3319, 2980, 2930, 1727, 1691, 1661, 1543, 1370 cm-1;
1H NMR: (250 MHz, CDCl3): d = 7.40–7.14 (m, 11H), 5.56 (br
d, J = 7.6 Hz, 1H), 5.13 (d, J = 12.25 Hz, 1H), 5.00 (d, J =
12.25 Hz, 1H), 4.60 (m, 1H), 3.39 (m, 3H), 3.14 (m, 2H), 2.90 (m,
1H), 1.50 (br s, 1H) 1.40 (s, 9H), 1.22 (d, J = 7.2 Hz, 3H); 19F
NMR: (235.4 MHz, CDCl3): d = -77.3 (d, J = 7.2 Hz); 13C NMR
(62.9 MHz, CDCl3): d = 176.0, 171.6, 156.0, 136.5, 136.3, 129.5,
128.6, 128.5, 128.1, 127.0, 125,5 (q, J = 282.7 Hz), 82.3, 67.0, 60.4,
58.0 (q, J = 29.6 Hz), 56.1, 38.6, 28.0, 19.4, 14.2; MS (70 eV): e/z
(%): 538 [M+ + 1] (1), 436 (20), 91 (100).
Synthesis of 18 is described as an example. The nitro group of
pure diasteroisomer 17a was hydrogenated as described above.
The crude material obtained after the hydrogenation (2.02 mmol,
616 mg) was dissolved in dry DMF (7 ml) and Boc-LPro-OH
(2.22 mmol, 477 mg), sym collidine (1.3 ml), followed by solid
HOBT (2.22 mmol, 299 mg) and solid EDC (2.22 mmol, 425 mg)
were added at 0 ◦C. The mixture was stirred overnight at room
temperature, quenched with HCl 1 N and extracted with ethyl
acetate. The organic layer was washed once with water, once with
HCl 1 N, once with sodium hydrogencarbonate and finally once
with brine. The organic layer was dried on Na2SO4 and filtered;
the solvent was removed in vacuo, and the crude product purified
by FC (AcOEt:-hexane 6:4), affording 692 mg of 18 (75%).
Procedure for the synthesis of 17a,b
To a solution of Cbz-LLeu-OH 15 (5.76 mmol, 1.53 g) in dry DMF
(15 ml), BOP (5.76 mmol, 2.55 g), dimethylamine hydrochloride
(5.76 mmol, 0.47 g) followed by DIPEA (17.28 mmol, 2.9 ml) were
added at rt and the solution stirred for two hours. 1 N HCl was
added, the mixture extracted with AcOEt and the organic phase
washed with saturated NaHCO3 and brine. The organic phase
was dried on Na2SO4, filtered and the solvent evaporated under
reduced pressure. The crude was purified by FC (hexane:AcOEt
1:1) affording 1.43 g (85%) of 16. A solution of 16 (4.90 mmol,
1.43 g) and 1 N HCl (4.90 mmol, 4.9 ml), in MeOH (30 ml), in the
presence of a catalytic amount of palladium hydroxide on carbon,
was stirred at rt for three hours, under a hydrogen atmosphere.
Then, the mixture was filtered on a Celite pad, and the solvent
removed in vacuo. The crude material obtained was dissolved in
DCM (29 ml) and a solution of 2 (4.90 mmol, 691 mg) in DCM
(2 ml) followed by an excess of NaHCO3 were added and the
mixture left to stir at rt for two hours. Water was added and the
mixture extracted with DCM, the collected organic layers dried
on Na2SO4, filtered and the solvent evaporated under reduced
pressure. The crude product was purified by FC (toluene:AcOEt
5:1) affording 1.2 g (3.9 mmol, 80% yield) of a 1:1 mixture of 17a
and 17b.
18. Rf : 0.41 AcOEt/n-Hexane 7:3 [a]23 = -44.5 (c = 1.1,
D
CHCl3); FT IR (film): nmax = 3380, 1632, 1529, 1471, 1114 cm-1;
1HNMR (400 MHz, CDCl3): d = 7.99 (br s, 1H), 4.15 (br s, 1H),
3.86 (m, 1H), 3.44 (m, 4H), 3.23 (m, 1H), 3.10 (s, 3H), 2.94 (s, 3H),
2.21 (m, 1H), 1.89 (m, 4H), 1.41 (s, 9H), 1.33 (m, 2H), 0.95 (m,
6H); 13C NMR (125.7 MHz, CDCl3): d = 176.7, 176.2, 156.1, 127.4
(q, J = 283.5 Hz), 81.5, 62.1, 59.9 (q, J = 26.1 Hz), 56.1, 44.3,
39.7, 37.3, 36.2, 32.4, 31.4, 28.7, 25.8, 25.4, 24.7, 23.8, 22.2; 19F
NMR (235.4 MHz, CDCl3): d = -75.0, J = 7.5 Hz; MS (70 eV):
e/z (%): 467[M+ + 1] (20), 367 (35), 294 (55), 70 (100), 57 (50), 43
(20).
18. (1.03 mmol, 480 mg) was dissolved in a 20% solution
of TFA in DCM, stirred for an hour at room temperature, and
the solvent was removed in vacuo. The residue was dissolved
in a minimal amount of AcOEt, washed with a 5% aqueous
solution of NaHCO3 5%, the organic layer dried on Na2SO4,
filtered and the solvent evaporated under reduced pressure. The
crude material obtained was coupled with Boc-LVal-OH following
the HOBt/EDC protocol described above affording 367 mg of
tetrapeptide 19 (63%).
19. Rf : 0.48 (AcOEt/n-Hexane 4:1); [a]23 = -39.0◦ (c =
17a. Rf : 0.46 Toluene/AcOEt 4:1; [a]23 = -29.3◦ (c = 0.9,
D
D
2.3, CHCl3); FT IR (film): nmax = 3440, 1637, 1501, 1441, 1394,
1164 cm-1; 1HNMR (400 MHz, CDCl3): d = 7.02 (br s, 1H), 5.18
(br d, J = 8.6 Hz,1H), 4.46 (m, 1H), 4.26 (br t, J = 7.3 Hz), 3.72
(m, 2H), 3.56 (m, 2H), 3.17 (m, 2H), 3.02 (s, 3H), 2.94 (s, 3H), 2.21
(m, 2H), 2.07 (m, 1H), 1.96 (m, 3H), 1.86 (m, 1H), 1.41 (s, 9H),
CHCl3); FT IR (film): nmax = 1650, 1582 cm-1; 1HNMR (400 MHz,
CDCl3): d= 4.61 (m, 2H), 3.78 (m, 1H), 3.67 (m, 1H), 3.00 (s, 3H),
2.93 (s, 3H), 2.19 (m, 1H), 1.98 (m, 1H), 1.34 (m, 1H), 1.20 (m,
1H), 0.91 (m, 6H),; 19F NMR (235.4 MHz, CDCl3): d = -76.5 (d,
J = 7.5 Hz); 13C NMR (125.7 MHz, CDCl3): d = 174.4, 124.7 (q,
J = 281.4 Hz), 74.5, 58.5 (q, J = 30.5 Hz), 57.3, 43.1, 36.4, 36.0,
24.4, 23.6, 21.2; MS (70 eV): e/z (%): 322 [M+ + Na] (100), 300
[M + 1] (31), 159 (97).
1.36 (m, 1H), 1.24 (m, 1H), 0.97 (d, J = 7.3 Hz), 0.91 (m, 9H); 13
C
NMR (100 MHz, CDCl3): d = 174.8, 172.3, 171.6, 155.9, 125.9 (q,
J = 284.4 Hz), 79.5, 60.2, 58.4 (q, J = 26.1 Hz), 56.9, 53.9, 47.6,
43.5, 37.8, 36.6, 35.9, 31.4, 28.3, 27.9, 25.1, 24.6, 23.5, 21.8, 19.4,
17.3; 19F NMR (470.6 MHz, CDCl3): d = 74.7 (d, J = 7.7 Hz); MS
(70 eV): e/z (%): 566 [M+ + 1] (20), 466 (10), 294 (25), 70 (100), 57
(70), 43 (20).
17b. Rf : 0.40 Toluene/AcOEt 4:1; [a]23 = -12.8◦ (c = 2.9,
D
CHCl3); FT IR (film): nmax = 1643, 1563 cm-1; 1HNMR (250 MHz,
CDCl3): d= 4.60 (dd, J = 13.1, 3.3 Hz, 1H), 4.42 (dd, J = 13.1,
9.7 Hz, 1H), 3.94 (m, 1H), 3.58 (dd, J = 10.9, 2.7 Hz, 1H), 2.99
(s, 3H), 2.96 (s, 3H), 2.80 (br s, 1H), 1.78 (m, 1H), 1.36 (ddd, J =
13.6, 10.9, 3.6 Hz, 1H), 1.11 (ddd, J = 13.6, 9.7, 3.1 Hz, 1H),
0.89 (m, 6H); 19F NMR (235.4 MHz, CDCl3): d = -75.1 (d, J =
7.5 Hz); 13C NMR (125.7 MHz, CDCl3): d = 174.1, 124.7 (q, J =
283.5 Hz), 74.9, 59.3 (q, J = 26.1 Hz), 56.0, 44.0, 36.5, 35.8, 24.0,
23.6, 21.1; MS (70 eV): e/z (%): 300 [M + 1] (15), 227 (80), 86 (55),
43 (100).
Typical procedure for the synthesis of final
Ac-LVal-L,DPro-W[CH(CF3)NH]Gly-LLeu-NMe2 tetrapetides
Synthesis of 11 is described as example. 19 was Boc-deprotected
and the resulting trifluoroacetate salt neutralized following the
procedure described above. The crude (0.12 mmol, 55 mg) was
dissolved in dry DCM and TEA (36 ml) was added at 0 ◦C.
After two minutes Ac2O (12 ml) was added and the mixture was
2294 | Org. Biomol. Chem., 2009, 7, 2286–2296
This journal is
The Royal Society of Chemistry 2009
©

stirred for one hour at 0 ◦C. The solution was washed once with
HCl 1 N, twice with NaHCO3 5% and then dried on Na2SO4.
The solvent was removed in vacuo and the crude purified by FC
(CHCl3/MeOH 98:2), affording 54 mg (89%) of 11.
MD was further minimized. An average structure was created for
each MD simulation.
Acknowledgements
11. [a]23D = -62.2 (c = 2.1, CHCl3); FT IR (film): nmax = 3325,
Politecnico di Milano and CNR are gratefully acknowledged for
economic support
1
2961, 2244, 1634, 1539, 1446, 1390 cm-1; HNMR (400 MHz,
CDCl3): d = 7.92 (br s, 1H), 6.15 (br d, J = 5.0 Hz, 1H), 4.65 (dd,
J = 8.9,6.3 Hz, 1H), 4.50 (m, 1H), 3.83 (dd, J = 10.9, 2.7 Hz, 1H),
3.78 (m, 1H), 3.70 (m, 1H), 3.51 (m, 1H), 3.34 (m, 1H), 3.01 (s,
3H) 2.98 (s, 3H), 2.10 (m, 10H), 1.45 (ddd, J = 13.7, 10.6, 3.9 Hz,
1H), 1.24 (ddd, J = 13.7, 10.0, 3.0 Hz, 1H), 0.93 (m, 13H); 13C
NMR (100 MHz, CDCl3): d = 175.6, 172.3, 171.2, 170.0, 125.7 (Q,
J = 281.8 Hz), 60.5, 58.5 (Q, J = 27.4 Hz), 56.3, 55.6, 47.7, 42.2,
36.7, 36.0, 31.4, 29.1, 25.1, 24.5, 23.6, 21.0, 19.3, 17.8; 19F NMR
(235.4 MHz, CDCl3): d = -76.8 (d, J = 5.0 Hz); MS (70 eV): e/z
(%): 508 [M+ + 1] (10), 367(30), 294(98), 70(100).
Notes and references
1 A. Loffet, J. Pept. Sci., 2002, 8, 1–7.
2 (a) G. L. Olson, D. R. Bolin, M. P. Bonner, M. Bo¨s, C. M. Cook, D.
C. Fry, B. J. Graves, M. Hatada, D. E. Hill, M. Kahn, V. S. Madison,
V. K. Rusiecki, R. Sarabu, J. Sepinwall, G. P. Vincent and M. E. Voss,
J. Med. Chem, 1993, 36, 3039–3049; (b) J. Gante, Angew. Chem., 1994,
106, 1780–1802; J. Gante, Angew. Chem. Int. Ed. Engl., 1994, 33, 1699–
1720; (c) D. Leung, G. Abbenante and D. P. Fairlie, J. Med. Chem.,
2000, 43, 305–341.
3 J.-L. Fauche`re and C. Thurieau, Adv. Drug Res., 1992, 23, 127–159.
4 A. F. Spatola, In Chemistry and Biochemistry of Amino Acids, Peptides
and Proteins, Vol 7 (Ed. B. Weinstein) Marcel Dekker, New York, 1983,
267-357.
5 (a) J. S. Morley, T. D. Hennessey and J. W. Payne, Biochem. Soc. Trans.,
1983, 11, 798–800; (b) A. B. Smith, III, R. Hirschmann, A. Pasternak,
M. C. Guzman, A. Yokoyama, P. A. Sprengeler, P. L. Darke, E. A.
Emini and W. A. Schleif, J. Am. Chem. Soc., 1995, 117, 11113–11123.
6 For a review on the trifluoethylamine unit, see: M. Sani, A. Volonterio
and M. Zanda, ChemMedChem., 2007, 2, 1693–1700.
NMR experiments
The NMR spectra were recorded on a Bruker AMX 600 spec-
trometer operating at a frequency of 600.13 MHz for 1H nucleus.
The chemical shifts (d) were measured in ppm and referenced
to external DSS signal for CDCl3 solution. Estimated accuracy
0.01 ppm. The experiments were performed at 25 ◦C. The
concentration of peptides was 4.5 mM. ROESY and COSYmqf
spectra were acquired in the phase sensitive TPPI mode, with
1K ¥ 512 complex FIDs, spectral width of 6329 Hz, recycling
delay of 1.2 s, 48 scans. The mixing times for ROESY experiments
ranged between 150 ms and 800 ms. TOCSY spectra were recorded
with the use of a MLEV-17 spin-lock pulse (field strength of
7550 Hz, 60 ms total duration, 2.5 ms of TRIM pulse). All
spectra were transformed and weighted with a 90◦ shifted sine-
bell squared function to 1K ¥ 1K real data points.13C NMR
spectra were obtained by means of one-bond and long range
correlation spectroscopy using HMQC and HMBC in the TPPI
mode. The NH temperature coefficients were obtained in CDCl3,
by monitoring the amide and amine NH chemical shifts over a
temperature range of 233 to 298 K. All changes in NH chemical
shifts (-Dd/DT) were linear over the above temperature range.
7 (a) W. C. Black, C. I. Bayly, D. E. Davies, S. Desmarais, J.-P. Falgueyret,
S. Le´ger, C. S. Li, F. Masse´, D. J. McKay, J. T. Palmer, M. D. Percival,
J. Robichaud, N. Tsou and R. Zamboni, Bioorg. Med. Chem. Lett.,
2005, 15, 4741–4744; (b) C. S. Li, D. Deschenes, S. Desmarais, J.-P.
Falgueyret, J. Y. Gauthier, D. B. Kimmel, S. Le´ger, F. Masse´, M. E.
McGrath, D. J. McKay, M. D. Percival, D. Riendeau, S. B. Rodan, M.
The´rien, V.-L. Truong, G. Wesolowski, R. Zamboni and W. C. Black,
Bioorg. Med. Chem. Lett., 2006, 16, 1985–1989; (c) W. C. Black and M.
D. Percival, ChemBioChem, 2006, 7, 1525–1535; (d) J. Y. Gauthier, N.
Chauret, W. Cromlish, S. Desmarais, L. T. Duong, J.-P. Falgueyret, D.
B. Kimmel, S. Lamontagne, S. Le´ger, T. LeRiche, C. Sing Li, F. Masse´,
D. J. McKay, D. A. Nicoll-Griffith, R. M. Oballa, J. T. Palmer, M. D.
Percival, D. Riendeau, J. Robichaud, G. A. Rodan, S. B. Rodan, C.
Seto, M. The´rien, V.-L. Truong, M. C. Venuti, G. Wesolowski, R. N.
Young, R. Zamboni and W. C. Black, Bioorg. Med. Chem. Lett., 2008,
18, 923–928; (e) P. D. O’Shea, C. Chen, D. Gauvreau, F. Gosselin, G.
Hughes, C. Nadeau and R. P. Volante, J. Org. Chem., 2009, in press,
http://pubs.acs.org/doi/pdfplus/10.1021/jo8020314.
8 J. D. Dunitz and R. Taylor, Chem. Eur. J., 1997, 3, 89–98.
9 (a) A. Volonterio, P. Bravo and M. Zanda, M., Org. Chem., 2000, 2,
1827–1830; (b) A. Volonterio, P. Bravo and M. Zanda, Tetrahedron
Lett., 2001, 42, 3141–3144; (c) A. Volonterio, S. Bellosta, P. Bravo, M.
Canavesi, E. Corradi, S. V. Meille, M. Monetti, N. Moussier and M.
Zanda, Eur. J. Org. Chem., 2002, 428–438; (d) M. Sani, P. Bravo, A.
Volonterio and M. Zanda, Colelct. Czech. Chem. Commun., 2002, 67,
1305–1319; (e) A. Volonterio, S. Bellosta, F. Bravin, M. C. Bellucci, L.
Bruche´, G. Colombo, L. Malpezzi, S. Mazzini, S. V. Meille, M. Meli,
C. Ramirez de Arellano and M. Zanda, M., Chem. Eur. J., 2003, 9,
4510–4522.
Molecular modeling
Molecular models were built using a Silicon Graphics 4D35GT
workstation running the Insight II & Discover software. Molecular
mechanics (MM) and molecular dynamics (MD) were carried out
using cvff force-fields. The starting geometry of the peptides was
generated usingstandard bond lengths and angles. The simulations
were performed in vacuo with relative permittivity e = 1.0. At
the first step we performed a minimization by Discover with
steepest-descendent algorithm followed by conjugate gradient
minimization for a maximum of 2000 iterations each or RMS
deviation of 0.001 Kcal/mol. Then we performed MD 100 ps
simulations at a constant temperature of 300 K with distance
constraints derived from the ROE cross peaks from the ROESY
spectra in CDCl3. Distance constraints with a force constant
10 M. Molteni, A. Volonterio and M. Zanda, Org. Lett., 2003, 5, 3887–
3890.
11 S. Bigotti, A. Volonterio and M. Zanda, Synlett, 2008, 958–962.
12 T. Billard, Chem. Eur. J., 2006, 12, 974–979.
13 For a review, see: F. A. Luzzio, Tetrahedron, 2001, 57, 915–945.
14 Nitro-alkene 2 was prepared starting from a commercial available
aqueous solution of fluoral hydrate: O. Klenz, R. Evers, R. Miethchen
and M. Michalik, J. Fluorine Chem., 1997, 81, 205–210See also:; M.
Molteni, R. Consonni, T. Giovenzana, L. Malpezzi and M. Zanda, J.
Fluorine Chem., 2006, 127, 901–908.
15 For a review of asymmetric conjugate additions to nitroalkenes, see: T.
A. Johnson, D. O. Jang, B. W. Slafer, M. D. Curtis and P. Beak, J. Am.
Chem. Soc., 2002, 124, 11689–11698 and references therein.
16 In a recent work concerning a tandem process involving an aza-Michael
reaction with fluorinated acceptors, we found that the polarity of the
˚
of 15 Kcal/A were applied with a range of lower and upper
˚
bound of 2.0–3.0, 3.0–4.0, 4.0–5.0 A in accordance with the
NOE intensities of strong, medium and weak, respectively. No
H-bonding constraints was used. Every structure obtained from
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2286–2296 | 2295
©

solvent as well as the structural features of the base govern remarkably
the diastereoselection of the process: M. Sani, L. Bruche´, G. Chiva,
S. Fustero, J. Piera, A. Volonterio and M. Zanda, Angew. Chem. Int.
Ed., 2003, 42, 2060–2063. However, it should be stressed that the two
processes are conceptually and mechanistically unrelated, since the
stereocontrol of the reaction described in the reference above arose
from an enolate protonation following the aza-Michael reaction, that
per se was not stereogenic.
Chem. Soc., 2005, 127, 4193–4198; (c) S. T. Phillips, G. Piersanti and P.
A. Bartlett, Proc. Natl. Acad. Sci. USA, 2005, 102, 13737–13742.
27 C. Schmuck, M. Heil, J. Scheiber and K. Baumann, Angew. Chem. Int.
Ed. Engl., 2005, 44, 7208–7212.
28 Y. Fu, J. Bieschke and J. W. Kelly, J. Am. Chem. Soc., 2005, 127, 15366–
15367.
29 (a) A. Aemissegger, V. Krautler, W. F. van Gunsteren and D. Hilvert, J.
Am. Chem. Soc., 2005, 127, 2929–2936; (b) V. Krautler, A. Aemissegger,
P. H. Hunenberger, D. Hilvert, T. Hansson and W. F. van Gunsteren,
J. Am. Chem. Soc., 2005, 127, 4935–4942; (c) S. L. Dong, M. Lowenec,
T. E. Schrader, W. J. Schreier, W. Zinth, L. Moroder and C. Renner, C.,
Chem. Eur. J., 2006, 12, 1114–1120.
17 a-Amino acid ester hydrochloride 3b was preliminarly neutralized by
treating with an aqueous solution of NaHCO3.
18 S. Bigotti, S. V. Meille, A. Volonterio and M. Zanda, J. Fluorine Chem.,
2008, 129, 767–774.
19 S. Fustero, G. Chiva, J. Piera, A. Volonterio, M. Zanda, J. Gonzalez
and A. Moran Ramallal, Chem. Eur. J., 2007, 13, 8530–8542.
20 For examples of conjugate additions to nitroolefins where a hydrogen
bond involving a nitro group is proposed in the transition state, see:
(a) D. Enders and A. Seki, Synlett, 2002, 26–28; (b) W. Wang, J. Wang
and H. Li, Angew. Chem. Int. Ed., 2005, 44, 1369–1371; (c) M. Wiesner,
J. D. Revell and H. Wennemers, Angew. Chem. Int. Ed., 2008, 47, 1871–
1874.
30 G. D. Rose, L. M. Gierasch and J. A. Smith, Adv. Protein Chem., 1985,
37, 1.
31 (a) T. S. Haque, J. C. Little and S. H. Gellman, J. Am. Chem. Soc., 1996,
118, 6975–6985; (b) J. D. Fisk, D. R. Powell and S. H. Gellman, J. Am.
Chem. Soc., 2000, 122, 5443–5447.
32 The stereochemistry of adducts 17a,b was determined by chemical
correlation with major diastereoisomer 4f as depicted below.
21 For an overview on the capacity of F atoms to undertake polar
interactions, rather than hydrogen bonds,see: K. Mu¨ller, C. Faeh and
F. Dietrich, Science, 317, 1881–1886and references therein.
22 Spectroscopic data revealed the presence of only one out of the two
possible isomers. Thermodynamically more stable trans geometry was
confidently assigned to 6 in analogy with the nitro-alkene 2.
23 In order to obtain a-substituted nitro-compounds like 7, we also tried
to alkylate compound 4e in the presence of different bases such as NaH,
tBuOK, Cs2CO3, NH4OAc. In all cases, we obtained the corresponding
oxime 20 derived from deoxygenation of the starting material. For a
related reaction, see: C. Czekelius and E. M. Carreira, Angew. Chem.
Int. Ed., 2005, 44, 612–615..
33 M. Rance, O. W. Sørensen, G. Bodenhausen, G. Wagner, R. R. Ernst
and K. Wu¨thrich, Biochem. Biophys. Res. Commun., 1983, 117, 479–
485.
34 L. Braunschweiler and R. R. Ernst, J. Magn. Reson, 1983, 53, 521.
35 A. A. Bothner-By, R. L. Stephensen and J. Lee, J.Am.Chem. Soc, 1984,
106, 811.
36 A. Bax, R. H. Griffey and B. L. Hawkins, J. Magn. Reson., 1983, 55,
301.
24 For a recent review concerning model systems for b-hairpins and b-
sheets, see: R. M. Huges and M. L. Waters, Curr. Opin. Chem. Biol.,
2006, 16, 514–524.
25 S. H. Gellman, Curr. Opin. Chem. Biol., 1998, 2, 717–725.
26 (a) S. T. Phillips, L. K. Blasdel and P. A. Bartlett, J. Org. Chem., 2005,
70, 1865–1871; (b) S. T. Phillips, L. K. Blasdel and P. A. Bartlett, J. Am.
37 It is noteworthy that for PMR-W[NHCH(CF3)]Gly peptides, secondary
structures were stabilized by the presence of the bulky CF3 group rather
than the presence of an intramolecular hydrogen bond involving the
aminic proton of our surrogate which was never observed (see ref. 9e).
38 A. Torganis, I. P. Gerothanassis, Z. Athanassiou, T. Mavromoustakos,
G. E. Hawkes and C. Sakarellos, Biopolymers, 2000, 53, 72–83.
2296 | Org. Biomol. Chem., 2009, 7, 2286–2296
This journal is
The Royal Society of Chemistry 2009
©