Pseudopeptide foldamers - The homo-oligomers of benzyl (4S,5R)-5-methyl-2-oxo-1,3-oxazolidine-4-carboxylate

Article abstract of DOI:10.1002/ejoc.200390027

A 2-oxo-1,3-oxazolidine-4-carboxylic acid was designed as a new, conformationally restricted building block for the construction of pseudopeptide foldamers. IR, ¹H NMR and CD techniques, implemented by detailed DFT computational modeling, were

Full text of DOI:10.1002/ejoc.200390027

FULL PAPER
Pseudopeptide Foldamers ؊ The Homo-Oligomers of Benzyl (4S,5R)-5-
Methyl-2-oxo-1,3-oxazolidine-4-carboxylate
Claudia Tomasini,[a] Valerio Trigari,[a] Simone Lucarini,[a] Fernando Bernardi,[a]
Marco Garavelli,*[a] Cristina Peggion,[b] Fernando Formaggio,[b] and Claudio Toniolo*[b]
Keywords: Conformation analysis / Density functional calculations / Foldamers / NMR spectroscopy / Oxazolidin-2-ones
A 2-oxo-1,3-oxazolidine-4-carboxylic acid was designed as a
new, conformationally restricted building block for the con-
conformation was found to be stabilized by intramolecular α-
C−H···O=C hydrogen bonds. This novel, acylurethane-based,
struction of pseudopeptide foldamers. IR, 1H NMR and CD ternary foldameric structure, if appropriately functionalized,
techniques, implemented by detailed DFT computational
modeling, were exploited to investigate the preferred three-
dimensional structure of benzyl (4S,5R)-5-methyl-2-oxo-1,3-
oxazolidine-4-carboxylate homo-oligomers synthesized to
holds promise as a robust template for a variety of applica-
tions.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
the pentamer level. The resulting poly(
L-Pro)n II like helical
Introduction
Molecules of a class of unnatural, synthetic oligomers
have been recently defined as foldamers by Gellman.[1] The
essential requirement for an oligomer to be included in the
foldamer family is to possess a well-defined, repetitive sec-
ondary structure, imparted by conformational restrictions
of the monomeric unit. In this area of research we described
a general method for the synthesis of 2-oxo-oxazolidine-4-
carboxylates.[2] These compounds contain a cyclic urethane
moiety, so that by coupling it with a carboxylic acid derivat-
ive an imido-type (acylurethane) function is obtained. This
latter group is characterized by a nitrogen atom connected
to an endocyclic and an exocyclic carbonyl group which
tend to adopt a trans disposition.[3] As a consequence of
this local rigidity, these oligomers will presumably be forced
to fold in an ordered conformation. As an initial step to
test this hypothesis, we have recently reported the synthesis
of very short homo-oligomers of benzyl (4S,5R)-5-methyl-
2-oxo-1,3-oxazolidine-4-carboxylate (-Oxd) (Figure 1).[4]
Here we present the synthesis of an expanded set of these
pseudopeptide homo-oligomers (to the pentamer level)
complemented by an extended DFT computational investi-
gation and a solution conformational analysis. We have ex-
Figure 1. Chemical formula of -Pro, -Oxd, and -pGlu; the rotat-
able bond is highlighted in each formula
perimentally demonstrated below that these oligo(acylure-
thanes) are real foldamers with a remarkable conforma-
tional preference for a poly(-Pro)n II type helix[5] and that
this 3D disposition can be predicted by computational ana-
lysis.
Results and Discussion
Synthesis and Characterization
Oxazolidin-2-ones are common heterocycles, which can
be easily obtained by reaction of vicinal amino alcohols
with a synthetic equivalent of carbon dioxide such as phos-
gene.[6] These molecules can be cleaved both under strong
acidic and under strong basic conditions, but are stable
enough to be handled under mild conditions. Enantiomer-
ically pure 4-substituted or 4,5-disubstituted oxazolidin-2-
ones have been extensively used as chiral auxiliaries[7] for a
wide range of reactions. Indeed, after acylation of the nitro-
gen atom, the side chain can be functionalized under a vari-
ety of conditions in order to create new stereogenic centers
with high stereoselectivity; e.g. alkylations,[8] and aldol[9]
and pericyclic reactions[10] have been successfully studied.
[a]
Department of Chemistry ‘‘G. Ciamician’’, Alma Mater
Studiorum, University of Bologna,
40126 Bologna, Italy
Fax: (internat.) ϩ 39-051/209-9456
E-mail: mgara@ciam.unibo.it
[b]
Institute of Biomolecular Chemistry, C.N.R., Department of
Organic Chemistry, University of Padova,
35131 Padova, Italy
Fax: (internat.) ϩ 39-049/827-5239
E-mail: claudio.toniolo@unipd.it
Eur. J. Org. Chem. 2003, 259Ϫ267  2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim 1434Ϫ193X/03/0102Ϫ0259 $ 20.00ϩ.50/0
259

M. Garavelli, C. Toniolo et al.
FULL PAPER
The preparation of benzyl (4S,5R)-5-methyl-2-oxo-1,3-
oxazolidine-4-carboxylate (H--Oxd-OBn) (1) (Scheme 1)
can be performed through two different ways: path A rep-
resents a general synthetic method that allows one to obtain
any 5-substituted benzyl (4S,5R)-2-oxo-1,3-oxazolidine-4-
carboxylate from an enantiomerically pure β-amino acid,[2]
while path B is a straightforward method from L-threon-
ine,[11] which allows one to synthesize efficiently only the 5-
substituted oxazolidin-2-ones which can be obtained from
commercially available amino acids. In this work we have
exploited the latter method to synthesize 1 on a multigram
scale.
Scheme 1. Synthetic routes for H--Oxd-OBn (1); reagents and con-
ditions: (a) LiHMDS (2 equiv.), dry THF (Ϫ60 °C); (b) I2 (1.2
equiv.), dry THF, (Ϫ60 °C); (c) Sn(OTf)2, dry CH2Cl2; (d) 1 
NaOH, DMF; (e) BnBr (1.1 equiv.), TEA, acetone
Scheme 2. Synthetic routes for the (-Oxd)n homo-oligomers; re-
agents and conditions: (a) DIEA (4 equiv.), DMAP (0.1 equiv.),
dry DMF, room temp., 16 h; (b) H2, Pd/C (10%), EtOAc, room
temp., 1 h; (c) CF3CO2C6F5 (1.25 equiv.), pyridine (1.1 equiv.), dry
DMF, room temp., 1 h
The acylation step is not straightforward. It was origin-
ally achieved by formation of the lithium salt of the oxazoli-
din-2-one with n-butyllithium, followed by addition of the
acid chloride.[12] Recently, a different procedure was envis-
aged to eliminate the use of n-butyllithium; the oxazolidin-
2-one may be efficiently acylated with a mixed anhydride
using triethylamine and a slight molar excess of lithium
chloride.[13] In our case we had to perform a homo-oligo-
merization, so that the acylation source was the 2-oxo-
oxazolidine-4-carboxylate itself. As the corresponding acyl
chloride is unstable, Boc--Oxd-OBn (Boc: tert-butyloxy-
carbonyl) was transformed into the corresponding pen-
tafluorophenyl ester Boc--Oxd-OPfp (2) under mild condi-
tions.[14] The coupling reaction was performed by treating
H--Oxd-OBn (1) with 2 in the presence of diisopropylethy-
lamine (DIEA) and 4-(dimethylamino)pyridine (DMAP)
(Scheme 2). According to the same procedure, homo-oligo-
mers up to the pentamer level were synthesized.
of conformational minima for the (-pGlu)n foldameric
structures (Figure 1). Therefore, we carried out our calcula-
tions on these systems at a fully correlated level using den-
sity functional theory (DFT) with both a well-tested func-
tional and a basis set.
A preliminary 1H NMR characterization of dimer 3,
trimer 6, and tetramer 9 of the Boc-(-Oxd)n-OBn series,
reported by us in a previous paper,[4] indicated that a homo-
oligomer composed by n -Oxd monomers has n Ϫ 1 down-
field shifted α-CH proton signals. These anomalous chem-
ical shift values were tentatively associated with the occur-
rence of weak α-CϪH···OϭC hydrogen bonds.
To determine the conformational minima for the (-
Oxd)n oligomers and to obtain a more accurate explanation
of their magnetic properties, in this work we carried out
fully unconstrained DFT optimizations for dimer 3, trimer
6, tetramer 9 and pentamer 12. The starting structures used
for DFT optimization and refinement of dimer 3 and trimer
DFT Computational Modeling
Molecular mechanics and molecular dynamics are the 6 were initially located by a preliminary AM1 optimized
conventional tools used for modeling oligomers and for surface scanning about the relevant free rotatable dihedral
conformational search and analysis. These are very conveni- angles of the oligomeric skeleton: N1ϪC1AϪC1ϪN2 for
ent approaches when the size of the molecular system under the dimer, and N1ϪC1AϪC1ϪN2 and N2ϪC2AϪC2ϪN3
investigation is progressively increased; however, a heavy for the trimer. Note that the imide bonds are not free to
dependence of the reliability of the results obtained on the rotate. Subsequently, the DFT-optimized structures for
quality of the force field used was often observed. This dimer 3 and trimer 6 were exploited as templates to gener-
drawback is particularly crucial in the description of weak ate the starting geometries for the DFT optimizations of
hydrogen bonds which can be involved in the stabilization the longer analogues 9 and 12.
of complex molecular architectures. In a previous work[15]
For dimer 3 the optimized bond lengths and angles be-
we showed that weak intramolecular hydrogen bonds (such tween the atoms C1A, H1A, O2D and C2D involved in the
as CϪH···OϭC) play an important role in the stabilization hypothetical α-CϪH···OϭC interaction have the following
260
Eur. J. Org. Chem. 2003, 259Ϫ267

Pseudopeptide Foldamers
FULL PAPER
˚
values: the optimized H1AϪO2D bond length is 2.34 A,
To further validate the optimized structure for dimer 3,
while the optimized H1AϪO2DϪC2D bond angle and the 1H NMR chemical shift (δ) values for its H1A and H2A
C1AϪH1AϪO2DϪC2D dihedral angle are 97.7 and 52.0°, protons were simulated[17,18] and compared to those ob-
respectively. These data are in agreement with those re-
ported in the literature for related systems; in particular, the
mean hydrogen bond length for this kind of interaction is
served in a CDCl3 solution. These data are reported in the
first two columns of Table 2. The experimental results show
a chemical shift of δ ϭ 4.59 ppm for proton H2A but also
a downfield signal at δ ϭ 5.49 ppm for proton H1A induced
by the deshielding effect of the H1A···O2D hydrogen bond.
Although consistently shifted to higher field by ca. 0.4 ppm
(this effect is due to the approximation of the simulation
method used here, as documented elsewhere),[15,17,18] the
simulated chemical shift values for H1A and H2A (δ ϭ
4.22 ppm and δ ϭ 5.12 ppm, respectively) do in fact exhibit
the same trend as that of the experimental data. In conclu-
sion, it is now possible to ascribe the anomalous downfield
chemical shift computed and observed for the proton H1A
in 3 to the noncovalent H1A···O2D interaction. Therefore,
we experimentally demonstrated that the presence of a weak
hydrogen bond originates a signal for the related proton to
lower field by ca. 1 ppm. On the other hand, if this interac-
tion is not taken into account in the computations, this pro-
ton signal would be observed in its normal range of chem-
ical shifts. We have previously estimated the strength for
such a weak α-CH···OϭC interaction to be about 1.5
[16]
˚
2.4 A and the bond angle may vary from 90 to 180°.
Therefore, the fully relaxed structure of 3 does seem to in-
volve an α-CϪH···OϭC interaction (see Figure 2 and the
first column of Table 1), as recently shown for the related
(-pGlu)2 oligomer.[15] Quite remarkably, this latter struc-
ture is almost identical to the one reported here (Figure 2),
thus providing an indirect validation of the method em-
ployed.
kcal/molϪ1 [15]
.
The lowest-energy minimum for trimer 6 was optimized
in a similar fashion (Figure 3). According to the procedure
outlined above, DFT optimizations for the longer analogues
(i.e. the homo-oligomers 9 and 12) were also performed. In
all these cases, after the full relaxation (i.e. an uncon-
strained optimization), we found that the atoms involved in
the hypothetical α-CϪH···OϭC interactions[4] possess the
Figure 2. DFT-optimized structures for the dimer 3 with partial
atom labeling; for clarity only the H-bonded hydrogen atoms are
shown
Table 1. Calculated geometrical parameters for dimer 3, trimer 6, tetramer 9 and pentamer 12
Dimer 3 Trimer 6 Tetramer 9
Pentamer (12)
˚
˚
˚
˚
Bond length [A]
Bond length [A]
Bond length [A]
Bond length [A]
H1AϪO2D 2.34
H1AϪO2D 2.38
H2AϪO3D 2.35
H1AϪO2D 2.38
H2AϪO3D 2.36
H3AϪO4D 2.37
H1AϪO2D 2.36
H2AϪO3D 2.37
H3AϪO4D 2.38
H4AϪO5D 2.38
Bond angle [°]
Bond angle [°]
Bond angle [°]
Bond angle [°]
C1AϪH1AϪO2D 108.7
H1AϪO2DϪC2D 97.7
C1AϪH1AϪO2D 106.0
C2AϪH2AϪO3D 107.1
H1AϪO2DϪC2D 97.0
H2AϪO3DϪC3D 96.5
C1AϪH1AϪO2D 106.1
C2AϪH2AϪO3D 105.3
C3AϪH3AϪO4D 105.1
H1AϪO2DϪC2D 96.5
H2AϪO3DϪC3D 95.9
H3AϪO4DϪC4D 96.9
C1AϪH1AϪO2D 106.9
C2AϪH2AϪO3D 105.0
C3AϪH3AϪO4D 103.1
C4AϪH4AϪO5D 104.3
H1AϪO2DϪC2D 96.8
H2AϪO3DϪC3D 96.0
H3AϪO4DϪC4D 96.3
H4AϪO5DϪC5D 97.2
Dihedral angle [°]
Dihedral angle [°]
Dihedral angle [°]
Dihedral angle [°]
C1AϪH1AϪO2DϪC2D 52.0
C1AϪH1AϪO2DϪC2D 56.8
C2AϪH2AϪO3DϪC3D 58.1
C1AϪH1AϪO2DϪC2D 57.9
C2AϪH2AϪO3DϪC3D 61.3
C3AϪH3AϪO4DϪC4D 59.0
C1AϪH1AϪO2DϪC2D 56.8
C2AϪH2AϪO3DϪC3D 61.1
C3AϪH3AϪO4DϪC4D 61.5
C4AϪH4AϪO5DϪC5D 57.9
Eur. J. Org. Chem. 2003, 259Ϫ267
261

M. Garavelli, C. Toniolo et al.
FULL PAPER
Moreover, we performed a calculation of chemical
shifts[17,18] for protons H1A, H2A and H3A in trimer 6,
H1A, H2A, H3A and H4A in tetramer 9, and H1A, H2A,
H3A, H4A and H5A in pentamer 12. As expected, the cal-
culated δ values for the protons involved in the noncovalent
α-CϪH···OϭC interaction are ca. 1 ppm higher than the
chemical shifts for the free protons. The simulated values
(Table 2) are very close to the experimental data (except for
the 0.4 ppm higher field shift consistently seen in the sig-
nals) and both sets of values show the same progress (Fig-
ure 6). Therefore, we may deduce that for each homo-oligo-
mer there are n Ϫ 1 (where n is the number of monomers
in each molecule) downfield proton signals and, as a con-
sequence, that there are also n Ϫ 1 α-CϪH···OϭC hydrogen
bonds. These interactions, although weak, do stabilize the
overall secondary structure allowing the molecule to adopt
a robust poly(-Pro)n II type helix (31 helix).[5] Front and
side views of the ternary helix optimized for pentamer 12
are reported in Figure 7 (part a and b, respectively).
Table 2. Calculated and experimental (CDCl3 solution) chemical
shifts (δ in ppm) for dimer 3, trimer 6, tetramer 9 and pentamer 12
Dimer 3
Trimer 6
Tetramer 9
Pentamer 12
Proton calcd. exp. calcd. exp. calcd. exp. calcd. exp.
H1A
H2A
H3A
H4A
H5A
5.12
4.22
5.49 5.02
4.59 5.32
4.17
5.47 4.97
5.62 5.28
4.56 5.29
4.16
5.46 5.00
5.63 5.25
5.57 5.20
4.54 5.26
4.06
5.48
5.62
5.58
5.58
4.55
correct geometrical features to allow for the onset of these
hydrogen bonds. Remarkably, the average values for the
H···O bond length and the H···OϭC bond angle are 2.37
˚
A and 96.7°, respectively (Figures 2Ϫ5 and Table 1), being
similar to those computed for dimer 3 and with the values
reported in the literature for related interactions and sim-
ilar systems.[15,16]
Although these computational findings do not provide
an ultimate proof for the ordered foldameric structure in
the longest homo-oligomers, we still believe that these data,
when combined with experimental results, do offer a signi-
ficant clue for the stability of the afore-mentioned foldamer
conformations. In addition, the experimental 1H NMR
analysis, when complemented and supported by computa-
tional modeling, does represent a valid tool to elucidate the
secondary structure of these molecules.
Solution Conformational Analysis
Information on the preferred conformation of the -Oxd
homo-oligomers in solution was obtained in structure-sup-
porting solvents, CDCl3 and methanol (MeOH), by FT-IR
1
absorption, H NMR and CD techniques.
Figure 8 illustrates the FT-IR absorption spectra (CϭO
stretching region) of the Boc/OBn-protected -Oxd series to
the pentamer level. A strictly comparable trend was ob-
served for the C-deprotected series (not shown). From liter-
ature data on imides and acylurethanes[19] and a visual in-
Figure 3. DFT-optimized structures for the trimer 6 with partial
atom labeling; for clarity only the H-bonded hydrogen atoms are
shown
Figure 4. DFT-optimized structures for the tetramer 9 with partial atom labeling; for clarity only the H-bonded hydrogen atoms are shown
262 Eur. J. Org. Chem. 2003, 259Ϫ267

Pseudopeptide Foldamers
FULL PAPER
Figure 5. DFT-optimized structures for the pentamer 12 with partial atom labeling; for clarity only the H-bonded hydrogen atoms
are shown
Figure 7. Front view (a) and side view (b) of the calculated pen-
tamer 12 helix; the terminally protecting groups have been removed
for clarity
Figure 6. Progress of the calculated (top) and experimental (CDCl3
spectra in this region from the COOBn (in the ester series)
solution) (bottom) chemical shifts for the homo-oligomers 3, 6, 9
and COOH (in the free carboxylic acid series) groups is
expected to be of minor significance and located near
and 12
1740[20] and 1710 cmϪ1 [19a]
respectively. From this FT-IR
,
spection of the spectra in Figure 8 it is evident that the ma- absorption analysis it may be concluded that there is no
jor contribution to all four bands in the 1830Ϫ1700 cmϪ1 evidence of any abrupt conformational change in a CDCl3
region is given by the carbonyl stretching absorptions of the solution as the pseudopeptide main chain elongates.
acylurethane moiety. In particular, the intensity of the two
In this Boc/OBn-protected -Oxd series the only poten-
1
bands near 1790 and 1720 cmϪ1 regularly increases with tially informative protons for a H NMR conformational
oligomer backbone lengthening. The contribution to the analysis are the α-CH protons. To obtain additional, more
Eur. J. Org. Chem. 2003, 259Ϫ267
263

M. Garavelli, C. Toniolo et al.
FULL PAPER
Figure 8. FT-IR absorption spectra in the 1900Ϫ1650 cmϪ1 region
of the monomer Boc--Oxd-OBn (M) and the higher homo-oligo-
mers 3, 6, 9, and 12 in a CDCl3 solution; peptide concentration
1 m
1
Figure 9. Plot of α-CH chemical shifts in the H NMR spectra of
the homo-pentamer Boc-(-Oxd)5-OBn (12) as a function of in-
creasing percentages of DMSO to the CDCl3 solution (v/v); peptide
concentration 1 m
detailed information on the preferred conformation in a
CDCl3 solution of Boc-(-Oxd)5-OBn (12), the longest and
most significant homo-oligomer of this series, we carried
out a 400-MHz 1H NMR investigation by use of the
DMSO dependence of an α-CH proton chemical shift. This
solvent has a strong hydrogen-bonding acceptor character
and, if bound to an acidic proton, is expected to dramatic-
ally move its chemical shift downfield.[21]
As discussed above, from Figure 9 it is evident that in a
CDCl3 solution the chemical shifts of all α-CH protons,
except that assigned to the C-terminal residue (H5A), clus-
ter in a region largely downfield compared to that common
to peptide α-CH protons (δ ഠ 4.5Ϫ5.0 ppm)[22] and espe-
cially to the α-CH protons of poly(-Pro)n II (δ ϭ
4.80 ppm).[23] We ascribed this unusual shift to the presence
of a 2-oxo-1,3-oxazolidine carbonyl group in the vicinity of
the N-terminal and internal α-CH protons. This α-
CϪH···OϭC interaction is clearly absent in the case of the
C-terminal α-CH proton, that therefore resonates in the ex-
pected spectral region. In our 1H NMR titration experi-
ments two classes of α-CH protons were observed: (1) The
first class (H1AϪH4A protons) includes protons whose
Figure 10. CD spectra in the 195Ϫ270 nm region of the monomer
Boc--Oxd-OH (MЈ) and the higher homo-oligomers 4, 7, 10, and
13 in an MeOH solution; peptide concentration 1 m
chemical shifts are insensitive to the addition of the per- tensity with increasing backbone length. In these spectra
turbing agent DMSO. (2) The second class (H5A proton) the positive lobe is lower in intensity than the negative lobe
includes that displaying a behaviour characteristic of at le- and the crossover point is at 212Ϫ216 nm. Also, these
ast partially exposed protons (moderate sensitivity of chem- curves give rise to a nearly isodichroic point in the vicinity
1
ical shift to solvent composition). The present H NMR of 217 nm. The CD spectra of the benzyl ester oligomers
spectroscopic data support the view that in a solvent of low (not shown) strictly compare with those reported in Fig-
polarity (CDCl3) the H1AϪH4A protons of the pentamer ure 10. As there is no published systematic investigation of
are inaccessible to the perturbing agent and therefore, most the complex semicyclic acylurethane nor of the related im-
probably, are intramolecularly hydrogen-bonded.
ide chromophore,[24] and in view of the lack of a precise
The electronic absorption spectra of the acylurethane correlation of the positions of the UV absorption bands
chromophore of the Boc-protected -Oxd homo-oligomers versus the wavelengths of the positive/negative maxima and
in an MeOH solution are characterized by a moderately cross-over points in the CD spectra of our homo-oligomers,
intense (εR ϭ 2000Ϫ2500) transition in the 200Ϫ250 nm we made no attempt to assign the observed Cotton effects
region centred at ca. 210 nm (not shown). The related chiro- to any specific electronic transition. However, the shape
spectroscopic properties are illustrated in Figure 10. While similarity and the regularly increasing intensity of the CD
the monomer exhibits a single, weak and positive band (at spectra with main-chain elongation to the pentamer point
ca. 215 nm) at Ͼ 200 nm, the CD spectra of the higher oli- to the persistence of the same backbone conformation for
gomers show dichroic doublets of regularly increasing in- these -Oxd oligomers in an MeOH solution. We also
264
Eur. J. Org. Chem. 2003, 259Ϫ267

Pseudopeptide Foldamers
FULL PAPER
checked the conformational stability of a representative Experimental Section
homo-oligomer, the tetramer, by recording the CD spectra
Synthesis and Characterization: Materials and reagents were of the
highest commercially available grade and used without further
purification. Reactions were monitored by thin-layer chromato-
graphy using Merck silica gel 60 F254 covered plastic plates. Com-
pounds were visualized by UV light and ceric ammonium molybd-
ate. Flash chromatography was performed using a Merck silica gel
in an MeOH solution as a function of heating from 20 to 50
°C (not shown). In this temperature range the conformation
appears to be remarkably stable as no change in the shape
and only a modest decrease (less than 10%) in the band
intensities were observed.
1
60 stationary phase. H and 13C NMR spectra were recorded with
a Varian Gemini 300 or a Varian Mercury 400 spectrometer. Chem-
ical shifts are reported in δ values relative to the solvent peak of
CHCl3, set at δ ϭ 7.27 ppm. Infrared spectra were recorded with
a Nicolet 210 FT-IR absorption spectrometer. Melting points were
determined in open capillaries and are uncorrected.
Conclusions
When joined in specific sequences, physicochemical in-
formation inherent in the α-amino acid monomeric units
enables the onset of functional folded peptide and protein
architectures. This property of biopolymers can provide a
valuable basis to our attempts to design novel materials
aimed at emulating catalytic processes, energy and electron
transfer mechanisms, and self-organizing phenomena.
In this work we designed, synthesized and determined[25]
the conformational preference of a novel pseudopeptide fol-
dameric series based on -Oxd building blocks character-
Benzyl (4S,5R)-5-methyl-2-oxo-1,3-oxazolidine-4-carboxylate (1,
Scheme 1): This oxazolidin-2-one, starting material for the syn-
thesis of the homo-oligomers, was obtained by reaction of -
threonine with triphosgene, according to the method described by
Hassner and co-workers.[11]
General Procedure for the Synthesis of the Homo-Oligomers of
Benzyl
(4S,5R)-5-Methyl-2-oxo-1,3-oxazolidine-4-carboxylate
(Scheme 2): The detailed procedures for the steps of protection,
deprotection, activation and coupling, along with the characteriza-
tion data for Boc--Oxd-OBn (1), and compounds 2, 3, 6, 7, 8 and
9, are also reported in ref.[4]
ized by a semi-extended helical conformation. As in the re-
[15]
lated poly(-Pro)n II[5] and (-pGlu)n
(Figure 1) helices,
the torsion angle is frozen in a skew conformation (from
Ϫ50 to Ϫ80°) by virtue of the steric restriction imposed by
the five-membered ring system, while the ψ and ω torsion
angles are close to trans (from 140 to 160°) and trans (180°),
respectively. All three are left-handed, elongated, ternary
helices with an α-C ···α-C distance (pitch) of 9.0Ϫ9.5 A.
Obviously, the stabilizing α-CϪH···OϭC intramolecular hy-
drogen bond characteristic of the (-Oxd)n and (-pGlu)n
helices is absent in the poly(-Pro) II helix. It may be an-
ticipated that, for this reason and for the stable anti disposi-
tion of the two carbonyl groups of each semicyclic acylure-
thane or imide system, the (-Oxd)n and (-pGlu)n polymers
Boc-( -Oxd)3-OH (4): Yield: 95% (0.95 mmol, 0.47 g); m.p. 132 °C.
L
1
[α]2D0 ϭ Ϫ142.5 (c ϭ 1.0, CH2Cl2). H NMR (400 MHz, CDCl3):
δ ϭ 1.50 (s, 9 H, tBu), 1.52Ϫ1.75 (m, 9 H, 3 Me), 4.56 (d, JH,H
3.2 Hz, 1 H, CHN), 4.70 (dq, JH,H ϭ 1.0, 6.6 Hz, 1 H, CHO),
4.75Ϫ4.91 (m, 2 H, 2 CHO), 5.49 (d, JH,H ϭ 1.0 Hz, 1 H, CHN),
3
ϭ
3
1
4
˚
3
3
5.65 (d, JH,H ϭ 1.0 Hz, 1 H, CHN), 6.61 (br. s, 1 H, OH) ppm.
13C NMR (100 MHz, CDCl3): δ ϭ 20.6, 21.2, 27.8, 60.7, 61.1, 62.0,
65.9, 73.1, 75.5, 85.0, 149.0, 151.5, 152.0, 152.3, 166.6, 168.3, 169.0
ppm. IR (Nujol): ν˜ ϭ 1791, 1718 cmϪ1 (CϭO). C20H25N3O12
(499.14): calcd. C 48.10, H 5.05, N 8.41; found C 48.06, H 5.09,
N 8.44.
1
Boc-(
L
-Oxd)3-OPfp (5): Yield: 90% (0.90 mmol, 0.60 g). H NMR
are not expected to exhibit the amide cis-trans dynamic
(300 MHz, CDCl3): δ ϭ 1.53 (s, 9 H, tBu), 1.64 (d, 3JH,H ϭ 6.3 Hz,
3 H, Me), 1.69 (d, JH,H ϭ 6.3 Hz, 3 H, Me), 1.76 (d, JH,H
6.0 Hz, 3 H, Me), 4.70 (dq, JH,H ϭ 1.8, 6.3 Hz, 1 H, CHO), 4.76
(dq, JH,H ϭ 2.1, 6.3 Hz, 1 H, CHO), 4.88Ϫ5.05 (m, 2 H, CHN ϩ
Ǟ
ǟ
equilibrium typical of the related poly(-Pro)n I
II heli-
3
3
ces.[5]
ϭ
3
As this phenomenon, particularly significant in short
homo-oligomers,[26] is responsible for the onset of multiple
coexisting conformers, it is evident that the (-Pro)n oligo-
peptides, unless ω-frozen by a bulky substitution in position
5 of the pyrrolidine ring,[27,28] cannot be safely exploited as
molecular rulers with well-predetermined distances. On the
other hand, the rigid (-Oxd)n homo-oligomeric system, as
its γ-lactam (-pGlu)n counterpart, if adequately func-
tionalized in position 5, is expected to be a very useful scaf-
fold in future investigations relating chemistry to biology
and material sciences. In this connection it is worth men-
tioning that a qualitative check on the (-Oxd)n homo-oli-
gomers points to a definite, although moderate, water solu-
bility.
Although neither of these findings does provide per se an
ultimate proof for the geometry of the most stable structure
of the (-Oxd)n homo-oligomers, still all these data, when
taken in conjunction with our recent results on the related
(-pGlu)n system,[15] provide a satisfactory interpretation of
this unique type of foldameric structure.
3
3
3
CHO), 5.50 (d, JH,H ϭ 1.8 Hz, 1 H, CHN), 5.67 (d, JH,H
ϭ
2.1 Hz, 1 H, CHN) ppm.
Boc-(L-Oxd)4-OH (10): Yield: 93% (0.93 mmol, 0.58 g); m.p. 136
°C. [α]2D0 ϭ Ϫ183.7 (c ϭ 1.0, CH2Cl2). 1H NMR (400 MHz,
3
CDCl3): δ ϭ 1.46 (s, 9 H, tBu), 1.54 (d, JH,H ϭ 6.0 Hz, 3 H, Me),
3
3
1.56 (d, JH,H ϭ 6.0 Hz, 3 H, Me), 1.60 (d, JH,H ϭ 5.7 Hz, 3 H,
3
3
Me), 1.62 (d, JH,H ϭ 5.7 Hz, 3 H, Me), 4.46 (d, JH,H ϭ 3.0 Hz,
3
1 H,CHN), 4.62 (dq, JH,H ϭ 2.1, 6.0 Hz, 1 H, CHO), 4.78 (dq,
3JH,H ϭ 3.0, 6.0 Hz,, 1 H, CHO), 4.85Ϫ4.98 (m, 2 H, 2 CHN),
5.45 (d, 3JH,H ϭ 2.1 Hz, 1 H, CHN), 5.61 (d, JH,H ϭ 2.1 Hz, 2 H,
3
CHN) ppm. 13C NMR (100 MHz, CDCl3): δ ϭ 20.5, 21.2, 27.7,
61.1, 61.2, 61.7, 61.8, 72.7, 75.3, 75.9, 76.0, 84.5, 148.9, 151.0,
152.2, 152.4, 166.7, 167.0, 168.3, 170.6 ppm. IR (Nujol): ν˜ ϭ 1790,
1718, 1700 cmϪ1 (CϭO). C25H30N4O15 (626.17): calcd. C 47.93, H
4.83, N 8.94; found C 47.89, H 4.78, N 8.97.
Boc-(
(300 MHz, CDCl3): δ ϭ 1.49 (s, 9 H, tBu), 1.59 (d, 3JH,H ϭ 6.3 Hz,
L
-Oxd)4-OPfp (11): Yield: 91% (0.91 g, 0.72 g). 1H NMR
3
3
3 H, Me), 1.68 (d, JH,H ϭ 6.0 Hz, 3 H, Me), 1.70 (d, JH,H
ϭ
3
5.7 Hz, 3 H, Me), 1.74 (d, JH,H ϭ 6.0 Hz, 3 H, Me), 4.65 (dq,
Eur. J. Org. Chem. 2003, 259Ϫ267
265

M. Garavelli, C. Toniolo et al.
FULL PAPER
3JH,H ϭ 3.0, 6.0 Hz, 1 H, CHO), 4.77 (dq, JH,H ϭ 2.1, 6.0 Hz, 1
3
under the same conditions. Cells with path lengths of 0.1, 1.0 and
10 mm (with CaF2 windows) were used. Spectrograde CDCl3
2.1 Hz, 1 H, CHN), 5.60 (d, JH,H ϭ 2.1 Hz, 1 H, CHN), 5.62 (d, (99.8% D) was purchased from Fluka.
3
H, CHO), 4.86Ϫ4.98 (m, 3 H, CHN ϩ 2 CHO), 5.47 (d, JH,H
ϭ
3
3JH,H ϭ 1.8 Hz, 1 H, CHN) ppm.
1H NMR: The 1H NMR spectra were recorded with a Bruker AM
400 spectrometer. Measurements were carried in deuteriochloro-
form (99.96% D; Aldrich) and dimethyl sulfoxide ([D6]DMSO)
(99.96% D6; Acros Organics) with tetramethylsilane as the in-
ternal standard.
Boc-(L-Oxd)5-OBn (12): Yield: 50% (0.50 mmol, 0.42 g); m.p.
128Ϫ131 °C. [α]2D0 ϭ Ϫ215.8 (c ϭ 1.0, CH2Cl2). 1H NMR
(400 MHz, CDCl3): δ ϭ1.52 (s, 9 H, tBu), 1.59 (d, 3JH,H ϭ 6.6 Hz,
3
3
3 H, Me), 1.60 (d, JH,H ϭ 6.3 Hz, 3 H, Me), 1.64 (d, JH,H
ϭ
3
6.9 Hz, 3 H, Me), 1.66 (d, JH,H ϭ 6.6 Hz, 3 H, Me), 1.70 (d,
CD: The CD spectra were obtained with a Jasco J-710 spectropola-
rimeter. Cylindrical fused quartz cells of 10, 1, 0.2 and 0.1 mm path
length (Hellma) were used. The values are expressed in terms of
[θ]T, the total molar ellipticity (deg ϫ cm2 ϫ dmolϪ1). Spectrograde
MeOH (Baker) was used as solvent.
3JH,H ϭ 6.3 Hz, 3 H, Me), 4.55 (d, JH,H ϭ 3.6 Hz, 3 H, CHN),
3
3
4.62Ϫ4.74 (m, 3 H, 3 CHO), 4.85 (dq, JH,H ϭ 1.8, 6.6 Hz, CHO),
4.90 (dq, JH,H ϭ 1.8, 6.3 Hz, CHO), 5.14 (d, JH,H ϭ 12.0 Hz, 1
H, OCHHPh), 5.27 (d, JH,H ϭ 12.0 Hz, 1 H, OCHHPh), 5.48 (d,
3
2
2
3JH,H ϭ 1.8 Hz, 1 H, CHN), 5.59 (d, 3JH,H ϭ 1.5 Hz, 2 H, 2 CHN),
5.63 (d, 3JH,H ϭ 1.8 Hz, 1 H, CHN), 7.24Ϫ7.48 (m, 5 H, Ph) ppm.
13C NMR (400 MHz, CDCl3): δ ϭ20.6, 21.0, 21.2, 27.8, 60.9, 61.0,
61.1, 61.2, 61.9, 68.7, 72.7, 75.2, 75.3, 75.7, 75.9, 84.5128.3, 128.5,
128.7, 133.8, 150.8, 151.7, 152.0, 152.2, 152.3, 1661.6, 166.8, 167.1,
168.4 ppm. IR (Nujol): ν˜ ϭ 1832, 1792, 1706 cmϪ1 (CϭO).
C37H41N5O18 (843.74): calcd. C 52.67, H 4.90, N 8.30; found C
52.69, H 4.98, N 8.27.
Acknowledgments
M. G., F. B., S. L., C. T., and V. T. thank MURST (Cofin 2000,
Cofin 2001 and 60%), the Alma Mater Studiorum Ϫ University of
Bologna (Funds for Selected Topics) and CINECA (Grant K1I-
BOZZ1) for financial support.
Boc-(L-Oxd)5-OH (13): Yield: 91% (0.91 mmol, 0.69 g); m.p.
129Ϫ132 °C. [α]2D0 ϭ Ϫ190.9 (c ϭ 0.7, CH2Cl2). 1H NMR
(400 MHz, CDCl3): δ ϭ 1.52 (s, 9 H, tBu), 1.58Ϫ1.75 (m, 15 H, 5
[1]
S. H. Gellman, Acc. Chem. Res. 1998, 31, 173Ϫ180.
[2]
C. Tomasini, A. Vecchione, Org. Lett. 1999, 1, 2153Ϫ2156.
3
3
Me), 4.56 (d, JH,H ϭ 3.6 Hz, 3 H, CHN), 4.69 (dq, JH,H ϭ 1.9,
[3]
G. Valle, C. Toniolo, G. Jung, Liebigs Ann. Chem. 1986,
1809Ϫ1822.
S. Lucarini, C. Tomasini, J. Org. Chem. 2001, 66, 727Ϫ732.
A. J. Hopfinger, in Conformational Properties of Macromolec-
ules, Academic Press, New York, 1973, pp. 182Ϫ189, and refer-
ences therein.
3
6.3 Hz, CHO), 4.78Ϫ4.98 (m, 4 H, 4 CHO), 5.50 (d, JH,H
ϭ
3
2.1 Hz, 1 H, CHN), 5.61 (d, JH,H ϭ 1.8 Hz, 1 H, CHN), 5.64 (d,
3JH,H ϭ 1.5 Hz, 2 H, 2 CHN) ppm. 13C NMR (100 MHz, CDCl3):
δ ϭ 20.6, 21.3, 27.9, 60.8, 61.2, 61.3, 62.0, 73.0, 75.4, 75.9, 76.0,
84.8, 149.1, 151.2, 151.9, 152.1, 152.3, 152.4, 166.7, 167.1, 168.4,
169.0 ppm. IR (Nujol): ν˜ ϭ 1792, 1719, 1701 cmϪ1 (CϭO).
C10H15NO5 (753.20): calcd. C 47.81, 4.68, N 9.29; found C 47.76,
H 4.72, N 9.35.
[4]
[5]
[6] [6a]
S. M. Newman, A. Kutner, J. Am. Chem. Soc. 1951, 73,
[6b]
4199Ϫ4204.
J. B. Hyne, J. Am. Chem. Soc. 1959, 81,
6058Ϫ6061. [6c] W. D. Lubell, H. Rapoport, J. Org. Chem. 1989,
54, 3824Ϫ3831.
[7]
[8]
D. J. Ager, I. Prakash, D. R. Schaad, Chem. Rev. 1996, 96,
Computational Methods. All calculations were carried out using the
tools available in the Gaussian 98[29] package on an SGI Origin
3800 multiprocessor system, using the DFT/B3LYP functional (i.e.
835Ϫ875.
[8a] D. A. Evans, J. Bartroli, T. L. Shih, J. Am. Chem. Soc. 1981,
[8b]
103, 2127Ϫ2129.
D. A. Evans, M. D. Ennis, D. J. Mathre,
Becke’s
three-parameter
hybrid
functional
with
the
[8c]
J. Am. Chem. Soc. 1982, 104, 1737Ϫ1739.
V. Nelson, T. R. Taber, Top. Stereochem. 1982, 13, 1Ϫ115.
D. A. Evans, J.
LeeϪYangϪParr correlation functional).[30] This functional has
been shown to properly describe both standard hydrogen bonds,[31]
as well as nonclassical, weakly bound hydrogen bonds (such as
CϪH···OϭC interactions).[32] In particular, according to the literat-
ure,[33] a mixed basis set was used for the determination of the weak
CϪH···OϭC hydrogen bonds: a 6-31ϩG(d) basis set for the CiA,
HiA, O(iϩ1)D, C(iϩ1)D atoms (i ϭ 1 for dimer; i ϭ 1, 2 for trimer;
i ϭ 1, 2, 3 for tetramer; and i ϭ 1, 2, 3, 4 for pentamer) and a 6-
31G(d) for all other atoms. Since the O(iϩ1)D lone pairs strongly
interact with HiA, the use of diffuse functions on heavy atoms is
fundamental. Chemical shift simulations were performed by using
[9] [9a]
D. A. Evans, J. M. Takacs, L. R. McGee, M. D. Ennis, D.
J. Mathre, J. Bartroli, Pure Appl. Chem. 1981, 53, 1109Ϫ1127.
[9b]
[9c]
D. A. Evans, Aldrichim. Acta 1982, 15, 23Ϫ32.
D. A.
Evans, D. L. Rieger, M. T. Bilodeau, F. Urpi, J. Am. Chem.
Soc. 1991, 113, 1047Ϫ1049.
[10] [10a] D. A. Evans, K. T. Chapman, J. Bisaha, J. Am. Chem. Soc.
[10b]
1984, 106, 4261Ϫ4263.
D. A. Evans, K. T. Chapman, J.
[10c]
Bisaha, J. Am. Chem. Soc. 1988, 110, 1238Ϫ1256.
D. A.
Evans, K. T. Chapman, D. T. Hung, A. T. Kawaguchi, Angew.
Chem. Int. Ed. Engl. 1987, 26, 1184Ϫ1186.
E. Falb, A. Nudelman, A. Hassner, Synth. Commun. 1993,
23, 2839Ϫ2844.
[11]
the standard tools available in the Gaussian 98 package.[17]
A
[12] [12a]
standard 6-31G(d) basis set was used for all the atoms. The com-
puted data do not directly yield the chemical shift value, but only
a value for the isotropic magnetic tensor. The chemical shift value
is obtained from the equation δH ϭ 32.18 Ϫ σH, where 32.18 is the
calculated isotropic magnetic tensor for the protons in tetramethyl-
silane, and σH is the calculated isotropic magnetic tensor for the
investigated proton. This procedure has been recently validated and
applied by us for a similar molecular system.[15]
S. S. Canan Koch, A. R. Chamberlin, J. Org. Chem. 1993,
58, 2725Ϫ2737. [12b] J. R. Gage, D. A. Evans, Org. Synth. 1990,
68, 83Ϫ91.
[13]
[14]
[15]
G.-J. Ho, D. J. Mathre, J. Org. Chem. 1995, 60, 2271Ϫ2273.
M. Green, M. Berman, Tetrahedron Lett. 1990, 31, 5851Ϫ5854.
F. Bernardi, M. Garavelli, M. Scatizzi, C. Tomasini, V. Trigari,
M. Crisma, F. Formaggio, C. Peggion, C. Toniolo, Chem. Eur.
J. 2002, 8, 2516Ϫ2525.
[16] [16a] F. H. Allen, O. Kennard, R. Taylor, Acc. Chem. Res. 1983,
[16b]
16, 146Ϫ153.
P. Chakrabarti, S. Chakrabarti, J. Mol. Biol.
FT-IR: The FT-IR absorption spectra were recorded with a
PerkinϪElmer 1720X spectrophotometer, nitrogen-flushed,
equipped with a sample-shuttle device, at 2 cmϪ1 nominal resolu-
tion, averaging 100 scans. Solvent (base-line) spectra were recorded
[16c]
1998, 284, 867Ϫ873.
V. Nagarajan, V. Pattabhi, Acta Crystallogr. 1997, D53,
G. Felcy Fabiola, S. Krishnaswamy,
[16d]
316Ϫ320.
734Ϫ742.
J. Bella, H. M. Berman, J. Mol. Biol. 1996, 264,
Z. S. Derewenda, L. Lee, V. Derewenda, J. Mol.
[16e]
266
Eur. J. Org. Chem. 2003, 259Ϫ267

Pseudopeptide Foldamers
FULL PAPER
L. Halab, F. Gosselin, W. D. Lubell, Biopolymers (Pept. Sci.)
2000, 55, 101Ϫ122.
[16f]
[27]
Biol. 1995, 252, 248Ϫ262.
T. Steiner, J. Chem. Soc., Perkin
Trans. 2 1995, 1315Ϫ1319. [16g] G. R. Desiraju, Acc. Chem. Res.
[16h]
[28]
[29]
1996, 29, 441Ϫ449.
T. Steiner, W. Saenger,J. Am. Chem.
L. Halab, W. D. Lubell, J. Pept. Sci. 2001, 7, 92Ϫ104.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M.
A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgom-
ery Jr., R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Mil-
lan, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J.
Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Po-
melli, C. Adamo, S. Clifford, J. Ochterski, G. A. Patersson, P.
Y. Ayala. Q. Cui, K. Morokuma, D. K. Malik, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, A.
G. Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith,
M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challac-
ombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, J. L.
Andres, C. Gonzales, M. Head-Gordon, E. S. Replogle, J. Po-
ple, Gaussian 98, Revision A.9, Gaussian Inc., Pittsburgh, PA,
1998.
[16i]
Soc. 1992, 114, 10146Ϫ10154.
G. R. Desiraju, The Weak
Hydrogen Bond in Structural Chemistry and Biology, Oxford
Science, Oxford, U. K., 1999.
[17] [17a] K. Wolinski, J. F. Hilton, P. Pulay, J. Am. Chem. Soc. 1990,
[17b]
112, 8251Ϫ8260.
K. Wolinski, A. J. Sadlej, Mol. Phys.
[17c]
1980, 41, 1419Ϫ1430.
R. Ditchfield, Mol. Phys. 1974, 27,
[17d]
789Ϫ807.
R. McWeeny, Phys. Rev. 1962, 126, 1028Ϫ1024.
[17e]
F. London, J. Phys. Radium 1937, 8, 397Ϫ409.
A. Bagno, Chem. Eur. J. 2001, 7, 1652Ϫ1660.
[18]
[19] [19a]
L. J. Bellamy, in The Infrared Spectra of Complex Molec-
ules, Methuen, London, 1966. [19b]W. L. Driessen, P. L. A. Ev-
erstijn, J. Coord. Chem. 1980, 10, 155Ϫ158.
[20]
A. Bavoso, E. Benedetti, B. Di Blasio, V. Pavone, C. Pedone,
C. Toniolo, G. M. Bonora, Macromolecules 1982, 15, 54Ϫ59.
[21] [21a]
K. D. Kopple, M. Ohnishi, A. Go, Biochemistry 1969, 8,
4087Ϫ4095. [21b] D. Martin, G. Hauthal, in Dimethyl Sulfoxide,
Van Nostrand-Reinhold, Wokingham, U. K., 1975.
D. S. Wishart, B. D. Sykes, F. M. Richards, J. Mol. Biol. 1991,
222, 311Ϫ333.
A. D. Becke, J. Chem. Phys. 1993, 98, 5648Ϫ5652.
[30]
[31] [31a]
L. Gonzalez, O. Mo, M. Yanez, J. Comput. Chem. 1997,
[22]
[23]
[31b]
18, 1124Ϫ1135.
585Ϫ599.
15837Ϫ15845.
R. A. Klein, J. Comput. Chem. 2002, 23,
J. J. Novoa, C. Sosa, J. Phys. Chem. 1995, 99,
[31c]
Y. Xu, P. Saweczko, H.-B. Kraatz, J. Organomet. Chem. 2001,
637Ϫ639, 335Ϫ342.
T. Polonski, J. Chem. Soc., Perkin Trans. 1 1988, 629Ϫ637.
D. J. Hill, M. J. Mio, R. B. Prince, T. S. Hughes, J. S. Moore,
Chem. Rev. 2001, 101, 3893Ϫ4011.
E. Benedetti, A. Bavoso, B. Di Blasio, V. Pavone, C. Pedone,
C. Toniolo, G. M. Bonora, Biopolymers 1983, 22, 305Ϫ317.
[32]
[33]
L. Brunel, F. Carre’, S. G. Dutremez, C. Guerin, F. Dahan, O.
Eisenstein, G. Sini, Organometallics 2001, 20, 47Ϫ54.
J. J. Novoa, B. Tarron, M.-H. Whangbo, J. M. Williams, J.
Chem. Phys. 1991, 95, 5179Ϫ5186.
[24]
[25]
[26]
Received June 7, 2002
[O02309]
Eur. J. Org. Chem. 2003, 259Ϫ267
267