Development of a functionalizable external β-turn mimic based on a cis-fused 1,7-naphthyridine scaffold

Article abstract of DOI:10.1002/ejoc.200200652

An intramolecular Diels-Alder strategy using 2(1H)-pyrazinones was applied to generate a substituted perhydro-1,7-naphthyridine ring system that served as a scaffold to construct the type VI β-turn mimic 2, featuring a cis-amide linkage between the centra

Full text of DOI:10.1002/ejoc.200200652

FULL PAPER
Development of a Functionalizable External β-Turn Mimic Based
on a cis-Fused 1,7-Naphthyridine Scaffold
Frederik J. R. Rombouts,[a] Wim M. De Borggraeve,[a] David Delaere,[b] Matheus Froeyen,[c]
Suzanne M. Toppet,[a] Frans Compernolle,[a] and Georges J. Hoornaert*[a]
Keywords: Bioorganic chemistry / Conformation analysis / Cycloaddition / Naphthyridine / Peptidomimetics
An intramolecular Diels−Alder strategy using 2(1H)-pyrazi-
Modeling studies indicated that the cis-fused bicyclic system
nones was applied to generate a substituted perhydro-1,7- adopts a conformation suitable for induction of a β-turn when
naphthyridine ring system that served as a scaffold to con-
the angular position 8a bears a non-H substituent. Extensive
struct the type VI β-turn mimic 2, featuring a cis-amide link-
NMR analysis of an 8a-methyl derivative 25 confirmed its β-
age between the central i+1 and i+2 residues. The synthesis turn-inducing properties in various solvents, including water.
permits mimicking of the amino acid side chains of the cent-
ral dipeptide, a unique feature for external-turn mimics.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
Introduction
tion conformations, thus lowering the entropy cost for bind-
ing and enhancing selectivity by preclusion of conformers
that give rise to undesired bioactivity. In a previous com-
munication[4] we reported the synthesis of the potential β-
turn mimic 1 (Figure 1), which meets the criteria of func-
tionalizability and rigidity. In a next phase of this investi-
gation, we wanted to impose additional conformational re-
strictions into 1 in order further to rigidify its presumed β-
turn-inducing structure. We thus envisioned the incorpor-
ation of the acetamide nitrogen atom in a six-membered
ring, which should be cis-annelated to the piperidinone core
in order to retain the spatial proximity to the carboxamide,
as shown in structure 2.
With helices and sheets, turns comprise the third major
motif of peptide and protein secondary structure. A turn
can be defined as the site at which a peptide changes its
overall direction. Different types of turns have been de-
scribed, including δ-, γ-, β-, α-, and π-turns corresponding
to loops involving two to six residues, respectively.[1] Of
these, the most common naturally occurring type is the β-
turn, which reverses direction over four residues, often with
a hydrogen bond between the carbonyl group of residue i
and the NH moiety of residue iϩ3. The prevalence of β-
turns in peptides and on protein surfaces suggests that they
play essential roles in molecular recognition events in
biological systems, such as receptorϪligand, enzymeϪ
substrate, and antigenϪantibody interactions. This has
raised the challenge of the development of functionalizable
β-turn mimics[2] to study these interactions or to enhance
in vivo absorption and the metabolic stability of peptide
probes.[3] Moreover, the incorporation of rigid β-turn ana-
logues might render peptides more active by limiting solu-
[a]
Laboratorium voor Organische Synthese, K. U. Leuven,
Celestijnenlaan 200F, 3001 Leuven, Belgium
Fax: (internat.) ϩ 32-16/327990
E-mail: losh@chem.kuleuven.ac.be
[b]
Groep Kwantumchemie, K. U. Leuven,
Figure 1. Type VI β-turn mimics 1 and 2 and the natural β-turn
Celestijnenlaan 200F, 3001 Leuven, Belgium
Fax: (internat.) ϩ 32-16/327393
E-mail: David.Delaere@chem.kuleuven.ac.be
Compounds 1 and 2 can be viewed as rigidified mimics
of cis-amide turns such as the type VI β-turn, which has
cis-proline at the iϩ2 position of a tetrapeptide entity. The
type VI β-turn has been implicated in the bioactive confor-
mations of several peptides.[5] Notably, cis-amide models
[c]
Laboratorium voor Medicinale chemie, K. U. Leuven,
Minderbroedersstraat 10, 3000 Leuven, Belgium
Fax: (internat.) ϩ 32-16/337379
E-mail: Mathy.Froeyen@rega.kuleuven.ac.be
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
1868
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejoc.200200652
Eur. J. Org. Chem. 2003, 1868Ϫ1878

Functionalizable External β-Turn Mimic Based on a cis-Fused 1,7-Naphthyridine Scaffold
FULL PAPER
Synthesis of 11
might also serve as mimics of cis-amide moieties other than
cis-proline, which, according to X-ray data, may occur
more frequently than previously thought.[6] In our current
approach, application of the 1,7-naphthyridine scaffold 2
for mimicking cis-amide β-turns is based on the following
criteria:
(a) The cis-amide linkage between the central residues;
this is constrained in a rigid bicyclic lactam system rep-
resenting the iϩ1 and iϩ2 residues of the tetrapeptide.
(b) The spatial proximity of the residues attached to the
N- and C-termini of the central dipeptide is enabled by the
cis disposition of the N1 atom and the carboxamide group.
(c) The similarity of the bicyclic lactam structure 2 with
that of other successful mimics of the type VI β-turn:[7] the
latter mimics were also based on a lactam structure but they
lacked the capacity to mimic various amino acid side
chains, in contrast to the broad functionalization enabled
by the variable R3 and R6 substituents in the general struc-
ture 2.
N-(p-Methoxybenzyl)pyrazinone 4 was generated by
cyclization of 2-[N-(4-methoxybenzyl)amino]acetonitrile
(HCl salt) 3 with oxalyl chloride (Scheme 2).[9] Subsequent
Stille coupling, with tetraphenyltin and [Pd(PPh3)4] as a
catalyst, provided the corresponding 3-phenyl derivative
5,[10] which was N-debenzylated by heating with trifluoro-
acetic acid to produce pyrazinone 6. This was converted
into the desired N-alkenylated pyrazinone 7 by heating with
5-bromo-1-pentene and Cs2CO3 in dioxane. Intramolecular
cycloaddition of pyrazinone 7 and subsequent conversion
of the resulting adduct 8 into bis(lactam) 9 were ac-
complished by heating 7 in bromobenzene at reflux tem-
perature, followed by hydrolysis of the imidoyl chloride in-
termediate in water-saturated ethyl acetate. The strained tri-
cyclic bis(lactam) 9 was converted into cis-fused 1,7-naph-
thyridine 10 by selective acid methanolysis of the tertiary
amide group, effected by heating with 2.5 equiv. of meth-
anesulfonic acid in boiling methanol. In the two final steps,
amino ester 10 was transformed into target diamide 11 by
N-acetylation of the free amino group, followed by amino-
lysis of the ester function by treatment with methylamine.
Results and Discussion
In the following sections we first describe our synthetic
approach to the 8a-H and 8a-substituted target molecules
2. Next, we examine the conformational behavior of 2, by
use both of molecular modeling and of detailed NMR
analysis. Finally the β-turn-inducing potential of 8a-substi-
tuted molecules is evaluated.
Synthesis
To construct the 1,7-naphthyridine scaffold we applied
our recently reported intramolecular DielsϪAlder reaction
strategy using N-pentenyl-substituted 2(1H)-pyrazinones
(Scheme 1; R3 ϭ Ph, R6 ϭ H).[8] Cycloaddition to form a
bridged imidoyl chloride intermediate, followed by hydroly-
sis, provides stable tricyclic compounds exhibiting both a
distorted tertiary amide and a nonstrained secondary amide
function. The nonplanar tertiary amide can in turn be sub-
jected to selective methanolysis to generate cis-fused bicyc-
lic ring systems of type 2. In a first approach we decided to
use the already reported substitution pattern (R3 ϭ Ph,
R6 ϭ H), but this did not provide the desired β-turn-in-
ducing properties. According to modeling studies, an ap-
propriate conformation should be imposed on the cis-fused
bicyclic system through the introduction of a substituent
R6 ϭ alkyl.
Scheme 2. a) (COCl)2, Et3N·HCl, PhCl, room temp., 2 d; b) SnPh4,
[Pd(PPh3)4], toluene, 120 °C, 7 d; c) CF3CO2H, reflux, 5 h; d)
Br(CH2)3CHϭCH2, Cs2CO3, dioxane, 50 °C, 7 d; e) PhBr, reflux,
overnight; f) EtOAc satd. with H2O, room temp., overnight; g)
MeOH, 2.5 equiv. CH3SO3H, reflux, overnight; h) Ac2O, 40 °C,
1 h; i) 33 wt% CH3NH2/EtOH, room temp., overnight
Synthesis of 25
The sequence described above, used for the preparation
of 11, suffers from a very critical N-alkenylation step, in
which cesium carbonate is used for selective N-alkenylation
of pyrazinone 6 (Scheme 2, step d). Indeed, all methods
other than the use of cesium carbonate tried out for this
crucial step were found to give mainly O-alkylation. To in-
vestigate the tolerance of the cesium carbonate method for
substituents attached to the 6-position of the pyrazinone,
we first attempted the reaction with the 3-methyl-substi-
tuted pyrazinone 15 (Scheme 3). This was prepared by treat-
ment of 2-[(4-methoxybenzyl)amino]propanenitrile 12 with
oxalyl chloride to form 13, followed by Stille coupling with
tetramethyltin (14) and trifluoroacetic acid catalyzed re-
moval of the N-(p-methoxybenzyl) protecting group
(Scheme 3). It is noteworthy that the signals of the two
methyl groups of 14 appear in the NMR spectrum as quad-
ruplets with a rare 7J value of ca. 1 Hz when the
Scheme 1
Eur. J. Org. Chem. 2003, 1868Ϫ1878
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1869

G. J. Hoornaert et al.
FULL PAPER
GaussianϪLorentz enhance function (lb: Ϫ2.0 Hz; gb:
0.5 Hz) is applied prior to the FID processing. Unfortu-
nately, the undesired O-alkenylated compound 16 was iso-
lated almost exclusively when 15 was heated with 5-bromo-
1-pentene and Cs2CO3 in dioxane, demonstrating that selec-
tive N-alkenylation of 6-substituted pyrazinones is not feas-
ible.
Scheme 4 . a) CH3CHO, KCN, NaHSO3, MeOH/H2O, 60 °C, over-
night; b) HCl bubbling, 0 °C, 15 min; c) (COCl)2, Et3N·HCl, PhCl,
room temp., 2 d; d) SnPh4, [Pd(PPh3)4], toluene, 120 °C, 7 d; e)
mCPBA, CH2Cl2, Ϫ15 °C to room temp., 30 min; f) reflux, over-
night; g) PhBr, reflux, 7 d; h) EtOAc satd. with H2O, room temp.,
overnight; i) MeOH, 2.5 equiv. CH3SO3H, reflux, overnight; j)
Ac2O, 40 °C, 2 h; k) 33 wt% CH3NH2/EtOH, room temp., over-
night
the enantiomers of amine precursor 24 by HPLC on a chi-
ral stationary phase, we did succeed in analytical-scale sep-
aration of the bis(lactam) enantiomers 23 (see Exp. Sect.).
Scheme 3. a) (COCl)2, Et3N·HCl, PhCl, room temp., 2 d; b)
Sn(CH3)4, [Pd(PPh3)4], toluene, 120 °C, 1 d; c) CF3CO2H, reflux,
10 h; d) Br(CH2)3CHϭCH2, Cs2CO3, dioxane, 50 °C, 5 d
Conformational Analysis of Amines 10 and 24
Amino esters 10 and 24 predominantly exist as two op-
posite cis-fused conformers 10A and 24B as shown by the
To solve this N-alkenylation problem, we envisaged a dif-
ferent strategy based on the use of 5-(phenylselenyl)pentan-
1-amine (17) as a substitute for direct incorporation of the
acid-sensitive pent-4-en-1-amine into the pyrazinone ring
system (Scheme 4). Amine 17 was converted into the corre-
sponding aminonitrile 18 through a Strecker reaction with
acetaldehyde, and the hydrochloride salt of crude 18 was
treated with oxalyl chloride to provide the 3,5-dichloropyra-
zinone 19. Subsequent Stille coupling with tetraphenyltin
and [Pd(PPh3)4] as a catalyst afforded the corresponding 3-
phenyl derivative 20, which was converted into 21 in high
yield by application of a one-pot oxidation/elimination pro-
cedure with mCPBA. Intramolecular cycloaddition of 21 to
form 22 proceeded more slowly than the analogous conver-
sion of 7 (typically 7 d in refluxing PhBr) but with the same
regioselectivity to provide bis(lactam) 23 upon hydrolysis of
the intermediate imidoyl chloride adduct. As observed for
bis(lactam) 9, 13C NMR analysis of 23 revealed the exist-
1
3J coupling values in the H NMR spectra observed be-
tween the angular proton 4a-H and its vicinal protons 4-H
and 5-H (Figure 2). The structure of 10A was apparent
from a large 3J-trans value with 5-Hax and a small 3J-gauche
value with both 4-H protons, while that of 24B was charac-
3
terized by a small J-gauche value of about 5 Hz with both
3
5-H protons and a large J-trans value with 4-Hax. These
conformational preferences were confirmed by modeling
calculations by MM3* optimization in the gas phase and in
CHCl3 (Macromodel 5.0). Interestingly, opposite confor-
mational preferences were demonstrated for the diamide
ence of a distorted nonplanar tertiary amide group (δC2
ϭ
179.0 ppm versus δC10 ϭ 171.2 ppm for the secondary am-
ide) implying a ketone-like character for the carbonyl moi-
ety and an increased basicity for N-3. Accordingly, selective
acid-catalyzed solvolysis of the tertiary amide could again
be effected by heating of 23 with 2.5 equiv. of methanesul-
fonic acid in methanol to furnish cis-1,7-naphthyridine 24
in almost quantitative yield (99%). The final conversion of
amino ester 24 into the diamide 25 was carried out under
the conditions described above for the preparation of the
8a-H analogue 11.
Even though the synthesis of 25 is entirely diastereoselec-
tive, it still delivers two enantiomers. Both are potential β-
turn mimics: the (4aR,6S,8aR) isomer, which mimics two -
amino acids upon deletion of the bridging elements, and the
(4aS,6R,8aS) isomer, which mimics two conformationally
Figure 2. Conformational equilibria of cis-fused 1,7-naphthyridines
favoring the opposite conformers 10A and 24B for the free amines
1
or 11B and 25A for the diamides, according to H NMR charac-
terization and molecular modeling; the conformations detected by
constrained -residues. While we were unable to separate NMR are boxed
1870  2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjoc.org
Eur. J. Org. Chem. 2003, 1868Ϫ1878

Functionalizable External β-Turn Mimic Based on a cis-Fused 1,7-Naphthyridine Scaffold
FULL PAPER
compounds 11 and 25 by the molecular modeling and the
1H NMR spectroscopic data given below. Apparently, con-
former 10B is disfavored relative to 10A by 1,3-diaxial re-
Molecular Modeling of Diamides 11 and 25
To find the most relevant minimum-energy confor-
pulsion between the ester group and C4, whereas in the mations of 11, a random conformer search[11] and energy
favored conformer 24B this effect is counterbalanced by the minimization were carried out with the Macromodel 5.0
equatorial position of the 8a-Me group with respect to the AMBER* or MM3* force field in combination with the
piperidine chair moiety (conformer 24A exhibits 1,3-diaxial GB/SA solvation model (water). Ruling out improbable
repulsions between 8a-Me and 2-Hax and 4-Hax). In the twist conformations, we identified three major structures
analogous diamide conformer 25B, such an equatorial 8a- 11AϪC in this way (Table 1). According to MM3* the low-
Me group in turn meets a strong repulsion with the co- est-energy conformer has the piperidinone ring in a half-
planar N-acetyl substituent, thus favoring the desired ‘‘clo- chair conformation (11B), while AMBER* finds the
sed’’ form 25A. This repulsive effect is largely removed in (slightly deformed) piperidinone boat conformation 11C to
the favored 8a-H diamide conformer 11B, so that a similar be lowest in energy. Conformer 11A, with the desired β-turn
but less severe repulsion (in relation to that in 8a-Me con- structure, was found to be much higher in energy relative
former 25B) between the NϪAc amide bond and the co- to the global minimum by MM3*, but only slightly higher
planar equatorial 8Ϫ8a linkage in disfavored form 11A according to AMBER*. To obtain improved energy data
comes into play.
for the representative conformers 11AϪC, they were sub-
Table 1. Conformers of 11 in water and corresponding energies found by random conformer search by AMBER* and MM3*; these
energies are compared to energies obtained by ab initio DFT calculations
Geometry optimization and hydration model [kJ/mol][a]
Conformer
AMBER*, GB/SA
MM3*, GB/SA
DFT[b][c]
DFT,[c] CPCM
DFT,[c] PCM
11A
11B
11C
1.05
12.83
0.00
17.06
0.00
1.66
18.70
11.84
0.00
20.12
0.00
11.93
22.59
0.00
17.74
[a]
[b]
[c]
The global minimum energy is set to zero. Gas phase. B3LYP/6-31G**.
Table 2. Conformers of 25 in water and corresponding energies found by random conformer search by the AMBER* and MM3* methods;
these energies are compared to energies obtained by ab initio DFT calculations
Geometry optimization and hydration model [kJ/mol][a]
Conformer
AMBER*, GB/SA
MM3*, GB/SA
DFT[b][c]
DFT,[c] CPCM
DFT,[c] PCM
25A
0.00
Ϫ[d]
45.53
31.77
0.00
2.49
31.90
1.32
0.00
22.54
25.11[e]
24.18
0.00
13.25
12.14
14.73
0.00
11.45
7.03
25B
25C[e]
25D
7.49
[a] The global minimum energy is set to zero. [b] Gas phase. [c] B3LYP/6-31G**. [d] Not found by the conformer search. [e] Gas-phase DFT
geometry optimization of 25C gave a local minimum of conformer 25B which was used in the CPCM and PCM hydration models.
Eur. J. Org. Chem. 2003, 1868Ϫ1878
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1871

G. J. Hoornaert et al.
FULL PAPER
mitted to DFT geometry optimization with the B3LYP
functional in conjunction with the 6-31G** basis set. The
results, summarized in Table 1, indicate that the lowest-en-
ergy conformer in the gas phase is 11C. When the CPCM
or PCM hydration models are used, however, the lowest-
energy conformer is 11B, in accordance with NMR analysis
(see below).
The data described above indicate that the hydrogen-
bonded conformer 11A is sterically disfavored relative to
the ‘‘open’’ forms 11B and 11C. However, the introduction
of groups at the 8a-position could counterbalance steric ef-
fects in favor of the A-type conformation, due to the repul-
sion of the 8a-substituent with the co-planar N-acetyl group
in the ‘‘open’’ forms of type B and C. This supposition was
supported by an analogous computational analysis for 8a-
methyl-1,7-naphthyridine 25 (Table 2). Random conformer
searches by both the AMBER* and the MM3* force-field
approaches yielded conformation 25A as the global mini-
mum, displaying properties conforming to the desired β-
turn induction, including the generation of a hydrogen
bond. The next two significantly different conformers ident-
ified by use of both force fields were 25D and 25C: the
former is characterized by a twist-boat form of the piperi-
dine moiety, while 25C exhibits a piperidinone boat struc-
ture. Interestingly, conformer 25B was not found by AM-
BER* but only when MM3* was used; however, a structure
representing a local minimum of 25B was generated by
DFT geometry optimization of 25C in the gas phase. After
similar geometry optimization of 25A, B, and D, the four
optimized structures were subjected to further hydration en-
ergy calculations, which revealed 25A as the most stable
conformer by 7Ϫ12 kJ/mol depending on the hydration
model used.
Figure 3. Solution conformation and relevant NOESY correlations
(gray) of 11
While conformer 11B is not desired for β-turn induction,
it seemed worthwhile to verify the criteria usually cited for
intramolecular hydrogen-bonding analysis by NMR: (1) the
temperature shift of the hydrogen-bonded proton is Յ Ϫ3
ppb/°C, and (2) if the solvent is changed, a large shift is
seen for free NH proton(s) and a small one for the hydro-
gen-bonded proton. By increasing the temperature from 298
to 343 K in steps of 5 K ([D6]DMSO) an NHCH3 tempera-
ture dependence of Ϫ2.9 and Ϫ3.8 ppb/K was determined
for the two N-Ac rotamers (Figure 4). The fact that one of
the NH proton signals seems to fulfil the criterion is decep-
tive, since the two signals coalesce at 333 K, giving rise to
second-order effects and a deviation from linearity. The
N7H proton signals show a greater temperature dependence
of Ϫ5.3 and Ϫ5.4 ppb/K. When [D6]DMSO was exchanged
for CDCl3, large shifts were observed for all NH protons.
NMR Analysis of Diamides 11 and 25
1H NMR spectra of 11 were run in CDCl3, [D6]DMSO
and D2O; the last spectrum gave the least overlap but no
NH signal. In each case a series of separate signals revealed
the existence of two slowly interconverting (cis,trans) rota-
mers about the N1-Ac amide linkage. From the NOE en-
hancement observed for 2-Heq upon presaturation of the N-
Ac protons in the NOE difference spectra, the trans rotamer
[(Z) form] was shown to be the main form in D2O but not
in [D6]DMSO, in which strong dipolar interactions with the
solvent may be involved. In all three solvents the cis-fused
Figure 4. NH temperature dependence observed for 11 and 25
In [D6]DMSO, the appearance of the spectrum of 25 was
naphthyridine moiety apparently adopts a similar confor- totally unlike that of 11. Only one N-acetyl rotamer was
mation corresponding to that of 11B, as demonstrated by observed, and the spectrum also displayed a different mul-
the relevant coupling constants for the angular proton 4a- tiplet structure for coupling of the angular proton 4a-H
3
H. The latter showed a small J-gauche value of ca. 5 Hz with the vicinal protons 5-H and 4-H. For example, 5-Hax
with both 5-H protons and a large 3J-trans value of ca. was observed as a triplet at δ ϭ 2.41 ppm, corresponding
14 Hz with 4-Hax, in accordance with the coupling values to geminal coupling with 5-Heq and trans-diaxial coupling
3
calculated for 11B.[12] The 2D NOESY spectrum in CDCl3 with 4a-H (2J·ഠ J ഠ 14 Hz). This indicates an axial posi-
or D2O also displayed a weak correlation between the phe- tion of the angular proton 4a-H with respect to the piperidi-
nyl ortho-protons and proton 8a-H (Figure 3), indicating a none ring. 5-Heq was observed as a broadened doublet, due
partial orthogonal orientation of the phenyl ring relative to to a very small (unresolved) coupling with 4a-H, at δ ϭ
the piperidinone ring in 11B.
1.98 ppm. A gauche coupling value of ca. 6 Hz was meas-
1872
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 1868Ϫ1878

Functionalizable External β-Turn Mimic Based on a cis-Fused 1,7-Naphthyridine Scaffold
FULL PAPER
ured between 4a-H and 4-Hax, which appeared as a triplet bones of 1 and 2 and the mono- and bicyclic lactam struc-
of triplets (4-Heq signal hidden). These coupling values tures 26 and 27 (Macromodel 5.0, MM3*-optimized struc-
clearly support 25A as the predominant conformation. tures), suggesting that these compounds may have similar
Since an intramolecular hydrogen bond was indicated by β-turn-inducing properties.[13] Like other turn mimics that
the conformational calculations carried out for favored con- conformationally constrain or isosterically replace the cen-
former 25A, diamide 25 was submitted to the NMR criteria tral dipeptide backbone, lactam compounds 1 and 2 can be
described above. When the solvent was changed from defined as external β-turns.[14] These can be compared to
[D6]DMSO to CDCl3, the signal of the NH proton of the internal β-turns, which try to mimic the whole β-turn tetra-
piperidinone ring was shifted from δ ϭ 7.88 ppm to δ ϭ peptide without explicitly considering the backbone ge-
6.42 ppm, while the signal of the NHMe proton did not ometry. Generally, external β-turn mimics are more rigid,
move significantly (from δ ϭ 7.80 ppm to δ ϭ 7.85 ppm). but are often unable to mimic amino acid side chains.
As expected for hydrogen bonding of the NHMe proton,
However, in comparison with the previously described
the NH temperature dependence was only Ϫ1.2 ppb/K for external mimics (e.g., 26 and 27), an asset of the naphthyri-
NHMe, versus Ϫ3.5 ppb/K for the piperidinone amide pro- dine-type mimic 2 is that it can be variably functionalized
ton.
at the stereogenic Cα2 and Cα3 positions bearing the angu-
Additional support for conformation 25A was obtained lar R3 and R6 substituents, thus allowing for modulation of
from the NOESY spectrum ([D6]DMSO). As illustrated in receptorϪligand interactions. From a molecular recognition
Figure 5, strong NOEs were found between (amongst perspective, the classification of β-turns as types IϪVII ac-
others): (1) the ortho-protons of the phenyl group and 4a- cording to the dihedral angles of the central residues is not
H/5-Heq, (2) NHMe and 5-Hax, (3) 2-Hax and 8a-CH3, and entirely satisfactory, since it does not adequately describe
(4) NCOCH3 and 2-Heq. These correlations confirm the the disposition of the Cα2 and Cα3 substituents and the
existence of conformer 25A.
orientation of the i and iϩ3 residues representing the posi-
tions at which the peptide chain would enter and exit the
β-turn, respectively. To solve this problem, the topographies
of β-turns and their mimics can be described in terms of a
single virtual torsion angle β, defined by C1, Cα2, Cα3, and
N4 of the tetrapeptide model and the interatomic distance
d between Cα1 and Cα4 (Ball et al.[15]). The collection of
naturally occurring β-turns (type I, II, ...) examined by Ball
has a broad distribution of β-values (roughly Ϯ100° with a
slight preference for small, positive β values). It therefore
seems relevant to ‘‘match’’ the β value of the mimic with
that of the target peptide. For type VI turns, however, a
much narrower β value (Ϫ19° Ͻ β Ͻ 20°) was found. We
can therefore use this interval as a boundary for the β value
of a type VI turn mimic. Thus, the DFT geometry-optim-
ized global minimum 25A was evaluated as a tight reverse
turn inducer based on three geometrical criteria (Figure 7):
(1) the virtual torsion angle β should be within 0Ϯ20°, (2)
the Cα1ϪC4 interatomic distance d should be less than or
Figure 5. Solution conformation ([D6]DMSO) and relevant
NOESY correlations (gray) of 25
Evaluation of β-Turn-Inducing Potential
Relevant structural features of mimics 1 and 2 may be
compared to those of rigidified cis-amide mimics con-
strained in a six-membered lactam ring, such as the mono-
cyclic lactam 26,[7a] and the 5(1H)-indolizinone mimic 27
(Figure 6).[7b] A good overlay was observed for the back-
˚
equal to 7 A, and (3) the distance between the N-acetyl
carbonyl oxygen atom and the methyl carboxamide proton
˚
should be smaller than 2.5 A and the NHϪO and HϪOC
angles greater than 120 and 90°, respectively (hydrogen
bonding). All of these criteria are fulfilled by the tetrapep-
tide model: the |β| value measured for (4aR,6S,8aR)-25A
was 16° [and concurrently, for (4aS,6R,8aS)-25A, Ϫ16°],
˚
the Cα1ϪCα4 interatomic distance was 5.5 A, and the
Figure 6. Compounds 1 and 2 in comparison with previously re-
ported type VI β-turn mimics 26 and 27
Figure 7. The virtual torsion angle β and the Cα1ϪCα4 interatomic
distance d measured for (4aR,6S,8aR)-25A
Eur. J. Org. Chem. 2003, 1868Ϫ1878
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1873

G. J. Hoornaert et al.
FULL PAPER
˚
NHϪOϭC distance was 2.0 A. The NHϪO and HϪOC tures were also found to adopt stable conformations, each
angles measured were 155 and 143°, respectively. Hence, with two β-turns. A molecular dynamics simulation was
these data clearly support the β-turn-inducing properties also carried out on linear AcϪAlaϪAlaϪ(4aR,6S,8aR)-
of 25A.
25ϪAlaϪAlaϪNHMe and AcϪAlaϪAlaϪ(4aS,6R,8aS)-
The picture described above is somewhat idealized, since 25ϪAlaϪAlaϪNHMe. This resulted in narrow distribution
several low-energy conformations equilibrate with 25A at profiles similar to those found for 27 by use of the param-
room temperature. To assess the contributions of all confor- eters described above (Figure 8). Moreover, 94% of the
mations that fulfil the conditions for reverse-turn induction, (4aR,6S,8aR)-25-hexapeptide and 81% of the (4aS,6R,8aS)-
a molecular dynamics simulation of 1600 ps at 300 K in 25-hexapeptide was found to be intramolecularly hydrogen-
˚
water was carried out for 25 by use of AMBER 6.0 and bonded (NHϪOϭC distance Ͻ 2.5 A; NHϪO angle Ͼ
the TIP3P water model. In this way, 8000 snapshots were 120°; HϪOC angle Ͼ 90°).
generated, from which the β values and Cα1ϪCα4 distances
d could be extracted. As can be seen from the results sum-
marized in Table 3, 66% of the conformations have a β
˚
within Ϯ20° and 87% have a d value below 7 A. According
to our proposed hydrogen-bond conditions, 25% of all con-
formations are hydrogen-bonded and 67% of the confor-
˚
mations have an NHϪOϭC distance of less than 4 A (with
the same angle conditions), the minimum distance usually
cited for significant interaction. These results indicate that
25 largely preserves its β-turn-inducing properties at 300 K
in water.
Table 3. Parameters extracted from a molecular dynamics simu-
lation of 25
Property
Compound 25
β within Ϯ20°
66%
Mean β (4aR,6S,8aR)-25A
Mean β (4aS,6R,8aS)-25A
(17 Ϯ 8)°
(Ϫ17 Ϯ 8)°
87%
˚
d within 7 A
˚
Figure 8. Interatomic distances CαiϪCαiϩ3 and pseudodihedral
angles CαiϪCαiϩ1ϪCαiϩ2ϪCαiϩ3 for both enantiomers of the
β-turn-induced by 25 in an Ala-hexapeptide; values determined by
an MD simulation (force field AMBER*, solvent H2O)
Mean d
6.1 A
H-bonded
NHϪOϭC interaction
25%
67%
Conclusion
In this context, it should be noted that in a confor-
mational study using a molecular dynamics simulation in
which several β-turn mimetics were built into cyclic Ala-
hexapeptides,[16] the rigid cis-amide 27 was shown to be
compatible with the opposing β-turn in the ring, which was
not the case for a biphenyl-based mimic. Also, in a Monte
Carlo simulation, 27 displayed the best β-turn-inducing po-
tential, including the generation of a H-bridge between the
i and iϩ3 residues. In this simulation, distribution profiles
of the following properties were generated: (a) the
donorϪacceptor distance (HNiϩ3)Ϫ(CϭOi) of the turn-sta-
bilizing bond, b) the CαiϪCαiϩ3 distance characteristic of
The intramolecular DielsϪAlder reaction of N-alkenyl-
2(1H)-pyrazinones, followed by cleavage of the strained tri-
cyclic adducts to form cis-fused 1,7-naphthyridines, can be
used to construct a novel type of rigid β-turn. Substitution
at the 8a-position of the cis-fused naphthyridine scaffold
proved to be of crucial importance for attaining the desired
conformation, as indicated by random conformer searches,
ab initio geometry optimizations, and molecular dynamics
conformational space probing around the global minimum.
The presumed existence of 25A as the predominant con-
former required for induction of β-turns and internal hy-
drogen bonding was confirmed by extensive NMR analysis.
In comparison with the previously described cis-amide
mimics 26 and 27, an additional asset of type 2 mimics is
that they can be variably functionalized at the stereogenic
Cα2 and Cα3 positions, which should allow for further
modulation of the receptorϪligand interaction.
˚
β-turns (ideal value: 4.1Ϫ4.8 A), according to Lewis et
al.,[17] and c) the pseudodihedral angle encompassing four
consecutive Cα atoms of the turn-forming amino acids or
dipeptide analogues (ideal value: 50° Ͼ Θ Ͼ Ϫ50°). The
ideal values were taken from the geometries of known β-
turns. A narrow distribution was considered important for
β-turn-inducing properties.
When an analogous molecular dynamics simulation was
Experimental Section
performed on
a
cyclic Ala-hexapeptide in which
(4aS,6R,8aS)-25 and (4aR,6S,8aR)-25 were incorporated
General Remarks: All compounds were analyzed with the analytical
(Macromodel 8.0, GB/SA hydration model), these struc- instruments described previously.[18] NMR spectra were calibrated
1874
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 1868Ϫ1878

Functionalizable External β-Turn Mimic Based on a cis-Fused 1,7-Naphthyridine Scaffold
FULL PAPER
with TMS, except for 11 in D2O (HOD) and 25 in [D6]DMSO
(CHD2SOCD3). For 11, only the 1H NMR spectroscopic data of
major isomers are given.
3,5-Dichloro-1-(4-methoxybenzyl)-2(1H)-pyrazinone (4): The prod-
uct was prepared as described in ref.[9], starting from 4-methoxy-
benzylamine (26.1 mL, 0.2 mol). Yield: 29.0 g (51%), yellow crys-
tals; m.p. 96Ϫ100 °C (EtOH). IR (KBr, cmϪ1): ν˜ ϭ 1662.8 (CϭO),
1588.7 (CϭN). 1H NMR (400 MHz, CDCl3): δ ϭ 7.28 (m, 2 H,
3Ј-H, 5Ј-H), 7.15 (s, 1 H, 6-H), 6.92 (m, 2 H, 2Ј-H, 6Ј-H), 5.03 (s,
2 H, CH2), 3.82 (s, 3 H, CH3) ppm. 13C NMR (100 MHz, CDCl3):
δ ϭ 159.9 (C-4Ј), 151.5 (C-2), 147.0 (C-3), 130.3 (C-2Ј, C-6Ј), 125.6,
125.1 (C-6, C-5), 123.8 (C-1Ј), 114.6 (C-3Ј, C-5Ј), 55.2 (CH3), 52.3
(CH2) ppm. MS EI: m/z (%) ϭ 284 (5) [Mϩ·], 121 (100)
[CH3OC6H4CH2ϩ]; exact mass calculated for C12H10Cl2N2O2
284.0119; found 284.0138.
Random Conformer Searches: The calculations were carried out
with Macromodel 5.0.[19] The Macromodel implementation of the
AMBER force field (denoted AMBER*)[20] was used with the
water GB/SA solvation model of Still et al.[21] The conformational
space was sampled by use of Goodman and Still’s internal coordi-
nate Monte Carlo search.[22] For both molecules (11 and 25), 5000
structures were generated and minimized to an energy convergence
˚
of 0.05 kJ/mol A by use of the PolakϪRibiere conjugate gradient
method implemented in Macromodel. Duplicate structures and
those greater than 50 kJ/mol above the global minimum were dis-
carded.
5-Chloro-1-(4-methoxybenzyl)-3-phenyl-2(1H)-pyrazinone (5): The
product was prepared as described in ref.[10], starting from 4
(10.0 g). Yield: 9.0 g (78%); yellow crystals; m.p. 84 °C (Et2O). IR
(KBr, cmϪ)1: ν˜
ϭ
1640.4 (CO), 1585.4 (CϭN). 1H NMR
Ab Initio Geometry Optimization: Gas-phase geometry optimiza-
tions were performed at the Density Functional Theory level in
combination with the B3LYP functional level in conjunction with
the d,p polarized basis sets 6-31G**. Hydration energies were cal-
culated by use of the polarizable continuum model (PCM) devel-
oped by Tomasi and co-workers.[23] A second solvation model used
is CPCM,[24] which is actually an implementation of the Conduc-
tor-like Screening Model (COSMO).[25] The hydration free energies
were calculated at the HF/6-31G** level. The United Atom Model
for Hartree Fock (UAHF) definition[26] was used for the construc-
tion of the solute cavity. No geometry optimizations in solvent were
performed for the hydration free energies; the calculated gas-phase
structures were used. All calculations were performed with the aid
of the Gaussian 98 package.[27]
(300 MHz, CDCl3): δ ϭ 8.36 (m, 2 H, ortho-H), 7.43 (m, 3 H,
meta-H, para-H), 7.29 (d, J ϭ 8 Hz, 2 H, 3Ј-H, 5Ј-H), 7.15 (s, 1 H,
6-H), 6.88 (d, J ϭ 8 Hz, 2 H, 2Ј-H, 6Ј-H), 5.02 (s, 2 H, CH2), 3.78
(s, 3 H, CH3) ppm. 13C NMR (75 MHz, CDCl3): δ ϭ 159.9 (C-4Ј),
154.3 (C-2), 152.0 (C-3), 134.8 [C-ipso (Ph)], 130.5, 129.2, 128.0 [C-
ortho, C-meta, C-para (Ph)], 130.3 (C-2Ј, C-6Ј), 126.3, 126.2 (C-1Ј,
C-5), 125.1 [C-6 (determined by DEPT)], 114.5 (C-3Ј, C-5Ј), 55.2
(CH3), 52.3 (CH2) ppm. MS EI: m/z (%) ϭ 326 (9) [Mϩ·], 121
(100) [CH3OC6H4CH2ϩ]; exact mass calculated for C18H15ClN2O2
326.0822; found 326.0829.
5-Chloro-3-phenyl-2(1H)-pyrazinone (6): Compound 5 (9.0 g) was
added to trifluoroacetic acid (30 mL). This solution was stirred at
reflux for 5 h [TLC monitoring: Rf (CH2Cl2) ϭ 0.31 (5), 0.10 (6)].
After evaporation of the solvent, crude 6 was purified by column
chromatography (silica gel, CH2Cl2 to 15% EtOAc/CH2Cl2). Yield:
5.0 g (88%); yellow crystals; m.p. 175 °C (EtOAc/CH2Cl2). IR
(KBr, cmϪ1): ν˜ ϭ 2692 (broad, NH), 1648 (CϭO), 1590.7 (CϭN).
1H NMR (400 MHz, [D6]DMSO):[30] δ ϭ 8.10 (m, 2 H, ortho-H),
7.67 (s, 1 H, 6-H), 7.40 (m, 3 H, meta-H, para-H) ppm. 13C NMR
(100 MHz, [D6]DMSO): δ ϭ 155.4 (C-2), 148.5 (C-3), 135.1 (C-
ipso), 130.6 (C-para), 129.3 (C-6), 128.9, 128.6 (C-ortho, C-meta)
ppm; C-5 signal not resolved. MS EI: m/z (%) ϭ 206 (100) [Mϩ·],
178 (74) [Mϩ· Ϫ CO], 143 (33) [Mϩ· Ϫ CO Ϫ Cl], 116 (33) [Mϩ·
Ϫ CO Ϫ Cl Ϫ HCN]; exact mass calculated for C18H15ClN2O2
206.0247; found 206.0254.
Molecular Dynamics of 25: The simulation was performed at con-
stant temperature (300 K) and constant pressure (1 atm) by use of
the SANDER classic module of AMBER 6.0 with the AMBER
force-field of Cornell et al.[28] This was slightly modified to handle
the unnatural residue, and Mulliken charges from the DFT calcu-
lations were used instead of the standard values. An initial solvent
˚
box of ca. 37 ϫ ca. 34 ϫ ca. 31 A, containing 866 TIP3P water
molecules,[29] was constructed around the residue. The system was
minimized until the rms energy gradient dropped below 0.1 kcal/
˚
mol·A. After a relaxation period, the simulation was subsequently
run for 1600 ps with SHAKE on all bond lengths and with a time
step of 0.002 ps. Periodic boundary conditions were used to treat
˚
the nonbonded interactions with a cut-off distance equal to 9 A.
5-Chloro-1-(4-pentenyl)-3-phenyl-2(1H)-pyrazinone (7): Cs2CO3
(1.9 g, 1.2 equiv.) was added to a solution of 6 (4.0 g) in dioxane
(150 mL). This mixture was stirred at 50 °C for 30 min, followed
by addition of 5-bromo-1-pentene (0.9 mL, 1.5 equiv.). After hav-
ing been stirred for one week, the mixture was filtered to remove
CsBr and unchanged Cs2CO3. After evaporation of the solvent,
crude 7 was purified by column chromatography (silica gel,
CH2Cl2). Yield: 3.8 g (72%); yellow oil. IR (NaCl, cmϪ1): ν˜ ϭ
Snapshots were taken every 0.2 ps, yielding 8000 conformations
for analysis.
Molecular Dynamics of Ala-Hexapeptides with 25: Molecular dy-
namics simulations on cyclo(AlaϪAla-25ϪAlaϪAla) and the lin-
ear peptide AcϪAlaϪAla-25-AlaϪAlaϪNHMe (all non-β-turn
mimic peptide bonds in the trans conformation) were performed
with the Macromodel 8.0 package with use of the AMBER* force
field. The run was set up with all standard parameters and the GB/
SA solvation model (water).[21] The system was minimized and
after an equilibration run of 200 ps, a 1000 ps dynamics run was
performed (timestep 1 fs). The criteria described above were moni-
tored. For the linear peptide, an extended and a near-ideal β-turn
structure were chosen as a starting conformation. The reported
data are from the near-ideal starting conformation. In the case of
the extended starting conformation, results were considerably po-
orer for a 1000 ps run with 200 ns equilibration. When a longer
equilibration time (1000 ps) was used, however, the results became
similar to the reported data, 91% of AcϪAlaϪAlaϪ(4aR,6S,8aR)-
25-AlaϪAlaϪNHMe and 87% of AcϪAlaϪAla-(4aS,6R,8aS)-25-
AlaϪAlaϪNHMe being hydrogen-bonded.
1
1650.1 (CϭO), 1583.5 (CϭN). H NMR (400 MHz, CDCl3): δ ϭ
8.35 (m, 2 H, ortho-H), 7.43 (m, 3 H, meta-H, para-H), 7.18 (s, 1
H, 6-H), 5.80 (ddt, J ϭ 18, 10, 7 Hz, 1 H, 4Ј-H), 5.06 (m, 2 H, 5Ј-
H), 3.93 (t, J ϭ 7 Hz, 2 H, 1Ј-H), 2.15 (q, J ϭ 7 Hz, 2 H, 3Ј-H),
1.90 (quint, J ϭ 7 Hz, 2 H, 2Ј-H) ppm. 13C NMR (100 MHz,
CDCl3): δ ϭ 154.3 (C-2), 152.0 (C-3), 136.5 (C-4Ј), 134.8 (C-ipso),
130.6, 129.2, 128.1 (C-ortho, C-meta, C-para), 126.2 (C-5), 125.9
(C-6), 116.2 (C-5Ј), 50.1 (C-1Ј), 30.5, 27.5 (C-2Ј, C-3Ј) ppm. MS
EI: m/z (%) ϭ 274 (100) [Mϩ·], 220 (92) [Mϩ· Ϫ C4H6], 191 (50)
[Mϩ· Ϫ C4H6 Ϫ CHO); exact mass calculated for C15H15ClN2O
274.0873; found 274.0873.
1-Phenyl-3,10-diazatricyclo[5.3.1.03,8]undecane-2,9-dione (9):
solution of 7 (1.0 g) in bromobenzene (20 mL) was stirred at reflux
A
Eur. J. Org. Chem. 2003, 1868Ϫ1878
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1875

G. J. Hoornaert et al.
FULL PAPER
overnight. The solvent was then evaporated, and the residue was
redissolved in water-saturated ethyl acetate. The mixture was stirred
overnight and pure 9 was collected by filtration. Yield: 0.80 g
(86%). For spectroscopic data see ref.[8]
m/z (%) ϭ 298 (4) [M·ϩ], 121 (100) [CH3OC6H4CH2ϩ]; exact mass
calculated for C13H12Cl2N2O2 298.0276; found 298.0280.
5-Chloro-1-(4-methoxybenzyl)-3,6-dimethyl-2(1H)-pyrazinone (14):
The product was prepared as described in ref.[10], starting from 13
(10.0 g). Yield: 8.3 g (89%), brown crystals (CH2Cl2), m.p. 113Ϫ114
°C (CH2Cl2). IR (KBr, cmϪ1): ν˜ ϭ 1647.5 (CϭO), 1572.3 (CϭN).
1H NMR (300 MHz, CDCl3): δ ϭ 7.13 (d, J ϭ 9 Hz, 2 H, 3Ј-H,
5Ј-H), 6.85 (d, J ϭ 9 Hz, 2 H, 2Ј-H, 6Ј-H), 5.25 (s, 2 H, CH2), 3.78
(s, 3 H, OCH3), 2.48 (q, J ϭ 1 Hz, 3 H, 3-CH3)[31], 2.40 (q, J ϭ
1 Hz, 3 H, 6-CH3)[31] ppm. 13C NMR (75 MHz, CDCl3): δ ϭ 159.3
(C-4Ј), 156.2 (C-2), 154.2 (C-3), 133.2 (C-1Ј), 128.3 (C-3Ј, C-5Ј),
126.7, 125.5 (C-5, C-6), 114.3 (C-2Ј, C-6Ј), 55.2 (OCH3), 48.1
(CH2), 20.7 (3-CH3), 16.6 (6-CH3) ppm. MS EI: m/z (%) ϭ 278
(11) [M·ϩ], 121 (100) [CH3OC6H4CH2ϩ]; exact mass calculated for
C14H15ClN2O2 278.0822; found 278.0826.
Methyl 8-Oxo-6-phenyldecahydro[1,7]naphthyridine-6-carboxylate
(10): Methanesulfonic acid (95 µL, 2.5 equiv.) was added to a solu-
tion of 9 (150 mg, 0.52 mmol) in MeOH. This solution was stirred
at reflux overnight. After evaporation of the solvent, 10% K2CO3
in water and CH2Cl2 were added to the residue. The organic phase
was separated and the aqueous phase was extracted three times
with CH2Cl2. The combined organic layers were dried with anhy-
drous K2CO3. Pure 10 was obtained after filtration and removal of
the solvent. Yield: 164 mg (97%). For spectroscopic data see ref.[8]
1-Acetyl-N-methyl-8-oxo-6-phenyldecahydro[1,7]naphthyridine-6-
carboxamide (11): Compound 10 (110 mg, 0.38 mmol) was dis-
solved in acetic anhydride (20 mL). After having been stirred at 40
°C for 2 h, the solution was concentrated to dryness and the residue
was redissolved in a solution of CH3NH2 in ethanol (20 mL,
33wt%). This solution was stirred at room temp. overnight and con-
centrated, and the residue was purified by chromatography on pre-
parative plates (5% MeOH/CH2Cl2) to yield 11 (118 mg,
0.36 mmol). Yield: 94%; white crystals, m.p. 115Ϫ120 °C (Et2O).
IR (KBr, cmϪ1): ν˜ ϭ 1666.2 (CϭO), 1546.4 (CϭO). (Z) Isomer
(74%): 1H NMR (400 MHz, CDCl3): δ ϭ 8.18 (s, 1 H, 7-H),
7.49Ϫ7.26 (m, 5 H, arom. H), 6.81 [m (q), 1 H, NHCH3], 5.30 (d,
J ϭ 5 Hz, 1 H, 8a-H), 3.58 (br. d, J ϭ 14 Hz, 1 H, 2-Heq), 3.06
(dd, J ϭ 14, 3 Hz, 1 H, 5-Heq), 2.83 (td, J ϭ 13, 2.5 Hz, 1 H, 2-
Hax), 2.79 (d, J ϭ 5 Hz, 3 H, NHCH3), 2.22 (dd, J ϭ 14, 5 Hz, 1
H, 5-Hax), 2.08 (m, 1 H, 4a-H), 2.05 (s, 3 H, COCH3), 1.88 (m, 1
H, 4-Heq), 1.69 (m, 1 H, 3-Heq), 1.39 (qt, J ϭ 13, 3.5 Hz, 1 H,
3-Hax), 1.23 (m, 1 H, 4-Hax) ppm. (E) Isomer (58%): 1H NMR
([D6]DMSO): δ ϭ 8.23 (s, 1 H, 7-H), 7.84 [q (broad), J ϭ 4 Hz, 1
H, NHCH3], 7.44Ϫ7.27 (m, 5 H, arom. H), 4.56 (d, J ϭ 5 Hz, 1
H, 8a-H), 4.25 (br. d, J ϭ 14 Hz, 1 H, 2-Heq), 3.01 (br. d, J ϭ
13 Hz, 1 H, 5-Heq), 2.64 (d, J ϭ 5 Hz, 3 H, NHCH3), 2.13 (m, 3
H, 2-Hax, 4a-H, 5-Hax), 1.94 (s, 3 H, COCH3), 1.70Ϫ1.10 (m, 4 H,
3-H, 4-H) ppm. (Z) Isomer (68%): 1H NMR (400 MHz, D2O): δ ϭ
7.61Ϫ7.49 (m, 5 H, arom. H), 5.29 (d, J ϭ 6 Hz, 1 H, 8a-H), 3.95
(br. d, J ϭ 14 Hz, 1 H, 2-Heq), 2.97 (td, J ϭ 13, 2 Hz, 1 H, 2-Hax),
2.87 (s, 3 H, NDCH3), 2.80 (dd, J ϭ 15, 6 Hz, 1 H, 5-Heq), 2.72
(dd, J ϭ 15, 5 Hz, 1 H, 5-Hax), 2.35 (m, 1 H, 4a-H), 2.27 (s, 3 H,
COCH3), 1.93 (m, 1 H, 4-Heq), 1.83 (m, 1 H, 3-Heq), 1.57 (qt, J ϭ
13, 3.5 Hz, 1 H, 3-Hax), 1.35 (qd, J ϭ 13, 3 Hz, 1 H, 4-Hax) ppm.
(Z) Isomer: 13C NMR (100 MHz, CDCl3): δ ϭ 171.9, 170.7, 170.5
(CϭO), 142.4 (C-ipso), 129.0, 128.1, 124.8 (C-ortho, C-meta, C-
para), 65.1 (C-6), 52.7 (C-8a), 43.6 (C-2), 38.1 (C-5), 33.0 (C-4a),
27.2 (NHCH3), 27.1, 25.4 (C-3, C-4), 21.4 (COCH3) ppm. (E) Iso-
mer: most signals not resolved. MS EI: m/z (%) ϭ 329 (6) [M·ϩ],
286 (18) [M·ϩ Ϫ Ac], 271 (100) [M·ϩ Ϫ CONHCH3], 229 (26) [M·ϩ
Ϫ CH2O Ϫ CONHCH3], 152 (62) [8 Ϫ H10NO2ϩ], 96 (18)
[C6H10Nϩ], 83 (16) [C5H9Nϩ]; exact mass calculated for
C18H23N3O3 329.1739; found 329.1735.
5-Chloro-3,6-dimethyl-2(1H)-pyrazinone (15): The same procedure
as employed for compound 6 was used, starting from 14 (5.0 g).
Yield: 1.9 g (67%); white crystals, decompose upon heating. IR
(KBR, cmϪ1): ν˜ ϭ 1651.5 (CϭO), 1532.7 (CϭN). 1H NMR
(300 MHz, CDCl3): δ ϭ 12.32 (br. s, 1 H, 1-H), 2.25 (s, 3 H, 3-
CH3 or 6-CH3), 2.23 (s, 3 H, 6-CH3 or 3-CH3) ppm. 13C NMR
(75 MHz, [D6]DMSO): δ ϭ 157.4 (C-2); C-3, C-5, C-6 signals not
resolved due to tautomerism, 20.1 (3-CH3), 18.4 (6-CH3) ppm. MS
EI: m/z (%) ϭ 158 (99) [M·ϩ], 130 (59) [M·ϩ Ϫ CO], 129 (100)
[M·ϩ Ϫ CO Ϫ H]; exact mass calculated for C6H7ClN2O 158.0247;
found 158.0265.
2-Chloro-3,6-dimethyl-5-(4-pentenyloxy)-2(1H)-pyrazinone
(16):
The same procedure as employed for compound 7 was used, start-
ing from 15 (1.0 g). Yield: 0.66 g (46%); colorless oil. IR (NaCl,
cmϪ1): ν˜ ϭ 1641.2 (C2ϭN1), 1543.2 (C3ϭN4) ppm. 1H NMR
(300 MHz, CDCl3): δ ϭ 5.84 (ddt, J ϭ 18, 10, 7 Hz, 1 H, 4Ј-H),
5.04 (m, 2 H, 5Ј-H), 4.31 (t, J ϭ 7 Hz, 2 H, 1Ј-H), 2.47 (s, 3 H, 3-
CH3), 2.40 (s, 3 H, 6-CH3), 2.23 (q, J ϭ 7 Hz, 2 H, 3Ј-H), 1.89
(quint, J ϭ 7 Hz, 2 H, 2Ј-H) ppm. 13C NMR (300 MHz, CDCl3):
δ ϭ 156.6 (C-2), 145.2 (C-3), 140.9 (C-5), 137.6 (C-4Ј), 137.1 (C-
6), 115.1 (C-5Ј), 65.9 (C-1Ј), 30.1 (C-3Ј), 27.9 (C-2Ј), 21.1 (3-CH3),
18.2 (6-CH3) ppm. MS EI: m/z (%) ϭ 226 (16) [M·ϩ], 158 (100)
[M·ϩ Ϫ C5H8], 130 (18) [M·ϩ Ϫ C5H8 Ϫ CO], 68 (10) [C5H8ϩ];
exact mass calculated for C11H15ClN2O 226.0873; found 226.0874.
5-(Phenylselanyl)pentan-1-amine (17): This compound was pre-
pared as described previously, starting from diphenyl diselenide
(25 g).[32]
2-{[5-(Phenylselanyl)pentyl]amino}propanenitrile (18): The product
was prepared by a previously described procedure,[33] starting from
the amine (7.0 g, 28.9 mmol) at 60 °C and with a reaction time of
12 h. The compound was used for the preparation of 19 without
purification.
3,5-Dichloro-6-methyl-1-[6-(phenylselanyl)pentyl]-2(1H)-pyrazinone
(19): HCl was bubbled for 15 min through an ethereal solution of
18 (ca. 28.9 mmol). The mixture was then concentrated to dryness
and the residue was suspended in dry CH2Cl2 (100 mL). Oxalyl
chloride (6.0 mL, 69.1 mmol) in dry CH2Cl2 (100 mL) was added
3,5-Dichloro-1-(4-methoxybenzyl)-6-methyl-2(1H)-pyrazinone (13):
The product was prepared as described in ref.[9], starting from 4-
methoxybenzylamine (26.1 mL, 0.2 mol). Yield: 26.8 g (45%); white to this suspension with stirring. After 30 min, dry triethylammon-
crystals, m.p. 112 °C (EtOH). IR (KBr, cmϪ1): ν˜ ϭ 1667.2 (CϭO),
ium chloride (19.7 g, 144.0 mmol) was added, and the mixture was
1567.6 (CϭN). 1H NMR (300 MHz, CDCl3): δ ϭ 7.16 (d, J ϭ stirred at room temperature under N2 for additional 2 d. After
9 Hz, 2 H, 3Ј-H, 5Ј-H), 6.87 (d, J ϭ 9 Hz, 2 H, 2Ј-H, 6Ј-H), 5.30 evaporation of the solvent and unchanged oxalyl chloride, the resi-
(s, 2 H, CH2), 3.79 (s, 3 H, OCH3), 2.45 (s, 3 H, 6-CH3) ppm. 13C
due was purified by column chromatography (silica gel, CH2Cl2)
NMR (75 MHz, CDCl3): δ ϭ 159.6 (C-4Ј), 153.1 (C-2), 143.7 (C- to afford 19. Yield: 8.4 g (72%) from 17; yellow crystals, m.p.
3), 136.0 (C-1Ј), 128.6 (C-3Ј, C-5Ј), 125.9, 123.8 (C-5, C-6), 114.5 68Ϫ71 °C (CH2Cl2). IR (KBr, cmϪ1): ν˜ ϭ 1667.5 (CϭO), 1567.8
1
(C-2Ј, C-6Ј), 55.3 (OCH3), 49.7 (CH2), 16.8 (6-CH3) ppm. MS EI: (CϭN). H NMR (300 MHz, CDCl3): δ ϭ 7.48 (m, 2 H, ortho-H)
1876
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 1868Ϫ1878

Functionalizable External β-Turn Mimic Based on a cis-Fused 1,7-Naphthyridine Scaffold
7.26Ϫ7.24 (m, 3 H, meta-H, para-H), 4.03 (m, 2 H, 1Ј-H), 2.91 (t, m/z (%) ϭ 270 (32) [M·ϩ], 186 (100) [M·ϩ Ϫ (CH2)3NCO], 158 (15)
FULL PAPER
J ϭ 7 Hz, 2 H, 5Ј-H), 2.46 (s, 3 H, CH3), 1.78Ϫ1.66 (m, 4 H, 2Ј-H,
[M·ϩ Ϫ (CH2)3NCO Ϫ CO]; exact mass calculated for C16H18N2O2
4Ј-H), 1.53 (quint, J ϭ 7 Hz, 2 H, 3Ј-H) ppm. 13C NMR (75 MHz, 270.1368; found 270.1371. Separation of the Enantiomers: The sep-
CDCl3): δ ϭ 152.5 (C-2), 143.4 (C-3), 135.2 (C-ipso), 132.5 (C- aration was achieved on a ‘‘Diacel OJ’’ column with chiral station-
ortho), 130.0 (C-5), 129.0, 126.9 (C-meta, C-para), 123.7 (C-6), 47.2
ary phase, with a solvent gradient: 50% hexane/50% ethanol to 20%
(C-1Ј), 29.5 (C-2Ј), 27.2, 27.3, 26.7 (C-3Ј, C-4Ј, C-5Ј), 16.4 (CH3) hexane/80% ethanol during 20 min. Flow: 1 mL/min.
ppm. MS EI: m/z (%) ϭ 404 (8) [M·ϩ], 247 [M·ϩ Ϫ SePh], 192 (28)
N,8a-Dimethyl-8-oxo-6-phenyldecahydro[1,7]naphthyridine-6-
[M·ϩ Ϫ C10H12Se], 178 (81) [M·ϩ Ϫ C11H14Se], 156 (22) [PhSeϩ],
carboxamide (24): The same procedure as employed for the prep-
aration of compound 10 was used, starting from 23 (100 mg).
77 (24) [Phϩ], 69 (100) [C5H9ϩ]; exact mass calculated for
C16H18Cl2N2OSe 403.9961; found 403.9948.
Yield: 111 mg (99%); slightly yellow crystals, m.p. 173Ϫ176 °C
(CH2Cl2). IR (KBR, cmϪ1): nu(tilde ) ϭ 1737.8 (CϭO ester),
5-Chloro-6-methyl-3-phenyl-1-[5-(phenylselanyl)pentyl]-2(1H)-
pyrazinone (20): This product was prepared by a general procedure
described in ref.[10], starting from 19 (5.0 g). Yield: 4.0 g (72%); yel-
low crystals, m.p. 80Ϫ82 °C (Et2O). IR (KBr, cmϪ1): ν˜ ϭ 1649.5
(CϭO), 1551.8 (CϭN). 1H NMR (300 MHz, CDCl3): δ ϭ
8.34Ϫ7.23 (m, 10 H, arom. H), 4.07 (m, 2 H, 1Ј-H), 2.92 (t, J ϭ
7 Hz, 2 H, 5Ј-H), 2.15 (s, 3 H, CH3), 1.80Ϫ1.58 (m, 4 H, 2Ј-H, 4Ј-
H), 1.56 (m, 2 H, 3Ј-H) ppm. 13C NMR (75 MHz, CDCl3): δ ϭ
154.9 (C-2), 148.3 (C-3), 135.1 (SeC-ipso), 134.2 (3-C-ipso), 132.6,
130.1, 130.0, 129.1, 128.9, 128.0, 126.8, 126.5 (C-ortho, C-meta, C-
para, C-5, C-6), 46.2 (C-1Ј), 29.6 (C-2Ј), 27.44, 27.38, 27.0 (C-3Ј,
C-4Ј, C-5Ј), 16.8 (CH3) ppm. MS EI: m/z (%) ϭ 446 (16) [M·ϩ],
289 (100) [M·ϩ Ϫ SePh], 234 (18) [M·ϩ Ϫ C4H7SePh], 221 (31)
[M·ϩ Ϫ C5H9SePh], 192 (31) [M·ϩ Ϫ C12H14OSe], 156 (13) [PhSeϩ],
77 (18) [Phϩ], 69 (57) [C5H9ϩ]; exact mass calculated for
C22H23ClN2OSe 446.0664; found 446.0655.
1656.0 (CϭO lactam). 1H NMR (400 MHz, CDCl3):
δ ϭ
7.43Ϫ7.29 (m, 5 H, arom. H), 6.46 (s, 1 H, 7-H), 3.76 (s, 3 H,
OCH3), 3.01 (ddd, J ϭ 14.5, 5.5, 1 Hz, 1 H, 5-Heq), 2.86 (m (ddt),
1 H, 2-Heq), 2.68 (ddd, J ϭ 13, 10, 3 Hz, 1 H, 2-Hax), 2.26 (dd,
J ϭ 14.5, 4 Hz, 1 H, 5-Hax), 2.20 (br. s, 1 H, 1-H), 1.78Ϫ1.69 (m,
2 H, 4-Heq, 4a-H), 1.63 (m, 1 H, 3-Heq), 1.51 [m (qt), 1 H, 3-Hax],
1.31 (s, 3 H, 8a-CH3), 1.15 [m (qd), 1 H, 4-Hax] ppm. 13C NMR
(75 MHz, CDCl3): δ ϭ 175.7 (CO2CH3), 172.1 (C-8), 142.1 (C-
ipso), 129.1, 128.2, 124.0 (C-ortho, C-meta, C-para), 64.4 (C-6), 56.9
(C-8a), 53.3 (OCH3), 42.7 (C-2), 37.7 (C-4a), 35.5 (C-5), 27.6 (C-
4), 26.1 (8a-CH3), 24.3 (C-3) ppm. Ms EI: m/z (%) ϭ 302 (2) [M·ϩ],
287 (10) [M·ϩ Ϫ CH3], 259 (9) [M·ϩ Ϫ C2H5N], 110 (100)
[C7H12Nϩ], 97 (39) [C6H11Nϩ]; exact mass calculated for
C17H22N2O3 302.1630; found 302.1634.
1-Acetyl-N,8a-dimethyl-8-oxo-6-phenyldecahydro[1,7]naphthyridine-
6-carboxamide (25): The same procedure as employed for the prep-
aration of compound 11 was used, starting from 24 (100 mg).
Yield: 114 mg (100%); white crystals, m.p. 267Ϫ269 °C (Et2O). IR
(KBr, cmϪ1): ν˜ ϭ 1665.9 (CϭO), 1546.3 (CϭO). 1H NMR
(400 MHz, [D6]DMSO, 295 K): δ ϭ 7.85 (s, 1 H, 7-H), 7.82 (br. q,
J ϭ 4 Hz, 1 H, NHMe), 7.45 (d, J ϭ 8 Hz, 2 H, ortho-H), 7.35 (t,
J ϭ 7 Hz, 2 H, meta-H), 7.25 (t, J ϭ 7 Hz, 1 H, para-H), 3.60 (br.
d, J ϭ 14 Hz, 1 H, 2-Heq), 3.24 (ddd, J ϭ 14, 11, 4 Hz, 1 H, 2-
Hax), 2.61 (d, J ϭ 5 Hz, 3 H, NHCH3), 2.41 (t, J ϭ 14 Hz, 1 H, 5-
Hax), 2.05 (s, 3 H, COCH3), 1.98 (br. d, J ϭ 14 Hz, 1 H, 5-Heq),
1.88 (tt, J ϭ 14, 5 Hz, 1 H, 4-Hax), 1.60Ϫ1.35 (m, 3 H, 4a-H, 3-
H), 1.40 (s, 3 H, 8a-CH3), 1.31 (br. d, J ϭ 14 Hz, 1 H, 4-Heq) ppm.
13C NMR: δ ϭ 172.8, 171.1, 169.5 (CϭO), 142.7 (C-ipso), 127.8,
127.0 (C-meta, C-para), 126.1 (C-ortho), 63.9 (C-6), 58.4 (C-8a),
42.3 (C-2), 34.0 (C-4a), 33.8 (C-5), 26.1 (NHCH3), 25.1 (C-4), 23.2
(COCH3), 22.7 (8a-CH3), 21.8 (C-3) ppm. MS EI: m/z (%) ϭ 343
(0.4) [M·ϩ], 285 (100) [M·ϩ Ϫ CONHCH3], 243 (79) [M·ϩ Ϫ CH2O
Ϫ CONHCH3], 215 (16) [C14H19N2ϩ], 198 (20) [C14H16Nϩ], 186
(27) [C12H12NOϩ]; exact mass calculated for C19H25N3O3
343.1896; found 343.1902.
5-Chloro-6-methyl-1-(4-pentenyl)-3-phenyl-2(1H)-pyrazinone (21):
mCPBA (70% with water, 0.4 g, 3 equiv.) in CH2Cl2 (20 mL) was
added dropwise to a cooled (Ϫ15 °C) solution of 20 (500 mg,
11 mmol) in CH2Cl2 (50 mL). This mixture was warmed to room
temp. and was then stirred for 30 min. After addition of DMS (98
µl, 2 equiv.) to work up the excess of reagent and DIPA (448 µl, 6
equiv.), elimination was effected by stirring at 60 °C overnight.
After evaporation of the solvent, crude 21 was purified by column
chromatography (silica gel, CH2Cl2). Yield: 297 mg (92%); yellow
crystals, m.p.: 78Ϫ82 °C (Et2O). IR (KBr, cmϪ1): ν˜ ϭ 1644.9 (Cϭ
1
O), 1545.9 (CϭN). H NMR (300 MHz, CDCl3): δ ϭ 8.34 (m, 2
H, ortho-H), 7.43Ϫ7.40 (m, 3 H, meta-H, para-H), 5.84 (ddt, J ϭ
18, 10, 7 Hz, 1 H, 4Ј-H), 5.15 (m, 2 H, 5Ј-H), 4.09 (m, 2 H, 1Ј-H),
2.52 (s, 3 H, 6-CH3), 2.21 (quart, J ϭ 7 Hz, 2 H, 3Ј-H), 1.82 (quint,
J ϭ 7 Hz, 2 H, 2Ј-H) ppm. 13C NMR (75 MHz, CDCl3): δ ϭ 155.0
(C-2), 148.2 (C-3), 136.6 (C-4Ј), 135.1 (C-5), 134.2 (C-ipso), 130.0,
128.9, 128.0 (C-ortho, C-meta, C-para), 126.5 (C-6), 116.1 (C-5Ј),
45.9 (C-1Ј), 31.0 (C-3Ј), 26.8 (C-2Ј), 16.7 (6-CH3) ppm. MS EI:
m/z (%) ϭ 288 (75) [M·ϩ], 273 (100) [M·ϩ Ϫ CH3], 234 (54) [M·ϩ
Ϫ C4H6], 220 (70) [M·ϩ Ϫ C5H8], 205 (32) [M·ϩ Ϫ C4H6 Ϫ CHO],
192 (27) [M·ϩ
Ϫ C5H8 Ϫ CO]; exact mass calculated for
C16H17ClN2O 288.1029; found 288.1034.
Acknowledgments
8-Methyl-1-phenyl-3,10-diazatricyclo[5.3.1.03,8]undecane-2,9-dione
(23): The same procedure as employed for the preparation of com-
pound 9 was used, starting from 21 (200 mg). Reaction time: 7 d.
Yield: 146 mg (78%); white crystals, m.p. 243Ϫ244 °C (EtOAc). IR
The authors thank the FWO (Fund for Scientific Research Ϫ Flan-
ders, Belgium) and the IUAP-4-11 funding by DWTC. We are
grateful to R. De Boer for HRMS measurements, and to Prof. Piet
Herdewyn of the Division of Medicinal Chemistry for providing
Macromodel 5.0 and AMBER 6.0 for the MC and MD calcu-
lations. F. R. thanks the K. U. Leuven and W. D. B. [Postdoctoral
Fellow of the Fund for Scientific Research Flanders (Belgium)
FWO Ϫ Vlaanderen] thanks the FWO for the fellowships received.
1
(KBr, cmϪ1): ν˜ ϭ 3164.4 (NH), 1695.4 (CϭO), 1686.0 (CϭO). H
NMR (400 MHz, CDCl3): δ ϭ 7.45Ϫ7.39 (m, 5 H, arom. H), 6.48
(s, 1 H, 10-H), 4.00 (dd, J ϭ 14, 6 Hz, 1 H, 4-Heq), 3.25 (td, J ϭ
13, 4 Hz, 1 H, 4-Hax), 2.61 (dd, J ϭ 13, 10 Hz, 1 H, 11Ј-H), 2.30
(m, 1 H, 7Ј-H), 2.19 (m, 1 H, 6-Hax), 2.08 (dd, J ϭ 13, 2.5 Hz, 1
H, 11-H), 1.80 (m, 1 H, 5-Hax), 1.70 (s, 3 H, 8-CH3), 1.58 (m, 1 H,
6-Heq), 1.35 (m, 1 H, 5-Heq) ppm. 13C NMR (100 MHz, CDCl3):
δ ϭ 179.0 (C-2), 171.2 (C-9), 135.3 (C-ipso), 128.9, 128.8, 127.3 (C-
ortho, C-meta, C-para), 63.6, 63.1 (C-1, C-8), 42.8 (C-4), 37.2 (C-
11), 36.0 (C-7), 23.9 (C-6), 16.3 (C-5), 15.6 (CH3) ppm. MS EI:
[1]
K.-C. Chou, Anal. Biochem. 2000, 286, 1Ϫ16 and references
cited herein.
[2]
For leading reviews on β-turn mimics and peptidomimetics see:
[2a]
V. J. Hruby, P. M. Balse, Curr. Med. Chem. 2000, 7,
Eur. J. Org. Chem. 2003, 1868Ϫ1878
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1877

G. J. Hoornaert et al.
FULL PAPER
[2b]
[15]
[16]
[17]
[18]
945Ϫ970.
R. M. J. Liskamp, Recl. Trav. Chim. Pays-Bas
J. B. Ball, R. A. Hughes, P. F. Alewood, P. R. Andrews, Tetra-
hedron 1993, 49, 3467Ϫ3478.
G. Müller, G. Hessler, H. Y. Decornez, Angew. Chem. Int. Ed.
2000, 39, 894Ϫ896.
P. N. Lewis, F. A. Momany, H. A. Scheraga, Biochim. Biophys.
Acta 1973, 303, 211Ϫ229.
F. J. R. Rombouts, D. A. J. Vanraes, J. Wynendaele, P. K.
Loosen, I. Luyten, S. Toppet, F. Compernolle, G. J. Hoornaert,
Tetrahedron 2001, 57, 3209Ϫ3220.
F. Mohamadi, N. G. J. Richards, W. C. Guida, R. Liskamp,
M. Lipton, C. Caufield, G. Chang, T. Hendrickson, W. C. Still,
J. Comput. Chem. 1990, 11, 440Ϫ467.
S. J. Weiner, P. A. Kollman, D. A. Case, U. C. Singh, C.
Ghio, G. Algona, S. Profeta, Jr., P. J. Weiner, J. Am. Chem.
1994, 113, 1Ϫ19. [2c] J. B. Ball, P. F. Alewood, J. Mol. Recognit.
[2d]
1990, 3, 55Ϫ64.
S. Hanessian, G. McNaughton-Smith, H.-
G. Lombart, W. D. Lubell, Tetrahedron 1997, 53,
12789Ϫ12854.
[3]
For some leading references on the incorporation of β-turn
[3a]
mimics see:
R. M. Freidinger, D. F. Veber, D. S. Perlow, J.
[3b]
R. Brooks, R. Saperstein, Science 1980, 210, 656Ϫ658.
G. Hinds, N. G. J. Richards, J. A. Robinson, J. Chem. Soc.,
M.
[19]
[3c]
Chem. Commun. 1988, 22, 1447Ϫ1449.
M. Kahn, Synlett
D. Gramberg, C. Weber, R. Beeli, J. In-
[3d]
1993, 3, 821Ϫ826.
glis, C. Burns, J. A. Robinson, Helv. Chim. Acta 1995, 78,
1588Ϫ1606. [3e] A. A. Virgilio, A. A. Bray, W. Zhang, L. Trinh,
M. Snyder, M. M. Morrissey, J. A. Ellman, Tetrahedron 1997,
53, 6635Ϫ6644.
W. M. De Borggraeve, F. R. Rombouts, E. V. Van der Eycken,
S. M. Toppet, G. J. Hoornaert, Tetrahedron Lett. 2001, 42,
5693Ϫ5695.
Recent examples are: [5a] Tuftsin: C. V. Valdeavella, H. D. Blatt,
[20] [20a]
[20b]
Soc. 1984, 106, 765Ϫ784.
S. J. Weiner, P. A. Kollman, D.
T. Nguyen, D. A. Case, J. Comput. Chem. 1986, 7, 230Ϫ252.
W. C. Still, A. Tempczyk, R. C. Hawley, T. Hendrickson, J.
Am. Chem. Soc. 1990, 112, 6127Ϫ6129.
[4]
[5]
[21]
[22] [22a]
G. Chang, W. C. Guida, W. C. Still, J. Am. Chem. Soc.
[22b]
1989, 111, 4379Ϫ4386.
M. Saunders, K. N. Houk, Y.-D.
Wu, W. C. Still, M. Lipton, G. Chang, W. C. Guida, J. Am.
Chem. Soc. 1990, 112, 1419Ϫ1427.
S. Miertus, E. Scrocco, J. Tomasi, J. Chem. Phys. 1981, 55,
117Ϫ129.
A. Klamt, G. Schuurmann, J. Chem. Soc., Perkin Trans. 2
1993, 799Ϫ805.
A. Klamt, J. Phys. Chem. 1995, 99, 2224Ϫ2235.
A. Klamt, V. Jonas, T. Bürger, J. C. W. Lohrenz, J. Phys. Chem.
A 1998, 102, 5074Ϫ5085.
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, R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Milliam,
A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi,
V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C.
Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala,
Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, 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, C. Gonzalez, M. Challacombe,
P. M. W. Gill, B. G. Johnson, W. Chen, M. W. Wong, J. L.
Andres, M. Head-Gordon, E. S. Replogle, J. A. Pople, Gaussian
98 (Revision A.5), Gaussian Inc., Pittsburgh, PA, 1998.
W. D. Cornell, P. Cieplak, C. I. Bayly, I. R. Gould, K. M.
Merz, Jr., D. M. Ferguson, D. C. Spellmeyer, T. Fox, J. W.
Caldwell, P. A. Kollman, J. Am. Chem. Soc. 1995, 117,
5179Ϫ5197.
[5b]
B. M. Pettitt, Int. J. Pept. Protein Res. 1995, 46, 372Ϫ380.
Decorin/DS-PGII: Y. J. Wang, P. G. Scott, J. Sejbal, G. Koto-
vych, Can. J. Chem. 1996, 74, 389Ϫ401.
Gramberg, C. Weber, R. Beeli, J. Inglis, C. Bruns, J. A. Robin-
[5c]
Somatostatin: D.
[23]
[24]
son, Helv. Chim. Acta 1995, 78, 1588Ϫ1606.
A. Jabs, M. S. Weiss, R. Hilgenfeld, J. Mol. Biol. 1999, 286,
291Ϫ304.
[6]
[25]
[26]
[7] [7a]
D. S. Kemp, E. T. Sun, Tetrahedron Lett. 1982, 37,
[7b]
3759Ϫ3760.
J.-P. Dumas, J. P. Germanas, Tetrahedron Lett.
1994, 35, 1493Ϫ1496. [7c] J. A. Robinson, D. Gramberg, Tetra-
[27]
[7d]
hedron Lett. 1994, 6, 861Ϫ864.
manas, J. Org. Chem. 1997, 62, 2847Ϫ2852.
P. A. Burney, M. M. Campbell, D. J. Osguthorpe, Bioorg. Med.
K. Kyonghee, J. P. Ger-
[7e]
P. K. C. Paul,
[7f]
Chem. Lett. 1992, 2, 141Ϫ144.
Am. Chem. Soc. 1999, 64, 3312Ϫ3321.
ies, A. B. Holmes, J. Chem. Soc., Perkin Trans. 1 2000,
L. Halab, W. D. Lubell, J.
[7g]
S. Derrer, J. E. Dav-
2957Ϫ2967.
[8]
F. J. R. Rombouts, W. De Borggraeve, S. M. Toppet, F. Com-
pernolle, G. J. Hoornaert, Tetrahedron Lett. 2001, 42,
7397Ϫ7399.
J. Vekemans, C. Pollers-Wiee¨rs, G. Hoornaert, J. Heterocycl.
Chem. 1983, 20, 919Ϫ923.
K. J. Buysens, D. M. Vandenberghe, G. J. Hoornaert, Tetra-
hedron 1996, 52, 9161Ϫ9178.
This method is misleadingly called a Monte Carlo conformer
search. Monte Carlo simulations also generate structures by
random displacements, but do not minimize the resulting struc-
tures. Instead, energies are calculated immediately and are used
to obtain thermodynamic properties.
[9]
[10]
[11]
[28]
[29]
W. L. Jorgensen, J. Chandrasekhar, J. D. Madura, R. W. Impey,
M. L. Klein, J. Chem. Phys. 1983, 79, 926Ϫ935.
[30] 1H and 13C NMR spectra were taken at 50 °C, after addition
[12]
[13]
Feature implemented in Macromodel, based on: C. A. G Haas-
noot, F. A. A. M. de Leeuw, C. Altona, Tetrahedron 1980, 36,
2783Ϫ2792.
Except for 26, which can be viewed as a bridged GlyϪGly di-
peptide, the bridged cis-amide mimics 1, 2 and 27 shown in
Figure 5 are the enantiomeric forms corresponding to the natu-
ral -amino acids after omission of the bridging elements.
of a drop of DCl.
[31]
Resolution-enhancing LorentzϪGauss function applied to
FID: lb: Ϫ2.0 Hz; gb: 0.5 Hz.
[32]
W. R. Bowman, P. T. Stephenson, N. K. Terret, A. R. Young,
Tetrahedron 1995, 51, 7959Ϫ7980.
[33]
T. Inaba, M. Fujita, K. Ogura, J. Org. Chem. 1991, 56,
1274Ϫ1279.
[14]
J. B. Ball, P. F. Alewood, J. Mol. Recognit. 1990, 3, 55Ϫ64.
Received November 24, 2002
1878
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 1868Ϫ1878