Fluorinated and iodinated templates for syntheses of β-turn peptidomimetics

Article abstract of DOI:10.1016/S0040-4020(02)01086-4

Various 2-fluoro-5-nitrobenzoic acids and the homocysteine derivative 2 have been combined in solid phase syntheses of the peptidomimetic types 3-6. NMR and CD data collected for some of these compounds indicate that a transannular SO to HN hydrogen bond stabilizes β-turn conformations for the sulfones and one of the sulfoxide epimers. An extensive library of compounds was made and studied to test this assertion.

Full text of DOI:10.1016/S0040-4020(02)01086-4

Tetrahedron 58 (2002) 8743–8750
Fluorinated and iodinated templates for syntheses
of b-turn peptidomimetics
*
Luyong Jiang and Kevin Burgess
Department of Chemistry, Texas A&M University, P.O. Box 30012, College Station, TX 77842-3012, USA
Received 22 April 2002; revised 23 August 2002; accepted 26 August 2002
Abstract—Various 2-fluoro-5-nitrobenzoic acids and the homocysteine derivative 2 have been combined in solid phase syntheses of the
peptidomimetic types 3–6. NMR and CD data collected for some of these compounds indicate that a transannular SO to HN hydrogen bond
stabilizes b-turn conformations for the sulfones and one of the sulfoxide epimers. An extensive library of compounds was made and studied
to test this assertion. q 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
obtaining protected, enantiomerically pure, homocysteine
starting materials, B.12–16 The work described here was
made possible by a gift of 100 g of the homocysteine
derivative 1 from Bristol Myers Squibb. In preliminary
work, now communicated, we transformed this into an
appropriately protected derivative 2, and made a small
library of compounds A (X¼S, SO {2 epimers}, and SO2;
n¼1). It was established that the cyclic thiols A did not show
any pronounced conformational bias towards b-turn con-
formations in DMSO solution. Curiously, the corresponding
sulfones (X¼SO2), and one of the sulfoxide epimers
(X¼SO) did have biases towards b-turn conformations in
solution.17
Fluoronitrobenzoic acid derivatives are useful electrophiles
for nucleophilic aromatic substitution reactions (SNAr), and
they have been used in diverse solid phase syntheses for
various purposes.1–5 In these labs, such substrates have
been used in solid phase macrocyclization reactions to form
b-turn mimics of the type A.6–10 We are interested in using
such compounds to mimic or disrupt protein–protein
interactions, especially those between the nerve growth
factor (NGF) and its high affinity receptor (TrkA).11 To that
end, many compounds have been prepared in this series,
mostly with X¼N or O.
The observations described above led us to propose that the
b-turn conformations could be stabilized by transannular
SO to iþ2NH hydrogen bonds, as illustrated in Fig. 1. We
also proposed that the absolute stereochemistry of the
sulfoxide chiral centers could be predicted via confor-
mational analyses. While stereochemistries assigned in this
way are not incontrovertible, the weight of evidence in their
favor is substantial, and becomes more compelling as the
number of compounds studied is increased. The preliminary
study focused on a series of 14 compounds, but here a more
extensive study is reported.
Until recently, a conspicuous omission in the studies of turn
analogs A has been the compounds where X¼S and n¼1,
i.e. homocysteine derivatives. They were not investigated
extensively due to the cost and difficulties associated with
Keywords: b-turn peptidomimetics; fluoronitrobenzoic acid derivatives;
nerve growth factor.
*
Corresponding author. Tel.: þ1-409-845-4345; fax: þ1-409-845-8839;
Figure 1. b-Turn conformations and SO to HN transannular H-bonds may
be intimately linked.
e-mail: burgess@tamu.edu
0040–4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(02)01086-4

8744
L. Jiang, K. Burgess / Tetrahedron 58 (2002) 8743–8750
Figure 2. Fluoronitrobenzoic acid derivatives used to prepare the
peptidomimetics.
This paper describes work designed to test if b-turn
conformations could be correlated with the sulfones and
one configuration of the sulfoxide stereoisomer for a much
larger set of compounds. Thus solid phase syntheses of a
library of 60 peptidomimetics 3–6 is described. To do this,
other electrophilic templates for the SNAr reaction were
used in addition to 2-fluoro-5-nitrobenzoic acid, the
template used in the work communicated (giving some of
the compounds 3 in Fig. 2). Specifically the difluorinated
and iodinated templates also shown in Fig. 2 were used;
these were chosen on the basis of commercial availability or
ease of preparation. The oxidation state of the sulfur was
also varied. Conformational analyses (various NMR
experiments and CD studies) have now been performed
for a total of 20 compounds. These experiments were
designed to reveal if b-turn conformations were always
observed for the sulfone and one stereoisomer of the
sulfoxides. For nine compounds, these physical data were
then compared with computer simulated (via quenched
molecular dynamics (QMD))18,19 conformers to identify
ones that are predicted to be highly populated, low in
energy, and match the NMR data well. Those experiments
facilitate correlations of the NMR data with the predicted
favorable conformations, to further gauge the degree of
confidence that can be ascribed to configurational assign-
ments in this series.
Scheme 2. Synthesis of the peptidomimetics 3–6.
2. Syntheses of peptidomimetics 3–6
Scheme 1 illustrates how the disulfide 1 was transformed
into the protected amino acid 2. Briefly, the trifluoroacetate
groups of 1 were removed under basic conditions, the amine
functionalities were FMOC-protected, then the disulfide
was reduced and the free thiol group was protected with a
trityl group.
Preparation of the cyclic macrocycles 3–6 (Scheme 2)
began with a series of HBTU/HOBt20 couplings following a
conventional FMOC approach21 to obtain the open chain
intermediates 7, which were then capped with the appro-
priate fluoronitrobenzoic acid derivative (as indicated in
Fig. 2). The S-Trt group was then removed with dilute
trifluoroacetic acid allowing the SNAr macrocyclization to
proceed. Cyclic thioethers of the series 3–6 were obtained
at that stage by cleavage from the resin. Subsequently,
oxidations to the sulfone and sulfoxide compounds were
performed in solution. Thus, the sulfones were prepared
using hydrogen peroxide in formic acid, while the
sulfoxides were formed via treatment with sodium periodate
in aqueous acetonitrile.
Scheme 1. Synthesis of 2.

L. Jiang, K. Burgess / Tetrahedron 58 (2002) 8743–8750
Table 1. Peptidomimetics 3–6 prepared as indicated in Scheme 2
8745
Compound
X
X1
X2
R1
R2
Crude HPLC purity (%)a
Isolated yield (%)
3a
3b
3c
3d
3e
3f
3g
3h
3i
3j
3k
3l
3m
3n
3o
3p
3q
3r
3s
S
S
S
S
S
S
S
S
S
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
(CH2)2CO2H
H
(CH2)4NH2
87
88
84
84
81
90
90
93
95
30
62
18
70
22
68
81
82
75
78
71
85
85
87
83
32
34
31
27
22
38
34
37
42
5
(CH2)4NH2
(CH2)4NH2
(CH2)3NHC(vNH)NH2
CH2CONH2
H
CH3
CH2C6H4OH
CH2C6H5
(CH2)4NH2
(CH2)4NH2
CH2C6H4OH
CH2C6H4OH
CH2C6H5
CH(CH3)C2H5
CH(CH3)C2H5
CH2CONH2
CH2C6H4OH
CH3
CH2C6H5
CH(CH3)C2H5
(CH2)2CO2H
(CH2)2CO2H
CH2C6H5
(R)-SO
(S)-SO
(R)-SO
(S)-SO
(R)-SO
(S)-SO
SO2
SO2
SO2
SO2
SO2
8
4
16
4
CH2C6H5
CH(CH3)C2H5
CH(CH3)C2H5
(CH2)2CO2H
H
CH(CH3)C2H5
CH(CH3)C2H5
CH2CONH2
CH2C6H4OH
CH3
CH2C6H5
13
21
26
27
19
16
31
28
32
27
(CH2)4NH2
(CH2)4NH2
(CH2)4NH2
(CH2)3NHC(vNH)NH2
CH2CONH2
H
CH3
CH2C6H4OH
CH2C6H5
3t
3u
3v
3w
3x
SO2
SO2
SO2
SO2
CH2C6H5
CH(CH3)C2H5
4a
4b
4c
4d
4e
4f
4g
4h
4i
S
S
S
S
S
S
SO2
SO2
SO2
SO2
SO2
SO2
F
F
F
F
F
F
F
F
F
F
F
F
H
H
H
H
H
H
H
H
H
H
H
H
(CH2)2CO2H
H
(CH2)4NH2
(CH2)4NH2
(CH2)4NH2
(CH2)3NHC(vNH)NH2
CH2CONH2
H
(CH2)4NH2
(CH2)4NH2
(CH2)4NH2
(CH2)3NHC(vNH)NH2
CH2CONH2
H
81
42
65
85
87
79
79
32
46
78
81
72
33
12
18
27
21
24
25
5
12
17
21
20
CH(CH3)C2H5
CH(CH3)C2H5
CH2CONH2
CH2C6H4OH
(CH2)2CO2H
H
CH(CH3)C2H5
CH(CH3)C2H5
CH2CONH2
CH(OH)CH3
4j
4k
4l
5a
5b
5c
5d
5e
5f
5g
5h
5i
S
S
S
S
S
S
SO2
SO2
SO2
SO2
SO2
SO2
H
H
H
H
H
H
H
H
H
H
H
H
F
F
F
F
F
F
F
F
F
F
F
F
(CH2)2CO2H
H
(CH2)4NH2
(CH2)4NH2
(CH2)4NH2
(CH2)3NHC(vNH)NH2
CH2CONH2
H
(CH2)4NH2
(CH2)4NH2
(CH2)4NH2
(CH2)3NHC(vNH)NH2
CH2CONH2
H
85
72
81
92
85
79
79
65
70
85
82
70
32
19
27
42
35
22
26
11
25
31
22
14
CH(CH3)C2H5
CH(CH3)C2H5
CH2CONH2
CH2C6H4OH
(CH2)2CO2H
H
CH(CH3)C2H5
CH(CH3)C2H5
CH2CONH2
CH2C6H4OH
5j
5k
5l
6a
6b
6c
6d
6e
6f
6g
6h
6i
S
S
S
S
S
S
SO2
SO2
SO2
SO2
SO2
SO2
I
I
I
I
I
I
I
I
I
I
I
I
H
H
H
H
H
H
H
H
H
H
H
H
(CH2)2CO2H
H
(CH2)4NH2
(CH2)4NH2
(CH2)4NH2
(CH2)3NHC(vNH)NH2
CH2CONH2
H
(CH2)4NH2
(CH2)4NH2
(CH2)4NH2
(CH2)3NHC(vNH)NH2
CH2CONH2
H
88
77
54
85
91
86
84
71
34
77
87
81
28
18
12
32
35
28
22
18
6
CH(CH3)C2H5
CH(CH3)C2H5
CH2CONH2
CH2C6H4OH
(CH2)2CO2H
H
CH(CH3)C2H5
CH(CH3)C2H5
CH2CONH2
CH(OH)CH3
6j
6k
6l
28
28
23
a
As assessed by UV detection at 254 nm.
Table 1 shows the compounds that were prepared, the HPLC
purities of the crude materials, and the isolated yields. Most
of the compounds shown in Table 1 are sulfides and
sulfones. Considerably fewer sulfoxides were prepared
since these formed as a mixture of epimers that were not
always separable by HPLC. The possibility that macro-
cyclization products 5 were formed via displacement of the
fluoride ortho to the nitro was easily excluded by observing
the large 3JFH coupling constants for the remaining fluorine
atom.

8746
L. Jiang, K. Burgess / Tetrahedron 58 (2002) 8743–8750
Table 2. Key NMR data for some sulfide peptidomimetics in the series 3–6
Parametera
Sulfides
Ideal b-I turn
3a
0.0
Fast
6.37
Medium
8.5
8.0
Medium
Medium
3g
0.0
Slow
6.6
Slow
8.0
9.0
Strong
Weak
3h
1.6
Slow
6.0
Slow
8.5
8.0
Strong
Weak
3i
4a
4.8
5a
5.1
Medium
3.5
Slow
6.5
8.0
Strong
Medium
6a
5.3
Slow
3.5
Slow
6.0
8.5
Weak
Strong
NHiþ2 Tc (2ppb/K)
NHiþ2 H/D exchange
NHiþ3 Tc (2ppb/K)
NHiþ3 H/D exchange
3JCa-NHiþ1 (Hz)
3JCa-NHiþ2 (Hz)
NHiþ1–NHiþ2
1.2
Fast
6.0
Fast
9.0
9.0
Strong
Weak
N/A
N/A
,3.0
Slow
4.0
9.0
Medium
Medium
Fast
3.25
Slow
6.0
8.0
Strong
Strong
NHiþ2–NHiþ3
a
Throughout, the terms ‘fast’, ‘medium’, and ‘slow’ in the H/D exchange experiments were relative values for particular compounds based on plots of the
extent of exchange versus time. The terms ‘strong’, ‘medium’, and ‘weak’ in ROESY spectra were relative values for particular compounds based on
counting contours in the spectra.
Table 3. Key NMR data for some sulfone peptidomimetics in the series 3–6
Parametera
Sulfones
Ideal b-I turn
3p
3.7
Slow
1.70
Slow
5.5
9.0
Medium
Medium
3v
1.4
Slow
1.4
Slow
4.2
8.5
Medium
Medium
3w
1.6
Slow
0.8
Slow
5.5
9.0
Medium
Medium
3x
4g
2.4
Slow
1.5
Slow
4.5
8.5
Medium
Medium
5g
2.0
Slow
2.5
Slow
5.0
9.0
Medium
Medium
6g
2.5
Slow
2.0
Slow
4.5
8.5
Medium
Medium
NHiþ2 Tc (2ppb/K)
NHiþ2 H/D exchange
NHiþ3 Tc (2ppb/K)
NHiþ3 H/D exchange
3JCa-NHiþ1 (Hz)
3JCa-NHiþ2 (Hz)
NHiþ1–NHiþ2
0.4
Slow
0.4
Slow
4.8
9.0
N/A
N/A
,3.0
Slow
4.0
9.0
Medium
Medium
Medium
Medium
NHiþ2–NHiþ3
a
Throughout, the terms fast, medium, and slow in the H/D exchange experiments were relative values for particular compounds based on plots of the extent of
exchange versus time. The terms strong, medium, and weak in ROESY spectra were relative values for particular compounds based on counting contours in
the spectra.
Table 4. Key NMR data for some sulfoxide peptidomimetics in the series 3–6
Parametera
Sulfoxide-R
Sulfoxide-S
Ideal b-I turn
3j
3.7
3l
3n
4.0
Fast
4.0
N/A
7.0
8.0
Strong
None
3k
3.7
Slow
1.7
Slow
5.5
9.0
Medium
Medium
3m
3o
4.0
Medium
2.0
Slow
5.0
8.0
Medium
Medium
NHiþ2 Tc (2ppb/K)
NHiþ2 H/D exchange
NHiþ3 Tc (2ppb/K)
NHiþ3 H/D exchange
3JCa-NHiþ1 (Hz)
3JCa-NHiþ2 (Hz)
NHiþ1–NHiþ2
5.4
Fast
4.6
Slow
1.6
Slow
5.5
8.0
N/A
N/A
,3.0
Slow
4.0
9.0
Medium
Medium
Fast
1.26
Slow
5.0
7.5
Weak
Strong
1.4
Medium
7.5
8.5
Strong
None
Medium
Medium
NHiþ2–NHiþ3
a
Throughout, the terms fast, medium, and slow in the H/D exchange experiments were relative values for particular compounds based on plots of the extent of
exchange versus time. The terms strong, medium, and weak in ROESY spectra were relative values for particular compounds based on counting contours in
the spectra.
3. Conformational analyses and assignments of absolute
configurations of the sulfoxides
strong bias to those conformations. Conversely, in Table 3,
all the seven sulfones studied give NMR data that
corresponds with ideal type I b-turn conformations. A
shared conformational bias towards this type of structure is
implied. In Table 4, the data for one series of sulfoxide
epimers 3j, 3l, and 3n do not compare well with the NMR
parameters for and ideal type I b-turn, but the same
parameters for the complementary ones 3k, 3m, and 3o
do. Throughout, the first series is epimeric with the
corresponding sulfoxide in the second series.
The assertion that type I b-turns are favored conformations
for the sulfones and one epimer of the sulfoxides is
supported by the NMR data compiled in Tables 2–4, CD
studies (as previously communicated) and extensive
molecular simulations, the end-results of which are
summarized in Fig. 3. Comparison of the NMR data
collected for the sulfides 3a, 3g, 3h, 3i, 4a, 5a, and 6a in
Table 2 show a wide variance for the parameters. This
implies the sulfides have different conformational prefer-
ences. There is no correspondence between the data for any
one of these compounds and the values expected for an ideal
type I b-turn implying these compounds do not have a
Assignment of the S-atom configurations of the sulfoxides is
difficult because the only physical technique that would
unambiguously assign this is single crystal X-ray analyses.
Unfortunately, the compounds did not form suitable

L. Jiang, K. Burgess / Tetrahedron 58 (2002) 8743–8750
8747
Figure 3. Low energy conformations of three (R)-sulfoxides, which do not show b-turn conformations, contrasted with three (S)-sulfoxides which do.
crystals. For this reason, definitive proof of the absolute
configurations indicated for the sulfoxides in Table 1 was
not obtained. To approach this problem, we therefore
investigated the unusual idea that configurational assign-
ments of the sulfoxides could be made via conformational
analyses. Thus data from the experiments to measure
temperature coefficients, H/D exchange rates, coupling
constants, ROE close contacts, and CD spectra were
interfaced with QMD simulations. These molecular simu-
lations were performed without introduction of any
constraints from NMR or other physical measurements.
They predict that the (R)-sulfoxides in this series do not
have preferences for type I b-turn conformations but the
(S)-sulfoxides do (see Fig. 3 for simulated low energy
conformations of six representative sulfoxides). When the
NMR data is compared with the simulated low energy
conformations, two key observations are made:
3–6 can be prepared via multi-step solid phase syntheses to
give products with high analytical purities. The NMR
studies show that the sulfides in this series have various
conformational preferences that are not easily recognized,
the sulfones tend to adopt type I b-turns, and one of the
sulfoxide epimers in each series also has a preference for
this type of conformation. Molecular simulations indicate it
is the (S)-sulfoxides that has a detectable preference for type
I turns, but the (R)-sulfoxides do not and our assignment of
configuration at sulfur is based on this. Configurational
assignments based on this type of correlation are not
definitive. However, for them to be incorrect, the molecular
simulations of both epimers of the sulfoxides would have to
be consistently wrong, and the true conformation of the
(R)-sulfoxide epimer would have to be a type I b-turn that
does not emerge from the simulations. We conceive this is
possible, but think it is highly unlikely. This is especially so
since the molecular simulations predict type I b-turn
conformations for the sulfones but not for the sulfides, and
both these observations are consistent with the NMR
and CD studies. Overall, this work illustrates a rare case
where conformational analyses can be used to deduce
stereochemical assignments.
† simulations of the (R)-sulfoxides indicated these com-
pounds did not have a bias towards b-turn structures, and
the favored (non-b-turn) conformations fit well with the
physical data for the epimer that did not populate b-turn
conformations;
† simulations of the (S)-sulfoxides indicated these com-
pounds did have a bias towards b-turn structures, and the
favored virtual b-turn conformations fit well with the
physical data for the epimer that did populate b-turn
conformations.
5. Experimental
5.1. General methods
4. Conclusions
All the a-amino acids used were of the L-configuration,
except where otherwise indicated. All chemicals were
obtained from commercial suppliers and used without
The data presented in Table 1 indicate that peptidomimetics

8748
L. Jiang, K. Burgess / Tetrahedron 58 (2002) 8743–8750
further purification. N 2-(5-Fluoro-2,4-dinitrophenyl)-L-ala-
nine amide (Marfey’s reagent), 2-(1H-benzotriazole-1-yl)-
1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU),
N-hydroxybenzotriazole (HOBt), di-iso-propylethylamine
(DIEA), trifluoroacetic acid (TFA), CH2Cl2, DMF, thionyl
chloride, piperidine and tri-iso-propylsilane (TIS) were
purchased from Aldrich. Fluoronitrobenzoic acid deriva-
tives chlorides were obtained by refluxing the corresponding
fluoronitrobenzoic acid derivative in thionyl chloride for
4 h. TentaGel S RAM Fmoc resin was a gift from Rapp
polymere. All protected amino acids used (except N-a-
Fmoc-S-tritylhomocysteine) were purchased from
Advanced ChemTech or Novabiochem. Fluoronitrobenzoic
acid derivatives (in Fig. 2) were prepared according to
literature precedent.22,23 Reverse phase high performance
liquid chromatography (HPLC) was carried out on Vydac
C-18 columns of the following dimensions: 25£0.46 cm2
for analysis, and 25£2.2 cm2 for preparative work. All
HPLC experiments were performed using gradient con-
ditions. Eluents used were solvent A (H2O with 0.1% TFA)
and solvent B (CH3CN with 0.1% TFA). Flow rates used
were 1.0 ml min21 for analytical, and 10 ml min21 for
preparative HPLC.
(compounds 1–4). Explicit atom representations were used
throughout the study. The residue topology files (RTF) for
all the peptidomimetics were built using QUANTA97
(Molecular Simulations Inc.).
QMD simulations were performed using the CHARMm
standard parameters. All four molecules were modeled as
neutral compounds in a dielectric continuum of 45
(simulating DMSO). Thus, the starting conformers were
minimized using 1000 steps of steepest descent (SD) and
3000 steps of the adopted basis Newton–Raphson method
(ABNR), respectively. The minimized structures were then
subjected to heating, equilibration, and dynamics simu-
lation. Throughout, the equations of motions were inte-
grated using the Verlet algorithm with a time step 1 fs, and
SHAKE was used to constrain all bond lengths containing
polar hydrogens. Each peptidomimetic was heated to
1000 K over 10 ps and equilibrated for another 10 ps at
1000 K, then molecular dynamics runs were performed for a
total time of 600 ps with trajectories saved every 1 ps. The
resulting 600 structures were thoroughly minimized using
1000 steps of SD followed by 3000 steps of ABNR until an
21
RMS energy derivative of #0.01 kcal mol21
was
˚
A
obtained. Structures with energies less than 3.0 kcal mol21
relative to the global minimum were selected for further
analysis.
5.2. NMR studies
NMR spectra of the peptidomimetics were recorded on a
Varian UnityPlus 500 spectrometer. The concentrations of
the samples were approximately 5 mM in DMSO-d6
The QUANTA97 package was again used to display,
overlay, and classify the selected structures into confor-
mational groups. The best clustering was obtained using a
grouping method based on calculation of RMS deviation of
a subset of atoms, in this study these were the ring backbone
atoms. Thus, threshold cutoff values between 0.56 and
1
throughout. One-dimensional (1D) H NMR spectra were
recorded with a spectral width of 8000 Hz, 16 transients,
and a 2.5 s acquisition time. Vicinal coupling constants
were measured from 1D spectra at 258C. Assignments of 1H
NMR resonance in DMSO were performed using sequential
connectivities. Temperature coefficients of the amide
protons were measured via several 1D experiments in the
temperature range 25–508C adjusted in 58C increments with
an equilibration time of more than 20 min after successive
temperature steps.
˚
0.75 A were selected to obtain families with reasonable
homogeneity. The lowest energy from each family was
considered as a typical representative of the family as a
whole. Additionally, a second approach was also used to
obtain a representation of each family. In this alternative
protocol, the coordinates of all the heavy atoms in each
family were averaged in Cartesian space. The protons were
re-built on those heavy atoms using standard geometries for
each atom type, then the resulting structures were
minimized using 50–100 steps of SD to smooth the bond
lengths and angles. Finally, inter-proton distances and
dihedral angles from both the lowest energy and the
averaged structure were calculated for comparisons with
the ROE data.
Two-dimensional (2D) NMR spectra were recorded at 258C
with a spectral width of 8000 Hz. Through-bond connec-
tivities were elucidated by COSY, and through-space
interactions were identified by ROESY spectra, recorded
in 512 t1 increments and 32 scans per t1 increment, with 2K
data points at t2. ROESY experiments were performed using
mixing times of 200, 300, 400 ms; normally a mixing time
of 300 ms was superior. The intensities of the ROESY
cross-peaks were assigned as S (strong), M (medium), and
W (weak) from the magnitude of their volume integrals.
5.4.1. Bis-N-a-Fmoc-homocysteine. A round bottom flask
with a magnetic stirrer was charged with 10.6 g (0.100 mol)
of sodium carbonate in 60 ml of water, and 2.00 g
(4.34 mmol) of bis-N-trifluoroacetyl homocysteine (1)
was added. The solution was stirred at 258C for 12 h,
dioxane (60 ml) was added to the solution, then 3.22 g
(9.55 mmol) of N-[(9-fluorenylmethoxycarbonyl)-oxy]suc-
cinimide (FmocOSu) in 20 ml of dioxane was added to the
mixture over 15 min. The mixture was stirred for 4 h at 258C
followed by adding citric acid until a pH of 4 was reached.
After removing most of the dioxane in vacuo, 250 ml of
ethyl acetate was added, and the layers were separated in a
separating funnel. The aqueous phase was extracted with
ethyl acetate (3£150 ml). The combined organic layer was
washed with water (3£200 ml), saturated NaCl (2£200 ml).
5.3. CD studies
CD measurements were obtained on an Aviv (model 62 DS)
spectrometer. For these experiments the cyclic peptido-
mimetics were dissolved in H2O/CH3OH (80:20 v/v)
(c¼0.1 mg ml21, 0.1 cm path length). The CD spectra
were recorded at 258C.
5.4. Quenched molecular dynamics (QMD) studies
CHARMm (version 23.2, Molecular Simulations Inc.) was
used for the molecular simulations performed in this work

L. Jiang, K. Burgess / Tetrahedron 58 (2002) 8743–8750
8749
The organic phase was dried over sodium sulfate, filtered,
and concentrated in vacuo to give a shallow yellowish solid.
Thoroughly drying the solid yielded 2.78 g of bis-N-a-
procedures,24 N 2-(5-fluoro-2,4-dinitrophenyl)-L-alanine
amide was coupled to the S-tritylhomocysteine to form a
L,L-diastereomer, analytical HPLC showed a homogeneous
single peak. Retention time 21.2 min indicating the high
enantiomeric purity of the N-a-Fmoc-S-tritylhomocysteine
(2). As a control, a sample of S-tritylhomocysteine which
was prepared according to literature,13 but was known to
be a mixture of enantiomers, was also coupled to the
N 2-(5-fluoro-2,4-dinitrophenyl)-L-alanine amide, analytical
HPLC showed two peaks, retention times are 21.2 and
22.1 min, respectively, with a ratio of 2:1, which were the
L,L- and L,D-diastereomers, indicating the racemization.
1
Fmoc-homocysteine as the product in 90% yield. H NMR
(300 MHz, DMSO-d6, 258C) d 7.90 (d, J¼7.2 Hz, 4H), 7.73
(d, J¼7.2 Hz, 4H), 7.70 (d, J¼7.2 Hz, 2H), 7.42 (t, J¼
7.2 Hz, 4H), 4.35–4.28 (m, 4H), 4.27–4.20 (m, 2H), 4.15–
4.06 (m, 2H), 2.83–2.69 (m, 4H), 2.21–2.07 (m, 2H), 2.04–
1.89 (m, 2H). 13C NMR (DMSO-d6, 75 MHz, 258C) d 174.2,
156.9, 145.9, 144.5, 141.4, 128.3, 127.8, 127.5, 125.9,
120.8, 120.6, 66.3, 64.5, 53.2, 50.8, 47.4, 34.8, 30.9.
MALDI MS: calcd for [M], 712.83, found 714.02.
5.4.2. N-a-Fmoc-homocysteine.
A
sample of 2.0 g
5.6. General experimental for syntheses of the peptido-
mimetics sulfides: synthesis of compound 3a
(2.8 mmol) of bis-N-a-Fmoc-homocysteine was dissolved
in 20 ml of DMF, and 4 ml of water was added. 2.16 g
(14.0 mmol) of dithiothreitol (DTT) was added to the
solution, stirred at 508C for 4 h. When TLC showed no
starting material remained, 250 ml of water was added to the
solution and extracted with ethyl acetate (4£150 ml). The
combined organic layer was washed with water (3£150 ml),
saturated NaCl (2£150 ml). The organic phase was dried
over sodium sulfate, filtered, and concentrated in vacuo to
give a yellowish oil. After flash chromatography with
solvents ethyl acetate/hexane (1:2), 1.69 g of N-a-Fmoc-
TentaGel S RAM Fmoc resin (100 mg, 0.220 mmol g21
)
was swelled in DMF (10 ml g21) in a polypropylene syringe
for 30 min, then rinsed with DMF (2£10 ml g21, for each
washing cycle throughout). The Fmoc protecting group on
the Rink handle was removed by treating the resin with 20%
piperidine in DMF (2£15 min). After the resin was rinsed
with DMF (3£), CH3OH (3£), and CH2Cl2 (3£), Fmoc-
homoCys(Trt)-OH (2) (4 equiv.), HBTU (4 equiv.), HOBt
(4 equiv.), and DIEA (6 equiv.) were added in 5 ml of DMF.
After 2 h of gentle shaking, a ninhydrin test on a small
sample of beads gave a negative result. The reaction mixture
was drained and the resin was rinsed with DMF (4£). The
above deprotection/coupling cycles were repeated to
introduce Fmoc-Lys(Boc)-OH and Fmoc-Glu(OtBu)-OH
consecutively. The 2-fluoro-5-nitrobenzoic acid moiety
was introduced to the N-terminus of the tripeptide-resin
by treating with 2-fluoro-5-nitrobenzoyl chloride (2 equiv.)
and iPr2NEt (4 equiv.) in 1 ml of CH2Cl2 for 1 h. The side-
chain protecting group (trityl) of homoCys was removed by
treatment with 3% TFA and 4% TIS in CH2Cl2 (5£5 min).
After, the resin was rinsed with CH2Cl2 (3£), CH3OH (3£),
and DMF (3£), the macrocyclization step was carried out by
treating the supported peptide with 5 equiv. K2CO3 in DMF
at 258C with gentle shaking for 36 h. The peptide-resin was
washed with DMF (2£), H2O (3£), DMF (3£), H2O (2£),
CH3OH (3£), CH2Cl2 (3£), and then dried in vacuo for 4 h.
The peptide was cleaved from the resin by treatment with a
5 ml mixture of 95% TFA, 2.5% TIS and 2.5% H2O for 2 h.
The cleavage solution was separated from the resin by
filtration. After most of the cleavage cocktail was
evaporated by passing N2, the crude peptide was precipi-
tated using anhydrous ethyl ether, then dissolved in H2O,
and lyophilized to give the crude product. Preparative
HPLC (Beckman System, 5–90% B in 40 min) was carried
out to provide a slight yellowish powder (3.8 mg, 32%) of
3a. 1H NMR (500 MHz, DMSO-d6, 258C), see Table of 1H
NMR (500 MHz) chemical shifts (ppm) for 3a. Analytical
HPLC: single main peak, retention time 14.2 min (5–70% B
in 30 min). MALDI MS: calcd for C22H30N6O8S, 538.5,
found 539.6 [MþHþ]. 560.5 [MþNaþ] and 577.4 [MþKþ].
1
homocysteine was obtained in a yield of 85%. H NMR
(300 MHz, DMSO-d6, 258C) d 7.91 (d, J¼7.2 Hz, 2H), 7.74
(d, J¼7.2 Hz, 2H), 7.68 (d, J¼7.2 Hz, 1H), 7.43 (t, J¼
7.2 Hz, 2H), 7.35 (t, J¼7.2 Hz, 2H), 4.32 (m, 2H), 4.24 (m,
1H), 4.15 (dd, J¼7.2, 16 Hz, 1H), 2.53 (m, 2H), 2.35 (t, J¼
7.2 Hz, 1H), 1.94 (m, 2H). 13C NMR (DMSO-d6, 75 MHz,
258C) d 174.6, 157.3, 144.7, 141.5, 128.4, 127.8, 126.0,
120.9, 66.5, 53.5, 47.5, 35.7, 21.6. ESI-MS: calcd for [M],
357.42, found 358.66 [MþHþ], 380.48 [MþNaþ].
5.4.3. N-a-Fmoc-S-tritylhomocysteine (2). N-a-Fmoc-
homocysteine (0.89 g, 2.50 mmol), triphenylmethanol
(0.65 g, 2.50 mmol) were dissolved in 200 ml of anhydrous
CH2Cl2, followed by adding 0.5 equiv. of trifluoroacetic
acid (96 ml, 1.25 mmol), stirred at 258C for 6 h. Pyridine
was added to neutralize the trifluoroacetic acid, the organic
phase was washed with water to remove the pyridine salt,
evaporating the CH2Cl2, and the crude product was purified
by flash chromatography with solvents ethyl acetate/hexane
(1:2.5). 0.9 g of 2 was obtained in a yield of 60%. [a]2D0¼
1
þ128 (c¼1.0, CH3OH). H NMR (300 MHz, DMSO-d6,
258C) d 7.91 (d, J¼7.2 Hz, 2H), 7.72 (d, J¼7.2 Hz, 2H),
7.56 (d, J¼7.2 Hz, 1H), 7.44 (t, J¼7.2 Hz, 2H), 7.36–7.20
(m, 17H), 4.31 (m, 2H), 4.23 (m, 1H), 4.00 (m, 1H), 2.19 (m,
2H), 1.70 (m, 2H). 13C NMR (DMSO-d6, 75 MHz, 258C) d
174.8, 156.9, 145.1, 141.5, 129.8, 128.7, 128.4, 127.8,
127.4, 126.0, 120.9, 66.8, 66.4, 53.6, 47.4, 30.3, 28.8.
ESI-MS: calcd for [M], 599.74, found [MþHþ], 600.58,
[MþNaþ], 622.5.
5.5. Determination of the enantiomeric purity of
N-a-Fmoc-S-tritylhomocysteine (2)
5.7. General experimental for syntheses of the peptido-
mimetics sulfones: synthesis of compound 3p
A sample of 20 mg of 2 was dissolved in 1 ml of ethyl
acetate, and then 33 ml (20 equiv.) of piperidine was added,
followed by precipitating with 2 ml of hexane. 10 mg of
S-tritylhomocysteine was obtained with high purity
(analytical HPLC checked). Following Marfey’s test
10 mg of 3a was dissolved in 2 ml formic acid (90% in
H2O), 0.4 ml H2O2 (30% in H2O) was added, stirred
overnight at 258C. Evaporated most of solvent under

8750
L. Jiang, K. Burgess / Tetrahedron 58 (2002) 8743–8750
2. Goldberg, M.; Smith, L.; Tamayo, N.; Kiselyov, A. S.
Tetrahedron 1999, 55, 13887–13898.
vacuum and dissolved in H2O, lyophilized to give the crude
product. Preparative HPLC (Beckman System, 5–90% B in
40 min) was carried out to provide a white powder (6.6 mg,
65%, based on 3a) of 3p. H NMR (500 MHz, DMSO-d6,
1
258C), see Table of H NMR (500 MHz) chemical shifts
(ppm) for 3p. Analytical HPLC: single main peak, retention
time 13.1 min (5–70% B in 30 min). MALDI MS: calcd for
C22H30N6O10S, 570.5, found 571.6 ([MþHþ]), 593.7
([MþNaþ]).
3. Kiselyov, A. S.; Eisenberg, S.; Luo, Y. Tetrahedron Lett.
1999, 40, 2465–2468.
1
4. Kiselyov, A. S.; Smith, II., L.; Tempest, P. Tetrahedron 1999,
55, 14813–14822.
5. Ouyang, X.; Chen, Z.; Liu, L.; Dominguez, C.; Kiselyov, A. S.
Tetrahedron 2000, 56, 2369–2377.
6. Feng, Y.; Wang, Z.; Jin, S.; Burgess, K. J. Am. Chem. Soc.
1998, 120, 10768–10769.
5.8. General experimental for syntheses of the peptido-
mimetics sulfoxides: synthesis of compounds 3j and 3k
7. Feng, Y.; Burgess, K. Chem.—A Eur. J. 1999, 5, 3261–3272.
8. Feng, Y.; Wang, Z.; Jin, S.; Burgess, K. Chem.—A Eur. J.
1999, 5, 3273–3278.
Compound 3a (30 mg) was dissolved in 8 ml CH3CN/H2O
(3:5 v/v), 240 ml (2 equiv.) of NaIO4 solution (10 mg ml21
in H2O) was added, stirred at 258C for 24 h, HPLC analysis
showed only two main peaks, retention time 11.0 and
12.1 min with ratio of 1:2. HPLC purification gave 7.0 mg
(23%, based on 3a) and 14.0 mg (45%, based on 3a) of 3j
and 3k. 1H NMR (500 MHz, DMSO-d6, 258C), see Table of
1H NMR (500 MHz) chemical shifts (ppm) for 3j and 3k.
MALDI MS: calcd for C22H30N6O9S, 554.5, found 555.4
([MþHþ]) for both 3j and 3k.
9. Feng, Y.; Burgess, K. Biotech. Bioengng Comb. Chem. 2000,
71, 3–8.
10. Feng, Y.; Pattarawarapan, M.; Wang, Z.; Burgess, K. J. Org.
Chem. 1999, 64, 9175–9177.
11. Burgess, K. Acc. Chem. Res. 2001, 34, 826–835.
12. Miyazaki, H.; Ohta, A.; Kawakatsu, N.; Waki, Y.; Gogun, Y.;
Shiraiwa, T.; Kurokawa, H. Bull. Chem. Soc. Jpn 1993, 66,
536–540.
13. Pearson, D. A.; Lister-James, J.; McBride, W. J.; Wilson,
D. M.; Martel, L. J.; Civitello, E. R.; Taylor, J. E.; Moyer,
B. R.; Dean, R. T. J. Med. Chem. 1996, 39, 1361–1371.
14. Adamczyk, M.; Fishpaugh, J. R.; Thiruvazhi, M. Tetrahedron:
Asymmetry 1999, 10, 4151–4156.
Acknowledgements
´
15. Bischoff, L.; David, C.; Roques, B. P.; Fournie-Zaluski, M.-C.
Support for this work was provided by Procter and Gamble,
The NIH (CA 82642) and by The Robert A. Welch
Foundation. We wish to thank Dr S. Srivastava at Bristol
Myers Squibb for supplying us with a sample of bis-N-
trifluoroacetyl homocysteine, Ms M. Pattarawarapan and
Mr S. Reyes for MS analyses, and Mr C. Park for help with
the QMD studies. TAMU/LBMS-Applications Laboratory
headed by Dr Shane Tichy provided mass spectrometric
support, and the Laboratory for Molecular Simulation’s (Dr
Lisa Thompson) supported our molecular simulations work.
The NMR instrumentation in the Biomolecular NMR
Laboratory at Texas A&M University was supported by a
grant from the National Science Foundation (DBI-9970232)
and Texas A&M University.
J. Org. Chem. 1999, 64, 1420–1423.
16. Shiraiwa, T.; Nakagawa, K.; Kanemoto, N. Chem. Lett. 2000,
468–469.
17. Jiang, L.; Burgess, K. J. Am. Chem. Soc. 2002. in press.
18. Pettitt, B. M.; Matsunaga, T.; Al-Obeidi, F.; Gehrig, C.;
Hruby, V. J.; Karplus, M. Biophy. J. Biophys. Soc. 1991, 60,
1540–1544.
19. O’Connor, S. D.; Smith, P. E.; Al-Obeidi, F.; Pettitt, B. M.
J. Med. Chem. 1992, 35, 2870–2881.
20. Knorr, R.; Trzeciak, A.; Bannwarth, W.; Gillessen, D.
Tetrahedron Lett. 1989, 30, 1927–1930.
21. Atherton, E.; Sheppard, R. C. Solid Phase Peptide Synthesis, A
Practical Approach. IRL: Oxford, 1989.
22. Park, C.; Burgess, K. J. Comb. Chem. 2001, 3, 257–266.
23. Yoshida, Y.; Barrett, D.; Azami, H.; Morinaga, C.;
Matsumoto, S.; Matsumoto, Y.; Takasugi, H. Bioorg. Med.
Chem. 1999, 7, 2647–2666.
References
24. Bruckner, H.; Keller-Hoehl, C. Chromatographia 1990, 30,
621–629.
1. Kiselyov, A. S.; Eisenberg, S.; Luo, Y. Tetrahedron 1998, 54,
10635–10640.