Two new β-strand mimics

Article abstract of DOI:10.1016/S0968-0896(98)00225-9

In a previous report, Nowick and co-workers described β-strand mimic A, which duplicates the structure and hydrogen-bonding pattern of one edge of a tetrapeptide in a β-strand conformation (Nowick, J. S.; Pairish, M.; Lee, I. Q.; Holmes, D. L.; Ziller, J.

Full text of DOI:10.1016/S0968-0896(98)00225-9

Bioorganic & Medicinal Chemistry 7 (1999) 29±38
Two New ꢀ-Strand Mimics
James H. Tsai, Amy Sue Waldman and James S. Nowick*
Department of Chemistry, University of California, Irvine, Irvine, CA 92697-2025, USA
Received 11 June 1998; accepted 4 September 1998
AbstractÐIn a previous report, Nowick and co-workers described b-strand mimic A, which duplicates the structure and hydrogen-
bonding pattern of one edge of a tetrapeptide in a b-strand conformation (Nowick, J. S.; Pairish, M.; Lee, I. Q.; Holmes, D. L.;
Ziller, J. W. J. Am. Chem. Soc. 1997, 119, 5413). b-Strand mimic A is composed of a 5-amino-2-methoxybenzoic acid unit linked to
a 5-hydrazino-2-methoxybenzamide unit by means of an acylhydrazine group. This paper introduces two related b-strand mimics
(B and C) and reports their comparison to b-strand mimic A. b-Strand mimic B is composed of a 5-amino-2-methoxybenzoic acid
unit linked by a diacylhydrazine group to a fumaramide unit; b-strand mimic C is composed of a 5-amino-2-methoxybenzoic acid
unit linked by a diacylhydrazine group to a peptide. b-Strand mimics A±C were connected to tripeptide (Phe-Ile-Leu) groups by
1
means of 1,2-diaminoethane diurea turn units to form arti®cial b-sheets 1±3. H NMR studies, involving ROESY, chemical shift,
coupling constant, and variable temperature experiments, reveal that 1±3 adopt hydrogen-bonded antiparallel b-sheet conforma-
tions and establish that all three templates are viable b-strand mimics. # 1999 Elsevier Science Ltd. All rights reserved.
Introduction
modulate b-sheet formation.1±17 These decoys must
mimic aspects of b-sheet structure, such as the place-
ment of the amino acid side-chains or the alternating
array of hydrogen-bond donors and acceptors provided
by one edge of a peptide in a b-strand conformation.
Leading examples of b-strand mimics have been
developed by Kemp and co-workers,18±20 Smith,
Hirschmann, and co-workers,6±8 Pallai, Rebek, and
co-workers,4,5 Schrader and Kirsten,21,22 and our own
research group.23±25 These b-strand mimics have the
potential to be used as inhibitors and antagonists of
processes involving b-sheet formation.
b-Sheet formation plays a critical role in many biologi-
cal processes associated with diseases and normal func-
tions. b-Sheet interactions between proteins have been
shown or hypothesized to be involved in cell signaling
and oncogene expression associated with the binding of
Ras and Rap by the serine/threonine kinase Raf,1 the
clustering of membrane ion channels by PDZ domains,2
the binding of lymphocyte function-associated antigen-1
(LFA-1) by the intercellular adhesion molecule-1
(ICAM-1),3 and the interaction between the CD4
receptor and the HIV viral protein gp120.4,5 To cleave
peptides, proteolytic enzymes, such as HIV-1 protease
and renin, form b-sheet-like networks of hydrogen bonds
with their peptide substrates.6±8 HIV-1 protease dimerizes
through four-interdigitating b-strands.9 The met repres-
sor, a protein involved in gene regulation, also functions
as a b-sheet dimer; in this case, the b-sheet that forms is
directly involved in binding the major groove of
DNA.10 Many proteins aggregate to form insoluble b-
sheet structures that are associated with Alzheimer's
disease,11±13 Creutzfeld±Jacob disease and other prion
diseases,14 and progressive neurodegenerative disorders
that are associated with trinucleotide (CAG) repeats.15±17
In 1996, we introduced 5-amino-2-methoxybenzamide
as a b-strand mimic.23 This mimic duplicates the
hydrogen-bonding functionality of one edge of a dipep-
tide. To duplicate a tripeptide, we extended the b-strand
mimic with a diacylhydrazine group to form a 5-amino-
2-methoxybenzoic hydrazide.25
An intriguing possibility for the development of new
therapeutic strategies for treating diseases associated
with b-sheet formation involves chemical decoys that
Key words: Amino acids and derivs; mimetics; peptides and polypep-
tides; solid phase synthesis.
*Corresponding author. Fax: +1 (949) 824-8571; e-mail: jsnowick@
uci.edu
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(98)00225-9

30
J. H. Tsai et al. / Bioorg. Med. Chem. 7 (1999) 29±38
We also introduced b-strand mimic A, which contains
a 5-amino-2-methoxybenzoic acid unit coupled to a 5-
hydrazino-2-methoxybenzamide unit by means of an
acylhydrazine linkage.24 This b-strand mimic duplicates
the hydrogen bonding functionality of one edge of a
tetrapeptide in a b-strand conformation. To evaluate
the e€ectiveness of this mimic, we synthesized arti®cial
b-sheet 1, which contains b-strand mimic A linked to a
peptide strand by means of a 1,2-diaminoethane diurea
turn unit.26,27 1H NMR chemical shift and NOE studies
revealed that arti®cial b-sheet 1 adopts a hydrogen-
bonded b-sheet structure in CDCl3 solution.24 The che-
mical shift studies showed that the NH groups are
hydrogen bonded, and the NOE studies showed that the
b-strand mimic is proximal to the peptide strand.
Although many NOEs characteristic of b-sheet struc-
ture were present, we did not observe an NOE between
the aniline-type NH proton (Hg) and the Ile NH proton
(Hc). The absence of this NOE suggested that these NH
groups were further apart than the 3.1 A separation that
is characteristic of b-sheets. We attributed the absence
of the Hc Hg NOE to the propensity of the acylhy-
drazine group of the b-strand mimic to adopt a twisted
geometry (C±N±N±C dihedral angle & 90ꢀ).24
these considerations in mind, we set out to compare b-
strand mimics B and C to b-strand mimic A. This paper
describes this comparison.
Results and Discussion
To evaluate these b-strand mimics, we synthesized and
studied arti®cial b-sheets 1±3, which incorporate b-
strand mimics A±C. 1H NMR studies involving
ROESY, chemical shift, coupling constant, and variable
temperature experiments in CDCl3 solution reveal
that 1±3 form hydrogen-bonded antiparallel b-sheet
structures.
Although b-strand mimic A is an e€ective template in
arti®cial b-sheet 1, our concern over its twist prompted
us to investigate alternative b-strand mimics. In this
paper, we report two alternative b-strand mimics (B and
C) and compare them to b-strand mimic A. b-Strand
mimic B contains a diacylhydrazine group in place of
the monoacylhydrazine group of b-strand mimic A and
a fumaramide group in place of the 5-hydrazino-2-
methoxybenzamide group. b-Strand mimic C also con-
tains a diacylhydrazine group and contains a peptide in
place of the 5-hydrazino-2-methoxybenzamide group.
X-ray crystallographic and theoretical data suggest that
the diacylhydrazine group should better be able to
adopt a linear geometry (C±N±N±C dihedral angle &
180ꢀ), allowing it to better mimic a peptide b-strand.28,29
The diacylhydrazine group also contains an amide NH
group, which should be a better hydrogen-bond donor
than the aniline NH group of b-strand mimic A. With
Synthesis of arti®cial ꢀ-sheets 1±3
Arti®cial b-sheets 2 and 3 were synthesized by solid-
phase methods in a fashion analogous to that which we
had previously reported for the solid-phase synthesis of
arti®cial b-sheet 1.30 Scheme 1 illustrates the prepara-
tion of 2. Resin-bound tripeptide 4 was constructed on
Merri®eld resin by standard solid-phase peptide synth-
esis methods.31 The Boc protective group of 4 was
removed by treatment with TFA and the free amino
group was then liberated by treatment with TEA. The

J. H. Tsai et al. / Bioorg. Med. Chem. 7 (1999) 29±38
31
Scheme 2.
1H NMR ROESY studies of arti®cial ꢀ-sheets 1±3
1H NMR nuclear Overhauser e€ect studies indicate that
arti®cial b-sheets 1±3 adopt antiparallel b-sheet con-
formations in CDCl3 solution. These studies were
performed in the rotating frame using the transverse-
ROESY (Tr-ROESY) method,33,34 and reveal 12±14
interstrand ROEs between the b-strand mimics and the
adjacent main-chains and side-chains of the peptide
strands in each of the arti®cial b-sheets.
Scheme 1.
Arti®cial b-sheets 1±3 each exhibit networks of inter-
strand ROEs between their b-strand mimics and their
tripeptide groups (Fig. 1). All the arti®cial b-sheets
exhibit interstrand ROEs between aromatic ring pro-
tons Hn and the Phe a-protons (Hm) and between
aromatic ring protons Hn and the Ile amide protons
(Hc). Analogous ROEs between b-strand mimic protons
Ho and the Leu a-protons (Hk) and between b-strand
mimic protons Ho and methylamide protons Ha are
present. Arti®cial b-sheets 1±3 also exhibit ROEs
between methylamide protons Ha and b-strand mimic
resulting amine was coupled with carbamoyl chloride 5
to form urea 6. The Boc protective group was removed
from urea 6 by treatment with TFA. The free amino
group was liberated by treatment with TEA and cou-
pled with isocyanate 7 to form diurea 8. The Boc pro-
tective group of 8 was removed with gaseous HCl,32 and
the resulting hydrazine was coupled to N,N-dimethyl-
fumaramic acid (trans-Me2NCOCH=CHCO2H) with
DCC to a€ord the resin-bound b-sheet. Aminolytic
cleavage from the resin followed by chromatographic
puri®cation a€orded arti®cial b-sheet 2 in 27% overall
yield.
0
protons Hp (or Hp and Hp ). Arti®cial b-sheet 3 shows
0
ROEs between Hp and Hp and the terminal methyl
group of the peptide Hj. Arti®cial b-sheets 2 and 3
exhibit ROEs between hydrazine protons Hg and the Ile
amide protons (Hc); the analogous ROE was not detec-
ted in arti®cial b-sheet 1. This di€erence between arti®-
cial b-sheets 2 and 3 and arti®cial b-sheet 1 is consistent
with a model in which the diacylhydrazine group of the
former better adopts a linear geometry than the mono-
acylhydrazine group of the latter.
Arti®cial b-sheet 3 was synthesized in a similar fashion
to arti®cial b-sheet 2, in 25% overall yield. Scheme 2
illustrates its preparation from diurea 8. The Boc pro-
tective group of 8 was removed by treatment with gas-
eous HCl and the free hydrazino group was then
liberated by treatment with TEA. The resulting hydra-
zine was coupled to Boc-valine with DCC to give 9. The
Boc protective group of 9 was removed by treatment
with gaseous HCl and the resulting hydrochloride salt
was treated with TEA and coupled with isobutyryl
chloride to a€ord the resin-bound b-sheet. Arti®cial b-
sheet 3 was then liberated from the resin by aminolysis
with methylamine.
The arti®cial b-sheets also exhibit interstrand ROEs
between the b-strand mimics and the side-chains of the
peptide strands (not shown in Figure 1). These include
ROEs between aromatic ring protons Hn and the Phe
and Ile side-chains, b-strand mimic protons Ho and Hp
and the Leu side-chains, and hydrazine protons Hg
and the Ile side-chains.

32
J. H. Tsai et al. / Bioorg. Med. Chem. 7 (1999) 29±38
exhibits a number of unexpected ROEs involving b-
strand mimic protons Ho and Hq (Hg±Hq, Hk±Hq, Ho±Hp,
0
and Ho±Hp ). These ROEs may re¯ect that Ho and Hq
are strongly coupled and that their respective resonances
do not wholly correspond to the individual protons.
1H NMR coupling constant studies of arti®cial ꢀ-sheets
1±3
1
The H NMR coupling constants between the NH and
a-protons re¯ect the conformation of a peptide; cou-
pling constants of less than 6 Hz are consistent with a-
helical structure while coupling constants of greater
than 7 Hz are consistent with b-strand structure.35
Consistent with a b-strand conformation, the Phe-Ile-
Leu tripeptide strands of arti®cial b-sheets 1±3 have
coupling constants of 7.2±9.5 Hz. The coupling constant
for Hb in 1 is slightly smaller than in 2 and 3 (Table 1).
This (minor) di€erence may indicate that the peptide
strand in 1 adapts to the twist in the adjacent b-strand
mimic by adopting a distorted b-strand conformation.
1H NMR chemical shift studies of arti®cial ꢀ-sheets 1±3
1
The H NMR chemical shifts of the NH and a-protons
of the peptide and peptidomimetic groups of 1±3 also
indicate that these compounds form b-sheet structures.
In CDCl3, and in other non-competitive solvents, pro-
tons that are hydrogen bonded typically appear several
ppm down®eld of similar protons that are not hydrogen
bonded. Non-hydrogen-bonded peptide amide protons
generally appear at about 6 ppm, while hydrogen-bon-
ded peptide amide protons typically appear at about
8 ppm.36 In compounds 1±3, Ha and Hc appear at 8.02±
8.62 ppm, while Hb appears at 6.07±6.40 ppm (Table 2).
These chemical shifts are consistent with a b-sheet con-
formation in which Ha and Hc lie on the hydrogen-
bonded edge of the peptide, and Hb lies on the non-
hydrogen-bonded edge.
Figure 1. Important ROEs between the b-strand mimics and peptide
main chains in arti®cial b-sheets 1±3.
Protons Hd cannot be directly compared to protons
Ha±Hc, because protons Hd are part of a peptide urea.
The chemical shifts of Hd in 1±3 (4.80±4.88 ppm) are
typical of what we have previously observed for similar
non-hydrogen-bonded ureas.26 Protons He are part of
an aromatic urea and appear at 9.95±10.10 ppm; these
shifts are consistent with values we have previously seen
for related hydrogen-bonded urea protons.23±25 Protons
Hg di€er among compounds 1±3: in 1, it is part of an N-
acyl-phenylhydrazine group, while in 2 and 3 it is part
of a diacylhydrazine group. For this reason, the chemi-
cal shift values for the two di€erent groups cannot be
compared to each other. They are, however, consistent
with those that we have previously observed for similar
hydrogen-bonded protons.24,25
Interresidue ROEs within the peptide strands of 1±3
provide further evidence for b-sheet structure. In each
of these compounds the interresidue ROEs between the
a-protons of one peptide residue and the NH protons of
the adjacent residue are strong; these ROEs are much
stronger than the intraresidue ROEs between the a-
protons and the NH protons of each residue. These
ROEs indicate that the peptides adopt b-strand
conformations.
Several weak (or very weak) ROEs are not wholly con-
sistent with b-sheet structure. These include ROEs
between Ha and Hb, Hb and Hc, Hc and Hd, Hd and Hn,
and Hf and Ho. Molecular modeling indicates that the
distances between these protons should be 4.0±4.5 A;
the intensity of these ROEs should be only a few percent
of those of the strongest ROEs. These ROEs may re¯ect
deviation from ideal b-sheet structure or conforma-
tional mobility, or may result from TOCSY artifacts,
which can arise if the 180ꢀ pulse of the Tr-ROESY
experiments is not well calibrated. Arti®cial b-sheet 2
1
The H NMR chemical shifts of the a-protons of the
peptide strands in 1±3 provide further evidence that
these compounds form b-sheet structures. In b-sheets,
the a-protons of each amino acid residue typically
appear several tenths of a ppm down®eld of the same
residue in a random coil.37,38 In random coils, Phe,
Ile, and Leu exhibit chemical shifts of 4.66‹0.10,

J. H. Tsai et al. / Bioorg. Med. Chem. 7 (1999) 29±38
33
Table 1. 1H NMR coupling constants (Hz) of the NH protons of 1±3a
Table 4. Temperature dependence of the 1H NMR chemical shifts
(ppb/K) of the NH protons of 1±3a
Hb(Leu)
Hc(Ile)
Hd(Phe)
Ha
8.5
5.3
5.5
Hb
Hc
Hd
He
Hf
Hg Hh
Hi
b-sheet 1
b-sheet 2
b-sheet 3
7.2
9.1
9.5
9.1
9.4
9.1
8.5
9.1
8.3
b-sheet 1
b-sheet 2
b-sheet 3
2.7b
4.0
2.1 0.5 1.5 2.2 3.3 3.1
3.3 0.1 1.2 3.0 5.8
1.2 0.5 1.6 1.5 6.3
Ð
Ð
4.1
Ð
Ð
1.4
aSpectra were recorded in 1 mM in CDCl3 solution.
Table 2. 1H NMR chemical shifts of the NH protons of 1±3a
Ha Hb Hc Hd He Hf Hg Hh Hi
8.02 6.07 8.14 4.80 9.95 9.55 8.45 8.10
aSpectra were recorded in 1 mM in CDCl3 solution at 10 ꢀC intervals
from 30 ꢀC to 30 ꢀC.
bTemperature-dependence of the NH shift data exhibited poor linear
correlation.
b-sheet 1
b-sheet 2
b-sheet 3
Ð
Ð
6.29
and Hi in 3 also exhibit a larger temperature dependence
( 4.0 to 4.1 ppb/K). Although these data may re¯ect
the involvement of other conformations or equilibria,
we hesitate to place too much emphasis on the inter-
pretation of these data: much of what is known
about temperature dependence of NH shifts has been
established for peptide amide NH groups, but 1±3 con-
tain not only peptide amide NH groups but also urea,
amine, acylhydrazine, and aromatic amide NH groups.
8.16 6.40 8.62 4.88 10.10 11.28 12.04
8.03 6.29 8.38 4.88 10.02 10.85 11.56
Ð
Ð
aSpectra were recorded in 1 mM in CDCl3 solution.
3.95‹0.10, and 4.17‹0.10 ppm, respectively.37,38 The
a-protons of arti®cial b-sheets 1±3 all appear down®eld
of these values (Table 3); the Phe a-protons appear at
5.00±5.35 ppm, the Ile a-protons appear at 4.14±
4.31 ppm, and the Leu a-protons appear at 4.41±
4.83 ppm. Caution must be exercised in comparing these
values to the random coil values, however, since the a-
proton of Phe is next to a urea group, and since the
chemical shifts of the arti®cial b-sheets were measured
in CDCl3, while those of the random coils were mea-
sured in water.
Conclusions
These synthetic and 1H NMR spectroscopic studies
indicate that b-strand mimics A±C can template b-sheet
structure in adjacent peptide strands and establish that
b-strand mimics B and C are viable alternatives to b-
strand mimic A. It is dicult to conclude whether B and
C are better templates than A, since all three do an
admirable job. A few slight di€erences exist between
arti®cial b-sheets 2 and 3 and arti®cial b-sheet 1 (addi-
tional ROEs, slight di€erences in coupling constants,
slightly greater down®eld shifting of Ha, Hc, Hk, Hl, and
Hm, a slightly smaller temperature dependence of Ha),
which suggest that B and C are slightly better than A.
These di€erences may re¯ect the greater propensity of
the diacylhydrazine groups of B and C to adopt linear
geometries and the better hydrogen-bonding abilities of
the amide-type protons Hg of these groups.
1
Variable-temperature H NMR studies of arti®cial ꢀ-
sheets 1±3
The temperature dependence of the chemical shifts of
amide protons re¯ects their state of hydrogen bonding.
Peptide amide protons that are not hydrogen bonded or
are locked in a hydrogen-bonded conformation display
small temperature dependencies ( 2 to 3 ppb/K) in
chloroform solution, while peptide amide protons that
equilibrate between hydrogen-bonded and non-hydro-
gen-bonded states display larger temperature depen-
dencies ( 4 to 8 ppb/K).39,40 In arti®cial b-sheet 1, all
protons exhibit small temperature dependencies of che-
mical shifts ( 0.5 to 3.3 ppb/K), with the exception of
methylamide proton Ha, which exhibits a temperature
dependence of 8.5 ppb/K (Table 4). These data are
consistent with a model in which arti®cial b-sheet 1
adopts a hydrogen-bonded b-sheet conformation, but in
which the hydrogen bond to the leucine methylamide
NH proton `frays' upon warming. In arti®cial b-sheets 2
and 3, the temperature dependence of Ha is less ( 5.3 to
5.5 ppb/K), suggesting that less fraying occurs and
that these b-sheets are more stable. In these two b-
sheets, Hg exhibits a large temperature dependence
( 5.8 to 6.3 ppb/K), possibly indicating a greater
degree of fraying of its hydrogen bond. Protons Hb in 2
Experimental
General. Commercially available reagents and solvents
were used without further puri®cation. Phosgene was
obtained from Fluka as a 20% (1.93 M) solution in
toluene. Merri®eld resin (1% divinylbenzene cross-
linked polystyrene) bearing Boc-protected amino acids
was obtained from Advanced ChemTech. Each solid-
phase synthesis was performed using a 50- or 100-mL
reaction vessel with a sintered-glass frit and a stopcock
at one end and a stoppered 14/20-ground joint at the
other end. The reaction vessels were shaken with an air-
driven, 270ꢀ rotary shaker. The resin was washed by
adding solvent, shaking for 1 min, and draining the
reaction vessel. The reaction vessels were drained using
nitrogen gas at ca. 5 psi to expel the solvent. Either a 3/1
(v/v) mixture of dichloromethane (CH2Cl2) and tri-
¯uoroacetic acid (TFA) containing 1 mg/mL of indole
or HCl gas was used to remove Boc protective groups.
A 9/1 (v/v) mixture of CH2Cl2 and triethylamine (TEA)
Table 3. 1H NMR chemical shifts of a-CH protons of 1±3a
Hk(Leu)
Hl(Ile)
Hm(Phe)
b-sheet 1
b-sheet 2
b-sheet 3
4.41
4.83
4.83
4.14
4.31
4.18
5.00
5.35
5.07
aSpectra were recorded in 1 mM in CDCl3 solution.

34
J. H. Tsai et al. / Bioorg. Med. Chem. 7 (1999) 29±38
was used to convert the resulting tri¯uoroacetic acid
salts to the free amines.
with a mixture of fumaric acid 9 (0.20 g, 1.4 mmol),
DCC (0.20 g, 9.7 mmol), and CH2Cl2 (20 mL) for 2 h. The
resin was washed with CH2Cl2 (3Â20 mL) and shaken
with 40% (w/w) methylamine in MeOH (25 mL) for 3 h.
The solution was collected and concentrated by rotary
evaporation. The residue was puri®ed by chromato-
graphy on silica gel (i-PrOH/CH2Cl2, 7/93) followed by
reverse-phase HPLC (C18 column, CH3CN/H2O, 60/40)
Arti®cial ꢀ-Sheet 2. A 100-mL solid-phase reaction
vessel was charged with 8.00 g of Boc-Leu Merri®eld
resin (0.50 mmol/gm, 4.0 mmol). The resin was washed
with CH2Cl2 (3Â80 mL) and CH2Cl2/TFA/indole solu-
tion (1Â80 mL) and shaken with CH2Cl2/TFA/indole
solution (80 mL) for 30 min. The resin was drained and
washed with CH2Cl2 (6Â80 mL), CH2Cl2/TEA solution
(2Â80 mL), and CH2Cl2 (6Â80 mL). The resin was sha-
ken with a mixture of N-Boc-Ile-OH (2.5 g, 10 mmol),
1.0 M solution of DCC in CH2Cl2 (10 mL, 10 mmol),
and CH2Cl2 (70 mL) for 2 h. The resulting white
suspension was ®ltered, and the residue was washed
alternately (3Â) with CH2Cl2 (80 mL) and MeOH
(80 mL) to give the resin-bound dipeptide as white beads.
to a€ord 49 mg (27%) of 2 as a white solid: IR (CHCl3)
1
3417, 3315, 2251, 1655, 1626, 1597 cm
;
1H NMR
(500 MHz, CDCl3) d 12.08 (d, J=8.5 Hz, 1H), 11.30 (d,
J=8.5 Hz, 1H), 10.11 (s, 1H), 8.86 (br s, 1H), 8.64 (d,
J=9.4 Hz, 1H), 8.38 (dd, J=8.9, 2.8 Hz, 1H), 8.19 (br q,
J=4.6 Hz, 1H), 7.63 (d, AB pattern, J=14.9 Hz, 1H),
7.51 (d, AB pattern, J=14.9 Hz, 1H), 7.39±7.32 (m,
3H), 7.20±7.16 (m, 3H), 7.10±7.08 (m, 2H), 6.98 (d,
J=9.0 Hz, 1H), 6.88 (d, J=7.4 Hz, 2H), 6.57 (d,
J=9.0 Hz, 1H), 5.35 (td, J=9.4, 5.6 Hz, 1H), 4.93 (d,
J=9.0 Hz, 1H), 4.83 (q, J=7.6 Hz, 1H), 4.31 (t, J=
8.6 Hz, 1H), 4.09 (s, 3H), 3.95±3.82 (m, 1H), 3.63±3.51
(m, 1H), 3.48 (t, J=6.3 Hz, 2H), 3.42±3.32 (m, 2H), 3.23
(s, 3H), 3.07 (s, 3H), 3.01 (dd, J=14.1, 5.4 Hz, 1H),
2.81±2.75 (m, 1H), 2.77 (d, J=4.6 Hz, 3H), 2.72±2.59
(m, 2H), 1.74±1.66 (m, 1H), 1.65±1.43 (m, 4H), 1.15±
1.04 (m, 1H), 0.91 (d, J=5.6 Hz, 3H), 0.90 (d,
J=5.5 Hz, 3H), 0.81 (d, J=6.8 Hz, 3H), 0.75 (t, J=
7.4 Hz, 3H); 13C NMR (125 MHz, CDCl3) d 172.4, 171.5,
170.8, 165.2, 158.6, 158.3, 157.1, 155.6, 152.3, 140.7,
136.9, 135.0, 133.1, 130.3, 129.3, 128.6, 128.3, 128.2,
127.5, 126.7, 123.3, 122.1, 118.7, 118.1, 111.5, 57.7, 56.3,
56.0, 51.5, 51.1, 48.8, 45.3, 43.0, 39.4, 38.1, 37.5, 35.9,
26.0, 25.1, 24.8, 22.7, 22.5, 17.6, 15.2, 11.2; HRMS
(FAB) m/e for C49H66N11O9 (M+H)+ calcd 952.5045,
found 952.5045. Anal. calcd for C49H66N11O9: C, 61.81;
H, 6.88; N, 16.18. Found: C, 61.70; H, 7.09; N, 15.99.
The beads were washed with CH2Cl2 (3Â80 mL) and
CH2Cl2/TFA/indole solution (1Â80 mL) and shaken
with CH2Cl2/TFA/indole solution (80 mL) for 30 min.
The resin was washed with CH2Cl2 (6Â80 mL), CH2Cl2/
TEA solution (2Â80 mL), and CH2Cl2 (6Â80 mL). The
resulting amine was shaken with a mixture of N-Boc-
Phe-OH (2.7 g, 10 mmol), 1.0 M solution of DCC in
CH2Cl2 (10 mL, 10 mmol), and CH2Cl2 (70 mL) for 2 h.
The resulting white suspension was ®ltered, and the
residue was washed alternately (3Â) with CH2Cl2
(80 mL) and MeOH (80 mL) to give resin-bound tripep-
tide 4 as white beads.
Resin 4 was washed with CH2Cl2 (3Â80 mL) and
CH2Cl2/TFA/indole solution (1Â80 mL) and shaken
with CH2Cl2/TFA/indole solution (80 mL) for 30 min.
The resin was washed with CH2Cl2 (6Â80 mL), CH2Cl2/
TEA solution (2Â80 mL), CH2Cl2 (6Â80 mL), and
DMF (3Â80 mL). The resulting amine was shaken with
a mixture of DMF (80 mL), TEA (0.50 mL, 6.8 mmol),
and carbamoyl chloride 5 (2.3 g, 6.6 mmol) for 24 h. The
resulting white suspension was ®ltered, and the residue
was washed alternately (3Â) with CH2Cl2 (80 mL) and
MeOH (80 mL) to a€ord resin 6 as white beads.
Arti®cial ꢀ-Sheet 3. A 50-mL reaction vessel was
charged with 0.43 g (dry weight) of resin 8 (0.52 mmol/g,
0.22 mmol), and the resin was washed with dry CH2Cl2
(3Â20 mL). The vessel was charged with CH2Cl2
(20 mL) and HCl gas was bubbled into the reaction
vessel through a pipette for 10 min. The solution was
drained, and the resin was washed with CH2Cl2
(2Â20 mL), CH2Cl2/TEA solution (2Â20 mL), and
CH2Cl2 (2Â20 mL). The resulting amine was shaken
with a mixture of DCC (0.20 g, 1.0 mmol), Boc-Val-OH
(0.20 g, 0.92 mmol), and CH2Cl2 (20 mL) for 2 h. The
resulting suspension was ®ltered and the residue was
washed alternately (3Â) with CH2Cl2 (20 mL) and
MeOH (20 mL) to a€ord resin 9 as yellow beads.
Resin 6 was washed with CH2Cl2 (3Â80 mL) and
CH2Cl2/TFA/indole solution (1Â80 mL) and shaken
with CH2Cl2/TFA/indole solution (80 mL) for 30 min.
The resin was drained and washed with CH2Cl2
(6Â80 mL), CH2Cl2/TEA solution (2Â80 mL), and
CH2Cl2 (6Â80 mL). The resulting amine was shaken
with a solution of isocyanate 7 (2.0 g, 6.5 mmol) in
CH2Cl2 (80 mL) for 2 h. The resulting suspension was
®ltered, and the residue was washed alternately (3Â)
with CH2Cl2 (80 mL) and MeOH (80 mL) to a€ord
10.5 g of resin 8 as yellow beads.
Resin 9 was washed with CH2Cl2 (3Â20 mL). The vessel
was charged with CH2Cl2 (20 mL) and HCl gas was
bubbled into the reaction vessel through a pipette for
10 min. The solution was drained, and the resin was
washed with CH2Cl2 (2Â20 mL), CH2Cl2/TEA solution
(2Â20 mL), and CH2Cl2 (2Â20 mL). The resin was sha-
ken with a mixture of isobutyryl chloride (0.50 mL,
7.2 mmol), TEA (0.50 mL, 7.2 mmol), and CH2Cl2
(20 mL) for 30 min. The resin was washed with CH2Cl2
(3Â20 mL) and shaken with 40% (w/w) methylamine in
MeOH (25 mL) for 3 h. The solution was collected and
concentrated by rotary evaporation. The residue was
A 50-mL reaction vessel was charged with 0.50 g (dry
weight) of resin 8 (0.38 mmol/gm, 0.19 mmol), and the
resin was washed with CH2Cl2 (3Â20 mL). The vessel
was charged with CH2Cl2 (20 mL) and HCl gas was
bubbled into the reaction vessel through a pipette for
10 min. The solution was drained, and the resin was
washed with CH2Cl2 (2Â20 mL), CH2Cl2/TEA (2Â20 mL),
and CH2Cl2 (2Â20 mL). The resulting amine was shaken

J. H. Tsai et al. / Bioorg. Med. Chem. 7 (1999) 29±38
35
puri®ed by chromatography on silica gel (i-PrOH/
CH2Cl2, 7/93) followed by reverse-phase HPLC (C18
column, CH3CN/H2O, 70/30) to a€ord 55 mg (25%) of
3 as a white solid: IR (CHCl3) 3421, 3309, 2251, 1657,
312.1193. Anal. calcd for C13H17N3O6: C, 50.16; H, 5.50;
N, 13.50. Found: C, 49.96; H, 5.69; N, 13.38.
A suspension of 5-NO2-2-MeO-C6H3-CONHNHBoc
(0.300 g, 0.964 mmol) and 10% palladium on activated
carbon (0.100 g) in MeOH was stirred under hydrogen
gas for two hours. The suspension was ®ltered through
Celite, the ®ltrate was concentrated by rotary evaporation,
and the residue was puri®ed by column chromatography
on silica gel (i-PrOH/CH2Cl2, 1/9) to a€ord 0.265 g
(98%) of 5-NH2-2-MeO-C6H3-CONHNHBoc as a white
1
1632, 1599 cm 1; H NMR (500 MHz, CDCl3) d 11.61
(d, J=7.0 Hz, 1H), 10.84 (d, J=7.5 Hz, 1H), 10.04 (s,
1H), 8.77 (d, J=2.8 Hz, 1H), 8.40 (d, J=9.8 Hz, 1H),
8.38 (dd, J=8.6, 2.4 Hz, 1H), 8.07 (br q, J=4.6 Hz, 1H),
7.39±7.26 (m, 3H), 7.24±7.21 (m, 3H), 7.14 (d, J=
7.0 Hz, 2H), 6.96±6.93 (m, 3H), 6.51 (d, J=9.0 Hz, 1H),
6.34 (d, J=9.3 Hz, 1H), 5.07 (td, J=8.9, 6.6 Hz, 1H),
4.96±4.92 (m, 2H), 4.83 (dt, J=9.1, 7.3 Hz, 1H), 4.18 (t,
J=9.2 Hz, 1H), 4.04 (s, 3H), 3.90±3.75 (m, 1H), 3.68±
3.66 (m, 2H), 3.56±3.51 (m, 1H), 3.48±3.42 (m, 2H), 2.99
(dd, ABX pattern, JAB=14.0 Hz, JAX=6.4 Hz, 1H),
2.84 (dd, ABX pattern, JAB=14.0 Hz, JAX=9.4 Hz,
1H), 2.76 (d, J=4.5 Hz, 3H), 2.71±2.60 (m, 2H), 2.44
(septet, J=6.9 Hz, 1H), 2.19±2.02 (m, 1H), 1.68±1.49
(m, 3H), 1.43±1.36 (m, 2H), 1.20 (d, J=6.9 Hz, 3H),
1.14 (d, J=6.8 Hz, 3H), 1.06 (d, J=6.7 Hz, 3H), 1.02 (d,
J=6.8 Hz, 3H), 0.95 (d, J=6.1 Hz, 6H), 0.97±0.90 (m,
1H), 0.67 (d, J=6.6 Hz, 3H), 0.51 (t, J=7.3 Hz, 3H);
13C NMR (125 MHz, CDCl3) d 176.7, 172.4, 171.6,
170.8, 166.1, 159.7, 157.4, 155.5, 152.2, 141.0, 136.8,
135.0, 130.3, 129.2, 128.4, 128.0, 127.4, 126.7, 123.2,
122.1, 118.7, 118.2, 111.4, 57.8, 56.3, 56.2, 55.5, 51.3, 50.8,
48.5, 45.3, 42.3, 38.6, 38.0, 35.6, 32.6, 26.1, 24.8, 24.6,
23.0, 22.6, 20.1, 19.2, 19.0, 18.4, 17.6, 15.2, 10.8; HRMS
(FAB) m/e for C52H74N11O9 (M+H)+ calcd 996.5671,
found 996.5659. Anal. calcd for C52H73N11O9: C, 62.69;
H, 7.39; N, 15.47. Found: C, 62.50; H, 7.46; N, 15.06.
1
solid: IR (KBr) 3356, 3248, 1736, 1657, 1618 cm 1; H
NMR (500 MHz, CDCl3) d 9.67 (br s, 1H), 7.50 (d,
J=2.9 Hz, 1H), 7.36 (br s, 1H), 6.78 (dd, J=8.6, 2.7 Hz,
1H), 6.74 (d, J=9.0 Hz, 1H), 3.84 (s, 3H), 3.63 (br s, 2H),
1.47 (s, 9H); 13C NMR (125 MHz, CDCl3) d 164.2, 155.1,
150.7, 140.5, 120.0, 119.6, 118.3, 112.8, 81.5, 56.5, 28.1;
HRMS (CI) m/e for C13H19N3O4 (M+), calcd 281.1375,
found 281.1370. Anal. calcd for C13H19N3O4: C, 55.51;
H, 6.81; N, 14.94. Found: C, 55.18; H, 6.70; N, 14.80.
An ice-cooled, biphasic mixture of 5-NH2-2-MeO-
C6H3-CONHNHBoc (0.150 g, 0.533 mmol), 40 mL of
CH2Cl2, and 40 mL of satd aq NaHCO3 was rapidly
stirred while a 1.93 M phosgene in toluene (1.0 mL,
1.9 mmol) was added as a single portion. [CAUTION:
PHOSGENE VAPOR IS HIGHLY TOXICÐUSE
HOOD.] The reaction mixture was stirred for 30 min,
the organic phase was collected, and the aqueous phase
was extracted with CH2Cl2 (2Â15 mL). The combined
organic layers were dried over MgSO4, ®ltered, and
concentrated by rotary evaporation to a€ord 0.17 g
(102%) of 7 as a white solid: IR (neat) 3367, 2270, 1738,
Carbamoyl chloride 5. An ice-cooled, biphasic mixture
of PhNH(CH2)2N(Boc)-(CH2)2CN24 (0.050 g, 0.17 mmol),
15 mL of CH2Cl2, and 15 mL of satd aq NaHCO3 was
rapidly stirred while a 1.93 M phosgene in toluene
(0.5 mL, 0.97 mmol) was added as a single portion.
[CAUTION: PHOSGENE VAPOR IS HIGHLY
TOXICÐUSE HOOD.] The reaction mixture was stir-
red for 30 min, the organic phase was collected, and the
aqueous phase was extracted with CH2Cl2 (2Â15 mL).
The combined organic layers were dried over MgSO4,
®ltered, and concentrated by rotary evaporation to
a€ord 0.061 g (100%) of 5 as a thin ®lm: IR (neat) 2249,
1
1659 cm 1; H NMR (400 MHz, CDCl3) d 9.53 (br s,
1H), 7.97 (d, J=2.8 Hz, 1H), 7.19 (dd, J=8.8, 2.8 Hz,
1H), 6.94 (d, J=8.8 Hz, 1H), 6.93 (br s, 1H), 4.00 (s,
3H), 1.51 (s, 9H); 13C NMR (125 MHz, CDCl3) d 162.9,
155.1, 155.0, 128.9, 128.2, 126.8, 124.4, 120.2, 112.4,
81.5, 56.4, 28.0; HRMS (CI) m/e for C14H17N3O5 (M+),
calcd 307.1168, found 307.1164. Anal. calcd for
C14H17N3O5: C, 54.72; H, 5.58; N, 13.67. Found: C,
55.06; H, 5.90; N, 13.33.
N,N-Dimethylfumaramic acid. A solution of fumaric
acid monoethyl ester (5.10 g, 35.4 mmol), 40% (w/w) aq
Me2NH (1.5 mL, 33 mmol), and 1-ethyl-3-(30-dimethyl-
aminopropyl) carbodiimide hydrochloride (7.50 g,
39.1 mmol) in THF was stirred for 8 h and then con-
centrated by rotary evaporation. The residue was puri®ed
by column chromatography on silica gel (EtOAc/hexanes,
3/1) to a€ord 2.3 g (45%) of ethyl N,N-dimethylfumar-
amate (trans-Me2NCOCH=CHCO2Et) as a colorless
oil: IR (neat) 1722, 1659, 1630 cm 1; 1H NMR (500 MHz,
CDCl3) d 7.41 (d, J=15.5 Hz, 1H), 6.79 (d, J=15.0 Hz,
1H), 4.26 (q, J=7.0 Hz, 2H), 3.14 (s, 3H), 3.05 (s, 3H),
1.32 (t, J=7.0 Hz, 3H); 13C NMR (100 MHz, CDCl3) d
165.6, 164.7, 133.6, 131.1, 61.0, 37.4, 35.7, 14.0; HRMS
(CI) m/e for C8H14NO3 (M+H)+, calcd 172.0973, found
172.0971. Anal. calcd for C8H13NO3: C, 56.13; H, 7.65;
N, 8.18. Found: C, 55.84; H, 7.67; N, 7.97.
1
ꢀ
1738, 1697; H NMR (500 MHz, CDCl3, 50 C) d 7.45±
7.33 (m, 5H), 3.87 (br s, 2H), 3.58 (t, J=6.7 Hz, 2H),
3.46 (br s, 2H), 2.59 (t, J=6.5 Hz, 2H), 1.40 (br s, 9H).
Isocyanate 7. A solution of 2-methoxy-5-nitrobenzoic
acid24 (15.0 g, 76.1 mmol), t-butyl carbazate, and 1-
ethyl-3-(30-dimethylaminopropyl) carbodiimide hydro-
chloride (16.1 g, 83.7 mmol) in THF was stirred for 8 h.
The reaction mixture was diluted with 500 mL of water,
and a yellow precipitate formed. The precipitate was
isolated by ®ltration and recrystallized (MeOH/H2O, 1/1)
to a€ord 20.5 g (87%) of 5-NO2-2-MeO-C6H3-CON-
HNHBoc as a white solid: IR (CHCl3) 3410, 1741, 1678
cm 1; 1H NMR (300 MHz, CDCl3) d 9.40 (br s, 1H), 9.10
(d, J=2.8 Hz, 1H), 8.37 (dd, J=9.6, 3.0 Hz, 1H), 7.11 (d,
J=9.4 Hz, 1H), 6.96 (br s, 1H), 4.13 (s, 3H), 1.51 (s, 9H);
13C NMR (125 MHz, CDCl3) d 162.0, 161.6, 155.1, 141.7,
128.6, 128.3, 120.2, 111.9, 81.9, 57.2, 28.1; HRMS (CI)
m/e for C13H18N3O6 (M+H)+, calcd 312.1195, found
A mixture of ethyl N,N-dimethylfumaramate (1.80 g,
10.5 mmol), THF (100 mL), and 150 mL of 0.5 M

36
J. H. Tsai et al. / Bioorg. Med. Chem. 7 (1999) 29±38
NaOH was stirred at 0ꢀC for 45 min. The aqueous layer
was washed with CH2Cl2 (3Â100 mL) and slowly acid-
i®ed to pH 2 with concentrated HCl while cooling with
an ice bath. The resulting aqueous solution was con-
centrated to dryness. The solid residue was extracted
with CH2Cl2, and the resulting suspension was ®ltered
and concentrated to dryness to a€ord 1.39 g (92%) of
N,N-dimethylfumaramic acid as a white solid: IR
1: 3 (s), 5 (w), 24 (m), 25 (m), 26 (m). 2: 3 (w), 4 (w), 6
(w), 12 (m), 22 (m). 3: 1 (s), 2 (w), 7 (m), 13 (w), 18 (m),
19 (w), 34 (w), 35 (w), 39 (w). 4: 2 (w), 6 (m). 5: 1 (w), 14
(s). 6: 2 (w), 4 (m), 9 (s), 12 (w), 20 (s), 36 (w). 7: 3 (m),
13 (w), 17 (w), 18 (s), 19 (w), 21 (w), 34 (w), 35 (w), 38
(w). 8: 23 (w), 27 (s). 9: 6 (m), 17 (w), 20 (s), 27 (w), 29
(m), 36 (w). 10: 15 (s). 11: 13 (s). 12: 2 (s), 6 (w), 16 (s).
13: 3 (w), 7 (w), 11 (s), 18 (m), 19 (m), 28 (m), 30 (m).
14: 5 (s), 22 (s). 15: 10 (s), 19 (m), 24 (w), 25 (w), 26 (w).
16: 12 (s), 23 (s). 17: 7 (w), 9 (w), 20 (m), 21 (s), 32 (m),
33 (m), 38 (w). 18: 3 (m), 7 (s), 13 (m), 19 (s), 28 (m), 30
(m). 19: 3 (w), 7 (w), 13 (w), 15 (m), 18 (s), 28 (m), 30
(m). 20: 6 (m), 9 (s), 17 (m), 32 (m), 33 (m), 36 (m), 37
(m). 21: 7 (m), 17 (s), 32 (m), 34 (w), 35 (w), 38 (s), 39
(m). 22: 2 (m), 14 (s). 23: 8 (m), 16 (s). 24: 1 (m), 15 (w),
25 (s), 26 (w). 25: 1 (m), 15 (w), 24 (s), 26 (m). 26: 1 (m),
15 (w), 24 (w), 25 (m), 31 (m). 27: 8 (s), 9 (w), 36 (m). 28:
13 (m), 18 (m), 19 (m), 30 (s). 29: 9 (s). 30: 13 (m), 18
(w), 19 (m), 28 (s). 31: 26 (m). 32: 17 (m), 20 (m), 21 (m),
33 (s), 34 (m), 36 (s), 38 (s), 39 (m). 33: 17 (m), 20 (w), 32
(s), 36 (m), 37 (m). 34: 3 (w), 7 (w), 21 (w), 32 (m), 35 (s),
39 (m). 35: 3 (w), 7 (w), 21 (w), 34 (s), 38 (m), 39 (m). 36:
6 (w), 9 (w), 20 (m), 27(w), 32 (s), 33 (s). 37: 20 (m), 33
(m). 38: 7 (w), 17 (w), 21 (s), 32 (s), 35 (m), 39 (s). 39: 3
(w), 21 (m), 32 (m), 34 (m), 35 (s), 38 (s).
1
(CHCl3) 2939, 2765, 2449, 1703, 1649, 1628 cm 1; H
NMR (300 MHz, CDCl3) d 10.00 (br s, 1H), 7.48 (d, AB
pattern, J=15.3 Hz, 1H), 6.80 (d, AB pattern, J=
15.0 Hz, 1H), 3.15 (s, 3H), 3.07 (s, 3H); 13C NMR
(125 MHz, CD3SOCD3) d 166.6, 164.1, 134.3, 130.8,
37.0, 35.2; HRMS (CI) m/e for C6H10N3O (M+H)+
calcd 144.0660, found 144.0661. Anal. calcd for
C6H9N3O: C, 50.35; H, 6.34; N, 9.79. Found: C, 49.95;
H, 6.26; N, 9.43.
1H NMR Tr-ROESY experiments. Tr-ROESY studies
of arti®cial b-sheets 1±3 were performed on a Bruker
DRX500 spectrometer using 10 mM samples in CDCl3
that were degassed by ®ve freeze±pump±thaw cycles on
a high-vacuum line (<0.001 mm Hg) and sealed under
vacuum. ROESY spectra were recorded at 30 ꢀC with a
mixing time of 300 ms using the Tr-ROESY pulse
sequence of Hwang and Shaka.33,34 A spin locking ®eld
strength of 5 kHz was employed to collect 2048 points in
the f2 dimension and 1024 points in the f1 dimension, a
relaxation delay of 1 s was used, and 16 transients were
collected for each t1 value. The data were processed
using Bruker XWINNMR software. A sine-squared
window function shifted by 60ꢀ was applied in both
dimensions, and forward linear prediction of 1024
points was performed in the f1 dimension to give a ®nal
matrix of 1024 by 1024 real points. An automatic base-
line correction was applied in both dimensions after
Fourier transforming the data. The ROESY cross-peaks
were tabulated as follows: The proton resonances for
each of the three arti®cial b-sheets were numbered in
order of decreasing chemical shifts, and the resonances
were assigned by means of the ROESY and COSY
experiments. (These assignments are shown on the dia-
grams that follow.) The cross-peaks in the f1 dimension
were tabulated for each resonance in the f2 dimension
and were characterized as strong (s), medium (m), or
weak (w) on the basis of their relative peak heights
(45±100%, s; 15±45%, m; 3±15%, w).
1H NMR Tr-ROESY cross-peaks for arti®cial ꢀ-sheet
2.
1: 2 (m), 5 (w), 8 (m), 9 (w), 31 (w), 37 (w). 2: 1 (m), 20
(m). 3: 4 (s), 6 (w), 21 (m), 22 (w), 24 (s). 4: 3 (s), 5 (m), 6
(w), 12 (w), 16 (m), 37 (w). 5: 1 (w), 4 (m), 16 (s), 17 (w),
19 (m), 27 (w), 31 (m), 34 (w), 35 (w), 37 (m). 6: 3 (w), 4
(w), 13 (s). 7: 8 (m), 18 (s), 26 (m), 29 (s), 32 (w), 36 (w).
8: 1(s), 7 (m), 9 (s), 18 (s), 25 (m), 26 (w), 36 (m). 9: 1
(w), 8 (s), 18 (m), 25 (s), 26 (w). 10: 14 (s). 11: 12 (s). 12:
4 (w), 11 (s), 16 (s), 17 (m), 27 (m), 28 (s), 36 (w). 13: 6
(s), 20 (s). 14: 10 (s), 17 (m), 21 (w), 24 (m). 15: 18 (m),
19 (s), 32 (w), 33 (w), 37 (w). 16: 4 (m), 5 (s), 12 (m), 17
(m), 27 (m), 28 (m). 17: 5 (w), 12 (w), 14 (m), 16 (m), 27
(w), 28 (m). 18: 7 (s), 8 (m), 9 (w), 15 (m), 32 (m), 33 (m),
36 (s). 19: 5 (m), 15 (s), 31 (m), 34 (m), 35 (w), 37 (s), 38
(m). 20: 2 (m), 13 (s). 21: 3 (m), 14 (w), 22 (m). 22: 3 (w),
21 (m). 23: 30 (s). 24: 3 (s), 14 (w). 25: 8 (m), 9 (s). 26: 7
(m), 8 (w), 9 (m), 36 (m). 27: 5 (w), 12 (m), 16 (m), 17
(w), 28 (s). 28: 12 (m), 16 (m), 17(m), 27 (s). 29: 7 (s). 30:
23 (s). 31: 1 (w), 5 (m), 19 (m), 34 (m), 35 (m), 37 (s), 38
(m). 32: 7 (w), 15 (m), 18 (m), 36 (s). 33: 15 (w), 18 (m),
36 (s). 34: 5 (w), 19 (m), 31 (m), 35 (s), 37 (m), 38 (s). 35:
1H NMR Tr-ROESY cross-peaks for arti®cial ꢀ-sheet
1.

J. H. Tsai et al. / Bioorg. Med. Chem. 7 (1999) 29±38
37
5 (w), 19 (w), 31 (m), 34 (s), 37 (m), 38 (m). 36: 7 (m), 8
(m), 12 (m), 18 (s), 26 (m), 32 (s), 33 (s). 37: 1 (w), 5 (m),
15 (w), 19 (m), 31 (s), 34 (m), 35 (s). 38: 19 (m), 31 (m),
34 (s). 35 (s).
4. Roberts, J. C.; Pallai, P. V.; Rebek, Jr. J. Tetrahedron Lett.
1995, 36, 691.
5. Boumendjel, A.; Roberts, J. C.; Hu, E.; Pallai, P. V.;
Rebek, Jr. J. J. Org. Chem. 1996, 61, 4434.
6. Smith, III, A. B.; Keenan, T. P.; Holcomb, R. C.;
Sprengeler, P. A.; Guzman, M. C.; Wood, J. L.; Carroll, P. J.;
Hirschmann, R. J. Am. Chem. Soc. 1992, 114, 10672.
7. Smith, III A. B.; Holcomb, R. C.; Guzman, M. C.; Keenan,
T. P.; Sprengeler, P. A.; Hirschmann, R. Tetrahedron Lett.
1993, 34, 63.
1H NMR Tr-ROESY cross-peaks for arti®cial ꢀ-sheet
3.
8. Smith, III A. B.; Hirschmann, R.; Pasternak, A.; Akaishi,
R.; Guzman, M. C.; Jones, D. R.; Keenan, T. P.; Sprengeler,
P. A.; Darke, P. L.; Emini, E. A.; Holloway, M. K.; Schleif,
W. A. J. Med. Chem. 1994, 37, 215.
9. Zutshi, R.; Franciskovich, J.; Shultz, M.; Schweitzer, B.;
Bishop, P.; Wilson, M.; Chmielewski, J. J. Am. Chem. Soc.
1997, 119, 4841.
10. Somers, W. S.; Phillips, S. E. V. Nature 1992, 359, 387.
11. Harper, J. D.; Lansbury, P. T. Annu. Rev. Biochem. 1997,
66, 385.
12. Verbeek, M. M.; Ruiter, D. J.; de Waal, R. M. W. Biol.
Chem. 1997, 378, 937.
13. Selkoe, D. J. Int. J. Peptide Protein Res. 1994, 53, 438.
14. Lansbury, P. T. Acc. Chem. Res. 1996, 29, 317.
15. Perutz, M. F.; Johnson, T.; Suzuki, M.; Finch, J. T. Proc.
Natl. Acad. Sci. USA 1994, 91, 5355.
16. Stott, K.; Blackburn, J. M.; Butler, P. J. G.; Perutz, M.
Proc. Natl. Acad. Sci. USA 1995, 92, 6509.
17. Paulson, H. L.; Fischbeck, K. H. Annu. Rev. Neuroscience
1996, 19, 79.
18. Kemp, D. S.; Bowen, B. R. Tetrahedron Lett. 1988, 29,
5081.
19. Kemp, D. S.; Bowen, B. R. Tetrahedron Lett. 1988, 29,
5077.
20. Kemp, D. S.; Bowen, B. R.; Muendel, C. C. J. Org. Chem.
1990, 55, 4650.
21. Schrader, T.; Kirsten, C. Chem. Commun. 1996, 2089.
22. Kirsten, C. N.; Schrader, T. H. J. Am. Chem. Soc. 1997,
119, 12061.
23. Nowick, J. S.; Holmes, D. L.; Mackin, G.; Noronha, G.;
Shaka, A. J.; Smith, E. M. J. Am. Chem. Soc. 1996, 118, 2764.
24. Nowick, J. S.; Pairish, M.; Lee, I. Q.; Holmes, D. L.;
Ziller, J. W. J. Am. Chem. Soc. 1997, 119, 5413.
25. Smith, E. M.; Holmes, D. L.; Shaka, A. J.; Nowick, J. S.
J. Org. Chem. 1997, 62, 7906.
26. Nowick, J. S.; Abdi, M.; Bellamo, K. A.; Love, J. A.;
Martinez, E. J.; Noronha, G.; Smith, E. M.; Ziller, J. W. J.
Am. Chem. Soc. 1995, 117, 89.
1: 2 (m), 5 (m), 18 (s), 32 (w), 35 (w), 38 (w), 41 (w). 2: 1
(m), 21 (m). 3: 4 (s), 22 (m), 23 (s), 26 (w). 4: 3 (s), 5 (m),
11 (w), 16 (m), 36 (w), 42 (w). 5: 1 (w), 4 (m), 11 (w), 16
(s), 17 (w), 20 (m), 35 (m), 41 (w). 6: 13 (s). 7: 18 (w), 19
(s), 29 (s), 37 (w), 39 (w). 8: 12 (s). 9: 12 (s). 10: 11 (s).
11: 4 (w), 5 (w), 10 (s), 16 (m), 17 (m), 27 (m), 28 (m).
12: 8 (s), 9 (s), 17 (m), 23 (m). 13: 6 (s), 21 (s). 14: 19 (m),
20 (m), 33 (w), 34 (w). 15: 18 (m), 31 (m), 32 (w), 37 (w),
38 (m). 16: 4 (m), 5 (s), 11 (m), 17 (s), 27 (m), 28 (m). 17:
5 (w), 11 (w), 12 (m), 16 (s), 27 (m), 28 (m). 18: 1 (s), 7
(w), 15 (m), 19 (s), 32 (m), 38 (m), 39 (m). 19: 7 (s), 14
(m), 18 (s), 33 (m), 34 (m), 38 (w), 39 (s). 20: 5 (m), 14
(s), 35 (m), 40 (w), 41 (m), 42 (w). 21: 2 (m), 13 (s). 22: 3
(m). 23: 3 (s), 12 (w), 26 (m). 24: 25 (s), 30 (m). 25: 24
(s), 30 (m). 26: 3 (w), 23 (m). 27: 11 (m), 16 (m), 17 (w),
28 (s). 28: 11 (m), 16 (m), 17 (m), 27 (s). 29: 7 (s), 37 (m).
30: 24 (m), 25 (m). 31: 15 (s), 37 (s). 32: 1 (w), 15 (w), 18
(m), 38 (s). 33: 14 (w), 19 (m), 39 (s). 34: 14 (w), 19 (m),
39 (s). 35: 1 (w), 5 (m), 20 (m), 41 (s), 42 (m). 36: 4 (w),
40 (s). 37: 7 (w), 15 (w), 29 (m), 31 (s). 38: 1 (w), 15 (m),
18 (m), 19 (w), 32 (s). 39: 7 (w), 18 (w), 19 (s), 33 (s), 34
(s), 42 (m). 40: 20 (m), 36 (s). 41: 1 (w), 5 (w), 20 (m), 35
(s), 42 (s). 42: 4 (w), 20 (m), 35 (s), 39 (s), 41 (s).
27. Nowick, J. S.; Mahrus, S.; Smith, E. M.; Ziller, J. W. J.
Am. Chem. Soc. 1996, 118, 1066.
28. Ramondo, F.; Bencivenni, L. J. Chem. Soc., Perkin Trans.
2 1995, 1797.
29. Reynolds, C. H.; Hormann, R. E. J. Am. Chem. Soc. 1996,
118.
30. Holmes, D. L.; Smith, E. M.; Nowick, J. S. J. Am. Chem.
Soc. 1997, 119, 7665.
31. Stewart, J. M.; Young, J. D. Solid Phase Peptide Synth-
esis; W. H. Freeman: San Francisco, 1969.
32. In an early attempt to synthesize arti®cial b-sheet 2, we
used TFA instead of HCl and obtained a product in which the
nitrile group hydrolyzed to the corresponding primary amide.
As a result, the overall yield of 2 was only 11%. In other stu-
dies, we have also observed that the nitrile group is sensitive to
hydrolysis by acids, and its sensitivity appears to be a general
problem.
Acknowledgements
This work was supported by the National Institutes of
Health Grant GM-49076. J. S. N. thanks the following
agencies for support in the form of awards: the National
Science Foundation (Presidential Faculty Fellow
Award), the Camille and Henry Dreyfus Foundation
(Teacher±Scholar Award), and the Alfred P. Sloan
Foundation (Alfred P. Sloan Research Fellowship).
References and Notes
1. Nassar, N.; Horn, G.; Herrmann, C.; Scherer, A.;
McCormick, F.; Wittinghofer, A. Nature 1995, 375, 554.
2. Doyle, D. A.; Lee, A.; Lewis, J.; Kim, E.; Sheng, M.;
MacKinnon, R. Cell 1996, 85, 1067.
33. Hwang, T. L.; Shaka, A. J. J. Am. Chem. Soc. 1992, 114,
3157.
3. Michne, W. F.; Schroeder, J. D. Int. J. Peptide Protein Res.
1996, 47, 2.
34. Hwang, T. L.; Shaka, A. J. J. Magn. Res., Ser. B 1993,
102, 155.

38
J. H. Tsai et al. / Bioorg. Med. Chem. 7 (1999) 29±38
35. Dyson, H. J.; Wright, P. E. Annu. Rev. Biophys. Biophys.
Chem. 1991, 20, 519.
38. Wishart, D. S.; Sykes, B. D.; Richards, F. M. Biochem-
istry 1992, 31, 1647.
36. Nowick, J. S.; Smith, E. M.; Noronha, G. J. Org. Chem.
1995, 60, 7386.
39. Ribeiro, A. A.; Goodman, M.; Naider, F. Int. J. Peptide
Protein Res. 1979, 14, 414.
37. Wishart, D. S.; Sykes, B. D.; Richards, F. M. J. Mol. Biol.
1991, 222, 311.
40. Gellman, S. H.; Adams, B. R. Tetrahedron Lett. 1989, 30,
3381.