Secondary structure peptide mimetics: Design, synthesis, and evaluation of β-strand mimetic thrombin inhibitors

Article abstract of DOI:10.1021/jm980354p

Constrained dipeptide mimetic templates were designed to mimic the secondary structure of peptides in a β-strand conformation. Two templates corresponding to the D-Phe-Pro portion of the thrombin inhibitor D-Phe-Pro- ArgCH₂Cl were synthesized a

Full text of DOI:10.1021/jm980354p

J . Med. Chem. 1999, 42, 1367-1375
1367
Secon d a r y Str u ctu r e P ep tid e Mim etics: Design , Syn th esis, a n d Eva lu a tion of
â-Str a n d Mim etic Th r om bin In h ibitor s
P. Douglas Boatman,*,§ Cyprian O. Ogbu,§ Masakatsu Eguchi,§,† Hwa-Ok Kim,§ Hiroshi Nakanishi,§,†
Bolong Cao,§ J . Paul Shea,§ and Michael Kahn*,†,§
Molecumetics Ltd., 2023 120th Avenue NE, Suite 400, Bellevue, Washington 98005-2199, and Department of Pathobiology,
University of Washington, SC-38, Seattle, Washington 98195
Received J une 8, 1998
Constrained dipeptide mimetic templates were designed to mimic the secondary structure of
peptides in a â-strand conformation. Two templates corresponding to the D-Phe-Pro portion of
the thrombin inhibitor D-Phe-Pro-ArgCH2Cl were synthesized and converted into nine
R-ketoamide and R-ketoheterocycle inhibitors of thrombin. Additionally, a template corre-
sponding to L-Phe-Pro was synthesized and converted to a thrombin inhibitor. The in vitro
inhibition of thrombin by these compounds was determined, and those corresponding to the
D-Phe-Pro were found to be more potent inhibitors than the L-Phe-Pro mimetic. The
R-ketoamides were found to be more potent than the R-ketoheterocycles but had much slower
on rates. By comparison of a series of R-ketoamide analogues, it is apparent that the there is
a preference for binding of bulky hydrophobic substituents in the P′ portion of the thrombin
active site. Three of the inhibitors (MOL098, MOL144, and MOL174) were screened against a
series of coagulation and anticoagulation enzymes and found to be selective for inhibition of
the coagulation enzymes. Two of the inhibitors were tested in in vitro models of intestinal
absorption and found to have low absorption potential. The compounds were then tested in
vivo in both rats and primates, and one of them (MOL144) was approximately 25% absorbed
in both species. This study has delineated the synthesis of constrained dipeptide â-strand
mimetics and validated the potential for compounds of this type as potent thrombin inhibitors
and possible drug leads.
In tr od u ction
thrombin which is key among the enzymes in the blood
coagulation cascade in thrombus formation.
The utilization of 20 amino acid side chains, displayed
on a limited array of topological templates (i.e., common
secondary structure motifs such as reverse turns,
â-strands, and R-helices) provides the elements required
to delineate ligand-receptor and enzyme substrate/
inhibitor interactions for a multitude of processes. We
have been engaged in the development of mimics of
these secondary structure motifs1,2 and in particular the
utilization of combinatorial libraries of these “privileged
templates” to probe biological structure-function rela-
tionships and to develop orally available pharmaceutical
agents.
The trypsin-like serine proteases form a large and
diverse family of enzymes which includes the enzymes
in the blood coagulation cascade. Sequencing of these
proteases has shown the presence of a homologous
trypsin-like core with insertions that can modify speci-
ficity and which are generally responsible for interac-
tions with other macromolecular components.3 With
regard to the recognition of proteolytic substrates and
proteinaceous inhibitors by their cognate enzymes,
inspection of numerous X-ray crystal structures has
highlighted the fact that an extended strand motif is
uniformly adopted by the inhibitor/substrate in the
active site.4 Inhibition of certain enzymes in this family
may be beneficial in ameliorating associated disease
states. One potential target is the serine protease
Thrombin exhibits remarkable specificity in the re-
moval of fibrinopeptides A and B of fibrinogen, through
the selective cleavage of two Arg-Gly bonds of the 181
Arg/Lys-Xaa sequences in fibrinogen.5 Fibrinogen is
subsequently cross-linked by factor XIIIa, which is also
activated by thrombin, giving an insoluble fibrin clot.
Thrombin also activates platelets by binding to a specific
PAR-1 receptor on the cell surface and cleaving between
residues Arg41 and Ser42 to produce a tethered ligand
which activates the receptor.6 Activation of platelets
results in shape change, platelet aggregation, and
release of secretory granules, all of which strengthen
the blood clot. Thrombin also proteolytically activates
factors V and VIII, thereby amplifying its own produc-
tion.
PPACK (D-Phe-Pro-ArgCH2Cl), a compound which
had earlier reached phase II clinical evaluation, has
been shown to be a potent inhibitor of thrombin,7 and
many thrombin inhibitors based on this tripeptide
sequence have been synthesized.8,9 The X-ray crystal
structure of PPACK-thrombin has been determined
and reveals that the peptide binds in an extended strand
conformation similar to that of thrombin’s natural
substrate, fibrinogen.10 The templates 1 and 2 (Chart
1) were designed as mimics for an extended â-strand
secondary structure. We have synthesized these tem-
plates with substituents chosen to match those of the
D-Phe-Pro portion of PPACK and converted these to a
number of potent and selective inhibitors of thrombin.
* To whom correspondence should be addressed.
§ Molecumetics Ltd.
† University of Washington.
10.1021/jm980354p CCC: $18.00 © 1999 American Chemical Society
Published on Web 03/31/1999

1368 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 8
Ch a r t 1. Bicyclic â-Strand Mimetic Templates
Boatman et al.
Sch em e 1a
a
Reagents and conditions: (a) PhCHO, Et3N, CH2Cl2, rt; (b)
LDA, -78 °C, then allyl bromide, -78 °C f rt; (c) 1 M HCl (aq),
MeOH, rt, 1 h; (d) Boc2O, NaHCO3, THF/H2O, rt, 3 days; (e) O3,
CH2Cl2/MeOH; (f) hydrazine, THF, reflux, 3 days; (g) Pt2O (cat.),
H2, MeOH, rt, 48 h; (h) HCHO, ethyl acrylate, reflux; (i) LiOH (2
equiv, 1 M aq), THF, rt, 1.5 h.
Resu lts a n d Discu ssion
Design a n d Mod elin g. The bicyclic structures 1 and
2 were chosen as structures that would rigidify the
peptide backbone of the peptidomimetic inhibitors and
place the encompassed functional groups in approxi-
mately the orientation of an idealized peptide. This
strategy has been used extensively in the production of
potential â-turn mimetics11 but has not been used for
the production of â-strand mimetics. The â-turn mimet-
ics of this type generally encompass the i+1 and i+2
residues of the turn and position the i and i+3 residues
for formation of a hydrogen bond between the respective
amide NH and carbonyl. However, bicyclic structures
such as 1 and 2 also orient the atoms encompassed by
the template in positions that are approximately the
same as those in an extended strand. We chose to exploit
this similarity.
Monte Carlo conformational searches were carried out
on the â-sheet templates 1 and 2 with R1 ) CH3CONH,
R2 ) CH3, R3 ) H, and Y ) NHCH3 in a simulated
water environment (GBSA solvation12) using MM2/
MMOD force fields.13 Force-field parameters for the
angles and bond lengths were adopted from Amber*
force field, and the N(sp2)-N(sp3) distance (1.41 Å) was
from X-ray crystal data.14,15 No dihedral force constants
were assigned for the C-N-N-C. Low-energy confor-
mations were compared with dialanine peptides with
ideal parallel and antiparallel â-sheet conformations,
(φ,ψ))(-119,113) and (-139,135), respectively. Boltz-
mann population pi for conformer i at room remperature
(T ) 300 K) was estimated from the energy difference
of conformer i from the global minimum energy, assum-
ing the energy barriers between the conformers are
negligible, i.e., pi ) [exp(-∆E/RT)]/[∑∞i-1 exp(-∆E/
RT)]. The summations in the denominator were trun-
cated at conformer j with ∆Ej ) 50 kJ /mol. Boltzmann
weighted average rmsd (root mean-square deviation)
values at the seven superimposable heavy atom posi-
tions were 0.4 and 0.5 Å for 1 and 0.5 Å and 0.5 Å for 2
against parallel and antiparallel â-strands, respectively.
More importantly, the templates 1 and 2 (R1 ) CH3,
R2 ) NH2, R3 ) H, and Y ) NH2) superimposed very
well against BPTI, which is believed to represent a
canonical loop proteolysis substrate mimic, at the P2-
P3 sites and against PPACK with a Boltzmann average
rmsd value of 0.2 and 0.4 Å, respectively.
of thrombin, we chose to synthesize the templates 3 and
4 (Chart 1) with substituents matching those of PPACK.
Syntheses of 2-oxo-3-(N-Boc-amino)-1-azabicyclo[4.3.0]-
nonane-9-carboxylic acids (3) have been published pre-
viously.16,17 The synthesis of racemic 3-(N-Bocamino)-
3-benzyl-2-oxo-1,6-diazabicyclo[4.3.0]nonane-9-car-
boxylate (4) is detailed in Scheme 1.18
The benzaldimine of phenylalanine methyl ester was
formed and alkylated with allyl bromide, and the imine
was hydrolyzed to give (()-R-allylphenylalanine methyl
ester, 5. Protection of 5 as the N-Boc derivative and
ozonolysis of the olefin followed by reductive workup
gave the aldehyde ester 7 needed for formation of the
heterocyclic core of the peptide mimetic. Treatment of
the aldehyde ester with hydrazine at room temperature
gave mixtures of cyclic hydrazone 8 and isomeric acyclic
hydrazones. TLC and NMR evidence indicated rapid
formation of the acyclic hydrazone and slow formation
of the heterocycle. It was found that heating the reaction
mixture for extended periods gave the desired cyclic
hydrazone 8 in good yield. The slow rate of this
cyclization is probably due to a combination of slow
reaction of the hindered ester and equilibration of the
acyclic hydrazone isomers. Reduction by sodium cyano-
borohydride mediated by p-toluenesulfonic acid19 and
basic hydrolysis of the cyanoborane intermediate was
inefficient and capricious. However, hydrogenation of
8 over Adam’s catalyst was more reproducible and
cleanly yielded the tetrahydropyridazinone 9. The 1,3-
dipolar cycloaddition of the iminium ion, formed by
treatment of 9 with formaldehyde, in the presence of a
large excess of ethyl acrylate gave a mixture of diaster-
eomeric and regioisomeric products from which the
desired bicyclic template could be isolated in 27% yield.
Hydrolysis of the ester gave the Boc-protected amino
acid template 4.
An inhibitor of thrombin, 12 (MOL098), was produced
by coupling the template with HArg(Mtr)CH2Cl (11)
followed by TFA deprotection (Scheme 2). Although the
thrombin inhibitory activity of 12 was good, rivaling
that of PPACK, the possibility of toxicity due to non-
specific alkylation of the highly reactive chloromethyl
ketone led us to pursue alternative inhibitory strategies.
We chose to synthesize a number of arginine derivatives
with activated electrophilic carbonyl groups in order to
Ch em istr y. For the production of potential inhibitors

Secondary Structure Peptide Mimetics
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 8 1369
Sch em e 2a
Sch em e 4a
a
Reagents and conditions: (a) benzothiazole, n-BuLi, Et2O, -78
°C; (b) TFA, CH2Cl2 (1:4), rt, 0.5 h.
Sch em e 5a
a
Reagents and conditions: (a) i-BuOCOCl, NMM, THF, -50
°C, 10 min then 11; (b) TFA, H2O, PhSCH3, rt.
Sch em e 3a
a
Reagents and conditions: (a) EDC, HOBt, i-Pr2NEt, THF, rt;
(b) Dess-Martin periodinane, CH2Cl2, rt; (c) TFA, H2O, PhSMe
(19:1:1), rt, HPLC separation.
Ta ble 1
corporate
code
compd
X
Y
Z
18a
18b
18c
18d
18e
18f
18g
18h
18i
MOL106
N
CONHcyclohexyl
CONHCH2Ph
CONHCH2CH2Ph
CONHCH2CH2Php-Cl
CONHCH2CH2Php-OMe NHC(dNH)NH2
2-thiazolyl
CH2NH2
a
Reagents and conditions: (a) (MeS)3CH, n-BuLi, THF, -78
N
N
N
N
N
N
NHC(dNH)NH2
NHC(dNH)NH2
NHC(dNH)NH2
°C; (b) HgCl2, HgO, MeOH, rt; (c) RNH2, EDC, HOBt, i-Pr2NEt,
THF, rt; (d) p-TsOH, THF, rt or TFA, CH2Cl2, rt.
MOL126
MOL150
take advantage of the well-characterized covalent in-
teraction of such groups with Ser195 of thrombin. The
R-ketoester functionality had been shown to function
well in this capacity in other active site inhibitors of
elastase20 and thrombin.21
BocArg(Mtr)-OH and BocLys(Cbz)-OH were homolo-
gated to the hydroxy acids 14a ,22b23 via the aminoal-
dehyde derivatives 13 (Scheme 3). The hydroxy acids
were coupled with several amines to produce the R-hy-
droxy amides 16 and 17a -d . The Boc groups were
selectively removed under acidic conditions to give the
hydroxyamines used for coupling with the â-strand
template.
Recently, R-ketoheterocycles have been employed as
activated carbonyl compounds that function as covalent
active site inhibitors of the serine proteases elastase24
and prolyl endopeptidase.25 After we had initiated our
investigations of inhibitors of this type, a study report-
ing highly potent R-ketoheterocycle inhibitors of throm-
bin based on the D-Phe-Pro-ArgX motif was published.26
Addition of 2-lithiothiazole and 2-lithiobenzothiazole to
aldehyde 13b was inefficient when stoichiometric
amounts of the lithium reagents were used in THF but
proceeded smoothly with a large excess of the reagent
in diethyl ether. Selective removal of both Boc protecting
groups using TFA in methylene chloride gave the amino
alcohols 18 suitable for coupling with amino acid
derivatives.
MOL127
MOL144
NHC(dNH)NH2
NHC(dNH)NH2
NHC(dNH)NH2
NHC(dNH)NH2
2-benzothiazolyl
MOL174 CH 2-benzothiazolyl
MOL168a CH 2-benzothiazolyl
a
Synthesized from the diastereomer of the template shown in
Scheme 5, i.e., template 3b.
HOBt gave the corresponding amides as mixtures of
four diastereomers (Scheme 5). Oxidation of the second-
ary alcohols with Dess-Martin periodinane gave the
ketones which were deprotected with TFA to give the
thrombin inhibitors as mixtures of two diastereomers.
The diastereomers were separated by HPLC to give the
inhibitors 19a -h listed in Table 1. In an analogous
manner, the diastereomer of the template 3b was
coupled, oxidized, and deprotected to produce the inhibi-
tor 19i. This inhibitor was produced in order to ascertain
the importance of the D-configuration for potent throm-
bin inhibition as precedented in the peptide analogues.
Biology. The results of in vitro assays of the inhibi-
tors against thrombin are summarized in Table 2. It is
clear that the templates mimic the peptide conformation
in the D-Phe-Pro-Arg thrombin inhibitors well as 12
showed thrombin inhibitory activity nearly identical to
that of PPACK. The R-ketoamide inhibitors (19a ,c,d )
are slow tight-binding inhibitors of thrombin, whereas
the R-ketoheterocycle inhibitors have a much higher on
rate. This is readily apparent during enzyme assays of
these compounds. Although all compounds were equili-
brated with the enzyme for 30 min prior to assay, the
R-ketoheterocycle inhibitors gave similar results when
Coupling of the amino alcohols 16, 17a -d , and 18a ,b
with the â-strand templates 3a and 4 using EDC and

1370 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 8
Boatman et al.
F igu r e 1. Modeled 18g bound to human R-thrombin active site based on the X-ray crystal structures of 18f,i.
Ta ble 2. Inhibition of Thrombin and Trypsin and Thrombin
potency; however, 19a suffers a 6400-2800-fold loss in
potency as compared to the other ketoamides in this
study. The additional loss in inhibitory potency may be
attributed to poorer interactions of the shorter nonaro-
matic cyclohexane group with the P′ residues of throm-
bin. X-ray crystallographic studies27 indicate that bulky
S′ groups can be accommodated by thrombin; however,
it appears that such groups require flexibility in order
to attain a conformation that will increase interactions
with P′ residues. The ketoamide inhibitors that incor-
porate benzyl and phenethyl groups in the S′ site have
much better inhibitory potency than 19a . The length of
the spacer seems to have little effect on the P′ interac-
tions as 19b (Y ) CONHbenzyl) and 19c (Y ) CON-
Hphenethyl) have similar thrombin Ki values. Substi-
tution on the phenyl ring is detrimental to thrombin
inhibition as both 19d ,e suffered losses in potency as
compared to 19c.
On Rate
corporate
code
thrombin
Ki (nM)
thrombin
Kon (1/M‚s)
trypsin
Ki (nM)
compd
PPACK
12
1.5a
ND
ND
ND
ND
0.073
0.027
ND
ND
ND
MOL098
MOL106
1.2a
2.4 × 106
18a
18b
18c
18d
18e
18f
18g
18h
18i
2000
0.073
0.071
0.31
0.45
2.4
0.65
0.85
10
ND
8.8 × 105
6.4 × 105
3.8 × 105
2.5 × 105
>107
MOL126
MOL150
MOL127
MOL144
MOL174
MOL168
>107
0.64
0.23
ND
>107
>107
a
IC50 (nM).
no preequilibration was used suggesting a much higher
on rate. The actual on rates were measured,27 and as
can be seen in Table 2, the ketoheterocycles are much
higher. The reasons for these kinetic differences are not
yet known; however, the R-ketoamides have a high
propensity for hydrate, hemiacetal, or hemiaminal (with
the guanidine side chain) formation in protic media. The
solution kinetics of the R-ketoamides may have some
effect on the kinetics of interactions with proteolytic
enzymes.28 Although the inhibitory potency of the
R-ketoamides is high, the slow on rate was expected to
be a significant factor in in vivo efficacy, and indeed the
in vivo results were poorer (see discussion below). 19i
suffered a greater than 10-fold loss in potency as
compared to 19h verifying that the configuration cor-
responding to D-phenylalanine at the P3 site is optimal
for binding of the template to thrombin. X-ray crystal-
lographic data shows that the hydrophobic phenyl ring
of 19i is in a solvated region of thrombin which probably
contributes significantly to the decrease in binding and
lower Ki.29
Although a large number of inhibitors of thrombin
have been synthesized and their thrombin complexes
studied by X-ray crystallography, little is known about
interactions of inhibitors with the P′ residues of throm-
bin. From the ketoamide inhibitors investigated in this
study, a few inferences can be made regarding P′
selectivity. 19a is a lysine version of the inhibitors with
a bulky cyclohexyl moiety near the scissile bond. From
previous studies30 it can be surmised that a change from
arginine to lysine results in a ca. 100-fold loss in
X-ray crystallography has shown that the R-ketohet-
erocycles bind with the heterocycle in a position similar
to that of the cyclohexyl in 19a . However, the interaction
of the aromatic S′ residue is much stronger than that
of the cyclohexyl moiety. We have used the X-ray crystal
structures of thrombin-bound 19f,i to model the interac-
tions of 19g bound to human R-thrombin (Figure 1). The
major P′ contributions to this increase in thrombin
binding appear to be a favorable stacking interaction
between the aromatic heterocycle and Trp60D and a
hydrogen bond to His57 of thrombin. 19g binds in the
active site and mimics the extended strand conformation
of PPACK bound to thrombin very well. The electro-
philic carbonyl is bound to Ser195 in the active site, and
the guanidyl side chain extends through the specificity
pocket and interacts with Asp189. The bicyclic core of
the mimetic is embedded in the apolar S2 region of the
active site, and the five-membered ring closely mimics
the proline of PPACK. The phenyl ring of 19g (D-Phe-
like portion) is positioned within the aryl D-53 binding
site and interacts with Leu99, Ile174, and Trp215.
12 and 19g,h were tested more extensively in in vitro
assays for inhibition of coagulation and anticoagulation
enzymes (Table 3). Comparison of the data for PPACK
with those of the constrained â-strand indicates that the
template imparts greater selectivity for thrombin than
the more flexible peptide. The selectivity of the R-keto-

Secondary Structure Peptide Mimetics
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 8 1371
Ta ble 3. Inhibition and Selectivity of Coagulation vs
Anticoagulation Enzymes
Exp er im en ta l Section
Precoated analytical TLC plates were purchased from EM
Science (silica gel 60 F254) and were visualized with UV light
or by treatment with phosphomolybdic acid (PMA). Flash
chromatography was carried out on EM Science silica gel 60
with the solvent indicated. Preparative HPLC separations
were accomplished on a Waters Symmetry C18 column oper-
ated at room temperature and eluted at a 1.5 mL/min flow
rate, using a linear gradient of 100% H2O containing 0.1%
TFA-90%H2O/MeCN containing 0.1% TFA over 30 min, with
diode array detection. 1H NMR data were obtained using a
Varian Unity 500 spectrometer. Mass spectral data were
obtained on a Fisons VG Quattro mass spectrometer operated
in the mode indicated. Solvents and reagents were purchased
from Aldrich Chemical Co. and used as received. THF was
distilled from sodium benzophenone prior to use.
12
19g
19h
enzymea
PPACKb (MOL098)b (MOL144)c
(MOL174)c
thrombin
factor Xa
1.5 (1)
1.2 (1)
165 (110) 385 (320)
factor VIIa 200 (133) 140 (117)
281 (187) 528 (440) 3320 (5110) 1250 (15000)
699 (466) 978 (815)
urokinase 508 (339) 927 (772)
0.65 (1)
270 (415)
270 (415)
0.085 (1)
19.3 (230)
200 (2300)
protein C
plasmin
415 (640)
600 (920)
495 (760)
251 (3000)
335 (4000)
93 (1100)
t-PA
106 (70)
632 (527)
a
Inhibitor concentration (selectivity ) Ki other/Ki thrombin).
IC50 (nM). c Ki (nM).
b
heterocycle inhibitors is even greater showing particu-
larly good selectivity for thrombin over the anticoagu-
lation enzymes (protein C, plasmin, urokinase, and
t-PA). This trend is especially evident in 19h where the
selectivity is greater than 1000-fold for thrombin versus
all of the anticoagulation enzymes tested.
Syn th esis of r-Allylp h en yla la n in e Meth yl Ester Hy-
d r och lor id e (5). To a mixture of L-phenylalanine methyl ester
hydrochloride (7.19 g, 33.3 mmol) and benzaldehyde (3.4 mL,
33.5 mmol) stirred in CH2Cl2 (150 mL) at room temperature
was added triethylamine (7.0 mL, 50 mmol). Anhydrous
magnesium sulfate (2 g) was added, and the mixture was
stirred for 14 h and then filtered through a 1-in. pad of Celite
with CH2Cl2. The filtrate was concentrated under reduced
pressure to ca. one-half of its initial volume and then diluted
with an equal volume of hexanes. The mixture was extracted
twice with saturated aqueous NaHCO3, H2O, and brine and
then dried over anhydrous Na2SO4 and filtered. Concentration
19g,d ,h were evaluated in an in vivo arterio-venous
shunt thrombosis model in baboons.31 This model allows
for evaluation in real time of 111In platelet deposition
on a dacron thrombogenic surface. Subsequently, fi-
brinogen deposition can also be evaluated. Administra-
tion of the agent was performed locally, immediately
upstream of the shunt via continuous iv infusion. Both
19g,h at 1 µg/min were very effective in blocking
platelet deposition, whereas 19d despite having a more
potent equilibrium Ki than 19g was not as effective
presumably due to the lower Kon.
1
of the filtrate under vacuum yielded 8.32 g of colorless oil. H
NMR analysis indicated nearly pure (>95%) phenylalanine
benzaldimine. The crude product was used without further
purification.
To a solution of diisopropylamine (4.3 mL, 33 mmol) stirred
in THF (150 mL) at -78 °C was added dropwise a solution of
n-butyllithium (13 mL of a 2.5 M hexane solution, 33 mmol).
The resulting solution was stirred for 20 min; then a solution
of phenylalanine benzaldimine (7.97 g, 29.8 mmol) in THF (30
mL) was slowly added. The resulting dark red-orange solution
was stirred for 15 min; then allyl bromide (3.1 mL, 36 mmol)
was added. The pale yellow solution was stirred for 30 min at
-78 °C and then allowed to warm to room temperature and
stirred an additional 1 h. Saturated aqueous ammonium
chloride was added, and the mixture was poured into ethyl
acetate. The organic phase was separated, washed with water
and brine, and then dried over anhydrous sodium sulfate and
filtered. Concentration of the filtrate under vacuum yielded
8.54 g of racemic R-allylphenylalanine benzaldimine as a
viscous yellow oil.
To a solution of R-allylphenylalanine benzaldimine (5.94 g,
19.3 mmol) stirred in methanol (50 mL) was added 5% aqueous
hydrochloric acid (10 mL). The solution was stirred at room
temperature for 2 h and then concentrated under vacuum to
an orange-brown caramel. The product was typically used
without purification but could be recrystallized from CHCl3/
hexanes to give white crystals (3.56 g, 62% from phenylala-
nine): 1H NMR (500 MHz, CDCl3) δ 8.86 (3 H, br s, NH3+),
7.32-7.26 (5H, m, ArH), 6.06 (1 H, dddd, J ) 17.5, 10.5, 7.6,
7.3 Hz, CHdCH2), 5.33 (1 H, d, J ) 17.5 Hz, CHdCH2), 5.30
(1 H, d, J ) 10.5 Hz, CHdCH2), 3.70 (3 H, s, COOCH3), 3.41
(1 H, d, J ) 14.1 Hz, ArCH2), 3.35 (1 H, d, J ) 14.1 Hz, ArCH2),
2.98 (1 H, dd, J ) 14.5, 7.3 Hz, CH2CHdCH2), 2.88 (1 H, dd,
J ) 14.5, 7.6 Hz, CH2CHdCH2).
Syn t h esis of N-(ter t-Bu t oxyca r b on yl)-r-a llylp h en yl-
a la n in e Meth yl Ester (6). To a solution of the crude D,L-R-
allylphenylalanine hydrochloride, 5 (565 mg, 2.21 mmol),
stirred in a mixture of THF (15 mL) and water (5 mL) was
added di-tert-butyl dicarbonate followed by solid sodium
bicarbonate in small portions. The resulting two-phase mixture
was vigorously stirred at room temperature for 3 days and then
diluted with ethyl acetate. The organic phase was separated,
washed with water and brine, and then dried over anhydrous
sodium sulfate and filtered. Concentration of the filtrate under
vacuum yielded a colorless oil that was purified by column
chromatography (5-10% EtOAc in hexanes gradient elution)
On the basis of these considerations, 19g,h were
evaluated for their bioavailability both in vitro and in
vivo. The potential for oral absorption was estimated
using IAM chromatography.32 The log(k′IAM/MW) values
calculated for 19g,h were 0.031 and -0.709, respec-
tively. It is difficult to extrapolate these values to
prediction of oral absorption since a structurally similar
set of compounds has not been evaluated using this
system.33 However, comparison of 19g,h indicates that,
of these two, 19h has a slightly higher potential for oral
bioavailability. Evaluation of transepithelial perme-
ability of 19g,h in the Caco-2 cell assay34 gave calculated
Papp values of 0.30 × 10-6 cm/s for 19g and 0.25 × 10-6
cm/s for 19h .35 On the basis of these data, the absorption
potential of both compounds would be considered low.
The bioavailability of 19g,h was evaluated in rats and
nonhuman primates. The compounds were administered
as solutions to conscious, catheterized animals via
intravenous and oral routes. Serial plasma samples
were analyzed for thrombin inhibitory activity and
pharmacokinetic parameters calculated by noncompart-
mental methods. The bioavailability of 19g approxi-
mated 25% in both rat and primate, while 19h bioavail-
ability approximated 2% in both species. Although
19g,h display similar characteristics in the in vitro
assays, 19g was consistently at least 10-fold more
bioavailable in vivo.
In summary, this work has delineated the use of
constrained templates that mimic the â-strand second-
ary structure of proteins. These templates have been
used for the production of a series of potent inhibitors
of thrombin. Further utilization of this approach for the
inhibition of proteases as well as other enzymes which
recognize â-strand motifs will be reported in due course.

1372 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 8
Boatman et al.
to yield 596 mg (84%) of N-(tert-butyloxycarbonyl)-R-allyl-
phenylalanine: TLC Rf ) 0.70 (silica, 20% EtOAc in hexanes);
1H NMR (500 MHz, CDCl3) δ 7.26-7.21 (3 H, m, ArH), 7.05
(2 H, d, J ) 6.1 Hz, ArH), 5.64 (1 H, dddd, J ) 14.8, 7.6, 7.2,
7.2 Hz, CHdCH2), 5.33 (1 H, br s, BocNH), 5.12-5.08 (2 H,
m, CHdCH2), 3.75 (3 H, s, COOCH3), 3.61 (1 H, d, J ) 13.5
Hz, ArCH2), 3.21 (1 H, dd, J ) 13.7, 7.2 Hz, CH2CHdCH2),
3.11 (1 H, d, J ) 13.5 Hz, ArCH2), 2.59 (1 H, dd, J ) 13.7, 7.6
Hz, CH2CHdCH2), 1.47 (9 H, s, COC(CH3)3).
Rf isomer of the bicyclic ester and 0.552 g (25% overall) from
a second chromatography of impure fractions: 1H NMR (500
MHz, CDCl3) δ 7.27-7.21 (3 H, m, ArH), 7.09 (2 H, d, J ) 6.5
Hz, ArH), 5.59 (1 H, br s, BocNH), 4.52 (1 H, dd, J ) 9.1, 3.4
Hz, CHCO2Et), 4.21 (2 H, m, OCH2CH3), 3.40 (1 H, d, J )
12.5 Hz, ArCH2), 3.32 (1 H, d, J ) 12.5 Hz, ArCH2), 3.10 (2 H,
m, NCH2), 2.79 (1 H, br m, NCH2), 2.66 (1 H, br m, NCH2),
2.66 (1 H, br m, CH2CH2), 2.54 (1 H, br m, CH2CH2), 2.46 (1
H, m, CH2CH2), 2.18 (1 H, m, CH2CH2), 1.44 (9 H, s, OC(CH3)3),
1.28 (3 H, t, J ) 7.0 Hz, OCH2CH3); MS (CI+, NH3) 418.4 (M
+ H+).
Syn th esis of 7-Ben zyl-7-[(ter t-bu toxyca r bon yl)a m in o]-
8-oxoh exa h yd r o-1H-p yr a zolo[1,2-a ]p yr id a zin e-1-ca r box-
yla te (4). To a solution of the ethyl ester 10 (1.383 g, 3.312
mmol) stirred in THF (30 mL) was added aqueous lithium
hydroxide (1 M, 6.6 mL). The resulting mixture was stirred
at room temperature for 1.5 h; then the reaction was quenched
with 5% aqueous citric acid. The mixture was extracted twice
with ethyl acetate; then the combined extracts were washed
with water and brine. The organic layer was dried over
anhydrous sodium sulfate, filtered, and concentrated under
vacuum to 1.30 g (100%) of white foam: 1H NMR (500 MHz,
CDCl3) δ 7.34-7.29 (3 H, m, ArH), 7.16-7.13 (2 H, m, ArH),
5.44 (1 H, s, BocNH), 4.72 (1 H, dd, J ) 9.1,5.7 Hz, CHCO2H),
3.36 (1 H, d, J ) 12.9 Hz, ArCH2), 3.09 (1 H, d, J ) 12.9 Hz,
ArCH2), 3.08, (2 H, m, NCH2), 2.65 (1 H, dd, J ) 15.4, 7.4 Hz,
NCH2), 2.55-2.37 (5 H, m), 1.44 (9 H, s, OC(CH3)3).
Syn th esis of 2-Ben zyl-2-[(ter t-bu toxyca r bon yl)a m in o]-
4-oxobu ta n oa te (7). Ozone was bubbled through a solution
of 2.10 g (6.57 mmol) of the olefin 6 stirred with solid NaHCO3
at -78 °C in a mixture of CH2Cl2 (50 mL) and methanol (15
mL) until the solution was distinctly blue in color. The mixture
was stirred an additional 15 min; then dimethyl sulfide was
slowly added. The resulting colorless solution was stirred at
-78 °C for 10 min and then allowed to warm to room
temperature and stirred for 6 h. The mixture was filtered, and
the filtrate was concentrated under vacuum to 2.72 g of viscous
pale yellow oil which was purified by chromatography (10-
20% EtOAc in hexanes gradient) to yield 1.63 g (77%) of pure
aldehyde 7 as a viscous colorless oil: TLC Rf ) 0.3 (silica, 20%
1
EtOAc in hexanes); H NMR (500 MHz, CDCl3) δ 9.69 (1 H,
br s, CHO), 7.30-7.25 (3 H, m, ArH), 7.02 (2 H, m, ArH), 5.56
(1 H, br s, NH), 3.87 (1 H, d, J ) 17.7 Hz, ArCH), 3.75 (3 H,
s, OCH3), 3.63 (1 H, d, J ) 13.2 Hz, CHCHO), 3.08 (1 H, d, J
) 17.7 Hz, ArCH), 2.98 (1 H, d, J ) 13.2 Hz, CHCHO), 1.46 (9
H, s, OC(CH3)3).
Syn th esis of (1S,7S)-7-Am in o-N-[(1S)-4-{[a m in o(im i-
n o)m et h yl]a m in o}-1-(2-ch lor oa cet yl)b u t yl]-7-b en zyl-8-
oxoh exa h yd r o-1H -p yr a zolo[1,2-a ]p yr id a zin e-1-ca r b ox-
a m id e (12). To a solution of the acid 4 (24 mg, 0.062 mmol)
and N-methylmorpholine (0.008 mL) stirred in THF (1 mL)
at -50 °C was added isobutyl chloroformate. The resulting
cloudy mixture was stirred for 10 min; then 0.016 mL (0.14
mmol) of N-methylmorpholine was added followed by a solu-
tion of HArg(Mtr)CH2Cl (11; 50 mg, 0.068 mmol) in THF (0.5
mL). The mixture was kept at -50 °C for 20 min and then
was allowed to warm to room temperature during 1 h. The
mixture was diluted with ethyl acetate and extracted with 5%
aqueous citric acid, saturated aqueous sodium bicarbonate, and
brine. The organic layer was dried over anhydrous sodium
sulfate, filtered, and concentrated under vacuum to 49 mg of
colorless glass. Separation by column chromatography yielded
12 mg of less polar diastereomer and 16 mg of the more polar
diastereomer: 1H NMR (more polar diastereomer, 500 MHz,
CDCl3) δ 7.93 (1 H, br s, NH), 7.39-7.31 (3 H, m, ArH), 7.16
(2 H, d, J ) 6.9 Hz. ArH), 6.52 (1 H, s, Mtr-ArH), 6.30 (1 H, br
s, NH), 5.27 (1 H, s), 4.74 (1 H, dd, J ) 9.1, 6.9 Hz, CHCONH),
4.42 (1 H, br d, J ) 6.8 Hz, CH2Cl), 4.33 (1 H, d, J ) 6.8 Hz,
CH2Cl), 3.82 (3 H, s, OCH3), 3.28 (1 H, d, J ) 13.3 Hz, ArCH2),
3.26-3.12 (4 H, m, NCH2, NHCH2), 2.98 (1 H, d, J ) 13.3 Hz,
ArCH2), 2.69 (3 H, s, ArCH3), 2.60 (3 H, s, ArCH3), 2.59-2.33
(4 H, m, NCH2CH2C, NCH2CH2CH), 2.25-2.10 (3 H, m,), 2.11
(3 H, s, ArCH3), 1.77 (1 H, br m), 1.70-1.55 (3 H, br m), 1.32
(9 H, s, OC(CH3)3).
The more polar diastereomer (16 mg, 0.021 mmol) was
dissolved in 95% TFA/H2O (1 mL), and the resulting solution
was stirred at room temperature for 48 h and then concen-
trated under vacuum. The residue was triturated with ether,
washed twice with ether and then dried under high vacuum
for 14 h. Purification of the product by HPLC yielded 5 mg of
pure compound 12: MS (EI+) m/z 477.9 (M+).
P r ep a r a tion of (3S)-3-[(ter t-Bu toxyca r bon yl)a m in o]-
N-ben yl-2-h yd r oxy-6-[[[(ter t-bu toxyca r bon yl)im in o]{[(4-
m eth oxy-2,3,6-tr im eth ylp h en yl)su lfon yl]a m in o}m eth yl]-
a m in o]h exa n a m id e (15a ). To a solution of the acid 14b (65
mg, 0.105 mmol), HOBT (17 mg, 0.126 mmol), and EDC (24
mg, 0.126 mmol) in THF (5 mL) was added benzylamine (14
µL, 0.126 mmol) followed by diisopropylethylamine (28 µL,
0.158 mmol). The reaction mixture was stirred at room
temperature overnight and diluted with 5% citric acid. The
organic layer was separated and the aqueous phase extracted
with EtOAc (3×). The combined extracts were washed with
saturated solution of NaHCO3 and brine, dried over Na2SO4,
Syn th esis of 4-[(ter t-Bu toxyca r bon yl)a m in o]-4-ben z-
yld ih yd r oh yd r op yr id a zin -3(2H)-on e (8). To a solution of
the aldehyde 7 (1.62 g, 5.03 mmol) stirred in THF (50 mL) at
room temperature was added hydrazine hydrate (0.32 mL, 6.5
mmol). The resulting solution was stirred at room temperature
for 10 min and then heated to reflux for 3 days. The solution
was allowed to cool to room temperature and then concentrated
under vacuum to 1.59 g of colorless foam. The foam was
dissolved in ethyl acetate and filtered through a 2-in. plug of
silica gel with ethyl acetate. Concentration of the filtrate under
vacuum gave 1.16 g (76%) of the cyclic hydrazone 8 as a white
1
foam: TLC Rf ) 0.7 (50% EtOAc in hexanes); H NMR (500
MHz, CDCl3) δ 8.55 (1 H, br s, NdCH), 7.32-7.26 (3 H, m,
ArH), 7.17 (1 H, br s, NNH), 7.09 (2H, m, ArH), 5.55 (1 H, br
s, BocNH), 3.45 (1 H, d, J ) 17.7 Hz, ArCH), 3.29 (1 H, d, J )
13.5 Hz, CH2CH), 2.90 (1 H, d, J ) 13.5 Hz, CH2CH), 2.88 (1
H, dd, J ) 17.7, 1.3 Hz, ArCH), 1.46 (9 H, s, OC(CH3)3); MS
(CI+, NH3) m/z 304.1 (M + H+).
Syn th esis of 4-[(ter t-Bu toxyca r bon yl)a m in o]-4-ben z-
yltetr a h yd r op yr id a zin -3(2H)-on e (9). The cyclic hydrazone
8 (21.28 g, 70.15 mmol) was dissolved in 200 mL of methanol,
and PtO2 (1.59 g, 7.01 mmol) was added. The flask was flushed
twice with H2 gas; then a balloon filled with H2 was attached,
and the mixture was stirred vigorously overnight. An ad-
ditional 0.80 g (3.52 mmol) of PtO2 was added, and stirring
under an H2 atmosphere was continued for another day. The
mixture was filtered through Celite with the aid of ethyl
acetate; then the filtrate was concentrated under vacuum to
a white foam. The foam was purified by flash chromatography
(50% ethyl acetate/hexanes eluent) to give 14.37 g (67%) of
the hydrazide 9 as a white foam: 1H NMR (500 MHz, CDCl3)
δ 7.41-7.35 (4 H, m, ArH, NHNH), 7.18 (2 H, m, ArH), 7.08
(1 H, br s, NHNH), 5.32 (1 H, br s, BocNH), 3.47 (1 H, d, J )
12.8 Hz, ArCH2), 3.38 (1 H, dd, J ) 12.8, 5.5 Hz, NCH2), 2.72
(1 H, d, J ) 12.8 Hz, ArCH2), 2.45 (1 H, ddd, J ) 15.1, 15.0,
5.5 Hz, CH2CH2), 2.38-2.30 (2 H, m, CH2CH2), 1.44 (9 H, s,
OC(CH3)3); MS (CI+, NH3) m/z 306.2 (M + H+).
Syn th esis of Eth yl 7-Ben zyl-7-[(ter t-bu toxyca r bon yl)-
a m in o]-8-oxoh exa h yd r o-1H-p yr a zolo[1,2-a ]p yr id a zin e-1-
ca r boxyla te (10). To a solution of the cyclic hydrazide 9 (4.07
g, 13.32 mmol) stirred in ethyl acrylate (200 mL) at 90 °C was
added formaldehyde (1.2 mL of a 37% aqueous solution). The
mixture was heated to reflux for 4 h and then allowed to cool
to room temperature and concentrated under vacuum to a
white foam. The products were separated by column chroma-
tography (5-10% acetone/CHCl3) to yield 0.851 g of the highest

Secondary Structure Peptide Mimetics
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 8 1373
and filtered. After concentration the crude product was puri-
fied by chromatography (EtOAc/Hx, 1:1) to give 33 mg (44%)
of the amide: 1H NMR (500 MHz, CDCl3) δ 9.83 (s, 1 H,
NHSO2), 8.34 (t, J ) 5 Hz, 1 H, NHCH2), 7.33 (m, 1 H, CONH),
7.29 (m, 5 H, aromatic), 6.53 (s, 1 H, m-benzene sulfonyl CH),
5.26 (d, J ) 9.5 Hz, 1 H, NHBoc), 4.85 (d, J ) 5 Hz, 1 H, OH),
4.45 (m, 2 H, ArCH2N), 4.19 (dd, J ) 5.5, 3.0 Hz, 1 H, CHOH),
3.99 (m, 1 H, CHNHBoc), 3.82 (s, 3 H, OCH3), 3.63 (m, 1 H,
one guanidine CH2CH2), 3.18 (m, 1 H one guanidine CH2CH2),
2.68 and 2.58 (each s, 3 H, o-benzene sulfonyl CH3), 2.12 (s, 3
H, m-benzene sulfonyl CH3), 1.73-1.62 (m, 4 H, guanidine
CH2CH2CH2), 1.49 (s, 9 H, guanidine Boc), 1.38 (s, 9 H,
NHBoc); MS (ES+) m/z 706.5 (M + H+).
yellowish-white solid: 1H NMR (500 MHz, CDCl3) consistent
with the expected product; however, individual peak assign-
ment was difficult due to signal broadening; MS (FAB+) m/z
506.4 (M + H+).
P r ep a r a tion of (3S)-3-Am in o-N-p h en eth yl-2-h yd r oxy-
6-[(im in o{[(4-m et h oxy-2,3,6-t r im et h ylp h en yl)su lfon yl]-
a m in o}m eth yl)a m in o]h exa n a m id e (17b). The amine 17b
was prepared in a fashion similar to 17a from 15b: MS (ES+)
m/z 520.4 (M + H+).
P r ep a r a tion of (3S)-3-Am in o-N-(4-ch lor op h en eth yl)-2-
h yd r oxy-6-[(im in o{[(4-m et h oxy-2,3,6-t r im et h ylp h en yl)-
su lfon yl]a m in o}m et h yl)a m in o]h exa n a m id e (17c). The
amine 17c was prepared in a fashion similar to 17a from
15c: MS (ES+) m/z 554.5 (M + H+).
P r ep a r a tion of (3S)-3-Am in o-N-(4-m eth oxyp h en eth yl)-
2-h yd r oxy-6-[(im in o{[(4-m eth oxy-2,3,6-tr im eth ylp h en yl)-
su lfon yl]a m in o}m eth yl)a m in o]h exa n a m id e (17d ). The
amine 17d . was prepared in a fashion similar to 17a from
15d : MS (ES+) m/z 550.5 (M + H+).
P r ep a r a tion of (3S)-3-[(ter t-Bu toxyca r bon yl)a m in o]-
N-p h en eth yl-2-h yd r oxy-6-[[[(ter t-bu toxyca r bon yl)im in o]-
{[(4-m e t h oxy-2,3,6-t r im e t h ylp h e n yl)su lfon yl]a m in o}-
m eth yl]a m in o]h exa n a m id e (15b). Amide 15b was prepared
in a fashion similar to 15a from the acid 14b and phenethyl-
1
amine: 26 mg (77%); H NMR (500 MHz, CDCl3) δ 9.84 (s, 1
H, NHSO2), 8.34 (t, J ) 5 Hz, 1 H, NHCH2), 7.28 (m, 3 H,
aromatic), 7.21 (m, 2 H, aromatic), 7.04 (m, 1 H, CONH), 6.55
(s, 1 H, m-benzene sulfonyl CH), 5.16 (d, J ) 8.5 Hz, 1 H,
NHBoc), 4.56 (d, J ) 5 Hz, 1 H, OH), 4.11 (dd, J ) 5.0, 3.0 Hz,
1 H, CHOH), 3.98 (m, 1 H, CHNHBoc), 3.84 (s, 3 H, OCH3),
3.66 (m, 1 H, one guanidine CH2CH2), 3.51 (m, 2 H, ArCH2-
CH2N), 3.17 (m, 1 H, one guanidine CH2CH2), 2.81 (t, J ) 7.5
Hz, 2 H, ArCH2CH2N), 2.71 and 2.65 (each s, 3 H, o-benzene
sulfonyl CH3), 2.14 (s, 3 H, m-benzene sulfonyl CH3), 1.68-
1.52 (m, 4 H, guanidine CH2CH2CH2), 1.49 (s, 9 H, guanidine
Boc), 1.39 (s, 9 H, NHBoc); MS (FAB+) m/z 720.6 (M + H+)
(FAB-) m/z 718.5 (M-H+).
P r ep a r a tion of (3S)-3-[(ter t-Bu toxyca r bon yl)a m in o]-
N-(4-ch lor op h en et h yl)-2-h yd r oxy-6-[[[(ter t-b u t oxyca r -
bon yl)im in o]{[(4-m eth oxy-2,3,6-tr im eth ylp h en yl)su lfon -
yl]a m in o}m eth yl]a m in o]h exa n a m id e (15c). Amide 15c
was prepared in a similar fashion to 15a from the acid 14b
and 4-chlorophenethylamine: 55 mg (75%); 1H NMR (500
MHz, CDCl3) δ 9.82 (s, 1 H, NHSO2), 8.34 (t, J ) 5 Hz, 1 H,
NHCH2), 7.25 (m, 2 H, aromatic), 7.12 (m, 3 H, aromatic and
CONH), 6.55 (s, 1 H, m-benzene sulfonyl CH), 5.20 (d, J ) 9.0
Hz, 1 H, NHBoc), 4.66 (d, J ) 5 Hz, 1 H, OH), 4.11 (m, 1 H,
CHOH), 3.99 (m, 1 H, CHNHBoc), 3.83 (s, 3 H, OCH3), 3.64
(m, 1 H, one guanidine CH2CH2), 3.47 (m, 2 H, ArCH2CH2N),
3.17 (m, 1 H, one guanidine CH2CH2), 2.78 (t, J ) 7.5 Hz, 2
H, ArCH2CH2N), 2.70 and 2.64 (each s, 3 H, o-benzene sulfonyl
CH3), 2.14 (s, 3 H, m-benzene sulfonyl CH3), 1.68-1.52 (m, 4
H, guanidine CH2CH2CH2), 1.49 (s, 9 H, guanidine Boc), 1.39
(s, 9 H, NHBoc); MS (ES+) m/z 754.6 (M + H+).
P r ep a r a tion of (3S)-3-[(ter t-Bu toxyca r bon yl)a m in o]-
N-(4-m eth oxyp h en eth yl)-2-h yd r oxy-6-[[[(ter t-bu toxyca r -
bon yl)im in o]{[(4-m eth oxy-2,3,6-tr im eth ylph en yl)su lfon yl]-
a m in o}m eth yl]a m in o]h exa n a m id e (15d ). Amide 15d was
prepared in a fashion similar to 15a from the acid 14b and
4-methoxyphenethylamine: 57 mg (78%); 1H NMR (500 MHz,
CDCl3) δ 9.83 (s, 1 H, NHSO2), 8.34 (t, J ) 5 Hz, 1 H, NHCH2),
7.11 (m, 2 H, aromatic), 7.04 (m, 1 H, CONH), 6.83 (m, 2 H,
aromatic), 6.54 (s, 1 H, m-benzene sulfonyl CH), 5.22 (d, J )
9.0 Hz, 1 H, NHBoc), 4.66 (d, J ) 5 Hz, 1 H, OH), 4.11 (m, 1
H, CH2OH), 3.96 (m, 1 H, CHNHBoc), 3.83 (s, 3 H, benzene
sulfonyl OCH3), 3.78 (s, 3 H, aromatic OCH3), 3.62 (m, 1 H,
one guanidine CH2CH2), 3.46 (m, 2 H, ArCH2CH2N), 3.19 (m,
1 H, one guanidine CH2CH2), 2.74 (t, J ) 7.5 Hz, 2 H, ArCH2-
CH2N), 2.70 and 2.64 (each s, 3 H, o-benzene sulfonyl CH3),
2.14 (s, 3 H, m-benzene sulfonyl CH3), 1.70-1.52 (m, 4 H,
guanidine CH2CH2CH2), 1.49 (s, 9 H, guanidine Boc), 1.39 (s,
9 H, NHBoc); MS (ES+) m/z 750.7 (M + H+).
P r ep a r a tion of (3S)-3-Am in o-N-ben zyl-2-h yd r oxy-6-
[(im in o{[(4-m e t h oxy-2,3,6-t r im e t h ylp h e n yl)su lfon yl]-
a m in o}m eth yl)a m in o]h exa n a m id e (17a ). To a solution of
15a (25 mg, 0.035 mmol) in THF (5 mL) was added of
p-toluenesulfonic acid monohydrate (24 mg, 0.124 mmol). The
reaction mixture was stirred at room temperature overnight
to give a baseline spot by TLC. The solution was concentrated
in vacuo and the residue washed twice with ether to give a
Syn th esis of N-[4-Am in o-5-(1,3-ben zoth ia zol-2-yl)-5-
h yd r oxyp en t yl]-N′-[(4-m et h oxy-2,3,6-t r im et h ylp h en yl)-
su lfon yl]gu a n id in e (18b). To a solution of benzothiazole
(1.55 mL, 14 mmol) stirred in anhydrous diethyl ether (60 mL)
at -78 °C under a dry argon atmosphere was added a solution
of n-butyllithium (2.5 M in hexane, 5.6 mL, 14 mmol) dropwise
over a period of 10 min. The resulting orange solution was
stirred for 45 min; then a solution of the arginal 13b (1.609 g,
2.819 mmol) in diethyl ether (5 mL) was slowly added. The
solution was stirred for 1.5 h; then saturated aqueous am-
monium chloride solution was added, and the mixture was
allowed to warm to room temperature. The mixture was
extracted with ethyl acetate (3 × 100 mL), and the combined
extracts were extracted with water and brine and then dried
over anhydrous sodium sulfate and filtered. Concentration of
the filtrate under vacuum yielded a yellow oil that was purified
by flash chromatography (30% then 40% ethyl acetate/hexanes
eluent) to yield 1.22 g (61%) of the hydroxybenzothiazoles (ca.
1
2:1 mixture of diastereomers by H NMR analysis) as a white
foam.
The mixture of hydroxybenzothiazoles (1.003 g, 1.414 mmol)
was stirred in CH2Cl2 (12 mL) at room temperature, and
trifluoroacetic acid (3 mL) was added. The resulting solution
was stirred for 1.5 h and then concentrated under reduced
pressure to brown gum. The gum was triturated twice with
diethyl ether; then the residue was dried under high vacuum
to give 0.923 g (89%) of 17b as a pale yellow foam: MS (EI+)
m/z 506.2 (M + H+).
Syn t h esis of N-[4-Am in o-5-(1,3-t h ia zol-2-yl)-5-oxo-
p en tyl]-N′-[(4-m eth oxy-2,3,6-tr im eth ylp h en yl)su lfon yl]-
gu a n id in e (18a ). 18a was synthesized in a fashion analogous
to that of 18b from the aldehyde 13b and 1,3-thiazole in 67%
overall yield.
Syn t h esis of (1S,7S)-7-Am in o-N-[(1S)-4-{[a m in o(im i-
n o)m eth yl]a m in o}-1-(1,3-ben zoth ia zol-2-ylca r bon yl)bu t-
yl]-7-benzyl-8-oxohexahydro-1H-pyrazolo[1,2-a ]pyridazine-
1-ca r boxa m id e (19g). The bicyclic acid 4 (151 mg, 0.387
mmol) and HOBt hydrate (71 mg, 0.46 mmol) were dissolved
in THF (5 mL), and diisopropylethylamine (0.34 mL, 1.9 mmol)
was added followed by EDC (89 mg, 0.46 mmol). After stirring
for 10 min a solution of the amine 18b (273 mg, 0.372 mmol)
in THF (1 mL) was added along with a THF (0.5 mL) rinse.
The mixture was stirred at room temperature for 15 h and
then diluted with ethyl acetate and extracted sequentially with
5% aqueous citric acid, saturated aqueous sodium bicarbonate,
water, and brine. The organic solution was dried over anhy-
drous sodium sulfate, filtered, and concentrated under vacuum
to 297 mg of a yellow glass: 1H NMR a mixture of four
diastereomeric amides; MS (ES+) m/z 877 (M+).
The crude hydroxybenzothiazole (247 mg, 0.282 mmol) was
dissolved in CH2Cl2 (5 mL), and Dess-Martin periodinane (241
mg, 0.588 mmol) was added. The mixture was stirred at room
temperature for 6 h and then diluted with ethyl acetate and
stirred vigorously with 10% aqueous sodium thiosulfate for
10 min. The organic solution was separated, extracted with

1374 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 8
Boatman et al.
saturated aqueous sodium bicarbonate, water, and brine, and
then dried over anhydrous sodium sulfate and filtered. Con-
centration of the filtrate under vacuum yielded 252 mg of
yellow glass: 1H NMR a mixture of two diastereomeric
ketobenzothiazoles.
The ketobenzothiazole (41 mg, 0.047 mmol) was dissolved
in 95% aqueous trifluoroacetic (0.95 mL) acid, and thioanisole
(0.05 mL) was added. The resulting dark solution was stirred
for 30 h at room temperature and then concentrated under
vacuum to a dark brown gum. The gum was triturated with
diethyl ether and centrifuged. The solution was removed, and
the solid remaining was triturated and collected as above two
more times. The yellow solid was dried in a vacuum desiccator
for 2 h and then purified by HPLC to give 2.5 mg of 19g: MS
(ES+) m/z 563.5 (M + H+).
Syn t h esis of (1S,7S)-7-Am in o-N-[(1S)-4-{[a m in o(im i-
n o)m eth yl]am in o}-1-(cycloh exylam in o)-2-(oxoacetyl)bu t-
y l]-7-b e n z y l-8-o x o h e x a h y d r o -1H -p y r a z o lo [1,2-a ]-
p yr id a zin e-1-ca r boxa m id e (19a ). 19a was synthesized in
a similar fashion from the acid 4 and the amine 16.
Syn t h esis of (1S,7S)-7-Am in o-N-[(1S)-4-{[a m in o(im i-
n o)m eth yl]a m in o}-1-{2-(p h en eth yla m in o)-2-oxoa cetyl}-
b u t yl]-7-b e n zyl-8-oxoh e xa h yd r o-1H -p yr a zolo[1,2-a ]-
p yr id a zin e-1-ca oxa m id e (19b). 19b was synthesized in a
similar fashion from the acid 4 and the amine 17a : MS (ES+)
m/z 536.6 (M + H+).
Syn t h esis of (1S,7S)-7-Am in o-N-[(1S)-4-{[a m in o(im i-
n o)m eth yl]a m in o}-1-{2-(p h en eth yla m in o)-2-oxoa cetyl}-
b u t yl]-7-b e n zyl-8-oxoh e xa h yd r o-1H -p yr a zolo[1,2-a ]-
p yr id a zin e-1-ca r boxa m id e (19c). 19c was synthesized in
a similar fashion from the acid 4 and the amine 17b.
Syn th esis of (1S,7S)-7-Am in o-N-[(1S)-4-{[a m in o(im i-
n o)m et h yl]a m in o}-1-{2-[(4-ch lor op h en et h yl)a m in o]-2-
oxoa cetyl}bu tyl]-7-ben zyl-8-oxoh exa h yd r o-1H-p yr a zolo-
[1,2-a ]pyr idazin e-1-car boxam ide (19d). 19d was synthesized
in a similar fashion from the acid 4 and the amine 17c: MS
(ES+) m/z 611.3 (M + H+); HPLC tR ) 19.8 min.
were determined by unweighted nonlinear least-squares fitting
to a first-order reaction in either GraFit (Erithacus Software
Ltd., London, England) or GraphPad Prism (GraphPad Soft-
ware, Inc., San Diego, CA). The determined initial velocities
were then nonlinear least-squares fitted against the concen-
trations of a tested compound using GraFit or Prism to obtain
Ki. The general format of these assays were 100 µL of a
substrate solution and 100 µL of inhibitor solution were placed
in a microplate well; then 50 µL of enzyme solution was added
to initiate the reaction.
In thrombin assays, 0.05-0.2 nM human thrombin (Sigma
T-6759) and 40 µM N-p-tosyl-Gly-Pro-Arg-pNA or 20 µM N-p-
tosyl-Gly-Pro-Arg-AMC were used in pH 8.0 Tris buffer (Tris,
50 mM; Tween 20, 0.1%; BSA, 0.1%; NaCl, 0.15 M; CaCl2, 5
mM).
In trypsin assays, 4 nM bovine trypsin (Sigma T-1005) and
40 µM N-p-tosyl-Gly-Pro-Arg-pNA or 20 µM N-p-tosyl-Gly-Pro-
Arg-AMC were used in 50 mM Tris, pH 8.0, 0.15 M NaCl, and
5 mM CaCl2 buffer containing 0.1% Tween 20 (v/v).
Refer en ces
(1) Kahn, M. Peptide Secondary Structure Mimetics: Recent Ad-
vances and Future Challenges. Synlett 1993, 3, 821-826.
(2) Wu, T.-P.; Yee, V.; Tulinsky, A.; Chrusciel, R. A.; Nakanishi,
H.; Shen, R.; Priebe, Kahn, M. The Structure of a Designed
Peptidomimetic Inhibitor Complex of R-Thrombin. Protein Eng.
1993, 6, 471-478.
(3) Magnusson, S. L.; Sottrup-J ensen, T. E.; Petersen; Wojciechows-
ka, G. D.; Claeys, H. Proteolysis and Physiological Regulation.
Miami Winter Symposia; Academic Press: New York, 1976; Vol.
11, pp 203-239.
(4) Bode, W.; Huber, R. Natural Protein Proteinase Inhibitors and
Their Interaction with Proteinases. Eur. J . Biochem. 1992, 204,
2554-2566.
(5) Blomback, B.; Blomback, M.; Hessel, B.; Iwanaga, S. Structures
of N-Terminal Fragments of Fibrinogen and Specificity of
Thrombin. Nature 1967, 215, 1445-1448.
(6) Coughlin, S. R. Thrombin Receptor Function and Cardiovascular
Disease. Trends Cardiovasc. Med. 1994, 4, 77-83.
(7) Kettner, C.; Shaw, E. D-Phe-Pro-Arg-CH2Cl, a Selective Affinity
Label for Thrombin. Thromb. Res. 1979, 14, 969.
Syn th esis of (1S,7S)-7-Am in o-N-[(1S)-4-{[a m in o(im i-
n o)m eth yl]a m in o}-1-{2-[(4-m eth oxyp h en eth yl)a m in o]-2-
oxoa cetyl}bu tyl]-7-ben zyl-8-oxoh exa h yd r o-1H-p yr a zolo-
[1,2-a ]pyr idazin e-1-car boxam ide (19e). 19e was synthesized
in a similar fashion from the acid 4 and the amine 17d : MS
(ES+) m/z 607.4 (M + H+); HPLC tR ) 18.2 min.
Syn th esis of (1S,7S)-7-Am in o-N-[(1S)-4-{[a m in o(im i-
n o)m et h yl]a m in o}-1-(1,3-t h ia zol-2-ylca r b on yl)b u t yl]-7-
ben zyl-8-oxoh exa h yd r o-1H-p yr a zolo[1,2-a ]p yr id a zin e-1-
ca r boxa m id e (19f). 19f was synthesized in a similar fashion
from the acid 4 and the amine 18a : MS (ES+) m/z 513.5 (M
+ H+).
Syn t h esis of (1S,7R)-7-Am in o-N-[(1S)-4-{[a m in o(im i-
n o)m eth yl]a m in o}-1-(1,3-ben zoth ia zol-2-ylca r bon yl)bu t-
y l]-7-b e n z y l-8-o x o h e x a h y d r o -1H -p y r a z o lo [1,2-a ]-
p yr id a zin e-1-ca r boxa m id e (19h ). 19h was synthesized in
a similar fashion from the acid 3a and the amine 18b: MS
(ES+): m/z 562.4 (M + H+); HPLC tR ) 19.8 min.
Syn t h esis of (1S,7R)-7-Am in o-N-[(1S)-4-{[a m in o(im i-
n o)m eth yl]a m in o}-1-(1,3-ben zoth ia zol-2-ylca r bon yl)bu t-
y l]-7-b e n z y l-8-o x o h e x a h y d r o -1H -p y r a z o lo [1,2-a ]-
p yr id a zin e-1-ca r boxa m id e (19i). 19i was synthesized in a
similar fashion from the acid 3b and the amine 18b: MS (ES+)
m/z 562.4 (M + H+).
In h ibition Assa ys. Both thrombin and trypsin inhibition
assays were performed at room temperature in 96-well mi-
croplates using either a Bio-Rad microplate reader (model
3550) or a Molecular Devices SpectroMax 250 or Labsystems
Fluoroskan Ascent fluorescence plate reader. Solutions of 1
mM of testing compounds in water or 10 mM in DMSO served
as the stock solutions for both of the inhibition assays. The
hydrolysis of the chromogenic substrate N-p-tosyl-Gly-Pro-Arg-
pNA (Sigma) was monitored at 405 nm, and the fluorogenic
substrate N-p-tosyl-Gly-Pro-Arg-AMC (Sigma) was monitored
at 345-nm excitation wavelength and 450-nm emission wave-
length. The reaction progress curves were recorded by reading
the plates, typically 100 times at 21-s intervals. Initial rates
(8) Das, S.; Kimball, S. D. Thrombin Active Site Inhibitors. Bioorg.
Med. Chem. 1996, 3, 999.
(9) Kimball, S. D. Challenges in the Development of Orally Bio-
available Thrombin Active Site Inhibitors. Blood Coag. Fibrin.
1995, 6, 511-519.
(10) (a) Bode, W.; Turk, D.; Karshikov, A. The Refined 1.9 Å X-ray
Crystal Structure of D-Phe-Pro-Arg-Chloromethyl Ketone-
inhibited Human R-Thrombin: Structure Analysis, Overall
Structure, Electrostatic Properties, Detailed Active-Site Geom-
etry, and Structure Function Relationships. Protein Sci. 1992,
1, 426. (b) Bode, W.; Mayr, I.; Baumann, U.; Huber, R.; Stone,
S. R.; Hofsteenge, J . The refined 1.9 Å Crystal Structure of
Human alpha-Thrombin: Interaction with D-Phe-Pro-Arg Chlo-
romethyl ketone and Significance of the Tyr-Pro-Pro-Trp Inser-
tion Segment. EMBO J . 1989, 8, 3467.
(11) Hannessian, S.; McNaughton-Smith, G.; Lombart, H.-G.; Lubell,
W. D. Design and Synthesis of Conformationally Constrained
Amino Acids as Versatile Scaffolds and Peptide Mimetics.
Tetrahedron 1997, 53, 12789.
(12) Still, W. C.; Tempczyk, A.; Hawley, R. C.; Hendrickson, T.
Semianalytical Treatment of Solvation for Molecular Mechanics
and Dynamics. J . Am. Chem. Soc. 1990, 112, 6127.
(13) Mohamadi, F.; Richards, N. G. J .; Guida, W. C.; Liskamp, R.;
Lipton, M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still, W.
C. Macromodel - An Integrated Software System for Modeling
Organic and Bioorganic Molecules Using Molecular Mechanics.
J . Comput. Chem. 1990, 11, 440.
(14) Marraud, M.; Dupont, V.; Grand, V.; Zerkout, S.; Lecoq, A.;
Boussard, G.; Vidal, J .; Collet, A.; Aubry, A. Modifications of the
Amide Bond and Conformational Constraints in Pseudopeptide
Analogues. Biopolymers 1993, 33, 1135.
(15) Dupont, V.; Lecoq, A.; Mangeot, J .; Aubry, A.; Boussard, G.;
Marraud, M. Conformational Perturbations Induced by N-
Amination and N-Hydroxylation of Peptides. J . Am. Chem. Soc.
1993, 115, 8898.
(16) Kim, H.-O.; Kahn, M. Short Synthesis of (3S,6S,9S)-2-Oxo-3-
(N-Boc-amino)-1-azabicyclo[4.3.0]nonane-9-carboxylic Acid Meth-
yl Ester: Tandem Cyclization Protocol. Tetrahedron Lett. 1997,
37, 6483-6484.
(17) Eguchi, M.; Kim, H. O.; Gardner, B. S.; Boatman, P. D.; Lee, M.
S.; Nakanishi, H.; Kahn, M. Synthesis of Dipeptide Secondary
Structure Mimetics. Proc. 15th Am. Pept. Symposium; Kluwer
Academic Publishers: Dordrecht, The Netherlands, 1998; pp
212-213.

Secondary Structure Peptide Mimetics
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 8 1375
(18) J ungheim, L. N.; Boyd, D. B.; Indelicato, J . M.; Pasini, C. E.;
Preston, D. A.; Alborn, W. E. Synthesis, Hydrolysis Rates,
Supercomputer Modeling, and Antibacterial Activity of Bicyclic
Tetrahydropyridazinones. J . Med. Chem. 1991, 34, 1732.
(19) Calabretta, R., Gallina, C., Giordano, C. Synthesis 1991, 536.
(20) Peet, N. P.; Burkhart, J . P.; Angelastro, M. R.; Giroux, E. L.;
Mehdi, S.; Bey, P.; Kolb, M.; Neises, B.; Schirlin, D. Synthesis
of Peptidyl Fluoromethyl Ketones and Peptidyl R-Keto Esters
as Inhibitors of Porcine Pancreatic Elastase, Human Neutrophil
Elastase, and Rat and Human Neutrophil Cathepsin G. J . Med.
Chem. 1990, 33, 394-407.
Ohuchi, S.; Okonogi, T. Peptidyl R-Keto Thiazole as Potent
Thrombin Inhibitors. Bioorg. Med. Chem. Lett. 1997, 7, 533-
538. (c) Tamura, S. Y.; Shamblin, M. B.; Brunck, T. K.; Ripka,
W. C. Rational Design, Synthesis, and Serine Protease Inhibitory
Activity of Novel P1-Argininoyl Heterocycles. Bioorg. Med. Chem.
Lett. 1997, 7, 1359-1364.
(27) Hermans, J . M.; Monard, D.; J ones, R.; Stone, S. R. Inhibition
of Acrosin by Serpins. A Suicide Substrate Mechanism. Bio-
chemistry 1995, 34, 3678-3685.
(28) Lewis, S. D.; Lucas, B. J .; Brady, S. F.; Sisko, J . T.; Cutrona, K.
J .; Sanderson, P. E. J .; Freidinger; Mao, S.-S.; Gardell, S. J .;
Shafer, J . A. Characterization of the Two-step Pathway for
Inhibition of Thrombin by R-Ketoamide Trransition State Ana-
logues. J . Biol. Chem. 1998, 273, 4843-4854.
(29) St. Charles, R.; Matthews, J . H.; Zhang, E.; Tulinsky, A.; Kahn,
M. The Bound Structure of Novel P3-P1′ â-Strand Mimetic
Inhibitors of Thrombin. J . Med. Chem. 1999, 42, 1376-1383.
(30) Weber, P. C.; Lee, S. L.; Lewandowski, F. A.; Schadt, M. C.;
Chang, C. H.; Kettner, C. A. Kinetic and Crystallographic
Studies of Thrombin with Ac-(D)-Phe-Pro-BoroArg-OH and Its
Lysine, Amidine, Homolysine, and Ornithine Analogues. Bio-
chemistry 1994, 34, 3750-3757.
(31) Scott, N. A.; Nunes, G. L.; King, S. B., 3rd; Harker, L. A.; Hanson,
S. R. Local Delivery of an Antithrombin Inhibits Platelet-
Dependent Thrombosis. Circulation 1994, 90, 1951-1955.
(32) Pidgeon, C.; Ong, S.; Liu, H.; Qiu, X.; Pidgeon, M.; Dantzig, A.
H.; Munroe, J .; Hornback, W. J .; Kasher, J . S.; Glunz, L.;
Szczerba, T. IAM Chromatography: An in Vitro Screen for
Predicting Drug Membrane Permeability. J . Med. Chem. 1995,
38, 590-594.
(21) Iwanowicz, E. J .; Lin, J .; Roberts, D. G. M.; Michel, I. M.; Seiler,
S. M. R-Hydroxy- and R-Ketoester Functionalized Thrombin
Inhibitors. Bioorg. Med. Chem. Lett. 1992, 2, 1607-1612.
(22) Brady, S. F.; Sisko, J . T.; Stauffer, K. J .; Colton, C. D.; Qiu, H.;
Lewis, S. D.; Ng, A. S.; Shafer, J . A.; Bogusky, M. J .; Veber, D.
F.; Nutt, R. F. Amide and R-Keto Carbonyl Inhibitors of
Thrombin Based on Arginine and Lysine: Synthesis, Stability
and Biological Characterization. Bioorg. Med. Chem. 1995, 3,
1063-1078.
(23) Hagihara, M.; Schreiber, S. L. Reassignment of Stereochemistry
and Total Synthesis of the Trombin Inhibitor Cyclotheonamide
B. J . Am. Chem. Soc. 1992, 114, 6570.
(24) (a) Edwards, P. D.; Meyer, E. F., J r.; Vijayalakshmi, J .; Tuthil,
P. A.; Andisik, D. A.; Gomes, B.; Strimpler, A. Design, Synthesis
and Kinetic Evaluation of a Unique Class of Elastase Inhibitors,
the Peptidyl R-Ketobenzoxazoles, and the X-rayCrystal Structure
of the Covalent Complex between Porcine Pancreatic Elastase
and Ac-Ala-Pro-Val-2-Benzoxazole. J . Am. Chem. Soc. 1992, 114,
1854. (b) Edwards, P. D.; Wolanin, D. J .; Andisik, D. W.; Davis,
M. W. Peptidyl R-Ketoheterocyclic Inhibitors of Human Neu-
trophil Elastase. 2. Effect of Varying the Heterocyclic Ring on
in Vitro Potency. J . Med. Chem. 1995, 38, 76-85.
(25) (a) Tsutsumi, S.; Okonogi, T.; Shibahara, S.; Patchett, A. A.;
Christensen, B. G. R-Ketothiazole Inhibitors of Prolyl Endopep-
tidase. Bioorg. Med. Chem. Lett. 1994, 4, 831. (b) Tsutsumi, S.;
Okonogi, T.; Shibahara, S.; Ohochi, S.; Hatsushiba, E.; Patchett,
A. A.; Christensen, B. G. Synthesis and Structure-Activity
Relationships of Peptidyl R-Keto heterocycles as Novel Inhibitors
of Prolyl Endopeptidase. J . Med. Chem. 1994, 37, 3492.
(26) (a) Costanzo, M. J .; Maryanoff, B. E.; Hecker, L. R.; Schott, M.
R.; Yabut, S. C.; Zhang, H.-C.; Andrade-Gordon, P.; Kauffman,
J . A.; Lewis, J . M.; Krishnan, R.; Tulinsky, A. J . Med. Chem.
1996, 39, 3039. (b) Akiyama, Y.; Tsutsumi, S.; Hatsushiba, E.;
(33) Ong, S.; Liu, H.; Qiu, X.; Bhat, G.; Pidgeon, C. Anal. Chem. 1995,
67, 755-762.
(34) Hidalgo, I. J .; Hillgren, K. M.; Grass, G. M.; Borchard, R. T.
Characterization of the Unstirred Water Layer in Caco-2 Cell
Monolayers Using a Novel Diffusion Apparatus. Pharm. Res.
1991, 8, 222-227.
(35) Hidalgo, I. J . Cultured Intestinal Epithelial Cell Models. Models
for Assessing Drug Absorption and Metabolism; Plenum Press:
New York, 1996; pp 35-50.
J M980354P