Synthesis of a series of potent and orally bioavailable thrombin inhibitors that utilize 3,3-disubstituted propionic acid derivatives in the P3 position

Article abstract of DOI:10.1021/jm970397q

As part of an effort to prepare efficacious and orally bioavailable analogs of the previously reported thrombin inhibitors 1a,b, we have synthesized a series of compounds that utilize 3,3-disubstituted propionic acid derivatives as P₃ ligands.

Full text of DOI:10.1021/jm970397q

J . Med. Chem. 1997, 40, 3687-3693
3687
Syn th esis of a Ser ies of P oten t a n d Or a lly Bioa va ila ble Th r om bin In h ibitor s
Th a t Utilize 3,3-Disu bstitu ted P r op ion ic Acid Der iva tives in th e P 3 P osition
Thomas J . Tucker,*,† William C. Lumma,† S. Dale Lewis,‡ Stephen J . Gardell,‡ Bobby J . Lucas,‡ J ack T. Sisko,†
J oseph J . Lynch,§ Elizabeth A. Lyle,§ Elizabeth P. Baskin,§ Richard F. Woltmann,§ Sandra D. Appleby,§
I-Wu Chen, Kimberley B. Dancheck, Adel M. Naylor-Olsen,| J ulie A. Krueger,‡ Carolyn M. Cooper,‡ and
J oseph P. Vacca†
Departments of Medicinal Chemistry, Biological Chemistry, General Pharmacology, Drug Metabolism, and
Molecular Design and Diversity, Merck Research Laboratories, West Point, Pennsylvania 19486
Received J une 16, 1997X
As part of an effort to prepare efficacious and orally bioavailable analogs of the previously
reported thrombin inhibitors 1a ,b, we have synthesized a series of compounds that utilize 3,3-
disubstituted propionic acid derivatives as P3 ligands. By removing the N-terminal amino group,
the general oral bioavailability of this class of compounds was enhanced without excessively
increasing the lipophilicity of the compounds. The overall properties of the molecules could be
drastically altered depending on the nature of the groups substituted onto the 3-position of
the P3 propionic acid moiety. A number of the compounds exhibited good oral bioavailability
in rats and dogs, and numerous compounds were efficacious in a rat FeCl3-induced model of
arterial thrombosis. Compound 7, the 3,3-diphenylpropionic acid derivative, showed the best
overall profile of in vivo and in vitro activity. Molecular modeling studies suggest that these
compounds bind in the thrombin active site in a manner essentially identical to that previously
reported for compound 1a .
Ch a r t 1
In tr od u ction
The serine protease thrombin plays a key role in the
blood coagulation cascade. Thrombin is responsible for
the conversion of fibrinogen to fibrin and the activation
of factor XIII which cross-links fibrin monomers, and it
is the most potent known stimulator of platelet ag-
gregation. As such, thrombin has become a key target
in the effort to develop novel antithrombotic agents. We
have recently reported in several papers2 our strategy
directed at the development of orally bioavailable non-
covalent thrombin inhibitors. Our early efforts were
based on derivatives of compound 1a (Chart 1) which
utilize diaryl or dicycloalkyl2 alanines in the P3 position.
Analogs of compound 1a that utilize D-diphenylalanine
as a P3 group were extremely potent thrombin inhibi-
tors, but lacked oral bioavailability.1,2 Conversion of the
diaryl groups to cyclohexyl rings as in compound 1b also
provided a series of very potent thrombin inhibitors and
compound 1b showed high levels of oral bioavailability
in several animal species.2 Unfortunately, compound
1b was limited by its high lipophilicity and the associ-
ated protein binding which appears to compromise the
in vivo antithrombotic efficacy of the compound. As an
alternative to the use of Boc-D-dicyclohexylalanine in
the P3 position, we became interested in compounds of
generic structure 2 which lacked the N-terminal amino
group. Derivatives of compound 2 could provide a
moderate increase in lipophilicity without reaching the
extreme level of lipophilicity inherent in compound 1b.
In theory, the removal of the N-terminal amino group
would have the negative effect of lowering the potency
of the inhibitors, as the N-H of this amino group is
involved in a key hydrogen-bonding interaction with the
carbonyl of Gly 216 on the thrombin â-sheet.1 Potent
tripeptide inhibitors of thrombin that lack the N-
terminal amino group have been reported, but all of
these compounds also contain an activated carbonyl or
boronic acid serine trap which helps to offset the potency
loss.3 However, we believed that since inhibitors such
as compound 1 were already of subnanomolar potency,
we could trade off some of this intrinsic potency in an
effort to optimize other properties of the inhibitors. In
this paper, we describe the synthesis of a series of potent
and orally bioavailable thrombin inhibitors based on the
removal of the N-terminal amino group.
Ch em istr y
The requisite 3,3-disubstituted propionic acids were
obtained commercially (5a ), synthesized via literature
6
methods (5b,4 5c,4 5d ,5 5f ), or synthesized via routes
developed in our laboratories (Table 1). The saturated
bridged tricyclic 5e was synthesized by reduction of the
corresponding unsaturated compound 5d . The 3-(2-
pyridyl)-3-phenylpropionic acid (5g) was prepared by
addition of lithio di(2-pyridyl) cuprate to ethyl cin-
namate followed by ester hydrolysis (Scheme 1, method
A). This cuprate approach was unsuccessful for the
other pyridine isomers. The isomeric 3-pyridyl-3-phen-
ylpropionic acids 5h -j were synthesized by Heck aryla-
† Department of Medicinal Chemistry.
‡ Department of Biological Chemistry.
§ Department of General Pharmacology.
Department of Drug Metabolism.
| Department of Molecular Design and Diversity.
X Abstract published in Advance ACS Abstracts, October 1, 1997.
S0022-2623(97)00397-X CCC: $14.00 © 1997 American Chemical Society

3688 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 22
Ta ble 1. 3,3-Disubstituted Propionic Acids
Tucker et al.
stereochemistry of the purified diastereomers was not
determined.
Resu lts a n d Discu ssion
Compounds 7-21 were evaluated for their thrombin
inhibitory potency and their selectivity versus human
trypsin, and the results are summarized in Table 2. The
3,3-diphenylpropionic acid-based compound 7 exhibited
surprisingly good thrombin inhibitory potency and was
only 20 times less potent then the D-diphenylalanine
lead structure 1.1 Compound 7 also showed moderate
177-fold selectivity versus human trypsin. Molecular
modeling studies7 with compound 7 suggest that the
compound fits in the active site in a manner essentially
identical to that reported previously1 for compound 1,
with the two phenyl rings of 7 occupying similar
positions to the two phenyl rings of compound 1.
Reduction of the aromatic rings to give compound 8
resulted in a 7-fold loss of potency, while the trypsin
selectivity remained essentially the same. The two
monophenyl/monocyclohexyl diastereomers 9 and 10
were of intermediate potency as compared to 7 and 8.
Fusion of the two aromatic rings via an unsaturated
two-carbon bridge provided the tricyclic compound 11,
which was equipotent with compound 7 but showed
greater selectivity versus trypsin. Replacement of the
saturated two-carbon bridge of 11 with a two-carbon
unsaturated bridge as in compound 12 resulted in a
6-fold loss of potency. Initially we were puzzled by this
observation, as the addition of lipophilicity to the back
of the S3 binding pocket should in theory result in an
improvement in inhibitory potency. However, molecular
modeling studies7 with compounds 11 and 12 suggest
that the two-carbon bridging element is probably not
to the rear of the S3 binding pocket but is more likely
to be across the front side of the two rings near the
enzyme-solvent interface. An unsaturated bridging
element would be better tolerated in such a polar
environment, and the data clearly suggest that this is
the case. Addition of a third lipophilic group to the
benzylic position of the P3 moiety as in compound 13
resulted in an 8-fold potency loss.
a
b
Cy refers to a cyclohexane ring. Yield indicated in the table
refers to the overall yield for the multistep preparation of each
compound. c Isolated as sodium salts.
tion of the pyridylpropenoates 3h ,i with the appropriate
aryl halides followed by saponification and reduction
(Scheme 1, method B). The acids 5h -j were isolated
as their sodium salts and were used as obtained in the
next reaction. Final compounds for biological testing
(Table 2) were prepared by simple amino acid coupling
of the 3,3-disubstituted propionic acids 5a -j with trans-
[4-[(tert-butoxycarbonyl)amino]cyclohexyl]methyl L-pro-
line amide1 (6), followed by BOC removal and reversed
phase preparatory LC purification (Scheme 2). In the
cases where two diastereomers were obtained (com-
pounds 9 and 10, 14 and 15, 16 and 17, 18 and 19, and
20 and 21), the individual diastereomers were separated
and purified via reversed phase preparatory LC. Pairs
of diastereomers are numbered in order of elution from
a reversed phase C18 preparatory LC column. Absolute
The 3-pyridyl-3-phenylpropionic acid derivatives 14-
19 showed clear differences in potency based on the
stereochemistry at the P3 benzylic carbon. In each pair
Sch em e 1a
a
Reagents: (a) 1. 2-bromopyridine, n-BuLi/THF, -78 °C, 2. CuBr(CH3)2S; (b) 1 M LiOH/1,2-dimethoxyethane; (c) Pd(OAc)2, tri-o-
tolylphosphine, iodobenzene or 3,4-(methylenedioxy)phenyl iodide, triethylamine/CH3CN, 100 °C, sealed tube, 72 h; (d) 1 M LiOH/DME;
(e) 60% NaH dispersion, THF; (f) Raney nickel/EtOH, Parr apparatus, 60 psi, 24-48 h.

Potent and Orally Bioavailable Thrombin Inhibitors
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 22 3689
Ta ble 2. 3,3-Disubstituted Propionic Acid-Based Thrombin Inhibitors
a
b
Cy refers to a cyclohexyl ring. Ki values for all serine proteases are the average of at least two determinations, where the variation
in the assay is (10%. Experimental protocols are provided in ref 2. c The yield as indicated in the table is the yield obtained for the
three-step coupling, deprotection, and purification/separation sequence as shown in Scheme 2. Diastereomers are labeled as 1 or 2
d
based on order of elution from a C18 reversed phase preparatory LC column.

3690 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 22
Tucker et al.
Sch em e 2a
of either lipophilicity or molecular weight to the mol-
ecules. Application of this theme to compound 17
provided the diastereomers 20 and 21. Once again, a
clear difference in enzyme potency between the diaster-
eomers was observed, with the more potent diastere-
omer 21 exhibiting a 9-fold potency enhancement over
compound 17. Compound 21 also showed excellent
selectivity of almost 9300-fold versus human trypsin.
These results are consistant with the explanation of the
pyridyl/phenyl P3 diastereomer binding described above
and reinforce our hypothesis. It is probable that, as
predicted, the more potent diastereomer 21 is of the
absolute configuration that more easily places the
(methylenedioxy)phenyl ring back into the usual D-Phe
S3 binding position, while also placing the pyridyl ring
into the front aryl-binding position in the S3 pocket.
a
Reagents (a) EDC, HOBT, triethylamine/DMF; (b) TFA-
Key compounds were evaluated for selectivity versus
a number of human-derived serine proteases, and the
results are detailed in Table 3. In general, the com-
pounds demonstrated good selectivity versus all serine
proteases tested. The values of the second-order rate
constant of inactivation (kon) are also provided in Table
3 for several of the inhibitors, and all examples were
shown to be fast binding inhibitors. Key compounds
were also evaluated for oral bioavailability in several
animal species. Results are summarized in Table 4. As
expected, the more lipophilic compounds 7, 8, and 11
all performed well in rats, while the more polar pyri-
dine-containing compounds 17 and 21 performed poorly
in rats and dogs, respectively. Pharmacokinetic param-
eters for compound 7 in rats and dogs are detailed in
Table 5. Compound 7 exhibited good overall pharma-
cokinetic behavior in both species. Especially notewor-
thy was the long half-life observed for 7 in both species.
CH2Cl2 or EtOAc saturated with HClg, 0 °C; (c) reversed phase
(C18) preparatory HPLC.
of pyridine diastereomers, there was one compound
which was clearly more potent than its counterpart.
These observations may be explained in part by our
previous crystallographic studies1 on compound 1. It
is likely that the diastereomer which more easily places
the phenyl ring back into the usual D-Phe binding
position and the pyridyl ring into the front of the S3
pocket corresponds to the more potent diastereomer.
The rear of the S3 binding pocket is lined by lipophilic
amino acid side chains and would likely prefer to be
occupied by a phenyl ring rather than a pyridyl ring.
Conversely, the pyridine ring would be more easily
accommodated in the front diaryl-binding position of the
S3 pocket, as this region lies close to the enzyme-solvent
interface and would provide a more polar environment
for the pyridine nitrogen. The 3-pyridyl compound 17
also showed a surprising enhancement in selectivity
ratio versus trypsin. On the basis of the above detailed
hypothesis of differential binding of the various pyri-
dine-containing diastereomers, we theorized that it
should be possible to add bulk to the phenyl ring of the
appropriate pyridine diastereomer to further enhance
the potency of these compounds. Due to the excellent
selectivity ratio observed for compound 17 versus trypsin,
we chose this pyridine isomer for further work. In
previous unpublished work8 on D-phenylalanine P3-
containing thrombin inhibitors, we observed that the
addition of the appropriate substitution to the 3- and
4-positions of the aromatic ring provided enhanced
potency due to optimization of lipophilic interaction with
the back of the S3 binding pocket. We found that 3,4-
methylenedioxy was among the best substituents for
this potency enhancement and did not add large amounts
Compounds 7, 8, 17, and 21 were evaluated for in vivo
antithrombotic efficacy in the rat carotid artery FeCl3-
induced thrombosis model. Results are summarized in
Table 6. The more polar/less protein-bound compounds
17 and 21 performed extremely well, with both showing
full antithrombotic efficacy (all compounds were dosed
iv over 120 min as a 10 µg/kg/min infusion). The most
lipophilic compound 8 exhibited a complete lack of
antithrombotic efficacy and also showed low plasma
levels as determined at the end of the experiment. The
diphenyl compound 7 showed moderate antithrombotic
activity, preventing occlusion in 3 of 5 rats. Compounds
7, 17, and 21 all showed good levels of parent compound
as determined at the end of the experiment. In general,
these results are consistent with our previously re-
ported2 correlation between lipophilicity/2 × APTT/
protein binding and in vivo antithrombotic efficacy.
These data further reinforce our premise2 that a perfect
Ta ble 3. Selectivity of Thrombin Inhibitors versus Various Serine Proteasesa
Ki (µM)
Ki thrombin
kon thrombin
plasma
Kallikrein
compd no.
(nM)
(×10-7 M-1 s-1
)
trypsin
TPA
plasmin
Xa
980
220
>1000
370
7
8
2
14
3
1.8
2.2
3.4
0.28
21.0
23.0
18.0
18.0
21.0
15.5
25.1
0.35
2.50
1.60
0.89
2.2
>1000
84
744
176
414
>1000
170
11
14
17
19
21
268
775
767
476
>1000
>1000
>1000
744
746
>1000
2.2
2.60
436
>500
>1000
1000
400
572
1000
a
Ki values for all serine proteases are the average of at least two determinations, where the variation in the assay is (10%. Experimental
protocols are given in ref 2.

Potent and Orally Bioavailable Thrombin Inhibitors
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 22 3691
Ta ble 4. Oral Bioavailability of Thrombin Inhibitorsa
F (%)
generation lead structure for further synthetic work. We
are continuing our efforts toward the synthesis of the
“ideal” thrombin inhibitor for clinical development and
will report on our further efforts in future publications.
compd no.
log P
ratb (10 mg/kg)
dogc (5 mg/kg)
7
8
11
12
17
21
0.37
1.96
0.27
ND
-0.64
-1.62
10
56
51
12
ND
<5
58
ND
ND
ND
8
Exp er im en ta l Section
Melting points were determined in open capillary tubes in
a Thomas-Hoover apparatus and are uncorrected. 1H NMR
spectra were recorded on a Varian Unity 400 spectrometer.
Chemical shifts are reported in δ (ppm) relative to tetra-
methylsilane. All reagents were of commercial synthetic grade
unless otherwise specified. Aldrich Sure-Seal dimethylforma-
mide was used as solvent in all amino acid coupling reactions.
Reversed phase preparatory HPLC purification and diastere-
omer separations were performed on a Waters Prep 4000,
using a Waters C18 Prep-Pac and various gradients.
ND
a
All experimental protocols for these experiments are given in
b
ref 2. Compounds were dosed orally to rats as a suspension in
1% methocel. c Compounds were dosed orally to dogs as a 1%
methocel suspension.
Ta ble 5. Pharmacokinetic Parameters for Compound 7 in Rats
and Dogsa
3,3-Dicycloh exylp r op ion ic Acid (5b). Compound 5b was
synthesized via the method described in ref 4. White crystal-
line solid, mp ) 122.0-123.5 °C, yield ) 65%. 400 MHz 1H
NMR (CDCl3): 0.88-1.32 (complex, 10H), 1.42 (m, 2H), 1.51-
1.82 (complex, 11H), 2.28 (d, 2H). FAB MS: M+ ) 239.
3-P h en yl-3-cycloh exylp r op ion ic Acid (5c). Compound
5c was synthesized via the method described in ref 4. White
dose
AUC
t1/2
Cl
log P (mg/kg) route (µM‚h) (min) (mL/min/kg) F (%)
Ratb
1.70
1.51
Dogc
3.81
0.37
1
10
iv
po
268
23.2
10.2
10
58
1
0.37
0.37
1
5
iv
po
175
156
crystalline solid, mp ) 97-99 °C, yield ) 38%. 400 MHz H
11.20
NMR (CDCl3): 0.73-1.28 (complex, 5H), 1.46 (t, 2H), 1.61 (m,
2H), 1.76 (q, 2H), 2.58 (m, 1H), 2.82 (m, 1H), 2.84 (m, 1H),
7.19 (complex, 5H). FAB MS: M+ ) 233.
a
Details of the experimental protocols are given in ref 2.
Compound was dosed orally as 1% methocel suspension, iv in
b
5H-Diben zo[a ,d ]cycloh ep ten e-5-a cetic Acid (5d ). Com-
pound 5d was synthesized via the method described in ref 5.
10% propylene glycol/saline. c Compound was dosed orally as 1%
methocel suspension, iv as a solution in 20% PEG/saline.
1
Tacky yellow solid, yield ) 70%. 400 MHz H NMR (CDCl3):
2.77 (d, 2H), 4.51 (t, 1H), 6.91 (s, 2H), 7.29 (complex, 8H).
10,11-Dih yd r o-5H-d iben zo[a ,d ]cycloh ep t a n e-5-a cet ic
Acid (5e). A solution of 1.00 g (4.00 mol) of compound 5d in
30 mL of absolute ethanol containing 1.00 equiv of HCl was
hydrogenated on a Parr appartus at 50 psi over 200 mg of 10%
Pd on carbon catalyst for 18 h. The catalyst was filtered off
and the filtrate concentrated to a clear oil. The oil was purified
via column chromatography over silica gel with 4:1 hexane/
EtOAc to give 930 mg of product isolated as its ethyl ester.
The ester was hydrolyzed by dissolving in 15 mL of 1 M LiOH/
15 mL of dimethoxyethene and stirring at room temperature
for 18 h. The reaction was acidified with aqueous 10% citric
acid and extracted twice with 30 mL portions of EtOAc.
Drying (anhydrous MgSO4) and concentration in vacuo pro-
vided 681 mg (82% for two steps) of desired product 5c as a
Ta ble 6. Antithrombotic Efficacy of Thrombin Inhibitors: Rat
Carotid Artery FeCl3-Induced Thrombosis Modela
2 × APTT
(µM)c
% freed
compd
no.b
no.
final plasma
r
occlusions rat human rat human concn (nM)
7
8
17
21
5
6
5
6
6
2/5
6/6
0/5
0/6
0/6
0.65
7.20
0.59
0.38
ND
0.67
7.10
0.47
0.30
0.28
7
6
56
ND
ND
7
1
35
ND
ND
1600
658
2890
4000
ND
argatroban
a
Details of the experimental protocol are given in ref 2 and
references cited therein. Compounds were infused at 10 µg/kg/
min iv for 120 min prior to the FeCl3 insult. Vehicle for compound
administration was saline. Vehicle control animals showed com-
plete occlusion. c The 2 × APTT value is defined as the concentra-
tion of inhibitor in plasma required to double the activated partial
b
1
pale yellow crystalline solid, mp ) 155-157 °C. 400 MHz H
NMR (CDCl3): 3.10 (m, 2H), 3.13 (d, 2H), 3.28 (m, 2H), 4.76
d
thromboplastin time. Experimental protocol for the determina-
tion of this parameter is detailed in ref 2.
(t, 1H), 7.17 (complex, 8H). FAB MS: M+ ) 253.
3-Meth yl-3,3-d ip h en ylp r op ion ic Acid (5f). Compound
5f was prepared via the method described in ref 6. Tacky
white solid, yield ) 20%. 400 MHz 1H NMR (CDCl3): 1.90 (5,
3H), 3.19 (s, 2H), 7.19 (m, 6H), 7.28 (M, 4H).
balance of physical properties, potency, and selectivity
is crucial to the design of clinically viable oral thrombin
inhibitors.
Meth od A: P r ep a r a tion of 3-P yr id yl-3-p h en ylp r op i-
on ic Acid s 5g-j. 3-(2-P yr id yl)-3-p h en ylp r op ion ic Acid
(5g). To a solution of 1.80 g (11.36 mmol) of 2-bromopyridine
in 10 mL of anhydrous ether (argon atmosphere) cooled to -78
°C was added dropwise 7.10 mL (11.36 mmol) of 1.6 M
n-butyllithium in hexanes (Aldrich). The resulting deep red
solution was stirred at -78 °C for 30 min. The cold lithiopy-
ridine solution was added dropwise to a suspension of 1.17 g
(5.68 mmol) of copper bromide-dimethyl sulfide complex
(Aldrich) in 20 mL of anhydrous ether cooled to 0 °C. The
resulting dark solution was stirred at 0 °C for 20 min and then
removed from the cooling bath. The dark solution was treated
dropwise with 500.00 mg (2.84 mmol) of ethyl cinnamate, and
the mixture stirred at room temperature for 1 h. The reaction
mixture was poured into 40 mL of 90% saturated NH4Cl/10%
concentrated aqueous NH4OH, and the resulting mixture
stirred vigorously at room temperature for 45 min. The layers
were separated, and the aqueous layer was extracted with 25
mL of ether. The combined ethereal layers were extracted
twice with 40 mL portions of 2 N HCl. The combined acid
extracts were neutralized (pH 7-8) with concentrated aqueous
NH4OH and extracted twice with 30 mL portions of ether. The
combined extracts were washed with brine, dried (anhydrous
In summary, we have synthesized a series of potent
and selective thrombin inhibitors based on the use of a
3,3-disubstituted propionic acid as a P3 ligand. The
series has demonstrated that it is possible to achieve
high levels of potency against thrombin despite the
removal of the N-terminal amino group which is in-
volved in a key hydrogen-bonding interaction with the
thrombin â-sheet. A number of the compounds showed
good oral bioavailability in rats, and compound 7 also
demonstrated good oral bioavailability and good phar-
macokinetic behavior in dogs. Several of the compounds
also exhibited good in vivo antithrombotic activity in
the rat carotid artery FeCl3-induced thrombosis model.
In general, the more polar compounds performed poorly
in the oral bioavailability models while performing well
in the in vivo antithrombotic model; the opposite trends
were observed for the more lipophilic compounds. Com-
pound 7 demonstrated the best overall profile of all the
compounds in this series and has become a second-

3692 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 22
Tucker et al.
MgSO4), and concentrated in vacuo to give a dark oil. The
crude oil was purified via gravity column chromatography over
silica gel with 3:1 hexanes/ethyl acetate to give 220 mg (28%)
of the ether ester as a pale yellow oil. 400 MHz 1H NMR
(CDCl3): 1.16 (t, 3H), 2.98 (dd, J ) 7, 17 Hz, 1H), 3.43 (dd, J
) 7, 17 Hz, 1H), 4.04 (q, 2H), 4.64 (m, 1H), 7.17 (complex, 3H),
7.26 (m, 4H), 7.56 (m, 1H), 8.58 (m, 1H).
silica gel with 3:2 ethyl acetate/chloroform to give 105 mg of
coupling product as a clear glass. The product was dissolved
in 1 mL of CH2Cl2/1 mL of trifluoroacetic acid (Aldrich), and
the reaction mixture stirred at room temperature for 18 h. The
reaction was concentrated to dryness, and the residue was
purified via reversed phase preparatory HPLC. Pure fractions
were combined and concentrated to dryness to give a clear
glass. The glass was suspended in ether with a few drops of
hexane, scraped, and triturated. Filtration gave 70 mg (50%)
of desired product 7 as a glassy white solid, mp ) 120-123
A 215.00 mg (0.85 mmol) sample of the ethyl ester was
dissolved in 3 mL of LiOH/3 mL of dimethoxyethane, and the
solution stirred at room temperature for 2 h. The reaction
mixture was diluted with aqueous 10% citric acid to pH ∼6.5
and the solution extracted twice with 10 mL portions of ethyl
acetate. The combined extracts were dried (anhydrous MgSO4)
and concentrated in vacuo to give 190 mg (99%) of the desired
°C. Elemental Anal. for C27H35N3O2‚1.60TFA‚0.45H2O:
C
(58.11 calcd, 58.11 found), H (6.06 calcd, 6.02 found), N (6.73
1
calcd, 6.90 found). 400 MHz H NMR (CDCl3): 0.91 (m, 4H),
1.14 (br s, 1H), 1.63 (m, 2H), 1.81 (br s, 1H), 1.96 (m, 2H),
2.38 (m, 1H), 2.76 (m, 1H), 3.02 (m, 2H), 3.23 (m, 1H), 3.43
(m, 1H), 4.08 (q, 1H), 4.24 (q, 1H), 4.35 (m, 1H), 4.50 (d, 1H),
4.63 (t, 1H), 6.89 (br s, 1H), 7.21 (complex, 10H). High-
resolution FAB MS: M+ calcd ) 434.28075, M+ obsd )
434.28204.
1
product as a yellow oil. 400 MHz H NMR (CDCl3): 3.09 (dd,
J ) 5, 16 Hz, 1H), 3.39 (q, 1H), 4.62 (m, 1H), 7.22 (complex,
7H), 7.65 (t, 1H), 8.58 (d, 1H). FAB MS: M + 1 ) 228.
Met h od B: Gen er a l P r oced u r e for t h e Syn t h esis of
Com p ou n d s 5h -j via Heck Ar yla tion . 3-(3-P yr id yl)-3-
p h en ylp r op ion ic Acid (5h ). A solution of 250.00 mg (1.53
mmol) of methyl 3-(3-pyridyl)propenoate, 5.00 mg (0.02 mmol)
of palladium acetate (Aldrich), 25.00 mg (0.08 mmol) of tri-o-
tolylphosphine (Aldrich), 172.00 µL (1.53 mmol) of iodobenzene
(Aldrich), and 223.00 µL (1.60 mmol) of triethylamine (Aldrich)
in 1 mL of acetonitrile was heated at 100 °C in a sealed tube
for 72 h. The reaction mixture was cooled to room tempera-
ture, and concentrated in vauco to an orange-red solid. The
solid was purified via gravity column chromatography over
silica gel with 2:1 hexanes/ethyl acetate to give 190 mg (52%)
of coupling product as a pale yellow oil. The ester was
hydrolyzed in 2 mL of LiOH/2 mL of dimethoxyethane over
∼2 h, acidified to pH ∼6.5 with 10% citric acid, and extracted
with ethyl acetate. The extract was dried (anhydrous MgSO4)
and concentrated in vacuo to give 177 mg (99%) of unsaturated
acid. The acid was dissolved in THF and treated with an
equimolar amount of 60% NaH dispension (Aldrich). After
stirring for 1 h, the reaction mixture was concentrated to give
the unsaturated acid sodium salt as a yellow foam-solid. The
salt was hydrogenated on a Parr apparatus in absolute ethanol
over 100 mg of Raney nickel (Aldrich) catalyst at 65 psi. The
reaction was monitored by HPLC until complete (24-48 h),
and the reaction mixture was filtered carefully (Raney nickel
is extremely pyrophoricsuse care!). The filtrate was concen-
trated in vacuo to give 157 mg (80%) of the sodium salt of the
The following compounds were prepared in the same man-
ner as compound 7.
N-(1-Oxo-3,3-d icycloh exylp r op yl)-L-p r olin e [tr a n s-(4-
Am in ocycloh exyl)m eth yl]a m id e (8). White solid, mp )
104-107 °C, yield ) 38%. FAB MS: M+ ) 447. Elemental
Anal. for C27H47N3O2‚HCl‚0.35H2O: C (66.39 calcd, 66.11
obsd), H (10.05 calcd, 9.95 obsd), N (8.60 calcd, 8.56 obsd). 400
1
MHz H NMR (CDCl3): 0.89-1.25 (complex, 13H), 1.29-1.95
(complex, 19H), 1.95-2.29 (complex, 6H), 2.43 (m, 1H), 2.99
(m, 1H), 3.09 (m, 2H), 3.44 (m, 1H), 3.59 (m, 1H), 4.61 (m,
1H), 7.46 (br s, 1H), 8.39 (br, NH2).
N-[1-Oxo-2-(5H-d iben zo[a ,d ]cycloh ep ten -5-yl)eth yl]-L-
p r olin e [tr a n s-(4-Am in ocycloh exyl)m eth yl]a m id e (11).
Amorphous white lyophilizate, yield ) 40%. High-resolution
FAB MS: M+ calcd ) 458.280753, obs ) 458.280112. 400 MHz
1H NMR (CDCl3): 0.99 (m, 2H), 1.21-1.89 (complex, 7H), 2.22
(m, 3H), 2.78 (m, 2H), 2.84 (m, 2H), 3.05 (m, 1H), 3.10 (m,
2H), 3.21 (m, 1H), 4.39 (br m, 1H), 4.72 (t, 1H), 6.85 (br s,
1H), 6.96 (s, 2H), 7.31 (complex, 8H), 8.31 (br, NH2). Elemen-
tal Anal. for C29H36N3O2‚HCl‚1.55H2O: C (calcd 62.51, obsd
62.49), H (calcd 7.05, obsd 7.05), N (calcd 7.45, obsd 7.50).
N-[1-Oxo-2-(10,11-d ih yd r o-5H-d iben zo[a ,d ]cycloh ep t-
5-yl)eth yl]-L-p r olin e [tr a n s-(4-Am in ocycloh exyl)m eth yl]-
a m id e (12). Amorphous white lyophilizate, yield ) 40%.
High-resolution FAB MS: M+ calcd ) 460.296403, obsd )
460.296646. 400 MHz 1H NMR (CDCl3): 0.91 (m, 2H), 1.36
(m, 3H), 1.74 (m, 4H), 1.88 (q, 1H), 2.02 (m, 2H), 2.19 (m, 1H),
2.96 (m, 2H), 3.04 (m, 2H), 3.09 (m, 4H), 3.15 (m, 1H), 3.31
(m, 1H), 3.38 (m, 2H), 4.40 (d, 1H), 4.75 (t, 1H), 7.11 (complex,
9H), 7.79 (br, NH2). Elem. Anal. for C29H37N3O2‚1.60TFA‚
0.20H2O: C (59.90 calcd, obsd 59.89), H (6.09 calcd, 6.06 obsd),
N (6.51 calcd, 6.79 obsd).
1
desired acid as a white solid. 400 MHz H NMR (DMSO-d6):
2.71 (m, 2H), 4.58 (m, 1H), 7.18 (complex, 5H), 7.61 (m, 1H),
8.09 (d, 1H), 8.20 (d, 1H), 8.49 (s, 1H). The sodium salt was
used as isolated in the next reaction.
The following compounds were also prepared via method
B.
3-(4-P yr id yl)-3-p h en ylp r op ion ic Acid (5i). Glassy solid,
yield ) 69%. 400 MHz 1H NMR (DMSO-d6): 2.62 (d, 1H), 2.76
(d, 1H), 4.48 (br m, 1H), 7.20 (complex, 5H), 8.20 (d, 2H), 8.52
(br m, 2H). Isolated as sodium salt.
3-(3-P yr id yl)-3-[3,4-(m eth ylen ed ioxy)p h en yl]p r op ion -
ic Acid (5j). Foam, yield ) 78%. 400 MHz 1H NMR (CD3-
OD): 2.84 (d, 2H), 4.54 (t, 1H), 5.89 (s, 2H), 6.76 (m, 3H), 7.35
(m, 1H), 7.78 (m, 1H), 8.36 (m, 1H), 8.45 (s, 1H). Isolated as
sodium salt.
Cou p lin g of 3,3-Disu bstitu ted P r op ion ic Acid s 5a -
d ,f-h w ith tr a n s-[4-[(ter t-Bu toxyca r bon yl)a m in o]cyclo-
h exyl]m eth yl L-P r olin e Am id e 6. Gen er a l P r oced u r e for
th e P r ep a r a tion of Com p ou n d s 7, 8, a n d 11-13. N-(1-
Oxo-3,3-d ip h en ylp r op yl)-L-p r olin e [tr a n s-(4-Am in ocyclo-
h exyl)m eth yl]a m id e (7). A solution of 70.14 mg (0.31 mmol)
of 3,3-diphenylpropionic acid, 100 mg of trans [4-[(tert-butoxy-
carbonyl)amino]cyclohexyl]methyl L-proline amide,1 46.00 mg
(0.34 mol) of HOBt (Aldrich), 48.00 µL (0.34 mmol) of triethyl-
amine (Aldrich), and 65.00 mg (0.34 mmol) of EDC (Pierce) in
1 mL of anhydrous DMF was stirred at room temperature for
18 h. The reaction mixture was diluted with 3× its volume of
10% citric acid, and the suspension was extracted with two
25 mL portions of ethyl acetate. The combined extracts were
washed with brine, dried (anhydrous MgSO4), and concen-
trated in vacuo to give the crude coupling product. The crude
product was purified via gravity column chromatography over
N-(1-Oxo-3-m eth yl-3,3-diph en ylpr opyl)-L-pr olin e [tr a n s-
(4-Am in ocycloh exyl)m eth yl]a m id e (13). White powder,
mp ) 147-150 °C, yield ) 45%. High-resolution FAB MS:
M+ calcd ) 448.29640, obsd ) 448.29648. Elemental Anal.
for C28H37N3O2‚HCl‚1.80H2O: C (calcd 65.11, obsd 65.06), H
1
(8.12 calcd, 7.91 obsd), N (8.14 calcd, 8.26 obsd). 400 MHz H
NMR (CDCl3): 0.96 (m, 2H), 1.55 (m, 3H), 1.78 (m, 4H), 1.96
(s, 3H), 2.19 (m, 3H), 2.99 (m, 2H), 3.04 (m, 2H), 3.11 (s, 2H),
3.21 (m, 2H), 4.42 (br s, 1H), 7.00 (br s, 1H), 7.21 (complex,
10H), 8.22 (br, NH2).
Cou p lin g of 3,3-Disu b st it u t ed P r op ion ic Acid (or
Sod iu m Sa lt s) 5e,g-l w it h tr a n s-[4-[(ter t-Bu t oxyca r -
bon yl)a m in o]cycloh exyl]m eth yl L-P r olin e Am id e (6) a n d
Sep a r a tion of th e P r od u ct Dia ster eom er s. N-[1-Oxo-3-
(4-p yr id yl)-3-p h en ylp r op yl]-L-p r olin e [tr a n s-(4-Am in ocy-
cloh exyl)m eth yl]a m id e (18 a n d 19). A mixture of com-
pounds 18 and 19 was prepared via an analogous coupling,
purification, and deprotection scheme as described above for
compound 7. The product mixture was separated via reversed
phase preparatory HPLC by taking numerous fractions across
the closely separated peaks. The earlier eluting peak fractions
were combined and concentrated to remove volatiles, and the
aqueous residue was placed on a lyophilizer overnight. Lyo-
philization provided 65 mg (30%) of compound 18 as a fluffy
white amorphous powder. Elemental Anal. for C26H34N4O2‚2.50

Potent and Orally Bioavailable Thrombin Inhibitors
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 22 3693
TFA‚0.45H2O: C (51.16 calcd, 51.14 obsd), H (5.18 calcd, 5.02
obsd), N (7.37 calcd, 7.33 obsd). High-resolution FAB MS: M+
calcd ) 435.27600, obsd ) 435.27479. 400 MHz 1H NMR
(CDCl3): 0.95 (m, 2H), 1.41 (m, 2H), 1.79 (m, 4H), 2.10 (m,
4H), 2.96 (m, 2H), 3.11 (m, 2H), 3.43 (q, 1H), 3.61 (m, 1H),
3.79 (m, 1H), 3.80-4.30 (br, 2H, NH2), 4.30 (m, 1H), 4.48 (m,
1H), 4.75 (m, 1H), 7.01 (t, 1H), 7.29 (complex, 5H), 7.58 (br m,
2H), 8.61 (br m, 2H).
17: Glassy amorphous solid, yield ) 30%. High-resolution
FAB MS: M+ calcd ) 435.27600, obsd ) 435.27405. Elemen-
tal Anal. for C26H34N4O2‚2.50TFA‚1.00H2O: C (50.47 calcd,
5047 obsd), H (5.26 calcd, 5.06 obsd), N (7.60 calcd, 7.68 obsd).
1
400 MHz H NMR (CDCl3): 0.85 (m, 2H), 1.31 (m, 3H), 1.64
(t, 2H), 1.96 (t, 2H), 12.92 (m, 2H), 3.04 (m, 2H), 3.31 (m, 1H),
3.39 (m, 1H), 4.40 (m, 1H), 4.78 (br t, 1H), 7.21 (complex, 5H),
7.61 (br, 1H), 8.02 (d, 1H), 8.21 (m, 1H), 8.63 (m, 1H), 8.78 (d,
1H).
The later eluting peak fractions were combined and treated
as above to give 25 mg (12%) of compound 19 as a glassy solid.
Elemental Anal. for C26H34N4O2‚2.50TFA‚2.40H2O: C (48.80
calcd, 48.79 obsd), H (5.46 calcd, 5.22 obsd), N (7.34 calcd, 7.48
obsd). High-resolution FAB MS: M+ calcd ) 435.27600, obsd
N-[1-Oxo-3-(3-p yr idyl)-3-[3,4-(m eth ylen ed ioxy)p h en yl]-
p r op yl]-L-p r olin e [tr a n s-(4-Am in ocycloh exyl)m et h yl]-
a m id e (Dia ster eom er s 20 a n d 21). 20: White amorphous
lyophilizate, yield ) 26%. FAB MS: M+ 1 ) 479. Elemental
Anal. for C27H34N4O4‚2.50 TFA‚2.50H2O: C (calcd 47.62, obsd
47.65), H (calcd 5.19, obsd 4.90), N (calcd 6.95, obsd 7.20). 400
MHz 1H NMR (CD3OD): 1.02 (m, 2H), 1.36 (m, 4H), 1.89-
2.14 (complex, 8H), 3.01 (m, 2H), 3.16 (m, 1H), 3.31 (m, 1H),
3.64 (m, 2H), 4.26 (m, 1H), 4.21 (m, 1H), 5.96 (s, 2H), 6.91
(complex, 3H), 7.90 (m, 2H), 8.42 (d, 1H), 8.70 (s, 1H).
1
) 435.27657. 400 MHz H NMR (CDCl3): 0.91 (m, 2H), 1.31
(m, 2H), 1.71 (d, 2H), 1.91 (m, 2H), 2.11 (m, 4H), 2.92 (m, 2H),
3.04 (m, 2H), 3.32 (m, H), 3.39 (m, 1H), 3.69 (m, 1H), 380-
4.30 (br, 2H) 4.09 (br m, 1H), 4.39 (m, 1H), 4.74 (br m, 1H),
7.29 (complex, 6H), 8.21 (br d, 2H), 8.61 (br m, 2H). Both
diastereomers were of >99% purity by HPLC Anal. at 210 and
254 nm.
21: White amorphous lyophilizate, yield ) 30%. FAB MS:
M+ 1 ) 479. Elemental Anal. for C27H34N4O4‚2.50 TFA‚
2.50H2O: C (calcd 47.67, obsd 47.55), H (calcd 5.19, obsd 4.97),
The following compounds were prepared in the same man-
ner as compounds 18 and 19.
N-(1-Oxo-3-ph en yl-3-cycloh exylpr opyl)-L-pr olin e [tr a n s-
(4-Am in ocycloh exyl)m eth yl]a m id e (Dia ster eom er s 9 a n d
10). 9: White solid, mp ) 116-119 °C, yield ) 39%. High-
resolution FAB MS: M+ calcd ) 440.32770, obsd ) 440.32599.
Elemental Anal. for C27H41N3O2‚1.25 TFA‚0.80H2O: C (calcd
59.39, obsd 59.32), H (calcd 7.41, obsd 7.47), N (calcd 7.04, obsd
1
N (calcd 6.95, obsd 7.00). 400 MHz H NMR (CD3OD): 0.89
(m, 2H), 1.18 (m, 2H), 1.29 (m, 2H), 1.62-1.98 (complex, 5H),
2.08 (m, 1H), 2.89 (m, 2H), 3.14 (m, 7H), 3.42 (m, 1H), 3.59
(m, 1H), 4.17 (m, 1H), 4.63 (m, 1H), 5.82 (s, 2H), 6.74 (complex,
3H), 7.78 (t, 1H), 8.34 (d, 1H), 8.55 (d, 1H), 8.63 (s, 1H).
1
7.40). 400 MHz H NMR: 0.85 (m, 2H), 0.95 (q, 1H), 1.09 (m,
Ack n ow led gm en t. The authors thank Dr. Graham
Smith, Ms. Patrice Ciecko, Mr. Ken Anderson, and Mr.
Matt Zrada for elemental analyses and log P determina-
tions. The authors also thank Dr. Harri Ramjit and Dr.
Art Coddington for mass spectroscopy data. The authors
also thank Dr. J oel Huff and Dr. J ules Shafer for their
support and encouragement and Ms. J ean Kaysen for
manuscript preparation.
2H), 1.29 (m, 6H), 1.45 (m, 1H), 1.78 (m, 1H), 1.87 (m, 6H),
2.01 (m, 2H), 2.07 (m, 1H), 2.72 (m, 4H), 2.80 (m, 1H), 2.95
(m, 2H), 3.42 (m, 2H), 4.41 (t, 1H), 6.22 (t, 1H), 7.19 (complex,
5H), 7.78 (br, NH2).
10: White solid, mp ) 120-122 °C, yield ) 66%, High-
resolution FAB MS: M+ calcd ) 440.32770, obsd ) 440.32799.
Elemental Anal. for C27H41N3O2‚1.25TFA‚0.65H2O: C (calcd
59.66, obsd 59.61), H (calcd 7.39, obsd 7.36), N (calcd 7.08, obsd
7.23). 400 MHz 1H NMR (CDCl3): 0.79 (m, 1H), 0.96 (m, 4H),
1.09 (m, 2H), 1.25 (m, 1H), 1.38 (m, 4H), 1.42 (d, 1H), 1.51 (q,
1H), 1.71 (m, 4H), 1.86 (d, H), 1.99 (m, 4H), 2.61 (m, 1H), 2.81
(m, 2H), 2.90 (m, 2H), 2.95 (m, 2H), 3.19 (m, 1H), 3.54 (m,
1H), 4.22 (m, 1H), 7.18 (complex, 6H), 7.59 (br, NH2).
N-[1-Oxo-3-(2-pyr idyl)-3-ph en ylpr opyl]-L-pr olin e [tr a n s-
(4-Am in ocycloh exyl)m eth yl]a m id e (Dia ster eom er s 14
a n d 15). 14: Amorphous white lyophilizate, yield ) 30%.
High-resolution FAB MS ) M+ calcd ) 4.35.27600, obsd )
435.27610. Elemental Anal. for C26H34N4O2‚2.50TFA‚
0.85H2O: C (calcd 50.66, obsd 50.68), H (5.24 calcd, 5.19 obsd),
N (7.67 calcd, 7.85 obsd). 400 MHz 1H NMR (CDCl3): 0.92
(m, 2H), 1.52 (m, 4H), 1.80 (m, 2H), 1.95 (m, 4H), 7.09 (m,
2H), 2.20 (m, 2H), 2.94 (m, 2H), 3.19 (m, 2H), 3.59 (m, 2H),
4.36 (m, 1H), 4.96 (m, 1H), 7.16 (t, 1H), 7.31 (complex, 5H),
7.61 (d, 1H), 8.06 (m, 1H), 8.11 (m, 1H), 8.78 (m, 1H).
15: Amorphous white lyophilizate, yield ) 27%. High-
resolution FAB MS: M+ calcd ) 435.27600, obsd ) 435.77710.
Elemental Anal. for C26H34N4O2‚2.50TFA‚0.75H2O: C (calcd
51.41, obsd 51.44), H (calcd 5.15, obsd 5.02), N (calcd 7.74, obsd
7.74). 400 MHz 1H NMR (CDCl3): 0.78 (m, 2H), 1.18 (m, 2H),
1.30 (m, 2H), 1.55 (m, 2H), 1.99 (m, 5H), 2.78 (m, 1H), 2.91
(m, 1H), 3.04 (d, 1H), 3.16 (m, 1H), 3.39 (m, 1H), 3.80 (m, 1H),
4.00 (t, 1H), 4.40 (m, 1H), 4.89 (m, 1H), 7.31 (complex, 5H),
7.49 (br t, 1H), 7.79 (d, 1H), 8.00 (m, 1H), 8.23 (m, 1H), 8.64
(d, 1H).
Refer en ces
(1) Tucker, T. J .; Lumma, W. C.; Mulichak, A. M.; Chen, Z.; Naylor-
Olsen, A. M.; Lewis, S. D.; Lucas, R.; Freidinger, R. M.; Kuo, L.
C. Design of Highly Potent Noncovalent Thrombin Inhibitors
that Utilize a Novel Lipophilic Binding Pocket in the Thrombin
Active Site. J . Med. Chem. 1997, 40, 830-832.
(2) Tucker, T. J .; Lumma, W. C.; Lewis, S. D.; Gardell, S. J .; Lucas,
R. J .; Baskin, E. P.; Woltmann, R.; Lynch, J . J .; Lyle T. A.;
Appleby, S. D.; Chen, I-W.; Dancheck, K. B.; Vacca, J . P. Potent
Non-Covalent Thrombin Inhibitors that Utilize the Unique
Amino Acid D-Dicyclohexylalanine in the P3 Position. Implica-
tions on Oral Bioavailability and Antithrombotic Efficacy. J .
Med. Chem. 1997, 40, 1565-1569.
(3) (a) Balasubramanian, N.; St. Laurent, D. R.; Federici, M. E.;
Meanwell, N. A.; Wright, J . J .; Schumacher, W. A.; Seiler, S. M.
Active Site-Directed Synthetic Thrombin Inhibitors: Synthesis
in Vitro and in Vivo Activity Profile of BMY 44621 and Analogs.
An Examination of the Role of the Amino Group in the D-Phe-
Pro-Arg-H Series. J . Med. Chem. 1993, 36, 300-303. (b) World
Patent Appl. 9425049, 9612499, 9509858; DuPont Merck Phar-
maceutical Co.
(4) Leclere, G.; Decker, N.; Schwartz, J . Synthesis and Cardiovas-
cular Activity of
a New Series of Cyclohexylaralkylamine
Derivatives Related to Perhexiline. J . Med. Chem. 1982, 25,
709-714.
(5) Beylin, V. G.; Chen, H. G.; Dunbar, J .; Goel, O. P.; Harter, W.;
Marlatt, M.; Topliss, J . G. Cyclic Derivatives of 3,3-Diphenyl-
alanine (DIP) (II), Novel R-amino acids for Peptides of Biological
Interest. Tetrahedron Lett. 1993, 34, 953-956.
N-[1-Oxo-3-(3-pyr idyl)-3-ph en ylpr opyl]-L-pr olin e [tr a n s-
(4-Am in ocycloh exyl)m eth yl]a m id e (Dia ster eom er s 16
a n d 17). 16: Glassy amorphous solid, yield ) 30%. High-
resolution FAB MS: M+ calcd ) 435.27600, obsd ) 435.27530.
Elemental Anal. for C26H34N4O2‚2.50 TFA‚1.45H2O: C (calcd
49.97, obsd 49.55), H (calcd 5.33, obsd 4.93), N (calcd 7.51, obsd
7.47). 400 MHz 1H NMR (CDCl3): 0.95 (m, 2H), 1.39 (m, 4H),
1.76 (m, 2H), 1.92 (m, 2H), 2.06 (m, 3H), 2.21 (m, 1H), 2.91
(m, 2H), 3.04 (m, 2H), 3.15 (m, 1H), 4.41 (m, 1H), 4.78 (t, 1H),
7.04 (t, 1H), 7.28 (complex, 5H), 7.90 (d, 1H), 8.25 (m, 1H),
8.59 (m, 1H), 8.71 (s, 1H).
(6) Wierzbicki, M.; Hugory, P.; Duhault, J .; Cocour, F.; Boulanger,
M.; Eur. Patent Appl. 518769A1; Adir et al., Fr.
(7) Molecular modeling was performed as follows: Models were
constructed using AMF/C View (AMF is Merck proprietary
software). The ligands were energy-minimized within the context
of the active site using OPTiMOL with the MMFF force field.
The amino group of the P1 cyclohexyl ring was held fixed in the
position corresponding to the crystal structure of 1.
(8) Merck Research Laboratories, unreported results.
J M970397Q