Discovery and initial structure-activity relationships of trisubstituted ureas as thrombin receptor (PAR-1) antagonists

Article abstract of DOI:10.1016/S0960-894X(01)00538-8

Thrombin is the most potent agonist of platelet activation, and its effects are predominantly mediated by platelet thrombin receptors. Therefore, antagonists of the thrombin receptor have potential utility for the treatment of thrombotic disorders. Screening of combinatorial libraries revealed 2 to be a potent antagonist of the thrombin receptor. Modifications of this structure produced 11k, which inhibits thrombin receptor stimulated secretion and aggregation of platelets.

Full text of DOI:10.1016/S0960-894X(01)00538-8

Bioorganic & Medicinal Chemistry Letters 11 (2001) 2691–2696
Discovery and Initial Structure–Activity Relationships of
Trisubstituted Ureas as Thrombin Receptor (PAR-1)Antagonists
James C. Barrow,a,* Philippe G. Nantermet,a Harold G. Selnick,a Kristen L. Glass,a
Phung L. Ngo,a Mary Beth Young,a Janetta M. Pellicore,a Michael J. Breslin,a
John H. Hutchinson,a Roger M. Freidinger,a Cindra Condra,b Jerzy Karczewski,b
Rodney A. Bednar,b Stanley L. Gaul,b Andrew Stern,b Robert Gouldb
and Thomas M. Connollyb
aDepartment of Medicinal Chemistry, Merck Research Laboratories, West Point, PA 19486, USA
bDepartment of Pharmacology, Merck Research Laboratories, West Point, PA 19486, USA
Received 3 April 2001; accepted 1 August 2001
Abstract—Thrombin is the most potent agonist of platelet activation, and its effects are predominantly mediated by platelet
thrombin receptors. Therefore, antagonists of the thrombin receptor have potential utility for the treatment of thrombotic dis-
orders. Screening of combinatorial libraries revealed 2 to be a potent antagonist of the thrombin receptor. Modifications of this
structure produced 11k, which inhibits thrombin receptor stimulated secretion and aggregation of platelets. # 2001 Elsevier Science
Ltd. All rights reserved.
The enzyme thrombin plays a central role in thrombosis
and hemostasis, not only by enzymatic activation of
coagulation factors, but also by direct stimulation of
several cell types such as vascular endothelial cells,
smooth muscle cells, and platelets.1 Thrombin is the
most potent agonist of platelet activation, and recently
the mechanism of this activation has been revealed. In
1991 Coughlin et al. characterized the first thrombin
receptor, which is expressed on platelets, endothelial
cells, smooth muscle cells, and others.2 This 7-trans-
membrane domain G-protein coupled receptor is acti-
comes from the observation that in cells expressing
PAR-1, addition of the pentapeptide SFLLR-NH2
(thrombin receptor activating peptide, ‘TRAP’) mimics
the effect of thrombin on those cells, albeit at higher
concentration.
Interest in developing antiplatelet agents for treatment
of thrombotic disorders led us to consider the thrombin
receptor as an attractive target.5 Potential advantages
over current therapies are that blockade of the receptor
would prevent both platelet secretion of prothrombotic
and inflammatory mediators as well as aggregation.
Further, the enzymatic activity of thrombin would
remain intact, allowing for its anticoagulant properties.
Modulation of the thrombin receptor on other cell types
(e.g., vascular endothelial and smooth muscle cells) may
also play a beneficial role. A significant challenge in
developing a thrombin receptor antagonist is the
requirement to compete with an intramolecular ligand;
however, the fact that the TRAP peptide is a weak
agonist and the observation that the activated receptor
has a limited lifetime6 suggests that it is possible.
vated by
a novel mechanism whereby thrombin
proteolytically cleaves part of the extracellular N-term-
inal domain of the receptor, revealing a new N-terminus
ending in the residues SFLLR. This sequence then acts
as a ‘tethered ligand’ and binds to a site on the receptor
in an intramolecular fashion, causing receptor activa-
tion.3 This protein was therefore named protease acti-
vated receptor-1 (PAR-1), and subsequent to this
discovery, three other receptors that work by a similar
mechanism, PAR-2, PAR-3, and PAR-4, have been
identified.4 Support for the ‘tethered ligand’ hypothesis
Two complementary approaches directed toward find-
ing thrombin receptor antagonists are modification of
the endogenous TRAP ligand and screening of a large,
*Corresponding author. Tel.: +1-215-652-4780; fax: +1-215-652-
3971; e-mail: james_barrow@merck.com
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00538-8

2692
J. C. Barrow et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2691–2696
diverse sample collection.7 Efforts to develop agonist
and antagonist peptides and peptidomimetics based on
the TRAP sequence have resulted in high-affinity
ligands,8 exemplified by 1a (Fig. 1).8e Screening of
combinatorial libraries9 revealed urea 2 as a novel, non-
peptide structure with sub-micromolar activity. Mod-
ification of 2 has resulted in the discovery of potent and
selective thrombin receptor antagonists.
To examine replacements for the isopropyl group on the
central urea, different primary amines were used to dis-
place bromide 4 and the resulting secondary amines 9
were coupled to isocyanate 1011 to give ureas 11
(Scheme 2). To vary the dichlorophenyl portion of the
lead, epoxides were prepared from a variety of alde-
hydes12 and opened with excess cyclobutylamine
(Scheme 3). Regioselectivity for amine attack at the
least hindered carbon of the epoxide was greater than
5:1 under these conditions. The resulting b-amino alco-
hols were coupled to isocyanate 10 to give the desired
urea analogues 15. This method works in parallel
synthesis format with many aldehydes and amines to
provide an array of 1-aryl-N-alkyl-amino alcohols in
good yield and purity.
Chemistry
Preliminary investigations involved truncation of the
structure as well as modifications around the central
trisubstituted urea. Examination of the screening data
from the other compounds in the combinatorial library
that generated 2 revealed that the trisubstituted urea
portion of the molecule contained most of the impor-
tant binding elements. Thus, a series of simple amides
that replace the terminal two ureas were prepared as
outlined in Scheme 1.
Several conformationally constrained derivatives were
also prepared to try and deduce the bioactive con-
formation of the somewhat flexible lead. Cyclic urea 19
was synthesized as shown in Scheme 4. Michael addi-
tion of 4-(methylamino-t-butylcarbonyl)-aniline to
acrylate 16 in the presence of Yb(OTf)3 afforded the
b-amino ester 17. Ester reduction, mesylation and
displacement with isopropylamine provided diamine 18
which was converted to the cyclic urea 19 upon treat-
ment with phosgene. Five-membered cyclic urea 31
(Fig. 3) was prepared by the opening of 3,4-dichloro-
phenyloxirane with the sodium salt of N-phenyl ethyl-
enediamine urea.
The acetophenone derivative 3 was selectively reduced
with (+)DIP-Cl10 to afford the alcohol 4 which was
converted to the amino alcohol 5 by heating in an excess
of isopropylamine. Triphosgene mediated urea forma-
tion between 5 and three different anilines 6 (n=0, 1,
and 2) gave 7 which, after removal of the tert-butyl
carbamate protecting group, were coupled to a variety
of carboxylic acids to give the amides listed in Table 1.
Scheme 2. Ar=3,4-dichlorophenyl: (a) CH2Cl2, 23 ꢀC.
Figure 1. Lead structures of thrombin receptor antagonists.
Scheme 3. (a) Trimethylsulfonium iodide, KOH, H2O, AcN, 60 ꢀC; (b)
cyclobutylamine, neat, 60 ꢀC; (c) CH2Cl2.
Scheme 1. Ar=3,4-dichlorophenyl: (a) (+)DIP-chloride, THF, 0 ꢀC;
(b) i-Pr2NH2, neat, 60 ꢀC; (c) triphosgene, Et3N, CH2Cl2; (d) HCl,
EtOAc; (e) RCO2H, EDC, HOAt, DMF.
Scheme 4. Ar=3,4-dichlorophenyl: (a) 4-(methylamino-t-butylcarbon-
yl)-aniline, Yb(OTf)3, THF, 60 ꢀC; (b) LiBH4, THF; (c) MeSO2Cl,
Et3N, CH2Cl2; (d) i-PrNH2, 70 ꢀC; (e) phosgene, NaHCO3, CH2Cl2.

J. C. Barrow et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2691–2696
2693
Proline, thioproline and pipecolic acid derivatives were
prepared according to the synthetic sequence described
in Scheme 5. Conversion of l-proline to its Boc deriva-
tive and coupling with N,O-dimethyl hydroxylamine
provided Weinreb amide 21. Lithiation of 3,4-dichloro-
iodobenzene under halogen–metal exchange conditions
and addition to 21 lead to aryl ketone 22. Reduction
with NaBH4 resulted in a 4:1 mixture13 of diaster-
eomeric benzylic alcohols 23a and 23b which could be
separated by flash chromatography. Boc removal and
coupling with 4-(methylamino-t-butylcarbonyl)-aniline
in the presence of triphosgene afforded urea 25. Stereo-
chemical assignment of the carbinol center was achieved
by conversion of 23a to the corresponding fused bicyclic
oxazolidinone 24 and its analysis by 1-D and 2-D NMR
techniques. 5-Methyl and 5-ethyl proline were prepared
according to the method of Kikugawa14 as a cis/trans
mixture, which could be separated by flash chromato-
graphy after installation of the 3,4-dichlorophenyl moiety.
derivative 29b. Removal of the benzyl group and coup-
ling with isocyanate 10 led to urea 30.
Results and Discussion
The compounds prepared in this study were evaluated
for their ability to inhibit the TRAP-stimulated
3
(EC50=1 mM8a) secretion of H radiolabeled serotonin
from human washed platelets and are expressed as IC50
values.16 As shown in Table 1, replacement of the two
terminal ureas of 2 with a variety of amides results in
compounds with similar potency to the lead with lower
molecular weight and greater solubility. The optimal
chain length appears to be n=1; however, several of the
heterocyclic derivatives could tolerate different chain
lengths indicating some receptor flexibility. Efforts to
further simplify the structure by removal of this amide
or otherwise substitute the adjacent phenyl ring were
accompanied by large losses of activity (data not
shown); therefore, the 4-benzylic tert-butyl amide (e.g.,
8e) was used for subsequent studies.
Pyrrolidine and piperidine derivatives 28a and 28b were
prepared according to Scheme 6. The Boc derivative of
3-hydroxypiperidine was converted to the correspond-
ing ketone. Lithiation of 3,4-dichloro-iodobenzene
under halogen–metal exchange conditions and addition
to the previous ketone provided racemic alcohol 27. Boc
removal and coupling with 4-(methylamino-t-butylcar-
bonyl)-aniline in the presence of triphosgene lead to
urea 28.
Surprisingly, any modification to the 3,4-dichlorophenyl
ring resulted in large losses in potency as shown in
Table 2. Even the sterically similar 3,4-dimethyl sub-
stitution (15g) as well as other dichlorinated patterns
were significantly less active than the parent compound.
The 3,4-dibromophenyl (15f) and 4-cyanophenyl (15i)
were the only other groups to show any promise in this
area.
Morpholino derivative 30 was prepared according to
Scheme 7.15 Alkylation of bromide 3 with N-benzyl-
ethanolamine resulted in the formation of hemiketal
29a. Reduction with NaBH4 to the corresponding diol
and cyclization with HBr provided the morpholino
Scheme 7. (a) N-Benzyl-ethanolamine, CH3CN; (b) NaBH4, MeOH;
(c) HBr, 80 ꢀC; (d) a-chloroethyl chloroformate, CH2Cl2, 0–40 ꢀC; (e)
MeOH, 65 ꢀC; (f) 4-(methylamino-t-butylcarbonyl)-aniline isocyanate
10, CH2Cl2.
Table 1. Terminal urea replacements
Scheme 5. Ar=3,4-dichlorophenyl: (a) Boc2O, NaOH, dioxane,
water; (b) MeNHOMe, pyBOP, i-Pr2NEt; (c) 3,4-dichloro-iodo-
benzene, nBuLi, Et2O, ꢁ78 ꢀC; (d) NaBH4, MeOH, 4:1 ratio 23a/23b;
(e) HCl(g), EtOAc; (f) 4-(methylamino-t-butylcarbonyl)-aniline, tri-
phosgene, Et3N, CH2Cl2; (g) carbonyldiimidazole, CH2Cl2.
n=0
n=1
n=2
R
No.
IC50
(mM)a
No.
IC50
(mM)a
No.
IC50
(mM)a
tert-Butyl
CMe2NH2
2,4-Difluorophenyl
3-Pyridyl
4-(SO2NH2)phenyl
4-Benzimidazole
8a
—
8b
—
8c
8d
>10
—
>10
—
8e
8f
8g
8h
8i
0.7
1.0
0.9
0.7
0.85
0.45
8k
—
8l
8m
8n
8o
>10
—
>10
1.4
0.7
1
1
8j
0.36
aValues represent inhibition of secretion of radiolabeled serotonin
from washed human platelets stimulated by 3 mMof the TRAP pep-
tide and represent the average of three or more determinations with
the average standard error of the mean <10%.
Scheme 6. Ar=3,4-dichlorophenyl: (a) Boc2O, CH2Cl2; (b) SO3-pyri-
dine, Et3N, CH2Cl2; (c) 3,4-dichloro-iodobenzene, n-BuLi, Et2O,
ꢁ78 ꢀC; (d) HCl(g), EtOAc; (e) 4-(methylamino-t-butylcarbonyl)-ani-
line, triphosgene, Et3N, CH2Cl2.

2694
J. C. Barrow et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2691–2696
The most fruitful area of study was substitution of the
urea nitrogen with small alkyl groups. Table 3 demon-
strates a progressive increase in activity as the size of the
group increases, but with limits as cyclopentyl (11g) was
much worse than cyclobutyl (15a) and tert-butyl (11l)
was not as potent as s-butyl (11j and 11k).
case a potential H-bond acceptor or donor has been
removed (the OH in the case of 19, and the NH in the
case of 31).
Through examination of constrained analogues, which
retained the OH and NH groupings, proline and thio-
proline derived ureas of type 25 were identified as
potent antagonists of the thrombin receptor (Table 4).
The conformation of a trisubstituted urea is determined
in part by the nature of the substituents on the two
nitrogen atoms. Regarding nitrogen N1, both urea con-
formations A and B (Fig. 2) have to be considered.
Larger, branched R groups usually favor conformation
A and, as seen in Table 3, lead to potency enhancement.
This effect may be the result of increased hydrophobic
interaction with the receptor and/or enforcement of the
bioactive conformation. Conformation A might also be
stabilized by an intramolecular H-bond between the OH
and N2HAr, which is not possible within conformation
B.17
Urea 25a (derived from l-proline) blocked TRAP sti-
mulated activation of the thrombin receptor with an
IC50 of 1.5 mM(comparable to analogues 11c and 11d).
The lack of activity observed with any other diaster-
eomer of 25a (25b–d) confirmed the binding specificity
between antagonist 25 and the receptor. Piperidine
derivative 25e and thioproline derivative 25g exhibited
decreased activity, suggesting a five-membered or smal-
ler ring as the proper conformational constraint. The
lack of activity with sulfoxide derivative 25h suggests a
hydrophobic receptor pocket which accommodates the
N-R substituent. It was shown in Table 3 that further
branching of the R group (S-s-butyl) is beneficial to
activity. In order to mimic this effect, we introduced
alkyl substituents at the 5-position on the proline ring.
As hoped, potency was enhanced: the (5S)-methyl deri-
vative 25i was found to be a 0.4 mMantagonist of PAR-
1 while the (5S)-ethyl derivative 25k was slightly less
active. The most potent proline derivative 25i was still
less potent than the linear analogue 11k, suggesting that
a better substitution pattern or conformational con-
straint might be found.
We set out to examine various conformationally con-
strained derivatives of 2 in order to elucidate the bioac-
tive conformation. Initial efforts to mimic conformation
A or B lead to the preparation of six-membered cyclic
urea 19 and five-membered cyclic urea 31, respectively
(Fig. 3). Both compounds were inactive, but in each
Table 2. Terminal aryl modificationa
Further studies uncovered piperidine derivative 28b
(Fig. 4), in which the placement of the binding elements
is slightly altered, as a 0.5 mMantagonist of the throm-
bin receptor. Once again, ring size modification was
detrimental to activity as the pyrrolidine analogue 28a
was much less potent (>40 mM). In an attempt to fur-
ther refine the structure, the hydroxyl group was inser-
ted in the six-membered ring to yield morpholino
derivative 30, which was also much less active than the
hydroxy-piperidine derivative 28b. This negative result
further supports the identification of the hydroxyl group
as an H-bond donor or acceptor with the receptor (vide
supra). Racemic 28b is significantly more potent than
No.
R
IC50
(mM)b
IC50
(mM)b
15a
15b
15c
15d
15e
3,4-Dichloro
2,4-Dichloro
2,3-Dichloro
3,5-Dichloro
3,4-Difluoro
0.23
10
>10
>10
10
15f
15g
15h
15i
3,4-Dibromo
3,4-Dimethyl
3-Cyano
1
>10
>10
6
4-Cyano
aAll compounds in this table are racemic except for 15a, which has the
(S) hydroxyl configuration.
bSee Table 1, footnote a.
Table 3. Trisubstituted urea modification
No.
R
IC50
(mM)a
No.
R
IC50
(mM)a
Figure 2. Possible ground state conformations of trisubstituted urea.
11a
11b
11c
11d
11e
11f
15a
11g
H
Methyl
Ethyl
n-Propyl
n-Butyl
Cyclopropyl
Cyclobutyl
Cyclopentyl
>40
40
3.8
1.1
>10
6.7
8e
11h
11i
11j
11k
11l
iso-Propyl
iso-Butyl
0.7
>10
1.0
(ꢂ)2-Pentyl
(S)sec-Butyl
(R)sec-Butyl
tert-Butyl
Phenyl
0.4
0.23
1.7
0.23
>10
11m
>10
aSee Table 1, footnote a.
Figure 3. Cyclic urea derivatives.

J. C. Barrow et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2691–2696
2695
its enantioenriched linear analogues 11c and 11d, so
this result points the way to further pharmacophore
elucidation and the design of more potent antago-
nists.
nities to bind. There is also the possibility of involve-
ment of the PAR-4 platelet thrombin receptor. Further
studies are required to precisely establish the impor-
tance of these and other potential factors. As a final
measure of in vitro activity, 11k was evaluated for its
ability to block platelet aggregation using 1 nMthrom-
bin as agonist.16 As shown in Figure 5, at the higher
doses, 11k delays aggregation for several minutes as
measured by increases in light transmittance. Given the
sub-optimal activity of the compound in the secretion
assay with thrombin as the agonist, the delay of aggre-
gation in this assay is encouraging.
Several of the compounds were examined in a binding
assay using radiolabeled 1b8e as the ligand (Kd=13 nM)
with gel filtered platelets. As demonstrated in Table 5,
the binding affinity paralleled the functional data,
thereby supporting the notion that these compounds are
indeed interacting with the platelet thrombin receptor.
One of the most potent compounds from this series,
11k, was examined in further detail. The compound was
shown not to be an inhibitor of thrombin’s catalytic
activity (Ki>10 mM) and did not significantly prevent
ADP or collagen induced platelet activation
(IC50ꢃ10 mM), implying that it is indeed mediating its
effects through the platelet thrombin receptor. In the
secretion assay, 11k was again tested but with 1 nM
thrombin as the agonist. Disappointingly, 11k was only
able to elicit a 38% decrease in secretion at 1 mMand no
further reductions were possible at even higher con-
centrations of antagonist.18 This behavior was also
noted for peptide antagonists such as 1a and the reason
for diminished activity versus thrombin could be due to
a number of factors. The compounds may exhibit dif-
ferent binding affinity for cleaved (thrombin stimulated)
and uncleaved (TRAP stimulated) receptors. Alter-
natively, this class of compounds may have a fast off-
rate, thereby giving the tethered ligand more opportu-
Conclusion
Screening of combinatorial libraries has identified the
novel, nonpeptide thrombin receptor antagonist 2.
Examination of structure–activity relationships revealed
that the central disubstituted urea could be replaced
with a variety of simple amides, whereas the trisub-
stituted urea was less tolerant to change. Several con-
strained variants were prepared with piperidine 28b
being the most promising, and refinement of the acyclic
lead provided the more potent antagonist 11k. In vitro
binding and functional assays of this compound suggest
that it is indeed acting via the thrombin receptor and
has some activity when thrombin is used to generate the
tethered ligand. This structural class represents a pro-
mising area for further study with the hope that further
potency enhancements will improve activity versus
thrombin-stimulated platelets.
Table 4. Proline derived ureas
Table 5. Binding data for selected compounds
No.
TRAP
IC50 (mM)
Bind
IC50 (mM)a
8e
8i
0.7
0.85
40
6.7
0.23
1.5
0.7
1.2
6.6
2.1
0.12
1.1
1.0
No.
X
R
Stereo (2,20,5)
n
IC50 (mM)a
11b
11f
11k
25a
25i
25a
25b
25c
25d
25e
25f
25g
25h
25i
CH2
CH2
CH2
CH2
CH2
CH2
S
H
H
H
H
H
H
H
H
S,S
S,R
R,S
R,R
S,S
S,R
S,S
S,S-rac
S,S,S
S,S,R
S,S,S
1
1
1
1
2
2
1
1
1
1
1
1.5
>10
>10
>10
4.3
>10
2.5
>10
0.4
0.4
aValues represent the average of four or more determinations with the
average standard error of the mean <10%.
SO
CH2
CH2
CH2
Me
Me
Et
25j
25k
>10
0.7
aSee Table 1, footnote a.
Figure 5. Inhibition of 1 nMthrombin induced platelet aggregation by
11k.
Figure 4. Piperidine and morpholine derivatives.

2696
J. C. Barrow et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2691–2696
H. C.; Addo, M. F.; Darrow, A. L.; Eckardt, A. J.; Hoekstra,
Acknowledgements
W. J.; McComsey, D. F.; Oksenberg, D.; Reynolds, E. E.;
Santulli, R. J.; Scarborough, R. M.; Smith, C. E.; White, K. B.
Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 12257. (c) McComsey,
D. F.; Hawkins, M. J.; Andrade-Gordon, P.; Addo, M. F.;
Oksenberg, D.; Maryanoff, B. E. Bioorg. Med. Chem. Lett.
1999, 9, 1423. (d) Nose, T.; Satoh, Y.; Fijita, T.; Ohno, M.;
Nakajima, M.; Inoue, Y.; Ogino, Y.; Costa, T.; Shimohigashi,
Y. Bull. Chem. Soc. Jpn. 1998, 71, 1661. (e) Bernatowicz, M. S.;
Klimas, C. E.; Hartl, K. S.; Peluso, M.; Allegretto, N. J.; Sei-
ler, S. M. J. Med. Chem. 1996, 39, 4879. (f) Zhang, H.-C.;
Derian, C. K.; Andrade-Gordon, P.; Hoekstra, W. J.;
McComsey, D. F.; White, K. B.; Poulter, B. L.; Addo, M. F.;
Cheung, W.-M.; Damiano, B. P.; Oksenberg, D.; Reynolds,
E. E.; Pandey, A.; Scarborough, R. M.; Maryanoff, B. E. J.
Med. Chem. 2001, 44, 1021.
The authors thank Mary Becker for help in the pre-
paration of this manuscript, N. N. Tsou and R. G. Ball
for X-ray crystallography analysis, and J. Murphy, S.
Pitzenberger, S. Varga, A. Coddington, C. Ross and H.
Ramjit for analytical support.
References and Notes
1. Golsack, N. R.; Chambers, R. C.; Dabbagh, K.; Laurent,
G. J. Int. J. Biochem. Cell Biol. 1998, 30, 641.
2. Vu, T. K.; Hung, D. T.; Wheaton, V. I.; Coughlin, S. R.
Cell 1991, 64, 1057.
3. (a) Coughlin, S. R. Proc. Natl. Acad. Sci. U.S.A. 1999, 96,
11023. (b) Dery, O.; Corvera, C. U.; Steinhoff, M.; Bunnett,
N. W. Am. J. Physiol. 1998, 274, C1429. (c) Hou, L.; Howells,
G. L.; Kapas, S.; Macey, M. G. Br. J. Haematol. 1998, 101, 1.
(d) Brass, L. F.; Molino, M. Thrombosis Haemostasis 1997, 78,
234. (e) Grand, R. J. A.; Turnell, A. S.; Grabham, P. W. Bio-
chem. J. 1996, 313, 353. (f) Hollenberg, M. D. TiPS 1996, 17,
3. (g) Coughlin, S. R. Nature 2000, 407, 258. (h) Macfarlane,
S. R.; Seatter, M. J.; Kanke, T.; Hunter, G. D.; Plevin, R.
Pharmacol. Rev. 2001, 53, 245. (i) O’Brien, P. J.; Molino, M.;
Kahn, M.; Brass, L. F. Oncogene 2001, 20, 1570.
4. Dery, O.; Bunnett, N. W. Biochem. Soc. Trans. 1999, 27 (2),
246.
5. (a) Ray, A.; Hedge, L. G.; Gupta, J. B. Thromb. Res. 1997,
87, 37. (b) Ogletree, M. L.; Natarajan, S.; Seiler, S. M. Per-
spect. Drug Discov. Des. 1993, 1, 527. (c) Pakala, R.; Liang,
C. T.; Benedict, C. R. Thromb. Res. 2000, 100, 89.
9. (a) Hutchins, S.; Chapman, K. Tetrahedron Lett. 1995, 36,
2583. (b) Hutchins, S.; Chapman, K. Tetrahedron Lett. 1994,
35, 4055. Screening was performed using the assay described in
ref 16.
10. Brown, H. C.; Ramachandran, P. V. Acc. Chem. Res. 1992,
25, 16. The stereochemical assignment of 4 was made by analogy.
11. Prepared as a 2 Msolution in dichloromethane from 4-
(methylamino-t-butylcarbonyl)-aniline and phosgene.
12. (a) Bentley, T. W.; Jones, R. V. H.; Larder, A. H.; Lock,
S. J. J. Chem. Soc., Chem. Commun. 1994, 2309. (b) Bentley,
T. W.; Jones, R. V. H.; Larder, A. H.; Lock, S. J. J. Chem.
Soc., Perkin Trans. 2 1998, 1407.
13. 9:1 for pipecolic acid, 10:1 for thioproline, 9:1 for cis 5-
methylproline and cis 5-ethylproline, and 1:1 for trans 5-
methyl proline.
14. Li, H.; Sakamoto, T.; Kikugawa, Y. Tetrahedron Lett.
1997, 38, 6677.
6. Hammes, S. R.; Coughlin, S. R. Biochemistry 1999, 38,
2486.
15. (a) Houssin, R.; Vaccher, M. P.; Henichard, J. P.; Lesieur,
D. Heterocycles 1992, 34, 1343. (b) Perrone, R.; Berardi, F.;
Leopoldo, M.; Tortorella, V.; Lograno, M.; Daniele, E.;
Govoni, S. J. Med. Chem. 1992, 35, 3045.
16. Cook, J. J.; Sitko, G. R.; Bednar, B.; Condra, C.; Mellott,
M. J.; Feng, D. M.; Nutt, R. F.; Shafer, J. A.; Gould, R. J.;
Connolly, T. M. Circulation 1995, 91, 2961.
7. (a) Ahn, H. S. A.; Arik, L.; Boykow, G.; Burnett, D. A.;
Caplen, M. A.; Czarniecki, M.; Domalski, M. S.; Foster, C.;
Manna, M.; Stamford, A. W.; Wu, Y. Bioorg. Med. Chem.
Lett. 1999, 9, 2073. (b) Alexopoulos, K. A.; Matsoukas, J.;
Tselios, T.; Roumelioti, P.; Mavromoustakos, T.; Holada, K.
Amino Acids 1998, 15, 211. (d) Kato, Y.; Kita, Y.; Nishio, M.;
Hirasawa, Y.; Ito, K.; Yamanaka, T.; Motoyama, Y.; Seki, H.
Eur. J. Pharmacol. 1999, 384, 197.
17. X-ray crystallographic analysis on compound 11k
ꢀ
˚
revealed an N–O distance of 2.83 A with an angle of 144
which would be appropriate for an intramolecular H-bond.
8. (a) Feng, D. M.; Veber, D. F.; Connolly, T. M.; Condra,
C.; Tang, M. J.; Nutt, R. F. J. Med. Chem. 1995, 38, 4125. (b)
Andrade-Gordon, P.; Maryanoff, B. E.; Derian, C. K.; Zhang,
18. It is possible, however, to achieve greater inhibition of
thrombin-stimulated secretion by using lower concentrations
of thrombin agonist.